From the forage pea (Pisum sativum var. arvense L.) marker databases, we selected 11 SSR markers transferability present on the genomes of P. sativum (Table 2). When we look at the population structure analysis results, optimal K value was found as 2 based on ad hoc procedure and Evanno method (Fig 1a, b and Fig 2). To further explain population structure and genetic variation of pea populations, we carried out Principal Component Analysis (PCA) based on genomic compositions. PCA results suggested two main clusters seperation but with some minor exceptions. Generally, 7 commercial cultivars and 11 populations of forage pea (Pisum sativum var. arvense L.) seperated each other as two main groups. Ulubatlı(10) and Özkaynak(11) cultivars exceptionally located in the center. In addition, Yerlisu(1), Ortadüzü(15) and Kars(18) populations were also located among 7 commercial cultivars contrary to expectations (Fig. 2). As another approach, genetic distances matrix were determined and Neighbor-joining (NJ) dendrograms were created to investigate genomic seperations of pea (Pisum sativum L.) populations. With the genetic association dendrogram created using NJ from the genetic similarity matrix, it was seen that 18 forage pea (Pisum sativum var. arvense L.) populations were generally divided into two main groups with small differences. Ulubatlı(10) cultivar were located close to the central part of the dendrogram, while Livioletta(17) and Kars(18) populations were among the cultivars as different groups (Fig. 3). In terms of genetic similarity, it was determined that the closest groups were Taşkent(3) and Töre(4) cultivars, and the most distant ones were Yerlisu(1) and Altınbulak(6) populations among all pea groups (Table 3).
We evaluated allele number, major allele frequency, gene diversity and polymorphism information content (PIC) parameters of 18 forage pea (Pisum sativum var. arvense L.) genotypes for genetic diversity analysis. Over all number of allele ranged from 6 to 11 with average of 6. Major allele frequncy ranged from 0,50 to 0,94 with average of 0,77, gene diversity ranged from 0,10 to 0,50 with average of 0,34, while PIC values ranged from 0,10 to 0,38 with average of 0,28 (Table 4). The 11 SSR markers successfully produced total 66 polymorphic bands by percentage of 89,2% for 18 forage pea populations. Primer “PSAD270” has the lowest number of polymorphic bands (3 bands), while primer “PSGAPA1a” has the highest number of polymorphic bands (11 bands) (Table 5).
Genetic diversity within the species allows the development of varieties with desired agronomic characteristics. In this respect, it is important to know the genetic diversity and genetic affinity of local varieties as well as commercial varieties. Because local varieties can play an important role in eliminating the deficiencies of commercial varieties. DNA based molecular markers are mainly used in the evaluation of genetic diversity [19,20]. Many marker techniques were developed for molecular characterization of pea (Pisum sativum L.) cultivars and populations [19,21,22] but SSR markers have been often preferred due to their high, level of polymorphism and reliability for pea (Pisum sativum L.) genetic diversity researches [1,20,23,24,25,26,27,28,29,30,31]. As first approach we carried out Population STRUCTURE Analysis based on Bayesian statistics and evaluated using the ad hoc procedure [11] and [12] methods with admixture model. STRUCTURE analysis results suggested that there were two main groups as cultivars and populations (Fig 1a,b, Fig 2) but this seperation was not sharp depend on genomic content. Generally, genomic differences were in the range of 0,30-0,60 with small deviations (Fig 2). [20] obtained similar results with K= 2 value from cultivar and breeding lines of pea (Pisum sativum L.) but with sharp seperation. [30] found that pea genotypes were classified to 2 clusters with population Structure Analysis. [32], reported that the classification of pea (Pisum sativum L.) accessions into 4 groups depend on model-based population structure analysis. As second approach, we carried out Principal Component Analysis (PCA) based on genomic compositions. PCA results suggested two main clusters seperation but with some minor exceptions. Generally, 7 commercial cultivars and 11 populations of forage pea (Pisum sativum var. arvense L.) seperated each other as two main