Population study reveals genetic variation and introgression of four deciduous oaks at the junction between Taihang Mountain and Yanshan Mountain

Oaks (Quercus spp.) are considered as model plants for studying plant evolution and natural gene introgression. However, interspeci�c hybridization often occurred in sympatric sibling specie, resulting in blurred interspeci�c boundaries and hindering the development of breeding. Beijing area is at the junction between Taihang Mountain and Yanshan Mountain. It is a key area for deciduous oaks native to China and an overlapping area of several oaks. It is urgently necessary to evaluate the genetic diversity and population structure of the 4 deciduous oak species to understand the degree of gene introgression and to screen out ideal breeding materials. In this study, we collected 11 populations of 4 oak species (Q. variabilis, Q. mongolica, Q. dentata and Q. aliena) in the junction between Taihang Mountain and Yanshan Mountain. By using the polymorphic SSR markers, we analyzed the genetic variation of the collected 400 individuals and investigated the population structure and found gene introgression events. Q. variabilis had a clearer genetic background as compared to the other three species and the arti�cial population of Q. variabilis might have been transplanted from other regions outside Beijing. Q. mongolica had a more frequent gene introgression with Q. dentata and Q. aliena in this area. In addition, our results provided DNA �ngerprints of the sampled 40 individuals according 9 SSR markers. This research reported useful SSR data for analyzing the genetic variation of oak species native to China, laying the foundation of whole genome sequencing and conducting an oak germplasm nursery with clear genetic background.


Introduction
Oaks (Quercus spp.) are widespread all over the world. Quercus is a large genus in Fagaceae and oaks have been cultivated for a long history, especially in Asia, Europe and America, for their longevity and renewability (Plomion et al. 2018). There are over 450 Quercus species in the world and 51 in China. Oaks spread over south and north of China and play an important role of ecology and economy. In Beijing, the junction between Taihang Mountain and Yanshan Mountain, several deciduous oaks are natively distributed such as Chinese cork oak (Q. variabilis), Mongolian oak (Q. mogolica), Q. dentata and Q. aliena. The widespread of oaks has been described as "the evolutionary success". One of the explanations of the success is propensity for hybridization, contributing to adaptive introgression .
Oak species have been formalised into 2 subgenera: subgenus Quercus, comprising Ponticae, Protobalanus, Virentes, Lobatae and Quercus; and subgenus Cerris, encompassing sections Ilex, Cerris and Cyclobalanopsis (Liu et al. 2021;Hipp et al. 2020;Hubert et al. 2014). China, as a Quercus diversity centre, is home to 35-62 species. 4 native oak species are distributed in the same area, but they belong to different subgenus. Q. variabilis belongs to Cerris section East Asia Cerris with slender leaves, while Q. mogolica, Q. dentata and Q. aliena belong to Quercus section Roburoids with oval and irregular leaves ).
Studies have shown that interspeci c hybridization and gene introgression are common among species of deciduous oaks, which is increasingly recognized as an important process across diverse lineages of plants (Mir et al. 2006;Crowl et al. 2020). However, the introgression also makes it di cult to gure out the differences between two species with close relationship, leading to uncertainty when collecting the germplasm resources. Studies on gene introgression among Quercus species have been carried out for decades, and the phenomenon of hybridization and introgression has been clearly found in oak species widely distributed in Europe and North America, such as Q. rubra, Q. suber, Q. alba and Q. robur (Löpez-Aljorna et al. 2007; Moran et al. 2012;Crowl et al. 2020). A study in Japan also demonstrate the interspeci c hybridization of Q. mongolica and Q. dentata (Nagamitsu et al. 2020). Zeng et al. (Zeng et al. 2011) found that Q. mongolica and Q. liaotungensis had a high degree of gene introgression in Beijing area, with low degree of genetic differentiation between populations, leading to unclear species morphological characterization. However, the gene introgression among Q. variabilis, Q. mongolica, Q. dentata and Q. aliena have not been reported, and the genetic background of these oak species in Beijing area is still unknown. The ambiguity of genetic background also leads to the inaccuracy of some existing studies in representing a particular oak species.
DNA molecular markers especially SSR (Simple Sequence Repeat) and SNP (Single Nucleotide Polymorphism) markers have been developed for many years and have been widely used in recognition of species or cultivars (Lupini et al. 2019;Zurn et al. 2020). In coconut, 74 polymorphic SSR markers were identi ed by genotyping-by-sequencing (GBS) analysis of 40 coconut accessions and different traits could be divided into several groups, which was consistent with the amplicons of SSR markers (Riangwong et al. 2020). Recently, SSR markers also have been used in the identi cation of tree peony varieties and 18 SSR markers were found to be associated with the oral scent traits (Luo et al. 2021). In oaks, although SSR markers have been used to characterize the populations worldwide for many years , there were few studies focusing on large amount of oak genetic resources at the junction between Taihang Mountain and Yanshan Mountain, a speci c region that near the north border for Q. variabilis and Q. aliena natural distribution and near the south border for Q. mongolica natural distribution in China. Using SSR markers, we can generate the DNA ngerprint of the individuals, which can provide an exclusive identity information for each individual.
DNA ngerprints can be used to reveal and compare the relationship among individuals within a species or different species, and can be effectively applied to genetic analysis and breeding (Nybom 1991) (Hilde Nybom 2014). In pea, DNA ngerprint and ISSR marker analysis indicated that pea distributed in two places had a tendency of differentiation (Stavridou et al. 2020). Thus, germplasm from different regions were collected by authors for breeding and genetic improvement. Meanwhile, modern ngerprinting is crucial for varietal protection and germplasm characterization even for protecting plant breeders' rights. 5 microsatellite markers were used to distinguish different varieties of rice, ultimately solving the problem of mixing of 2 rice cultivars (Ganopoulos et al. 2011). As there is no oak cultivar native to China, germplasm is very important for plant breeding. It is urgent to analyze the genetic variation and introgression of native oaks in China, especially in Beijing area, ngerprinting the Chinese oak germplasm resources, thus we can conduct the oak germplasm nursery with clear genetic background.
Here, we collected 400 individuals of four oak species, including individuals both in natural and arti cial forests in the junction between Taihang Mountain and Yanshan Mountain. By using 9 SSR markers that were previously developed with high polymorphism, we analyzed the genetic variation among 4 oak species (Q. variabilis, Q. aliena, Q. dentata and Q. mongolica) and the genetic distance within one species from different area. Giving ngerprint of these oak individuals, we provided valuable data for further germplasm resource collecting and laid the foundation of the construction of oak germplasm resource nursery using oaks with clear genetic background, providing experimental basis for subsequent molecular breeding and other molecular biological and genetic studies of oaks native to China.

