Sympatric genome size variation and hybridization of four oak species as determined by flow cytometry genome size variation and hybridization

Abstract The Quercus species serve as a powerful model for studying introgression in relation to species boundaries and adaptive processes. Coexistence of distant relatives, or lack of coexistence of closely relative oak species, introgression may play a role. In the current study, four closely related oak species were found in Zijinshan, China. We generated a comprehensive genome size (GS) database for 120 individuals of four species using flow cytometry‐based approaches. We examined GS variability within and among the species and hybridization events among the four species. The mean GSs of Q. acutissima, Q. variabilis, Q. fabri, and Q. serrata var. brevipetiolata were estimated to be 1.87, 1.92, 1.97, and 1.97 pg, respectively. The intraspecific and interspecific variations of GS observed among the four oak species indicated adaptation to the environment. Hybridization occurred both within and between the sections. A hybrid offspring was produced from Q. fabri and Q. variabilis, which belonged to different sections. The GS evolutionary pattern for hybrid species was expansion. Hybridization between the sections may be affected by habitat disturbance. This study increases our understanding of the evolution of GS in Quercus and will help establish guidelines for the ecological protection of oak trees.

Flow cytometry (FCM), which is a fast and effective tool to estimate GS, has been successfully applied to ploidy identification, cell cycle analysis, and species identification, including hybrids, rarely occurring cytotypes, and aneuploids (Francis et al., 2008;Hanusova et al., 2014;Vit et al., 2014;Zhang et al., 2019). It has been used successfully in plant genetic variation studies, genetic analyses, and breeding, as well as in studies of reproductive ecology, evolution, and plant system classification (Bilinski et al., 2018;Galbraith, 2004;Sharma et al., 2019;Spaniel et al., 2019). In the 1980s, Galbraith et al. (1983) developed a fast, efficient, and convenient method for isolating plant nuclei, which meant that FCM could be more widely used in botanical research, especially to determine plant GS.
Quercus L. is a genus that contains many economically and ecologically important tree species found in the northern hemisphere (Aldrich & Cavender-Bares, 2011). From an evolutionary point of view, Quercus is a good material for studying the species boundaries and adaptive evolution (Porth et al., 2016;Yuan et al., 2018;Manuel et al., 2020;Petersson et al., 2020). Quercus is famous for its remarkable natural hybridization, which makes it difficult to identify species owing to the formation of hybrids (Song et al., 2015).

F I G U R E 1
The geographical location of the study area and plant materials (a) Location of the research area in China; (b) Google Earth high-resolution image of Zijin Mountain; (c) Features of four Quercus leaves Additionally, the frequent hybridization events of oak trees lead to shaping the community assembly and structure, as well as to the evolution of species and the generation of new species (Cannon & Scher, 2017;Wetherbee et al., 2020). Closely related species rarely co-exist in the same forest land, because hybridization and introgression lead to species merging over time, eliminating their co-existence (Cavender-Bares & Pahlich, 2009;Pollock et al., 2015). Even when oaks occur in sympatry, there is significant gene flow. These may include the introgression events that lead to adaptation.
In this study, we found four closely related species of Quercus in Zijinshan, China that allowed us to study whether hybridization occurs between sympatric Quercus. The four oaks belonged to two sections in the genus Quercus. Quercus acutissima and Quercus variabilis belong to the Section Cerris, which is called the Old world clade, while Quercus fabri, and Quercus serrata var. brevipetiolata belong to the Section Quercus, which is called the New world clade (Manos et al., 2001). Members of the same sections easily form hybrids (Cottam et al., 1982), and there is an especially high frequency of hybrid formation within the Quercus section. We investigated whether hybridization occurs between sections. First, we determine the GSs of the four oak species, and then we determined whether there were F I G U R E 2 Quercus and Petunia hybrida fluorescence intensities. (a) Q. acutissim; (b) Q. variabilis; (c) Q. Fabri; (d). Q. serrata var. brevipetiolata. Peak 1: Oak nuclei at the G 0 /G 1 phase; peak 2, P. hybrida nuclei at the G 0 /G 1 phase. DI, mean nuclear DNA fluorescence index (oak/P. hybrida) variations in the GSs of these oak species. Finally, we analyzed the species hybridization within and between the two sections and determined the evolution of GS in these four oak species. (1 < DBH≤10), and trees (DBH > 10) were collected randomly and evenly in this forest stand.

