Genome sequencing and population genomics modeling provide insights into the local adaptation of weeping forsythia

Abstract

Understanding the genetic basis underlying the local adaptation of nonmodel species is a fundamental goal in evolutionary biology. In this study, we explored the genetic mechanisms of the local adaptation of Forsythia suspensa using genome sequence and population genomics data obtained from specific-locus amplified fragment sequencing. We assembled a high-quality reference genome of weeping forsythia (Scaffold N50 = 7.3 Mb) using ultralong Nanopore reads. Then, genome-wide comparative analysis was performed for 15 natural populations of weeping forsythia across its current distribution range. Our results revealed that candidate genes associated with local adaptation are functionally correlated with solar radiation, temperature and water variables across heterogeneous environmental scenarios. In particular, solar radiation during the period of fruit development and seed drying after ripening, cold, and drought significantly contributed to the adaptive differentiation of F. suspensa. Natural selection exerted by environmental factors contributed substantially to the population genetic structure of F. suspensa. Our results supported the hypothesis that adaptive differentiation should be highly pronounced in the genes involved in signal crosstalk between different environmental variables. Our population genomics study of F. suspensa provides insights into the fundamental genetic mechanisms of the local adaptation of plant species to climatic gradients.

Keywords: Molecular ecology; Population genetics.

Fig. 1. Gene flow between groups and sampling localities of Forsythia suspensa.

a Sampling localities of 15 populations of F. suspensa, where colors distinguish groups; b maximum-likelihood tree and migration events among five groups of F. suspensa. The migration events are colored according to their weight

Fig. 2. Population genetic structure of Forsythia suspensa.

a Population genetic structure based on all SNPs estimated by ADMIXTURE analysis with K = 5. Each bar represents an individual, and the x-axis location specifies the sampling location. b The phylogenetic tree of all individuals of F. suspensa with F. viridissima and J. nudiflorum as outgroups was constructed using the neighbor-joining method in the program MEGA. c Each point represents an individual genotype, where colors distinguish groups. The amount of variation explained by PC1, PC2, and PC3 is 5.14%, 1.76% and 1.39%, respectively

Fig. 3. Manhattan plots showing signatures of selection in the BayeScan results.

Each dot represents a SNP, and a red dot represents a SNP under selection, with a cut-off value of a posterior probability above 0.76 and q value below 0.05

Fig. 4. Redundancy analysis of Forsythia suspensa showing the relative contribution of each environmental variable shaping the spatial genetic structure.

The biplot depicts the eigenvalues and lengths of eigenvectors for the RDA