Over the past decade, the geographic distributions of intraspecific cpDNA variations have been examined in a variety of plant species to elucidate their postglacial migration history (Petit et al. 2002; Mclachlan et al. 2005; Ikeda et al. 2009; Bai et al. 2010; Kikuchi et al. 2010; Lafontaine et al. 2010; Chou et al. 2011; Higashi et al. 2012; Sede et al. 2012; Cosacov et al. 2013). A principal goal of comparative phylogeography is to infer biogeographic history from recurrent patterns in the geographic distribution of genetic variation in co-distributed species (Bermingham and Moritz 1998; Moritz and Faith 1998; Avise 2000). Comparative phylogeography of woody plants has been especially useful in reconstructing Quaternary forest distributions, particularly in Europe and eastern North America (Palme et al. 2003, 2004; Heuertz et al. 2004; Grivet et al. 2006; Maliouchenko et al. 2007; Saeki et al. 2011), and similar types of studies on Japanese plants have been performed on tree species (Okaura and Harada 2002; Kanno et al. 2004; Okaura et al. 2007; Iwasaki et al. 2012) and alpine plants in temperate zones (Ohi et al. 2003; Fujii and Senni 2006; Ikeda and Setoguchi 2006, 2007; Ikeda et al. 2006, 2008a,b). However, to date, few comparative phylogeographic studies have been conducted for herbaceous species in the Japanese archipelago (Toyama and Yahara 2009).
The Japanese archipelago was connected to the Eurasian continent by four landbridges as a result of the lower sea level during the last glacial era (18,000–20,000 years ago): Sakhalin, the Kuriles, the Korean peninsula, and the Ryukyu Islands. These connections might have affect on the genetic structures and distributions of many plants (Hotta 1974). Phylogeographic studies on the Japanese archipelago elucidated several genetic disjunct distribution patterns. For example, genetic differentiation exists between populations along the Sea of Japan and the Pacific Sea sides of Japan (Okaura and Harada 2002; Tsumura et al. 2007a,b; Hiraoka and Tomaru 2009; Iwasaki et al. 2012). It might be thought that there are some mountain ranges from northeast to southeast in the Japanese mainland, probably acting as physical barriers to the migration or gene flow of many plant species (Tsukada 1980). The mountain ranges creates a climate boundary, dividing the main Japanese islands into zones with wet winter with heavy snow along the Japan Sea sides and with dry winter with small precipitation along the Pacific sides (Hiraoka and Tomaru 2009). Fagus crenata and Cryptomeria japonica have shown the genetic divergence between the sides of Japan Sea and the sides of the Pacific sea. Furthermore, a clearly phylogeographic break exists between the northern and central parts and/or western parts of Japan (Tsuda and Ide 2005, 2010; Fujii and Senni 2006; Ikeda et al. 2006, 2008a; Tsumura et al. 2007a,b; Sugahara et al. 2011). The most likely explanation for this phylogeographic pattern might be the multiple colonization events following different glacial episodes with the Japanese archipelago, during repeated glacial and interglacial cycles. Some alpine plants such as Pedicularis chamissonis and Primula cuneifolia have shown the divergence pattern between the northern parts and central parts. Fujii and Senni (2006) concluded that these divergence times corresponded to the period from the middle of the Pliocene (Tertiary) to the middle of the Pleistocene (Quaternary) in these two species. Some wood plants such as Chamaecyparis obtuse and Aesculus turbinate have shown the possibility that there have been refugia of high mountains in the western parts. Finally, phylogeographic divergence was detected between Hokkaido and Honshu islands (Aizawa et al. 2007; Hu et al. 2010; Ohsawa et al. 2011). Previous phylogeographic studies of Japanese broad-leaved trees (Fagus crenata, Okaura and Harada 2002; Betula maximowicziana, Tsuda and Ide 2005; Quercus mongolica var. crispula, Okaura et al. 2007) generally indicate more recent bottlenecks in the extant populations on Hokkaido than those in Honshu. The populations on Hokkaido probably recolonized from several southern refugia in Honshu during warmer postglacial periods. The genus Hosta has probably diversified because of speciation in the archipelago. Fujita (1976) classified 18 species and seven varieties in the genus Hosta. The congeneric species, Hosta sieboldiana and H. albomarginata of the Hosta genus, are the most common and widespread herbaceous species in the Japanese archipelago (Fig. 1). Hosta sieboldiana and H. albomarginata are not so close species in the genus Hosta. Hosta sieboldiana belongs to the section Helipteroides although H. albomarginata belongs to the section Nipponosta. However, the distributions of these two species are similar in Japan, and their overall morphologies and life forms are also similar. So, we think that these two species are good materials for comparative phylogeography. These two species are mostly pollinated by bumblebee species. The flowers of H. sieboldiana are self-compatible but require pollinators for seed production (Takahashi et al. 1993). Those of H. albomarginata are also weakly self-compatible and a facultatively xenogamous species (Suzuki et al. 2002). The seeds with wings of these two species are dispersed by wind. The common pollinator might visit the two species at the overlapping flowering seasons and the hybridization might occur. Although these species sometimes co-occur, they differ in their habitat preferences; H. sieboldiana mostly occupies cliffs around waterfalls and steep clines, whereas H. albomarginata widely inhabits wetlands. Furthermore, H. sieboldiana is endemic to Japan, while H. albomarginata is distributed in both Japan and Russia (Tamura and Fujita 2013). In addition to the Japanese archipelago, it occurs south of the Amurskaya and Primorsky Territories of the Eurasian continent, Sakhalin, and the Kurils of Russia (Gage et al. 2006).
Because the morphologies of these two species are very diverse, intraspecific taxa have been recognized. Populations of H. sieboldiana along the Sea of Japan have been treated as H. sieboldiana, although other populations were regarded as another species, H. montana (Maekawa 1940). Maekawa (1972) suggested that H. sieboldiana and H. montana can be discriminated in the length of peduncle, the color of perianth, the opened angle of the corolla, and the white powder of the back side of leaves. Hosta albomarginata has also been regarded as more than one species; the populations from the northern part of the Japanese archipelago have been treated as H. rectifolia Nakai, while those from the Kansai areas to the Chugoku areas of the archipelago have been treated as H. rhodeifolia F. Maek and the populations in the other regions of the archipelago have been regarded as H. albomarginata (Maekawa 1940). Maekawa (1940) insisted that the tube height of the perianth is shorter than the tube width of perianth in the northern populations of H. albomarginata. However, these characters could not discriminate between the northern populations and the western populations of H. albomarginata. Therefore, Fujita (1976) regarded these three species as one species, H. albomarginata, because the range of variations in morphological characters within H. albomarginata is not so large. These intraspecies subdivisions might reflect the population genetic structures within these species, which may have been generated by climate oscillations during the last glaciations (Qiu et al. 2009a,b; Lee et al. 2013). Moreover, the inference of different biogeographic histories for the two species has been based on the molecular marker (Albach et al. 2006; Maliouchenko et al. 2007; Toyama and Yahara 2009; Saeki et al. 2011). Therefore, we focused on the geographic pattern of cpDNA sequence variations in the two congeneric species, H. sieboldiana and H. albomarginata, to infer the phylogeographies of these species and to compare their distribution range histories. This study also serves as an example for elucidating different phylogeographic structures among closely related species differing in ecological habitats.