groups. Ulubatlı(10) and Özkaynak(11) cultivars exceptionally located in the center. In addition, Yerlisu(1), Ortadüzü(15) and Kars(18) populations were also located among 7 commercial cultivars contrary to expectations (Fig 3, 4). In this study, we also drew the NJ tree dendrogram for the 18 forage pea (Pisum sativum var. arvense L.) genotype in addition to the PCA analysis. Pea populations were generally divided into two main groups with small differences. Ulubatlı(10) was located close to the central part of the dendrogram, while Livioletta(17) and Kars(18) populations were among the cultivars as different groups (Fig 4.). When the PCA and NJ analysis results are evaluated in general, while both 7 and Kars(18) populations are close to cultivars, Ulubatlı (10) cultivar is located in completely different regions in terms of their genetic structures. The closest genetic relationship was obtained between the two cultivars Taşkent(3) and Töre (4) cultivars, and the most distant ones were Yerlisu(1) and Altınbulak(6) populations among all forage pea (Pisum sativum var. arvense L.) groups (Table 3). The main reason why some genotypes were exceptionally found in different regions in PCA and NJ analysis is that these genotypes used in the study have genetic diversities based on distance. These results are similar to those of other studies. The UPGMA dendrogram as used SSRs showed two main groups for 11 commercial cultivars of pea (Pisum sativum L.) [26]. [30], found that pea genotypes were clearly seperated into three major clusters using PCA but with small exceptions again [1] constructed NJ tree using the Dice coefficient and SSR information and seperated pea (Pisum sativum L.) accessions into 3 groups. The genetic similarity among the 19 pea (Pisum sativum L.) was graphicall represented by UPGMA dendrogram using 5 SSR markers. The dendrogram showed two main groups[4].
As the last approach, we evaluted genetic diversity. The 11 SSR markers used in the present study showed polymorphic bands in 7 cultivars and 11 populations of forage pea (Pisum sativum var arvense L.). The primers produced a total of 66 alleles. These polymorphic alleles were ranged from 6 to 11 with average of 6. Primer PSAD270 (the number of 5) has the lowest number of polymorphic bands (3 bands), while primer PSGAPA1a (the number of 1) has the highest number of polymorphic bands (11 bands) (Table 4 and 5). This results is consistent with some previous studies. [33] reported that the mean number of alleles was close to 6 in a study on 20 P. sativum accessions and [19], determined the mean number of alleles per locus was 5 using 148 pea (Pisum sativum L.) accessions. [10], obtained the mean of alleles per locus was 5,9 in 20 pea (Pisum sativum L.) varieties and 57 wild pea (Pisum sativum L.) accessions using 10 SSR markers. [34], notified an avarage of 4,69 alleles per SSR locus in a research investigated with Pakistani pea. [35], stated that the mean number of alleles was 3,49. In addition, several studies have also reported that a wide range (3,4-9,9) in the average number of alleles per locus in pea [19,24,23,33,34,35,36,37,38,39]. This variability of the mean of allele frequency in per locus based on the type of marker system and the number of genotype [37]. In the present study, in all pea (Pisum sativum L.) groups the highest Polimorphic Information Content (PIC) value was detected as 0,38 but the least value was 0,10. The means of the PIC value was 0,28 (Table 4). Various PIC values reported in different researches conducted by pea (Pisum sativum L.) cultivars [10,23,38,40,41]. In a study conducted by 8 pea (Pisum sativum L.) cultivars, avarage PIC value reported as 0,62 using 188 polymorphic bands [23]. [29] notified an avarage PIC value of 0,29 for SSR markers varied from 0,01 to 0,56. Similarly, PIC value was found as the mean of 0,50 with range of 0,32-0,63 [34]. In another study conducted by [20], The minimum PIC value was found to be 0,095 while maximum PIC value was found to be 0,500 with avarage of 0,349. The PIC values is the demonstration of marker effectiveness. Therefore, it varies according to the number of markers and genotypes used in studies. Gene diversity states heterozgous. In this research this parameter found to be the mean of 0,34 with range from 0,10-0,50. Moreover, we determined Major allele frequncy as average of 0,77, ranged from 0,50-0,94. (Table 4).