Phylogeography of Asparagus schoberioides Kunth ( Asparagaceae ) in Japan

To describe the phylogeographic structures of Asparagus schoberioides Kunth (Asparagaceae) in Japan, we investigated its nucleotide sequence variations with respect to its geographic distribution pattern. Sequencing of the internal transcribed spacer (ITS) 1 region in 29 samples of A. schoberioides revealed 20 polymorphic nucleotide sites. As a result, the 29 samples of A. schoberioides fell into 15 distinct haplotypes and phylogenetic analyses revealed these haplotypes fell into two major clades, Clade 1 and Clade 2. The haplotypes of Clade 1 were distributed chiefly along the Pacific Ocean side of Japan, while those of Clade 2 occurred mainly along the Japan Sea side. This result suggests that A. schoberioides has migrated via two routes in Japan.


Introduction
The genetic and geographic structures of natural plant populations are a consequence of ecological factors and historical events.Consequently, the amount of genetic variation within a species depends on its history and life strategy.Comparison of nucleotide sequences are the best way to analyze the history of plant populations, and phylogeographic studies employing nuclear DNA data have been used to test ever more sophisticated historical models [1][2][3].In particular, nucleotide divergences of the nuclear DNA (nrDNA) showed relative higher values than those of chloroplast DNA (cpDNA) [4], although recombination of nrDNA is biased the concepts of homology [5].
Asparagus schoberioides Kunth is a perennial herb bearing small flowers with yellowish green perianth in a raceme-like inflorescence and reddish colored berries.This species ranged from Far East Russia, northern China, Korea, Sakhalin, Japan and Taiwan [6,7].In Japan, its distribution extends from the northern-most part of Hokkaido down to Shikoku and Kyushu [6]. A. schoberioides is one of the most closely related species to the garden asparagus, A. officinalis L. [8], which is the most economically important species in this genus.While the garden asparagus is not native to Japan, it is cultivated there and has escaped from the crop fields [6].Recently, Ochiai et al. [9] and Ito et al. [10] reported successfully generating of interspecific hybrids between A. schoberioides and the garden asparagus.A natural hybrid between A. schoberioides and the garden asparagus has not been found to date, but the possibility of genetic introgression from crop to wild relatives suggests the escape of the garden asparagus from the crop fields may pose a substantial ecological risk for the environment and its biodiversity [2,[11][12][13][14][15][16].
In Japan, molecular approaches have been used to analyze the intraspecific genetic variation of a number of different plant species (e.g., Abies mariesii [17]; Aucuba chinensis and A. japonica [18]; Fagus crenata [19][20][21]; Purimula cunefolia [22]; Quercus serrata and its allied species [23]; Stachyurus praecox [24], Alpinia japonica, Arachnioides sporadosora, A. aristata, Daphne kiusiana, Elaeocarpus sylvestris var.ellipticus, Prunus zippeliana [25]).Most of these studies were conducted based on the chloroplast DNA (cpDNA) sequence.Similarly, in our previous study, we used primers designed by Taberlet et al. [26] and Nishizawa and Watano [27] to sequence approximately 3000 bp of 16 regions in the cpDNA of some Asparagus species [8].However, we detected little variation in these regions and concluded that the cpDNA is not suitable for phylogenetic analyses of this species.The internal transcribed spacer (ITS) region is a powerful tool for resolving historical relationships among populations of widespread plants [28].In fact, many researches with various species using ITS region had been reported in recent years [29][30][31][32].For example, Yokoyama et al. [33] analyzed the ITS region of Mitchella undulata Sieblod et Zucc. in Yakushima Island, they detected intraspecific variations that suggested this population had undergone rapid morphological modification.Thus, analysis of this region may permit plant relationships to be evaluated down to the intraspecific level.
The aims of this study were to describe the phylogeographic structures of A. schoberioides especially in Japan on the basis of its ITS region and to analyze the distribution of these genetic variations relative to the distribution of A. schoberioides.

Plant Materials
Twenty-nine samples of Asparagus schoberioides were examined in this study (Table 1). A. kiusianus Makino and A. officinalis were selected as outgroups on the basis of phylogenetic analyses of the genus Asparagus [8].Vouchers for all species sampled in this study have been deposited in the Herbarium, Graduate School of Science, Tohoku University (TUS) and the Herbarium of Tsumura Laboratory (THS).

