Population structure in Quercus suber L. revealed by nuclear microsatellite markers

Abstract

Quercus suber L. is a sclerophyllous tree species native to the western Mediterranean, a region that is considered highly vulnerable to increased temperatures and severe dry conditions due to environmental changes. Understanding the population structure and demographics of Q. suber is essential in order to anticipate whether populations at greater risk and the species as a whole have the genetic background and reproductive dynamics to enable rapid adaptation. The genetic diversity of Q. suber has been subject to different studies using both chloroplast and nuclear data, but population structure patterns remain unclear. Here, we perform genetic analyses on Q. suber using 13 nuclear microsatellite markers, and analysed 17 distinct locations across the entire range of the species. Structure analyses revealed that Q. suber may contain three major genetic clusters that likely result from isolation in refugia combined with posterior admixture and putative introgression from other Quercus species. Our results show a more complex structure scenario than previously inferred for Q. suber using nuclear markers and suggest that different southern populations contain high levels of genetic variation that may contribute to the resilience of Q. suber in a context of environmental change and adaptive pressure.

Keywords: Conservation; Cork oak; Glacial refugia; Population genetics; West Mediterranean.

Figure 1. Map of sampling sites.

Map showing the distribution of Quercus suber around the western Mediterranean and the location of the 17 sampling sites.

Figure 2. Clustering plots from the analyses of the full data set and the neutral data set.

Q-value plots from the STRUCTURE analysis of the 11 neutral loci for K = 2 (A), K = 3 (B) and K = 4 (C); Q-value plots from the MavericK analysis of 11 neutral loci for K = 3 (D).

Figure 3. Fst distance dendrogram.

Pairwise Fst consensus dendrogram for the 11 neutral loci data set, calculated with the UPGMA method from 1,000 bootstrap replicates.

Figure 4. PCA and CA of the 11 neutral SSR loci data set.

(A) Principal components analysis (PCA) showing axes 1 and 2; PC1 and PC2 explain 6. 85% of the variance (PC1:4.4%, PC2: 2.45%). (B) Correspondence analysis (CA) showing axes 1 and 2. Population labels as shown in Table 2.