| FCM measurement of the nuclear DNA content
Approximately 200 mg of oak leaves was used for the FCM analysis, and Petunia hybrida (2C = 2.85 pg) from the Nanjing Forestry University nursery was included as the internal standard plant (Marie & Brown, 1993 and 2.2 μl β-mercaptoethanol should be added to each ml of buffer before use (Favre & Brown, 1996;Marie & Brown, 1993

| Data analyses
The 2C DNA content was calculated using the following formula: The conversion ratio between the DNA mass and Base logarithm was 1 pg = 978 Mb (Doležel et al., 2003).
The DNA content data for 120 individuals were subjected to standardized processing. The average value and coefficient of variation (CV) were calculated and the variance in GS variation was analyzed using SPSS 19.0 software (SPSS, Inc.) and R studio 3.6.2. We set the GSs of the individual to be a variable, calculated the Euclidean genetic distances among the 120 samples and set 4 clustering categories. Finally, the hierarchical clustering was selected to realize the analysis with R studio 3.6.2.

| GSs of the four oak species
Owing to the low nuclear yields of mature leaves, we optimized Marie's Buffer for each of the four oak species. The CV values of the four oak species were all within the normal range, and the GSs of the individual species overlapped (Table S1). The results for the mixed samples prepared using oak and P. hybrida are shown in Figure 2. The oak and petunia formed narrow and high DNA peaks, respectively, and the petunia DNA peaks were generally lower than those of the oak peaks (Table 1).  Figure 3,

| GS variation among the four oak species' populations
To assess the GS variation among the four oak species' populations, repeated-measures ANOVAs were used (Table 2). There were signif- acutissima (Figure 4).

| Clustering analysis of GS variation within and between the four Quercus species
In this study, only GS variation parameters were used for cluster analysis without other indicator ( Figure 5)

| GS variation in different evolutionary branches of angiosperms
It has been reported that the GSs of angiosperms are significantly non-normal, and some lineages have extremely large GSs (Leitch & Leitch, 2013). In order to explore the characteristics of GSs of

| The GS evolution of four oak species
The GSs in 10,770 angiosperms have been estimated and ranged be- and Cerris are about five times larger than the Arabidopsis genome, which has a GS that ranges from 1.84 to 2.00 pg (Aldrich & Cavender-Bares, 2011). Our results corroborate those observed in earlier studies (Kremer et al., 2007;Leitch et al., 2019). However, the estimates for deciduous species calculated in this study were larger (1.93 vs. 1.64 pg/2C) than those calculated previously (Kremer et al., 2007).
A possible explanation was that GS evolution was unidirectional, resulting in a model for overall growth (Bennetzen & Kellogg, 1997).

