SSR polymorphism
Several studies about the genetic diversity and relationships of pistachio and wild Pistacia genotypes were performed using SSR markers (Vendramin et al. 2009; Baghizadeh et al. 2010; Pazouki et al. 2010; Arabnezhad et al. 2011; Kolahi-Zonoozi et al. 2014; Zaloglu et al. 2015; Motalebipour et al. 2016; Khodaeiaminjan et al. 2018). However, these studies are limited due to the low number of P. vera cultivars or SSR markers. The first development SSR study was reported using several pistachio cultivars by Ahmad et al. (2003, 2005). However, because there is no sufficient SSR markers for identification of genetic relationship and diversity, Kafkas et al. (2006) studied polymorphism of a wide germplasm of pistachio and comparison of different molecular markers such as AFLF, ISSR and RAPD. Recently, genetic characterization of the domestic Iranian pistachio cultivars was performed using SSR markers (Khadivi et al. 2018). To date, there is no genetic diversity study available with adequate number of SSR markers in large scale of pistachio cultivars with different origin.
The average allele value detected from previous studies were 3.30 (Ahmad et al. 2003), 2.90 (Ahmad et al. 2005), 2.75 (Baghizadeh et al. 2010), 1.96 (Pazouki et al. 2010), 2.80 (Arabnezhad et al. 2011), 2.75 (Kolahi-Zonoozi et al. 2014), 3.60 (Zaloglu et al. 2015), 4.50 (Motalebipour et al. 2016), 4.2 (Khodaeiaminjan et al. 2018), 2.73 (Karcı et al. 2020). In the present study, 576 alleles with an average of 6.26 alleles per locus derived from 92 SSR loci were detected. The reasons of the lower number of mean alleles in previous findings are insufficient number of markers, low number of cultivars and limited genetic variation within individuals. SSR loci derived from the coding and non-coding regions can be beneficial in molecular fingerprinting of P. vera L. if a study has high genetic diversities among accessions (Topçu et al. 2015; Khodaieminjan et al. 2018; Karcı et al. 2020).
In the current study, average PIC values were between 0.07 and 0.87 with an average of 0.58. The average PIC value in this study was higher than 0.33 and 0.44, respectively (Kolahi-Zonoozi et al. 2014; Baghizadeh et al. 2010), while it was lower than 0.64 (Khadivi et al. 2018). In the present study, average He and Ho values were 0.63 and 0.58, respectively. while Khadivi et al. (2018) reported them as 0.22 and 0.44, Kolahi-Zonoozi et al. (2014) reported as 0.35 and 0.49. Average values of Ho and He and PIC were 0.53, 0.56, and 0.51, respectively, in 18 P. vera cultivars (Khodaeiaminjan et al. 2018). The similar situation was observed a total of 51 EST-SSR loci developed from pistachio cv. Siirt and they generated He=0.40, Ho=0.38 and PIC=0.34 (Karcı et al. 2020). On the other hand, average Ho value was identified from 18 SSR loci as 0.64 by Arabnezhad et al. (2011) while average Ho value was calculated from four SSRs as 0.52 by Baghizadeh et al. (2010). In addition, He values were 0.45 and 0.75, respectively (Arabnezhad et al. 2011; Baghizadeh et al. (2010). As a result, polymorphism information content and observed and expected heterozygosities demonstrated variability due to using more or less polymorphic SSR loci and also higher or less narrow genetic diversity rate in the population.
Identification of cultivars using SSR markers is very efficient tool to select in early phase in different breeding programs (Ashkenazi et al. 2001). Some markers in this study can identify most important commercial cultivars. For example: Kerman (USA) and Siirt are the most cultivated female varieties in United States and Turkey can be identified by CUPBaPa1606 and CUPOhBa2127, respectively; while Peter and Kaska are the most cultivated male varieties can be identified by CUPOhBa1592 and CUPVEST2697, respectively. Locus-specific SSR markers are quite practical to determine resistance to disease in rice (Melaku et al. 2018), triploids in apple (Mazeikiene et al. 2019) and sugarcane sex (Pan et al. 2006). Thus, a total of 61 cultivar-specific loci that can identify 34 pistachio cultivars and genotypes presented in this study can be useful in future pistachio breeding programs.
Genetic relationships among pistachio cultivars
The results of SSR-based structural genetic analysis and UPGMA clustering were similar to the genetic relationships of pistachio cultivars were different origins (Turkey, Iran, Italy, Greece, Syria, and Tunisia) in the present study. Kafkas et al. (2006) divided 69 pistachio cultivars into three major clusters (Iranian, Mediterranean and outgroups) by using AFLP, RAPD and ISSR markers. Mediterranean cluster was subdivided to Turkish, Syrian and the closest genotypes to Siirt cv. according to genetic distance. In addition, the similar findings related with clustering were identified by Motalebipour et al. (2016) and Khodaeiaminjan et al. (2018). In genetic characterization studies performed by Khadivi et al. (2018) and Vendramin et al. (2009) Mumtaz and Ohadi were clustered in different groups. However, contrary to their findings, in the current study, they were categorized into the same cluster which agreed with the reports of Kafkas et al. (2006) and Kolahi-Zonoozi et al. (2014). In the current study, Cluster-I can be defined as Iranian cultivars and related genotypes. Cluster-II consisted of Siirt cultivar and genotypes closer to it although Cluster-III included Turkish and Syrian cultivars. Turkish cultivars Kaska, Bagyolu, Atli, and Uygur were replaced in Cluster-IV closer to Turkish and Syrian groups in the UPGMA analysis, while Structure analysis were clearly separated them as separate groups. Male-22 and Male-13 genotypes in Cluster-V were clustered in different group from others (Table 1, Fig. 1). The study performed by Kafkas et al. (2006) supported our results. Kafkas et al. (2006) conducted their study using AFLP, RAPD and ISSR markers and these technics have many disadvantages according to SSR marker technic such as applicable of the automation, less reliability and repeatable. Thus, although the obtained results in this study demonstrated similarity to their research, AFLP, RAPD and ISSR markers are not practical for getting results fast in the laboratory. Consequently, the most polymorphic and cultivars-specific SSR loci obtained from this study can be preferred for identifying cultivars in early stage of nursery and breeding programs for geneticist, breeders and farmers.