Evolution of East Asia's Arcto-Tertiary relict Euptelea (Eupteleaceae) shaped by Late Neogene vicariance and Quaternary climate change

Abstract

Background: The evolutionary origin and historical demography of extant Arcto-Tertiary forest species in East Asia is still poorly understood. Here, we reconstructed the evolutionary and population demographic history of the two extant Euptelea species in China (E. pleiosperma) and Japan (E. polyandra). Chloroplast/nuclear DNA sequences and microsatellite loci were obtained from 36 Euptelea populations to explore molecular structure and diversity in relation to past and present distributions based on ecological niche modelling (ENM). Time-calibrated phylogenetic/phylogeographic inferences and niche-identity tests were used to infer the historical process of lineage formation.

Results: Euptelea pleiosperma diverged from E. polyandra around the Late Miocene and experienced significant ecological differentiation. A near-simultaneous diversification of six phylogroups occurred during the mid-to-late Pliocene, in response to the abrupt uplift of the eastern Tibetan Plateau and an increasingly cooler and drier climate. Populations of E. pleiosperma seem to have been mostly stationary through the last glacial cycles, while those of E. polyandra reflect more recent climate-induced cycles of range contraction and expansion.

Conclusions: Our results illustrate how Late Neogene climatic/tectonic changes promoted speciation and lineage diversification in East Asia's Tertiary relict flora. They also demonstrate for the first time a greater variation in such species' responses to glacial cycles in Japan when compared to congeners in China.

Fig. 1

a Distribution of chlorotypes in 26 populations of E. pleiosperma (China) and 10 populations of E. polyandra (Japan) (see Additional file 2: Table S1 for population codes). Chlorotypes shared among populations are denoted by colour, while population-specific ones (‘E’) are white. The red and black dashed lines represent, respectively, the Sino-Himalayan/Sino-Japanese Forest boundary and the species’ boundary across the East China Sea region. The black dotted lines delimitate the four phylogroups of E. pleiosperma (i.e. population groups that share closely related chlorotypes). The black dotted line in Japan marks the boundary between ‘southern’ and ‘central’ phylogroups in E. polyandra. Abbreviations: TGMR, Three Gorges Mountain Region. b Dark-shaded areas in the inset indicate current mainland and island configurations, and light-shaded areas indicate exposed coastal areas and sea basins of East Asia during Late Pleistocene sea-level alterations owing to glaciations (modified after Park) [125]. c tcs-derived network of genealogical relationships between the 35 chlorotypes of the two species. Each circle denotes a single chlorotype with size proportional to frequency. Small open circles represent missing chlorotypes. The four clades (phylogroups) of E. pleiosperma identified by tcs were denoted as C1, C2, C3, C4, while those two clades (phylogroups) of E. polyandra were represented by J1 and J2. The baseline map was created by us using ArcGIS v.10.2.2

Fig. 2

a Distribution of ITS ribotypes in the 36 populations of Euptelea (see Additional file 2: Table S1 for population codes). Red and black dashed lines are identified in Fig. 1. b tcs-derived network of genealogical relationships between the 10 ribotypes of the two species. Each circle denotes a single ribotype, with size proportional to frequency. Small open circles represent missing ribotypes

Fig. 3

a Colour-coded grouping of the 36 Euptelea populations according to the three structure clusters (I–III) (see Additional file 2: Table S1 for population codes). Red and black dashed lines are identified in Fig. 1. b Histogram of the structure analysis for the model with K = 3 (according to both lnP(D) and ΔK; see Additional file 7: Figure S2). Each vertical bar represents one individual. The assignment ratio of each individual into one of the three clusters is shown along the y-axis. Each cluster is represented by a distinct colour

Fig. 4

a beast-derived chronogram of Ranunculales based on cpDNA (rbcL, matK) + 26S nrDNA sequences with calibration points denoted by nodes 1–4 (see text and Table 2 for details); and b the crown age of Euptelea and its phylogroups based on cpDNA (psbA-trnH, rpoB-trnC, rpL16, petN-trnC, matK, rbcL) sequences. Blue bars on each node indicate 95 % highest posterior density (HPD) confidence intervals for time estimates (in million yr ago, Ma). Posterior probabilities (PP > 0.5) are shown above nodes. Mean divergence times and 95 % HPDs are summarized in Table 2. Chlorotypes are represented by letter codes (H1–35). Codes of subclades (phylogroups) are identified in Fig. 1

Fig. 5

Potential distribution probability (in logistic value) of occurrence for Euptelea in East Asia. a At the present (0 kya); b at the Last Glacial Maximum (LGM: c. 21 kya) and c during the Last Interglacial (LIG: c. 130 kya). Presence records of E. pleiosperma in China (n = 99) and E. polyandra in Japan (n = 371) are plotted as purple and dark blue points in the maps, respectively. See text for details

Fig. 6

Niche identity test plots for five non-correlated (a, b) and nineteen (c, d) BIOCLIM variables between the two species of Euptelea as quantified by the standardized Hellinger distance (I) and Schoener’s D [91]. The vertical line in each plot represents the observed values of niche similarity (both D and I) while the histograms represent those of null distributions

Fig. 7

Results of the niche identity tests for five non-correlated (a, b) and nineteen (c, d) BIOCLIM variables between western and central-eastern populations of E. pleiosperma. The vertical line in each map represents the observed values of niche similarity (both D and I) while the histograms represent those of null distributions