The primary objective of conservation of genetic resources is to preserve the broad-based genetic diversity within each of the species that has a known or potential value to ensure their availability to current and future generations. The genetic diversity of a plant is structured at different spatial scales (for example, geographic areas, populations, between neighboring individuals) and largely designated by species' life-history characteristics, environmental impacts and demographic history (Engelhardt et al. 2014; Penas et al. 2016). Thus, conservation management plans often require knowledge of population dynamics, relative levels of genetic diversity within genetic structure of the species (Perez-Collazos et al. 2008). The importance of genetic diversity in maintaining biodiversity and evolutionary processes and in conservation biology studies of rare and endemic plant species has been recognized by researchers for decades (Laikre 2010).
In this study, genetic variation was investigated within and among C. amaena populations using ISSR markers. In general, endemic plant species tend to maintain low genetic diversity than widespread species (Ellstrand and Elam 1993). Contrarily, the genetic diversity of C. amaena was high at both population (P = 66.18 %, h = 0.252, I = 0 .372) and species (P = 78.43%, h = 0.306, I= 0.447) levels. These results may indicate that C. amaena did not have a history of severe or prolonged population bottlenecks sufficient to cause loss of genetic diversity. Similarly, many endemic species with high genetic diversity have been reported. For instance; Centaurea nivea (P= 91.88%, h = 0.296, I= 0.451), C. lycaonica (P:90.62%, h:0.2706, I:0.4148), Verbascum alyssifolium (P= 99.74%, h = 0.2651, I= 0.4206), Teucrium leucophyllum (P= 99.31%, h = 0.263, I= 0.418), Lilium regale (PPB: 97.3%, H: 0.198, I: 0.333) (Sözen and Özaydin 2009; Uysal et al. 2012; Wu et al. 2015; Hilooglu and Sozen 2017; Sozen et al. 2017).
The level of genetic diversity of C. amaena appears to be similar to that of the other endemic Centaurea species, although direct comparison is difficult when using different marker systems (eg AFLP, SSRs, allozyme). Freville et al. (2001b) investigated the genetic diversity of C. corymbosa via microsatellites and determined heterozygosity (He) values in the range of 0.36-0.62. By isozyme analysis of seven endemic Centaurea species, it was determined that heterozygosity values varied between 0.126 for C. cineraria subsp. cineraria to 0.276 for C. todari (Bancheva et al. 2006). A considerable amount of genetic variation was identified in endemic species Centaurea horrida (He=0.603–0.854) by using SSR markers (Mameli et al. 2008). In narrow endemic species of Centaurea tentudaica, quite high levels of genetic diversity were detected (P95 = 60.61, He = 0.287) by allozyme analysis (Moreyra et al. 2021). Mameli et al. (2008) suggested that high values of genetic diversity observed in these Centaurea species might have played a role in their survival in a challenging and stressful environment.
Reproductive biology of a species plays an important role in determining genetic variation at both the species and population levels. For instance, outcrossing taxa have the greatest diversity, while autogamous taxa have the lowest diversity (Hamrick and Godt 1996). Atasagun et al. (2018) determined that the breeding system of C. amaena was facultative xenogamous. This may be one of the reasons for high level of genetic diversity.
The greatest amount of genetic diversity in C. amaena was found within the population rather than among populations as estimated by Nei’s gene diversity (82.35%), Shannon’s information index (83.66%), and AMOVA (75.10%). Similar results have been previously reported in various studies of the following endangered species; C. horrida (Mameli et al. 2008), C. nivea (Sözen and Özaydin 2009), C. lycaonica (Uysal et al. 2012).
GST values above 0.30 indicate a high level of genetic differentiation, while GST values between 0.05-0.15 indicate a low level of genetic differentiation between populations. In C. amaena, the GST value was determined as 0.176, indicating moderate level of genetic differentiation among the populations. A wide variety of FST and GST values have been obtained from studies on Centaurea taxa (Table 4). High genetic diversity and low population differentiation in endemic and rare plants have been attributed to a number of factors; insufficient time to reduce genetic diversity following isolation, population size reduction and significant gene flow (Maguire and Sedgley 1997; Zawko et al. 2001).
Populations tend to diverge when gene flow has a low value, whereas when gene flow has a high value, populations tend to remain uniform (Geraci Anna et al. 2012). The Nm value indicates whether genetic drift can produce substantial genetic variation between populations. If Nm is high (≥1), gene flow is strong enough to avoid significant differentiation caused by genetic drift (Slatkin and Barton 1989). The value of effective gene flow (Nm) of C. amaena was found as 2.329, which indicates sufficient to avoid population differentiation due to random genetic drift.
A total of 48 genotypes of C. amaena from 2 populations were examined for genetic diversity by using the 10 ISSR primers in this study. The cophenetic correlation coefficient (r) among the populations was determined as 0.7821 using the normalized Mantel. This value shows that the dendrogram represented the similarity matrix very well and present analyzes were reliable. In the similarity analysis using the UPGMA method based on Dice similarity coefficient, two populations were obviously differentiated. In the UPGMA tree, two main clusters were observed. The first cluster was composed of P1 population individuals, whereas the second cluster had only P2 population individuals. It was observed that individuals belonging to each population were grouped together. It has been stated that this tree topology may be affected by the genetic structure of populations, which may be associated with genomic forces such as mutations, deletions and insertions (Filiz et al. 2014).
PCA analysis of C. amaena revealed the cumulative sum of the first two eigen values as 26.29%. Once the first two or three principal axes were able to explain 25% or more of the total variation, PCA may be more useful technique for grouping individuals with a scatterplot presentation (Mohammadi and Prasanna 2003). Similar to the phylogenetic tree, the results of PCA revealed that individuals of P1 and P2 were scattered from one another.
The Structure analysis of C. amaena (K=2) genotypes revealed that each population represented an independent unit, as all individuals were clustered according to their population status. This pattern was also supported by UPGMA and PCoA analysis, in which the genotypes clustered similarly. It also shows that the populations had a simple pedigree and that the genetic exchange between each pair of populations was low. Similar results were also observed in studies with endemic plant species with small and isolated populations (Petrova et al. 2017; Wang et al. 2020).
For analysis of variance, genotypes were classified according to 2 sub-populations from Structure analysis. The AMOVA analysis revealed the total variation among the populations as 24.89% and total variation within the populations as 75.10%. The variation rate within the populations was found to be significantly high (75.10%). Estimated FST value (Fst = 0.248, p < 0.001) value was found to be close to the mean level of among-population differentiation in endemic and narrow species (Nybom 2004).