Genome Sequencing Reveals Hidden Lineages in China’s Wild Medicinal Forsythia

Wild populations of Forsythia suspensa, a traditional Chinese medicinal shrub, have long been harvested for their dried fruit, known as lianqiao, yet the breeding of elite varieties has been severely hindered by a lack of genomic understanding of its wild genetic resources. Now, a team led by researchers at the Hebei Academy of Agricultural and Forestry Sciences, Beijing Forestry University, and the China Academy of Chinese Medical Sciences has sequenced the whole genomes of 149 wild individuals sampled from 30 populations spanning the species’ core distribution range in China. The study, published in BMC Plant Biology, provides the first genome-wide portrait of the evolutionary history, genetic differentiation, and local adaptation of this economically important species.

Whole-genome resequencing works by aligning millions of sequencing reads from each sampled individual against a reference genome, then calling single nucleotide polymorphisms (SNPs) that differ among individuals. These SNPs serve as dense markers across the entire genome, allowing researchers to reconstruct phylogenetic relationships, quantify genetic diversity, and detect regions that have been shaped by natural selection. The team used this approach to generate a comprehensive SNP dataset that underpins all subsequent analyses in the study.

Phylogenetic reconstruction and population-structure analyses identified four major genetic lineages across the sampled range. Among these, the Shandong (SD) lineage exhibited pronounced genetic differentiation from the other lineages, standing apart as the most genetically distinct group. This deep separation suggests that the SD lineage has followed an independent evolutionary trajectory, potentially shaped by geographic isolation and unique local environmental pressures that differ from those acting on the remaining populations.

Demographic modeling using fastsimcoal2, which simulates allele frequency spectra under alternative evolutionary scenarios, revealed a long-term population contraction driven by Pleistocene climatic oscillations. These glacial-interglacial cycles repeatedly compressed suitable habitat, reducing effective population sizes across the species. However, the modeling also uncovered a transient rebound in the SD lineage, a recovery that preserved critical ancestral variation that might otherwise have been lost during the contraction. This finding highlights the SD lineage as a reservoir of genetic diversity with direct conservation value.

Partial Mantel tests, which partition the relative contributions of geographic and environmental distance to genetic differentiation, demonstrated that isolation by distance (IBD) primarily drove population differentiation across the range. In other words, populations that are geographically closer tend to be more genetically similar, reflecting limited seed and pollen dispersal. This spatial pattern facilitates the formation of regional ecotypes, locally adapted forms that arise when gene flow is too restricted to homogenize populations facing different selective pressures.

The most striking result emerged from a selective-sweep analysis of the highly differentiated SD lineage. Selective sweeps occur when a beneficial mutation rises rapidly in frequency, dragging nearby linked variants along with it and leaving a characteristic signature of reduced diversity and elevated differentiation in the genome. The team identified 315 candidate regions under selection in the SD lineage, and functional enrichment analysis showed that these regions are significantly enriched in plant-pathogen interaction genes and the phenylpropanoid biosynthesis pathway.

The phenylpropanoid pathway is central to the synthesis of numerous secondary metabolites, including the lignans and phenolic acids that underpin the medicinal activity of F. suspensa. The enrichment of selective-sweep candidates in this pathway, alongside plant-pathogen interaction genes, suggests that natural biotic stress acts as a key driver for the differential accumulation of medicinal components among lineages. In the SD lineage, sustained pathogen pressure may have favored alleles that simultaneously enhance disease resistance and alter the profile of bioactive compounds, linking adaptation to the very traits that make the species valuable.

These findings carry direct implications for germplasm conservation and breeding. The authors recommend systematic grid-based germplasm collection to capture the full breadth of wild diversity, with particular attention to the highly differentiated SD lineage as a potentially valuable genetic resource. They further propose the development of specific molecular markers within the 315 candidate regions for marker-assisted selection, a breeding strategy that uses DNA markers to track beneficial alleles without waiting for slow phenotypic evaluation. Such markers could accelerate the development of high-yielding and disease-resistant elite varieties.

By connecting Pleistocene demographic history, spatial genetic structure, and pathway-level adaptation, the study reframes F. suspensa not as a uniform wild resource but as a mosaic of locally adapted lineages shaped by climate and biotic stress. As demand for lianqiao continues to grow, the genomic map now available offers breeders and conservationists a rational basis for protecting and exploiting the wild diversity that underpins this traditional medicine.

Beyond its immediate breeding implications, the study offers a template for how genomic resources can be brought to bear on other traditional medicinal plants whose wild populations remain poorly characterized. Many such species share a similar profile with F. suspensa: restricted geographic ranges, fragmented habitats, long generation times, and a pharmacopoeia of secondary metabolites whose accumulation varies among regions. The analytical pipeline assembled here, from genome-wide SNP discovery through demographic modeling, environmental association testing, and selective-sweep detection, could be applied to these relatives to identify the lineages and loci most relevant to both conservation and crop improvement.

