Phylogenomics Of Carpinus And Ostrya (Betulaceae)

Carpinus Linnaeus and Ostrya Scopoli are two sister genera in the family Betulaceae and trees of two genera comprise the important forest components in the Northern Hemisphere.Two genera include around 58 species with main distributions in Asia.In the previous researches,phylogenetic studies of the two genera are mainly based on few individuals of each species,few species and several chloroplast fragments and/or ITS fragments with the limited information and usually conflicting relationships.The evolutionary relationship of the species within two sister genera are not accurately undstood by using traditional method.In recent years,as the development of the nextgeneration sequencing technologies,phylogenomics has been prospered,which provides a powerful way to study the complicated phylogenetic relationships.Based on a large number of samples collected from China and abroad,we carried out the phylognenomic study of Carpinus and Ostrya for the first time.We expected to interpret the phylogenetic relationships between and within two genera in three levels: morphology,DNA barcoding and genomics at the population levle.Gene flows between these species were also examined to resolve the phylogenetic conflicts.The results were summarized as follows:1)A total of 1,330 trees were sampled from 243 populations 44 species of two genera(Carpinus: 36,Ostrya: 8)for DNA barcoding at the population level.Phylogenetic results based on the DNA barcoding data(trnL-F and ITS)indicated that 44 species could be divided into four clusters,largely consistent with the traditional morphological classification.Cluster A includes all eight Ostrya species with the nutlets completely covered by bracts in the infructescence.Cluster B comprises sect.Distegocarpus of the genus Carpinus with the slightly exposed nutlets in the infructescence.Cluster C and Cluster D include species of sect.Carpinus,in which nutlets are totally exposed.Species within Cluster C contain an obvious lobes at the base of inner bracts,different from those in the Cluster D.Howerer,relatinships between four clauses are not well solved.2)To reconstruct a more convincing phylogenetic tree for these species of two genera,a high-quality O.rehderiana genome as the reference for the following whole genome sequencing analysis was assembled firstly.Based on phylogenetic results of DNA barcoding data at the population level,a total of 170 individuals from 33 species,representing most species within each of above four major clusters were sequenced.All different datasets were generated based on the reads mapping file.Two methods were used to reconstruct the phylogenetic trees: the joint analysis and the coalescent-based methods.Nine datasets were used in the joint analysis,including the whole genome SNP set,the whole chloroplast genome,the coding region,the first two codons,the third codon,the four degenerate sites,the intron regions,the intergenic regions and the repeat sequences.Two datasets(single copy gene coding regions and intron regions)and two softwares(ASTRAL and MP-EST)were used to establish phylogenetic trees.Four stable clusters(Cluster A,(Cluster C,Cluster C))were recovered from all the constructed trees,which are consistent with phylogenetic analyses of two barcoding DNAs and morphological traits.The multiple individuals from each species could be identified as the monophyletic for most species,while those from four groups of species could not be discerned as the species-monophyletic.O.chisosensis nested within O.knowltonii and C.dayongina within C.polyneura.Individuals of C.mollicoma,C.omeiensis and C.rupestris were mixed together a group without the clear divisions.Similarly,the individuals from C.stipulata,C.turczaninowii and C.fargesiana could not comprise the separate lineages.3)A strong gene flow was detected between major clusters and some species,especially those with the conflicting phylogenetic relationships through multiple methods to detect the hybridization/introgression such as the examinations of genetic structure and gene flow.Multiple lines of evidence suggested that hybridization/introgression contributed mainly to the phylogenetic conflicts found for species within these two genera.4)In the combination of ancestral distribution simulations,divergent time assessments,and demograpghic history simulations,we found that the ancestor of the two genera was distributed in Asia and differentiated approximately 20 million years ago(Mya).Most of the species in Ostrya were divergenced before 8 Mya,while most of the Carpinus species formed after 5 Mya,which suggested most members in the former genus are relic species in the Miocene and most in Carpinus are ‘young’ species.In addition,we also found the consistent population peak for all species in two genera during 2~3Mya.Therefore,the temperate vegetation replaced tropical vegetation in the Pliocene period(5.3-2.5 Mya)and some species have the alternative chances to occupy widespread areas in this age.The Ostrya species prefer to the marmer and wetter climates than the Carpinus species.The climatic changes in this stage probably promoted the origin of numerous species in the Carpinus and probably leaded to the mass extinction in Ostrya.