Double migration of the endangered Tricyrtis formosana (Liliaceae) in Japan

Abstract

The Ryukyu Islands of Japan are a biodiversity hotspot due to geographical and historical factors. Tricyrtis formosana is a perennial herbaceous plant that commonly found in Taiwan. But only a few populations have been identified in a limited habitat on Iriomote Island, while populations of unknown origin occur near human settlements in an area on the main island of Okinawa. To better understand these populations of the phylogenetic uniqueness and intrinsic vulnerability, we conducted comparative analyses including (1) phylogeny and population structure with MIG-seq data, (2) photosynthesis-related traits of plants grown under common conditions and (3) transcriptome analysis to detect deleterious variations. Results revealed that T. formosana was split into two clades by the congeners and that Iriomote and Okinawa populations independently derived from ancestral Taiwanese populations in each clade. Photosynthetic efficiency was lowest in the Iriomote population, followed by Okinawa and Taiwan. Transcriptome analysis showed that the Iriomote population accumulated more deleterious variations, suggesting intrinsic vulnerability. These results indicate that each T. formosana population in Japan is phylogenetically unique and has been independently dispersed from Taiwan, and that the Iriomote population presents a high conservation difficulty with a unique photosynthesis-related characteristic and a larger amount of deleterious variations.

Figure 1

Locations of sampling sites of T. formosana with a legend indicating the number of samples per locations for MIG-seq. The gray shading in Taiwan represents areas with elevations of 2000 m or higher, which may act as barriers to the distribution of T. formosana. Shorelines were extracted from GSHHG, and the contour lines of 2000 m were generated from Digital Elevation Model of ETOPO 2022. These vector data were integrated, and the map was created using QGIS 3.22.

Figure 2

ML phylogenetic tree of T. formosana based on 233 SNP loci obtained by MIG-seq analysis using TVMef model. Labels (A–I) correspond to those in Fig. 1.

Figure 3

SPAD (chlorophyll content) and cross-sectional leaf length.

Figure 4

ETR (electron transportation rate) calculated at each PAR (photosynthetically active radiation) = 130–1640 mol quanta m⁻² s⁻¹. NT Northern Taiwan, MT Middle Taiwan, ET Eastern Taiwan, IR Iriomote, OK Okinawa population.

Figure 5

Box plots of genetic characteristics of the Taiwanese (northern, middle and eastern), Iriomote and Okinawa populations based on transcriptome analysis. Different alphabets indicate significant differences. (A) Genetic diversity based on the number of heterozygous synonymous SNVs per kb on the longest coding sequences, (B) ratios of deleterious amino acid variations in heterozygous SNVs estimated using PROVEAN, (C) ratios of non-synonymous SNVs calculated for each gene, (D) ratio of nonsense SNVs in the total non-synonymous SNVs.

Figure 6

Population structure of the 28 populations of T. formosana estimated by STRUCTURE analysis. The mean Fst at K = 3; Cluster 1 = 0.5319, Cluster 2 = 0.2824, Cluster 3 = 0.6326.