Sample distribution and plant materials
Brie y, in this study, we chose 6 representative regions for sampling in Beijing, which have abundant oak resources (see Supplementary Table S1 for details). Q. variabilis, Q. aliena, Q.dentata and Q. mongolica were studied in this work. Sampling sites are mostly forest parks rich in forest resources. Except for the Q. Genetic differentiation among oak populations and optimal population classi cation The analysis of the population structure was assessed in STRUCTURE 2.3.4 (Pritchard et al. 2000) using a Bayesian clustering approach setting parameters with a burn-in period of 100,000 iterations and 100,000 MCMC iterations after burn-in. In this study, we used "admixture model" and "allele frequencies correlated" during the modeling process. This approach revealed genetic structure by assigning individuals or prede ned groups to K clusters. Different K values that ranged from 1 to 10 were used to infer the number of clusters for 10 replicate runs. ΔK (ΔK=mean Genetic diversity of 11 oak population and PCoA analysis Na, Ne, Ho, He and xation index (F), Fst and gene ow (Nm) among 11 populations were calculated by GenAlEx6.5. The software was used to conduct the PCoA of the four oak species and analyze the length range of the ampli ed alleles among different species. According to the length range of the ampli ed alleles, histogram was showed by using SigmaPlot 12.5.

Polymorphism of the chosen SSR primers
As previous studies have validated that the 9 pairs of SSR markers have high polymorphism and stability (Wang 2012), we chose these SSR markers to identify the genetic background of the four oak species (Q. variabilis, Q. mongolica, Q. dentata, and Q. aliena). Meanwhile, the results of LD analysis showed 9 locus are in linkage equilibrium. First, we collected 400 individual samples (including 136 of Q. variabilis, 113 of Q. mongolica, 77 of Q. dentata, and 74 of Q. aliena) from 6 regions of Beijing (see Supplementary Table  S1 for detail). We used these 400 samples to validate the polymorphism of the 9 SSR markers. The 9 SSR markers ampli ed 215 alleles in total, with an average of 23.9 alleles per locus ( Table 1). Each SSR marker at least ampli ed 13 alleles for all the individuals. The observed heterozygosity (Ho) ranged from 0.45 to 0.82, with an average of 0.61. For all the observed loci, the Ho value was lower than the expected (He), which was ranged from 0.81 to 0.96 with an average of 0.87 (Table 1).

