Genetic identification and evolutionary trends of the seagrass Halophila nipponica in temperate coastal waters of Korea

Although seagrass species in the genus Halophila are generally distributed in tropical or subtropical regions, H. nipponica has been reported to occur in temperate coastal waters of the northwestern Pacific. Because H. nipponica occurs only in the warm temperate areas influenced by the Kuroshio Current and shows a tropical seasonal growth pattern, such as severely restricted growth in low water temperatures, it was hypothesized that this temperate Halophila species diverged from tropical species in the relatively recent evolutionary past. We used a phylogenetic analysis of internal transcribed spacer (ITS) regions to examine the genetic variability and evolutionary trend of H. nipponica. ITS sequences of H. nipponica from various locations in Korea and Japan were identical or showed very low sequence divergence (less than 3-base pair, bp, difference), confirming that H. nipponica from Japan and Korea are the same species. Halophila species in the section Halophila, which have simple phyllotaxy (a pair of petiolate leaves at the rhizome node), were separated into five well-supported clades by maximum parsimony analysis. H. nipponica grouped with H. okinawensis and H. gaudichaudii from the subtropical regions in the same clade, the latter two species having quite low ITS sequence divergence from H. nipponica (7–15-bp). H. nipponica in Clade I diverged 2.95 ± 1.08 million years ago from species in Clade II, which includes H. ovalis. According to geographical distribution and genetic similarity, H. nipponica appears to have diverged from a tropical species like H. ovalis and adapted to warm temperate environments. The results of divergence time estimates suggest that the temperate H. nipponica is an older species than the subtropical H. okinawensis and H. gaudichaudii and they may have different evolutionary histories.


Introduction
Seagrasses are a polyphyletic group of monocotyledonous angiosperms that evolved in the marine environment approximately 100 million years ago (Mya) [1][2][3][4]. Seagrasses have adapted successfully to marine environments and play important roles in coastal and estuarine ecosystems, providing food, habitat, and shelter to a wide variety of marine animals [5][6][7]. Although seagrasses are distributed in nearly all coastal areas of the world, the global species diversity of seagrasses is extremely low (approximately 72 species) compared to terrestrial angiosperms [8][9][10].
Seagrasses can adapt to either tropical or temperate thermal regimes. There is roughly the same number of temperate and tropical genera and species, while a few genera and species occur in both climate zones [9]. The marine members of the family Hydrocharitaceae, which includes the seagrass genera Enhalus, Thalassia, and Halophila, do not usually occur where minimum temperatures are less than 20˚C [11,12], and consequently the genera in this family have been considered largely tropical [9,13]. Although most Halophila species occur in tropical or subtropical regions, some Halophila species such as H. australis are distributed in temperate regions [13]. H. nipponica also occurs in warm temperate areas of the northwestern Pacific [14][15][16][17].
H. nipponica was first observed in temperate regions of the Japanese archipelago approximately 100 years ago, and was subsequently treated as H. ovalis because species classification and identification in the genus Halophila are controversial due to morphological similarity and variability [1,12,14,18]. This temperate Halophila species was described as a new species, H. nipponica, in 2006 based on morphology [14]. In the same year, H. japonica was described using both morphology and genetics [19]; subsequently, H. japonica was treated as a synonym of H. nipponica [15]. The first observation of H. nipponica on the southern coast of the Korean peninsula occurred in 2007, with its identification based on morphology [16]. Since then, many meadows of H. nipponica have been observed along the warm southern coast of Korea [17]. H. nipponica is now known to be widely distributed in warm temperate Korean and Japanese waters, and considered endemic to Korea and Japan [15,16]. Since H. nipponica occurs in temperate coastal waters of the northwestern Pacific, but is not present in the tropical west Pacific and Indian Oceans, this species has been considered a temperate-adapted Halophila species [14,16]. However, growth of H. nipponica is minimal at water temperature less than 15˚C, and no growth reduction in high summer water temperature was observed, implying that this species still possess a tropical seasonal growth pattern [17].
The genus Halophila contains approximately 20 species and consists of five sections, a taxonomic rank between the genus and the species, based on morphological differences [10,13,14]. Most species in the genus Halophila are in section Halophila, which contains species with a pair of petiolate leaves borne on short erect lateral shoots [13,14]. All other species are in the sections Microhalophila (H. beccarii), Spinulosae (H. spinulosa), Tricostata (H. tricostata), and Americanae (H. engelmannii and H. baillonis) [13]. Although identification of Halophila species has been established by various taxonomic studies [14], molecular genetic studies proposed that Halophila species such as H. johnsonii and H. hawaiiana should be treated as conspecific with H. ovalis [20,21]. Therefore, further taxonomic and molecular genetic studies are required for more accurate species classification of the genus Halophila.
The internal transcribed spacer (ITS) region of nuclear ribosomal DNA (nrDNA) is phylogenetically informative and useful in understanding the evolutionary and biogeographic relationships among closely related taxa [3,15,[20][21][22][23][24][25]. DNA sequences of the ITS region evolve rapidly and may vary among species within a genus or among populations of the same species [23,24,26,27]. Phylogenetic analyses of the ITS region of nrDNA have been conducted to investigate the taxonomic status and to infer biogeographic trends in the genus Halophila, suggesting that a molecular phylogenetic study of this region is useful to differentiate major taxonomic groups within the genus Halophila [3,15,20,21,25]. In this study, we conducted a phylogenetic analysis of the ITS region of nrDNA to assess genetic variability among H. nipponica collected from populations across its known range in Korean and Japanese coastal waters and to elucidate position of the species H. nipponica among the taxonomic groups of the genus Halophila. The evolutionary trend of H. nipponica was inferred using DNA sequences of the ITS region from all assessed Halophila species.

