Phylogeography of Populus koreana reveals an unexpected glacial refugium in Northeast Asia

Abstract

The genetic structure of temperate plants in the northern hemisphere was significantly influenced by the Quaternary climate oscillations. A species' biological characteristics and ecological niche are significant elements that can affect its phylogeographic history. We adopted the cold-tolerant, anemophilous and anemochorous tree, Populus koreana, as a model species to examine the impact of historical climate changes and biological characteristics on the evolutionary history of vegetation in Northeast Asia throughout the Quaternary period. The results showed that there is moderate genetic differentiation and a lack of phylogeographic structure among populations of P. koreana based on nuclear microsatellite and plastid markers. Demographic analyses and ecological niche modeling suggested that P. koreana is likely to have experienced a bottleneck around the last glacial maximum (LGM), followed by a rapid and continued range expansion coupled with a northward migration from the LGM to the mid-holocene (MH), present, and 2050. Notably, there were several separate refugia present throughout the range of P. koreana in Northeast Asia during the LGM. These include two widely recognized refugia located in the Changbai Mountains and the southern Korean Peninsula. We also unexpectedly found a previously unknown one in the northern Greater Khingan Mountains. Our study contributes to the understanding of the phylogeographic history of plant species in Northeast Asia, providing novel insights into the Greater Khingan Mountains as glacial refugia for a cold-tolerant tree species. These findings provide valuable insights into the Quaternary historical patterns of temperate forests in East Asia.

Keywords: Glacial refugia; Greater khingan; Northeast asia; Populus koreana; Quaternary.

Figure 1

(a) The main distribution range of Populus koreana. (b) Map of the 40 sampled P. koreana populations. Supplemental Table S1 contains the code and coordinates of each population.

Figure 2

Using 11 nSSR loci and DIYABC2.0, a schematic representation of the seven demographic scenarios for P. koreana was examined, along with model parameters.

Figure 3

Population genetic structure of P. koreana based on 11 nSSRs (n = 424). (a) The histogram illustrates the outcomes of the STRUCTURE assignment test with a value of K = 2. The representation of each individual is denoted by a vertical bar, which signifies the cumulative assignment probabilities to the two groups. The utilization of black lines serves the purpose of demarcating distinct populations. (b) The construction of a phylogenetic tree representing all populations of P. koreana based on DA. The genetic clusters found by STRUCTURE analysis are demarcated by branch colors. Please refer to Supplemental Table S1 for the corresponding population codes. (c) Principal coordinate analysis (PCoA) was conducted on a dataset consisting of 40 populations. The results revealed that Coord1 accounted for 33.20% of the variation, while Coord2 explained 15.68% of the variance.

Figure 4

(a) Chloroplast haplotype network of 42 haplotypes discovered in 40 P. koreana populations (n = 398). The size of the circles represents the relative frequency of each haplotype, while the presence of red dots signifies the absence of certain haplotypes. Bars indicate the number of mutations between haplotypes. (b) Haplotype frequency distribution in 40 P. koreana populations, colors correspond to the haplotype colors in (a).

Figure 5

Ecological niche modeling outputs of P. koreana (a)−(e) under (a) LIG, (b) LGM, (c) MH, (d) present and (e) future conditions, respectively. Warmer colors indicate higher probabilities of occurrence, and orange and yellow indicate medium and low probabilities, respectively.