Population Structure And Gene Introgression
A neighbor-joining tree was conducted to show the genetic relationship among all the observed individuals. The results showed that Q. mongolica might be the rst evolved species among these species and Q. aliena and Q. dentata evolved separately. Most of the individuals of Q. variabilis had the longest genetic distance with the other three species. Notably, a few individuals of Q. variabilis were grouped with Q. dentata and Q. mongolica (Fig. 1), indicating that possible gene introgression has happened among species. Meanwhile, Q. dentata, Q. aliena, and Q. mongolica had more individuals grouped with each other, which suggested that gene introgression happened more frequently among these three species, con rming that these three species had closer relationship as compared to Q. variabilis.
The population structure of 400 oak individuals was also con rmed by principal coordinates analysis (PCoA). We rst used all the samples to do the PCoA analysis and the results showed that the rst coordinate separated Q. variabilis and the other three species while the other three species were grouped together (Fig. 2a). In addition, PCoA analysis by using the samples for each species in different regions could not generate the signi cant group ( Supplementary Fig. S1), the results show that there was no distinct genetic differentiation among the different populations from different regions in Beijing area for each oak species.
The 400 oak individuals were evaluated for population strati cation. The assumed number of clusters was set from K = 1 to 10. The Bayesian analysis by STRUCUTRE Harvester showed that the individuals were suitable to be divided into four groups (K = 4) (Supplementary Fig. S2). The STRUCTURE results showed that Q. variabilis (dark blue) has a relative clear genetic background but still a few individuals had gene introgression with Q. dentata (cyan) (Fig. 2b). This result was consistent with PCoA analysis and phylogenetic tree. Q. mongolica had abundant genetic diversity and had gene introgression individuals from Q. aliena and Q. dentata, indicating that Q. mongolica, had a closer genetic relationship with Q. aliena and Q. dentata as compared to Q. variabilis as we could also see the population structure when K = 2 and K = 3 ( Supplementary Fig. S3).
The regions where we collected samples are located on the edge of the North China Plain (Fig. 3), and they are near the borders of natural distribution of Q. mongolica and Q. aliena. Using the data processed from STRUCTURE, we drew the pies on the map of Beijing area to show the genetic background of all the populations. Southwest of Beijing had abundant species diversity of deciduous oaks, while northeast of Beijing we mainly collected Q. mongolica (Fig. 3).
We then used the Unweighted Pair-Group Method and the Arithmetic (UPGMA) algorithm to build the phylogenetic tree of populations of 4 species from different regions in Beijing area. Populations from one species were grouped together and Q. mongolica had a closer genetic relationship with Q. aliena than that with Q. dentata (Fig. 4). Notably, JF-Qv population, which was an arti cial population, had a further genetic distance with other three Q. variabilis populations, indicating that the individuals might transplanted from other area outside Beijing.