DNA Extraction, Amplification, and Sequencing
Total DNAs were isolated from 200 -300 mg of phylloclades with a Plant Genomic DNA Mini Kit (VIOGENE, Sunnyvale, USA) used according to the manufacturers' protocols.The isolated DNA was resuspended in TE and stored at -20˚C until use.
For phylogenetic analysis, we amplified the ITS1 region with primers designed by White et al. [34].Double-stranded DNA was amplified by incubation at 94˚C for 2 min followed by 40 cycles of incubation at 94˚C for 1.5 min, 48˚C for 2 min, and 72˚C for 3 min, with a final extension at 72˚C for 15 min.After the amplification, reaction mixtures were subjected to electrophoresis in 1% low-melting-temperature agarose gels to purify of the amplified products.We sequenced the purified PCR products using a DYEnamic ET-terminator Cycle Sequencing Kit (Amersham Pharmacia) and a Model 373A auto-mated sequencer (Applied BioSystems) according to the manufacturers' instructions.For sequencing, we used the same primers as those used for amplification.

Data Analysis
ITS region sequences were aligned with the CLUSTAL X program [35].Phylogenetic analysis and a test of clade support were conducted using the PAUP* program (version 4.0b10; [36]).Maximum parsimony analyses were carried out via a heuristic search with TBR branch swapping and MULPERS option.Multiple islands of the most parsimonious trees [37] were identified using the heuristic option with 100 random sequence additions.To estimate confidence levels of monophyletic groups, the bootstrap method with 1000 replications were employed [38].

Results
We determined the ITS1 region sequence in 29 samples of Asparagus schoberioides and two outgroup species, namely A. kiusianus and A. officinalis.The ITS1 region of all A. schoberioides plants was 249 bp in length.Variable sites in this region are shown in Table 2. Fifteen ITS1 haplotypes were obtained from A. schoberioides, and the sequence of each haplotype has been deposited in the DDBJ/EMBL/GenBank international DNA data bank (Table 1).There are no indels among the ITS sequences of A. schoberioides and its allied species.
When we used the ITS1 sequence data set in phylogenetic analyses, we obtained 78 most parsimonious trees of the 28 steps with a consistency index (CI) of 0.71 and a retention index (RI) of 0.92.One of the most parsimonious trees is shown in Figure 1.
In our phylogenetic tree, all A. schoberioides samples formed the monophyletic group and two clades (Clade 1 and Clade 2 hereafter) in A. schoberioides were recognized.The relationship between the clades and the localities of the individuals is indicated in Table 1.The correspondence between haplotypes and samples is shown in Figure 2. The geographic distribution of the samples is shown in Figure 3. Clade 1 consists of the following fourteen samples: China, South Korea, Hok-kaido1, Hokkaido3, Hokkaido4, Hokkaido5, Hokkaido6, Miyagi1, Miyagi2, Saitama, Yamanashi, Nagano1, Na-gano2 and Shizuoka.Clade 2 contains the following fifteen samples: Russia, Hokkaido2, Iwate, Niigata1, Nii gata2, Niigata3, Nagano3, Kyoto1, Kyoto2, Kyoto3, Hyogo, Shimane, Yamaguchi, Fukuoka and Nagasaki.Thus, in our phylogenetic analysis, the samples from Hokkaido and Nagano extend into the two lineages that form Clades 1 and 2 (Figures 1, 3).Apart from these samples, Clade 1 consists of samples from the Pacific