Locations of Halophila nipponica and plant sample collection
H. nipponica is found on the southern coast of Korea and in temperate coastal waters of the Japanese archipelago with the exception of Hokkaido, the northernmost island of Japan ( Fig  1). H. nipponica shoots were collected for DNA extraction from five seagrass meadows on the southern coast of Korea at 5-m intervals using SCUBA (Fig 1). Fifty-four H. nipponica samples, including 20 from An-do Island (HN01-HN20), 8 from Sorok-do Island (HN21-HN28), 12 from Namhae Island (HN31-HN42), 10 from Koje Island (HN51-HN60), and 4 from Geomun-do Island (HN61-HN64), were collected for sequencing of ITS regions (Table 1). Plant samples of H. ovalis (n = 3) and H. minor (n = 3) were collected in Trang, Thailand. No specific permissions to collect research samples were required at the study sites, and the field study did not involve endangered or protected species. After collection, samples were cleaned with distilled water, desiccated, and stored at room temperature in silica gel for later DNA extraction.
Portions of each collection were preserved as herbarium voucher specimens, and deposited in the lab and the Herbarium of Kyungpook National University.

DNA extraction, PCR, and sequencing
Dried leaf tissue was ground in liquid nitrogen, and then genomic DNA was extracted using a DNeasy plant mini-kit (Qiagen, Valencia, USA), following the manufacturer's protocol. DNA extraction was checked using 1.5% agarose gel electrophoresis followed by ethidium bromide staining. Concentrations of genomic DNA were quantified using a NanoDrop (ND-1000) spectrophotometer.
Internal transcribed spacer (ITS) sequences in the nuclear ribosomal DNA (nrDNA), ITS-1, 5.8S nrDNA, and ITS-2, were amplified using the primer pairs ITS-1 (forward) and ITS-4 (reverse). All PCR reactions were performed using a PTC-100 thermal cycler (Bio-Rad, USA). The amplifications were done using QIAGEN Taq polymerase mixed manually with 10× PCR buffer, MgCl 2 , and dNTPs. The thermal cycling conditions were composed of an initial denaturation step at 94˚C for 2 min, then 35 cycles at 94˚C for 1 min, 50˚C for 1 min, 72˚C for 2 min, and a final extension time of 10 min at 72˚C. The annealing process was conducted for 1 min at 50˚C. PCR products were separated by 1.5% agarose gel electrophoresis followed by staining with ethidium bromide. Bands were excised from the agarose gel and purified using a QIAQuick Gel Purification kit (QIAGEN). DNA sequencing reactions were performed using ABI BigDye Terminator v3.1 cycle sequencing kits following the manufacturer's protocol. DNA sequences were obtained from an ABI 3730xl DNA analyzer.

