Dipterocarpoidae genomics reveal their demography and adaptations to Asian rainforests

Abstract

Dipterocarpoideae species form the emergent layer of Asian rainforests. They are the indicator species for Asian rainforest distribution, but they are severely threatened. Here, to understand their adaptation and population decline, we assemble high-quality genomes of seven Dipterocarpoideae species including two autotetraploid species. We estimate the divergence time between Dipterocarpoideae and Malvaceae and within Dipterocarpoideae to be 108.2 (97.8‒118.2) and 88.4 (77.7‒102.9) million years ago, and we identify a whole genome duplication event preceding dipterocarp lineage diversification. We find several genes that showed a signature of selection, likely associated with the adaptation to Asian rainforests. By resequencing of two endangered species, we detect an expansion of effective population size after the last glacial period and a recent sharp decline coinciding with the history of local human activities. Our findings contribute to understanding the diversification and adaptation of dipterocarps and highlight anthropogenic disturbances as a major factor in their endangered status.

Fig. 1. Genomic statistics for our focal Dipterocarpoideae species.

The Asian rainforest where Dipterocarpoideae species dominate the emergent and canopy layers (a), and statistics for the assembled genomes of seven focal species (b–h). a Most Dipterocarpoideae trees at/above the canopy layer are taller than 50 m, with the height of the tower being 80 m (photo was taken by Hui Chen from Xishuangbanna, China). b–h Synteny analysis showed that most conserved syntenic blocks only exist within each set of four homologous chromosomes. A: Gene density. B: GC content. C: Repeat density.

Fig. 2. Phylogenomics and genome evolution of our focal species.

Results of phylogenomic (a) and whole genome duplication (WGD) (b) analysis. a The phylogenetic tree was constructed using the genomes of 19 Dipterocarpoideae species and two Malvaceae species (with our focal Dipterocarpoideae species highlighted in red and their samaras are shown to distinguish different species) (Supplementary Fig. 4b). The black dots represent the time calibrations for estimating the divergence time from Bansal et al. , Ng et al. , Bell et al. and Vega et al. , respectively. WGD events are represented by an orange circle (WGD-1) and orange triangles (WGD-2). b Ks distributions of the seven Dipterocarpoideae species are represented by different colors, and abbreviations of species are shown in Supplementary Table 2.

Fig. 3. Results of comparative genomics.

Comparative genomics between  five focal Dipterocarpoideae species and five temperate tree species (a), results of in vitro functional characterization of two antioxidation-related genes (b), the number of genes in some significantly expanded and contracted gene families (see Supplementary Table 15, c). a The number of genes with significant signals of positive selection and the number of significantly expanded/contracted gene families are represented by yellow, blue and red, respectively. MRCA the most recent common ancestor. b The synthesized standards (at the concentration of 10 ng/ml) and reaction products (treatments with enzyme added for three replicates) for each gene (Supplementary Table 12) were identified using LC–MS. c Abbreviations of species are shown in Supplementary Table 2, and source data are provided as a Source Data file.

Fig. 4. Demographic history and genetic load of H. hainanensis and H. reticulata.

Demographic history of H. hainanensis and H. reticulata (a), genetic load metrics of H. hainanensis and H. reticulata (b, c), and box plots of frequency distribution and total lengths of homozygous fragments in different length categories in the genome of each sampled tree of H. hainanensis and H. reticulata (d). a Demographic dynamics of H. hainanensis and H. reticulata were inferred with three different values of generation time (15, 20, and 30 years) using SMC++, and the Last Glacial Period (LGP) and the Last Glacial Maximum (LGM) is highlighted in light cyan and cyan, respectively. b, c Boxplots show the accumulation of the derived deleterious mutations (DDMs) and the derived major-effect mutations (MEMs) in individuals sampled from populations of H. hainanensis and H. reticulata. Boxplots (b–d) for both H. hainanensis and H. reticulata display biologically independent samples (n = 30 and 32) as data points. The center lines, box edges, and whiskers represent medians, the 25% and 75% quartiles, and the upper and lower distribution of 1.5 times of quartile range. Source data of b–d are provided as a Source Data file.