Discussion
The phylogenetic analyses in this study indicate that Japanese Asparagus schoberioides could not be detected the haplotype of the garden asparagus, suggesting that there are no genetic introgressions from the crops or the escaped individuals of the crop fields into wild relatives.Moreover, our results suggest that the current distribution in A. schoberioides has employed two routes of expansion.One is along the Japan Sea side while the other is along the Pacific Oceanic side.This is consistent with the geographical features of Japan, which has a lofty backbone range that basically runs along the axis of Honshu and divides the island into two areas, namely, the Pacific Oceanic and the Japan Sea sides.The range affect the climate of both sides and may have blocked or limited the migration of plants to the opposite sides.This is supported by studies of plant intraspecies variation between the two sides.For example, the leaf of Euptelea polyandra Siebold et Zucc.growing on the Japan Sea side is broader than that of the same species growing on Pacific Oceanic side [39].Yonekura and Ohashi [40] reported similar observations with regard to Bistorta tenuicaulis (Bisset et Moore) Nakai.Moreover, Fujii et al. [20] and Okura and Harada [21] found that a phylogenetic tree constructed on the basis of the cpDNA variability of Fagus crenata was roughly divided into two clades that were composed of populations growing on the Pacific Oceanic side and the Japan Sea side, respectively.Thus, the expansion of A. schoberioides may similarly have been limited by the mountain range, leading to two different lineages on the Pacific Oceanic and the Japan Sea sides.
With regard to our phylogenetic analyses, Clade 1 showed relatively little mutation accumulation, whereas many more nucleotide substitutions were observed in Clade 2 (Figure 1).This discrepancy suggests that Clade 1 had different colonization or migration time from Clade 2. Clade 1 consists of samples from the Pacific Oceanic side of Japan, South Korea and China.Since only one sample each from South Korea and China were employed, the genetic diversity of A. schoberioides in these regions could not be detected.There are two phytogeographic hypotheses of Japanese A. schoberioides belonging to Clade 1.One is that they had migrated from the mainland of China or the Korean Peninsula into Japan, at which point it expanded to Hokkaido and the Pacific Oceanic side.In this study, nucleotide variation of Clade 1 is very small, suggesting that there may have experienced extinct events in the western and the Japan Sea sides of Japan.The other is that they had migrated from the mainland of China into Hokkaido, and ex-panded rapidly to the Pacific Oceanic side from Miyagi to Shizuoka of Honshu.In this case, they had migrated from the mainland of China into Korean Peninsula independently.However, it is not certain which hypothesis to support from our results.
In contrast to Clade 1, our phylogenetic analyses suggest that a common ancestor of Clade 2 originated from more ancient times than that of Clade 1. Clade 2 consists of samples from the Japan Sea side of Japan and Sakhalin, and does not include samples from the mainland of China or the Korean Peninsula, although not enough samples were collected from the Asian Continent to be certain of this.Therefore, the relationships between the Japanese and Asian Continent populations of A. schobe-rioides remain unclear at present.
Our phylogenetic analyses also indicate that within each clade, the phylogenetic relationship and geographic distribution do not correlate.For example, some neighboring locations in Japan belong to different phylogenetic position within the clade.Examples of this are the three samples from Niigata and the three Kyoto samples in Clade 2 (Figure 1).This suggests that A. schoberioides populations in Niigata and Kyoto consist of individuals that have experienced different histories.Furthermore, northern samples such as those from Sakhalin, Hokkaido and Iwate also appeared in various phylogenetic positions in Clade 2 (Figure 1).These results indicate that the migration of A. schoberioides along the Japan Sea side of Japan has occurred repeatedly.
Thus, our phylogeographic analysis has outlined the different histories of A. schoberioides between the Pacific Oceanic and the Japan Sea sides in Japan.What has aided the rapid migration of samples in Clade 1 and the frequent colonization of samples in Clade?One answer may lie in the fact that A. schoberioides bears reddish berries that may be bird-dispersed [41][42][43][44].This may aid the rapid dispersal of seeds of A. schoberioides throughout the two routes of Japan.
In general, a morphologically recognized species may be regarded as a composite of subtly defined cryptic species each of which has equal status [45].In fact, such a definition takes the systematic approach to an extreme which would appear to be unworkable in plants.For example, Yatabe et al. [46] concluded that there are some cryptic species of Asplenium nidus L. (Aspleniaceae) in West Jawa on the basis of sequence variations using rbcL of cpDNA.In this study, A. schoberioides had nucleotide differentiation between Clade 1 and Clade 2, suggesting that there are cryptic species with the geographical features of Japan.

Conclusions
There are no genetic introgression between A. schoberioides and A. officinalis (garden asparagus) and two migration routes of A. schoberioides exist in Japan.Such research, when applied to A. schoberioides and to other Japanese plant taxa, will provide new insights into the phytogeography of Japanese plants.

Figure 1 .
Figure 1.One of the 78 most parsimonious trees of 29 samples of Asparagus schoberioides and outgroups.The numbers above the branches indicate the synapomorphic characters; the bootstrap values are indicated below branches (only those more than 50% are indicated on the tree).Arrowheads indicate branches that do not appear in the strict consensus tree.Oceanic side of northern and central Honshu in Japan, while Clade 2 is comprised of samples from the Japan Sea side of Honshu (Figures 1, 3).Moreover, some of the samples from the neighboring locations in Japan fall into different phylogenetic position within the clade (Figures 1).

Figure 2 .
Figure 2. Schematic of the correspondence between haplotypes and samples in Figure 1.