| Intraspecific and interspecific variation in GS
In this study, considerable GS variability was found both within and among the four oak species. The GS estimate within the species of Section Cerris (1.08-fold) was greater than that of section Quercus when adapting to such an environment, resulting in large GS variability (Bilinski et al., 2018;Li et al., 2017). In addition, a comparison of mean GS estimates across two species in Section Cerris showed that the mean GS in Q. acutissima (1.87pg) is smaller than in Q. variabilis (1.92 pg). Some of the phenotypic traits, such as plant height, seed mass, cell size, and cell cycle time, may also facilitate GS variability (Benor et al., 2011;Kang et al., 2014;Knight et al., 2005). The differences in these phenotypic features may be related to growth rate, leaf anatomy, and photosynthesis. Previous continuous observations of the seedlings of biennial Q. acutissima and Q. variabilis revealed that the growth rate of the latter is less than that of the former . This is in line with the theory that plans with larger genomes have lower growth rates (Kang et al., 2014;Knight et al., 2005).
The use of GS may not be very useful for classification at higher taxonomic levels, but it is particularly valuable at the species level (Liu et al., 2020;Qiu et al., 2018;Zonneveld, 2008;Zonneveld et al., 2005). The GS variation is approximately 20% across species in a single genus. In woody plants, GS and chromosomal structure are highly conserved. Therefore, the interspecific variation between genomes is greater than the intraspecies variation (Chen et al., 2014). In this study, an analysis of variance showed that most of the variation is interspecific variation. The GS variation among populations, except for those of Q. serrata var. brevipetiolata and Q. fabri, showed significant differences. Several explanations for the interspecific variation have been proposed, such as repeated cycles of polyploidy, which is supported by genomic and isozyme evidence (Bowers et al., 2003;Otto & Whitton, 2000;Wendel, 2000). Earlier studies indicated that the interspecific variation in GS results partly from the appearance of extra B chromosomes, which are caused by the irregular segregation of additional chromosomes during mitosis (Piscor & Parise-Maltempi, 2015;Zoldos et al., 1998) (Dzialuk et al., 2007;Zoldos et al., 1998).
When the number of chromosome changes results from fission and fusion, then the evolution of the chromosomes may result in recombination between populations. Here, we speculated that the variation in GS was owing to hybridization. We suspect that ZJ116 and ZJ105 were hybrid individuals using a clustering analysis with GS expansion (2.024 pg and 2.06 pg). They may be hybrids resulting from crosses between two species in Section Quercus. The white oaks are wind-pollinated and unable to discriminate pollen from other species in the same section. In addition, we speculated that hybrid offspring having expanded GSs were produced from Section Quercus and Section Cerris. Cluster analysis showed that there were always Q. fabri and Q. serrata var. brevipetiolata individuals interspersed in the Q. variabilis population. The phenomenon of hybridization within different groups is rare and blocked by reproductive isolation, but it exists and has been reported (Burgarella et al., 2009). In general, a GS increases after polyploidization, but it may undergo a decrease in noncoding DNA sequences, leading to a reduced GS after polyploidization (Li et al., 2013). Furthermore, an increase or decrease in DNA repeat sequences during oak hybridization leads to variations in GS and is the main reason for GS changes in angiosperms.

F I G U R E 6
The total numbers of samples and random samples in each group. The phylogeny is adapted from Angiosperm Phylogeny Group IV

| Ecological protection proposals for oak trees
Interspecific hybridization and the introgression of Quercus leads to a series of systematic evolutionary and ecological results, such as community recombination and structural adjustment (Aldrich & Cavender-Bares, 2011;McVay et al., 2017;Song et al., 2015). In addition to the impact on community succession, the hybridization and introgression of oaks are conducive to increasing the genetic diversity and the rapid transformation and fixation of adaptive genes among species.

| CON CLUS IONS
In this study, the mean GSs of Q. acutissima, Q. variabilis, Q. fabri, and Q. serrata var. brevipetiolata were 1.87, 1.92, 1.97, and 1.97 pg, respectively, which were within a reasonable range. Thus, there was a low level of intraspecific variation in GSs among Q. acutissima, but it was relatively high among individuals of the other three oak species. Furthermore, there was a high level of interspecific variation among the four oak populations. The oaks in the same section produced hybrid introgression. Additionally, a hybrid offspring was produced from Q. fabri and Q. variabilis, which belong to different sections. The pattern of GS evolution for hybrids species is expansion. This study on GS described a valuable complementary method for studying genetic variation in oak species and has significance in guiding the ecological protection of oaks.

ACK N OWLED G M ENTS
This study was funded by the National Natural Science Foundation

CO N FLI C T O F I NTE R E S T
None declared.