Dna Fingerprints Of Oak Species In Beijing Area
As there was no oak germplasm nursery with clear genetic background in China, here we provided the ngerprints of the observed 400 oak individuals. According to above results, 10 trees were selected for each species to facilitate future studies. The rst ve are the single genetic background, and the last ve are complex. The DNA ngerprints of these 40 individuals were represented with numbers that showed the allele length and letters that indicated the ampli ed SSR alleles (Supplementary Table S3 Here we conducted ne-scale study to reveal the genetic relationship of oak species at the junction between Taihang Mountain and Yanshan Mountain, taking a detail look at the genetic diversity within and among species, and also observed the gene introgression among these oak species. We collected 400 individuals of oak species in the area, which is the north border of North China Plain. In this area, several oak species are naturally distributed including Q. variabilis, Q. aliena, Q. dentata and Q. mongolica, which are species native to China. As there is no improved cultivar for any of these oak species native to China, it is meaningful to detect the genetic variation of these individuals, preparing for germplasm nursery using accessions with clear genetic background. Previous studies have screened some SSR markers with high polymorphism for oak species (Kampfer et al. 1998;Guo 2018). Here, we used these SSR markers and further con rmed the polymorphism of the picked 9 markers. For the individuals of the 4 oak species in Beijing area, these SSR markers exhibit high polymorphism and 215 alleles were detected in total and 23.9 in average (Table 1) Table 3), suggesting that some individuals in JF-Qa population might be transplanted from Shangfang Mountain. Gene ow determines the genetic structure and survival potential of future populations of a species (Hamrick et al. 1992), previous study reported a lower gene ow (Nm = 3.648) among the population of Q. variabilis than our study (Shi et al. 2017), which objectively re ected Q. variabilis distributed in the junction between Taihang Mountain and Yanshan Mountain had more survival potential. In addition, SF-Qa natural population has more abundant genetic diversity than JF-Qa arti cial population (Table 2), further proving that JF-Qa might be generated from the same provenance.
The phylogenetic tree combining with UPGMA analysis and the STRUCTURE analysis showed that Q. variabilis is evolved separately from the other three species (Figs. 1, 2 & 4), which is consistent with previous studies using SNP markers and limited individuals within one species ). As Hipp et al. ) reported, Q. mongolica, Q. aliena and Q. dentata belong to Quercus section Roburoids, while Q. variabilis belongs to Cerris section East Asia Cerris together with Q. accutisima and Q. dentata evolves separately from Q. mongolica and Q. aliena. The introgression of Q. variabilis and Q. accutisima have already been con rmed by using high-quality genomic resources (Fu et al. 2022).Our results also demonstrate that Q. mongolica and Q. aliena have a closer genetic relationship as compared to Q. dentata (Figs. 1, 2 & 4). The gene ow between BH-Qm and YM-Qm reached 17.61, much higher than the other 3 species. Considering the geographical distance between Baihua Mountain and Yunmeng Mountain, the two Q. mongolica populations might be evolved together and separated in recent time. , it could not be ignored when analyzing the oak population in Beijing area. The STRUCTURE analysis and the phylogenetic tree showed that gene introgression individuals were grouped with corresponding species and Q. mongolica has the most individuals that have gene introgression with Q. dentata and Q. aliena (Figs. 1 & 2). Q. variabilis has the least gene introgression individuals, suggesting that Q. variabilis has a phylogenetically distant relationship with the other three species. Presumably due to the lower successful rate of interspeci c hybridization with the other three species in natural conditions. This result provides guidance for distant hybridization breeding of Q. variabilis, which should take the hybridization compatibility into consideration. In addition, as previously reported that Q. variabilis has distinct owering time compared to the other 3 species (Liu 2020), which might also cause the less chance for interspeci c hybridization (Schermer et al. 2020). Previous study on Q. dentata and Q. aliena showed that the male owers of Q. dentata the female owers of Q. aliena in this area had a certain overlap (Liu et al. 2018), suggesting these 2 oaks had frequent outcrossing.
The junction between Taihang Mountain and Yanshan Mountain is not only near the north border of the natural distribution of Q. variabilis and Q. aliena but also near the south border of the natural distribution of Q. mongolica. Q. mongolica distributed on the mountains with an altitude over 600 m. The north part of this distribution area is high while the south part is low topographically with warm temperate subhumid continental monsoon climate. Considering the pollen dispersal of oak species and the northwest wind direction in spring in the junction between Taihang Mountain and Yanshan Mountain, it could be explained that southwest of Beijing had abundant species diversity of deciduous oaks, which may also due to the warmer climate (Fig. 3). Fine-scale spatial genetic structure of the oak species could be further we observed the leaf shape of Q. dentata were similar to Q. mongolica ( Supplementary Fig. S5), which were quite different from the leaf of Q. variabilis. Further studies could use SNP markers to link the morphological traits or environmental adaptions (eg. temperature, drought, salt and etc.) and the genotype and more oak species native to China should be identi ed by molecular markers. With the clear identi ed genetic background, we could take a use of these genetic resources for breeding. DNA ngerprints could give the putative cultivar an identity, which is very important for intellectual property protection. Here we selected 10 individuals for each tree species and provided their DNA ngerprints through STRUCTURE, PCoA and phylogenetic tree analysis (Supplementary Table S3). Further study could be done to get the link between DNA ngerprints and their morphological traits by using more SSR markers or SNP markers like studies in some other species (Riangwong et al. 2020;Luo et al. 2021), making it clearer to focus on some speci c trait breeding.
In conclusion, we detected the genetic variation of natural and arti cial population of four oak species native to China and con rmed the genetic relationship and gene introgression of Q. variabilis, Q. mongolica, Q. aliena and Q. dentata. Our work laid the foundation for collecting oak germplasm and screened some individuals with clear genetic background for whole genome sequencing and cross breeding.

Declarations
All authors declare that they have no competing interest of this work.