Global climate change has increased concerns regarding biodiversity loss. However, many key conservation issues still required further research, including demographic history, deleterious mutation load, adaptive evolution, and putative introgression. Here we generated the first chromosome-level genome of the endangered Chinese hazelnut, Corylus chinensis, and compared the genomic signatures with its sympatric widespread C. kwechowensis–C. yunnanensis complex. We found large genome rearrangements across all Corylus species and identified species-specific expanded gene families that may be involved in adaptation. Population genomics revealed that both C. chinensis and the C. kwechowensis–C. yunnanensis complex had diverged into two genetic lineages, forming a consistent pattern of southwestern-northern differentiation. Population size of the narrow southwestern lineages of both species have decreased continuously since the late Miocene, whereas the widespread northern lineages have remained stable (C. chinensis) or have even recovered from population bottlenecks (C. kwechowensis–C. yunnanensis complex) during the Quaternary. Compared with C. kwechowensis–C. yunnanensis complex, C. chinensis showed significantly lower genomic diversity and higher inbreeding level. However, C. chinensis carried significantly fewer deleterious mutations than C. kwechowensis–C. yunnanensis complex, as more effective purging selection reduced the accumulation of homozygous variants. We also detected signals of positive selection and adaptive introgression in different lineages, which facilitated the accumulation of favorable variants and formation of local adaptation. Hence, both types of selection and exogenous introgression could have mitigated inbreeding and facilitated survival and persistence of C. chinensis. Overall, our study provides critical insights into lineage differentiation, local adaptation, and the potential for future recovery of endangered trees.

Ping’ou hybrid hazelnut is one of the most profitable tree nuts in China, but economically important cultivars must first be genetically validated to meet industrial demand. Traditional approaches used for cultivar identification are mainly trait-based and unreliable. Previous approaches at the DNA level, focusing on the identification of species or/and varieties that originated in China, were not used widely in hybrid hazelnut because there was no proper standard sample. In this research, a multiplexed fingerprinting test was conducted to allow for hazelnut cultivar identification using SSR markers derived from European hazelnut. Twenty-seven SSR markers were used to fingerprint 57 genetically unique Ping’ou hybrid hazelnut and related wild species. All markers showed a high level of polymorphism, as indicated by mean values for observed heterozygosity (H_o = 0.84), expected heterozygosity (H_e = 0.80), and polymorphism information content (PIC = 0.78). A total of 301 alleles were detected, and the number of effective alleles varied from 6 for KG817 and GB818 to 18 for B654, with an average of 11.2 alleles per locus. Moreover, the Shannon’s information index (I) ranged from 1.293 for BR215 to 2.385 for B654, with an average of 1.908. The neighbor-joining tree, principal coordinate analysis, and Bayesian analysis revealed clear separation between hybrid cultivars and wild forms (Cluster/group I), as well as the differentiation within hybrid genotypes (Clusters/groups II and III). Additionally, the NJ dendrogram demonstrated a further split within Clusters/group III (III a and III b). Altogether, with the comparable SSR information of the European hazelnut cultivar ‘Barcelona’, the newly developed marker sets can assist in the germplasm identification of hazelnut cultivars and reproductive materials. Importantly, these combined SSR loci can be applied to characterize the genetic relationships and population structures among wild genotypes and hybrid cultivars, which will then provide information to guide hazelnut breeding based on their genetic background.

M. laxiflora is an endangered plant that grows in the Yangtze River floodplain of China and often suffers from flooding stress. Due to the lack of a reference genome for M. laxiflora, the molecular regulatory mechanism of waterlogging stress in this plant remains unclear. In this study, we report a high-quality reference genome of M. laxiflora with a size of 1.29 Gb. A total of 23,666 gene-encoding proteins and 5457 ncRNAs were predicted in this reference genome. A comparative genome analysis revealed that 902 and 4299 gene families significantly expanded and contracted, respectively, in M. laxiflora. The expansions of the 902 gene families were significantly related to the “response to stress”, “response to abiotic stimulus”, and “response to oxygen-containing compounds” pathways. In the M. laxiflora genome, 101 MylAP2/ERF genes were identified and divided into five subgroups. Several MeJA-, ABA-, and hypoxia-responsive elements were found in the promoter regions of these MylAP2/ERF genes. According to the transcriptome data analysis, 74 MylAP2/ERF genes responded to flooding stress. Moreover, three genes (MylAP2/ERF49/78/91) that belong to the same branch as the RAP2.2 gene exhibited different expression trends under flooding stress. Our results provide valuable information on the molecular regulatory mechanism of flooding stress in M. laxiflora.

M. laxiflora is an endangered plant distributed in the Yangtze River floodplain of China, which often suffers from flood hazards during its growth and development. Due to the lack of a reference genome for M. laxiflora , the molecular regulatory mechanism of its waterlogging stress remains unclear. Here, we report for the first time the high-quality reference genome of M. laxiflora by using HiFi sequencing and Hi-C. The total assembly size of the M. laxiflora genome was 1.29 Gb with a scaffold N50 size of 29.5 Mb. A total of 23,666 genes encoding proteins and 5,457 ncRNA were predicted in this reference genome. In the M. laxiflora genome, a total of 101 MylAP2/ERF genes were identified and divided into 5 subgroups. Many MeJA responsive elements, ABA responsive elements, and hypoxia responsive elements were found in the promoter regions of these MylAP2/ERF genes. Based on transcriptome data analysis, a total of 74 MylAP2/ERF genes can respond to flooding stress. Meanwhile, it was found that three genes (MylAP2/ERF49/78/91) which belong to the same branch with RAP2.2 gene exhibited different expression trends under flooding stress. Our results provide valuable information on the molecular regulatory mechanism of flooding stress in M. laxiflora .