The demographic signal of long-term contraction deserves particular attention in the context of ongoing climate change. Pleistocene oscillations compressed suitable habitat for F. suspensa over tens of thousands of years, and the modeling suggests that effective population sizes declined accordingly. Modern warming may impose a comparable, if faster, compression of climatically suitable areas, and populations that survived past contractions, such as the SD lineage with its preserved ancestral variation, may again serve as refugial reservoirs. Conservation planning that prioritizes such reservoirs could safeguard the evolutionary potential of the species as environmental conditions continue to shift across its range.

The finding that isolation by distance dominates population differentiation also carries practical weight. Because geographically proximate populations are genetically similar, germplasm collected from nearby sites may capture largely redundant variation, whereas sampling across the full geographic extent of the species is needed to represent its diversity. This is precisely the logic behind the authors’ recommendation of systematic grid-based collection. The same spatial structure means that translocating material among regions for breeding purposes should be approached carefully, since mixing divergent lineages can disrupt locally adapted gene combinations even as it introduces useful variation.

The environmental dimension of the study, though secondary to geographic distance in the partial Mantel tests, is not negligible. The researchers screened nineteen bioclimatic variables from the WorldClim database for multicollinearity and retained four weakly correlated variables for their isolation-by-environment analysis. This careful variable selection matters because strongly correlated climate layers can inflate the apparent contribution of environment to genetic differentiation. The residual environmental signal, once geographic distance is accounted for, points to climatic factors that may act in concert with biotic stress to shape regional adaptation, a hypothesis that future reciprocal transplant or common garden experiments could test directly.

The enrichment of selective-sweep candidates in plant-pathogen interaction genes is especially intriguing given the biology of forsythia in cultivation. Fruit-bearing plants grown outside their native range often encounter novel pathogen communities, and alleles that confer resistance in wild populations may prove valuable when transferred into cultivated backgrounds. The co-enrichment of the phenylpropanoid biosynthesis pathway suggests that resistance and metabolite accumulation are not independent traits but may share a common regulatory basis, since phenylpropanoid-derived compounds frequently participate in plant defense responses. If this coupling holds, selecting for disease resistance could simultaneously influence the medicinal chemistry of improved varieties, a possibility that breeders will want to evaluate as marker-assisted programs mature.

Methodologically, the study illustrates the value of combining complementary inference frameworks. Phylogenetic and structure analyses delineated the four lineages, fastsimcoal2 compared fifteen demographic models using the Akaike Information Criterion to identify the most likely history, PSMC provided historical effective population size trajectories, and Dsuite quantified introgression among populations. No single method captures the full complexity of population history, and the convergence of evidence across these approaches strengthens confidence in the central conclusions. The supplementary materials, which include detailed sampling information, per-lineage diversity statistics, and the complete list of candidate genes within the 315 selective-sweep windows, provide a foundation for follow-up work by other groups.

For the broader community studying medicinal plant genomics, the identification of candidate regions tied to bioactive compound pathways represents a starting point rather than an endpoint. Confirming the functional effects of alleles within these regions will require association studies linking genotypes to metabolite profiles across diverse germplasm, and ultimately experimental validation of candidate gene function. The open-access status of the article and its accompanying data should facilitate such efforts, allowing researchers across institutions to build on the population framework established here.

In sum, the study moves F. suspensa from a largely uncharacterized wild resource to a genomically mapped species with clearly identified conservation priorities and breeding targets. The SD lineage emerges as the standout finding, both as a genetically distinct evolutionary unit and as a reservoir of variation shaped by pathogen-driven selection in a pathway of direct medicinal relevance. How quickly these insights translate into improved varieties will depend on sustained investment in germplasm collection, marker development, and the collaborative infrastructure needed to connect wild diversity with cultivated improvement.

Subject of Research: Whole-genome resequencing reveals population structure and local adaptation in wild Forsythia suspensa

Article Title: Whole-genome resequencing reveals population structure and local adaptation in wild Forsythia suspensa

Article References: Wu, N., Cao, L., Song, L., Wang, Y., Sun, H., Li, X., Yan, P., Jiang, T., Wang, Y., Wang, S., Dong, W., & Liu, L. (2026). Whole-genome resequencing reveals population structure and local adaptation in wild Forsythia suspensa. BMC Plant Biology. https://doi.org/10.1186/s12870-026-10052-4

Image Credits: AI Generated

DOI: 10.1186/s12870-026-10052-4

Keywords: Forsythia suspensa, whole-genome resequencing, population structure, local adaptation, selective sweep, phenylpropanoid biosynthesis, germplasm conservation, marker-assisted selection, Whole-genome, resequencing, reveals, population