Molecular analysis of ITS sequences of Halophila species
ITS sequences of Halophila species obtained in this study have been submitted to GenBank (http://www.ncbi.nlm.nih.gov/genbank/), and accession numbers were presented in Table 1 and S1 Table. Nucleotide sequence analyses were performed using BioEdit (Ver. 7.1.3) software for sequence compilation and alignment. Gaps were treated as missing data. Additional ITS sequences of H. nipponica from Japan and other Halophila species were obtained from the NCBI/GenBank database and included in the alignment (S1 Table). Approximately 130 ITS sequences of Halophila species were retrieved from the NCBI/GenBank database. Identical ITS=sequences of the same species at adjacent geographical locations were excluded, and the remaining 47 ITS sequences were included in the alignment to analyze phylogenetic relationships among species of the genus Halophila.
A maximum parsimony (MP) tree of ITS regions was obtained using the MEGA 5.1 program [28]. Neighbor-joining (NJ) analysis was performed using the maximum composite likelihood model with 1000 bootstrap replications [29]. The topologies of the phylogenetic trees from the MP and NJ analyses were almost identical, except for differences in bootstrap support values at some nodes. Thus, phylogenetic results from only the MP analysis are presented in this study. In the MP analysis, heuristic searches were performed with the Tree-Bisection- Evolutionary trends of Halophila nipponica Reconnection (TBR) branch-swapping algorithm. Support for the nodes of the MP tree was determined by calculating bootstrap values based on 1000 replications. Mean similarities between clades, and between species within clades, were calculated using the Kimura 2-parameter model and the numbers of sequence differences were counted using the estimation of pairwise distance model in the MEGA 5.1 program [30].
Relative divergence times of the clades in section Halophila, in which all species have a pair of petiolate leaves at each rhizome node, were estimated based on ITS sequence divergence to understand the evolutionary trend of H. nipponica using NETWORK 4.6 (http://www.fluxusengineering.com/sharenet.htm) software. Relative divergence times of morphologically similar Halophila species to H. nipponica such as H. okinawensis, H. gaudichaudii, and H. ovalis were also estimated based on ITS sequence divergence. To obtain estimates of the timing of divergence among Halophila species based on ITS sequences, we employed a consensus approach for ITS sequence diversity. Values between 1.72 × 10 −9 and 1.71 × 10 −8 mutations/site/year, which have been used as the range for ITS mutation rates in herbaceous plants, were employed as nucleotide mutation rates [31,32].

Phylogenetic position of Halophila nipponica
Halophila species in section Halophila were separated into five well-supported clades with 87-100% bootstrap support in the MP analysis (Fig 2). H. nipponica was grouped with H. okinawensis and H. gaudichaudii from subtropical regions into Clade I of section Halophila with 99% bootstrap support (Fig 2). In Clade I, H. nipponica and H. okinawensis were separated from H. gaudichaudii, and all H. nipponica samples from the Korean and Japanese coastal waters were included in a subgroup of Clade I (Fig 2).  (Fig 2). H. decipiens and H. stipulacea formed well-supported monophyletic groups, which received 100% bootstrap support in the MP analysis (Fig 2). Clade IV consisted of the single species, H. stipulacea, and Clade V consisted of the single species, H. decipiens. Although H. decipiens has a wide distributional range in tropical and subtropical waters, separation by geographical location was not supported by ITS sequence analysis (Fig 2). Halophila species which have complex phyllotaxy (H. engelmannii, H. beccarii, H. spinulosa, and H. tricostata) were clearly separated from species with simple phyllotaxy in section Halophila, except for one sample of H. australis (AB436923), in the phylogenetic tree (Fig 2). H. spinulosa and H. tricostata were grouped with 100% bootstrap support in the MP analysis and the group of H. beccarii and H. engelmannii was also supported well, with a bootstrap value of 99% (Fig 2).

ITS sequence similarity of Halophila nipponica
Similarities of ITS sequences among the five clades of section Halophila ranged from 91.7 to 95.9% (Table 2). Clade I, which includes H. nipponica, had the greatest similarity to Clade II, which includes H. ovalis, and the lowest similarity to Clade IV, which consists of a single species, H. stipulacea ( Table 2). Divergence of ITS sequences among species in section Halophila occurred primarily in the ITS-1 and ITS-2 regions, while only three polymorphic sites were found in the 5.8S region among the species analyzed (S1 Fig). Among the Halophila species in Clade I (H. nipponica, H. okinawensis, and H. gaudichaudii), five and nine polymorphic sites, which were mostly C/T mutations, were found in the ITS-1 and ITS-2 regions, respectively (S1 Fig). ITS sequences of H. nipponica from various locations in Korea and Japan were identical or showed less than 0.5% sequence divergence (3-bp difference) (Fig 3; S2 Table). ITS sequences of H. nipponica showed 1.1-1.6% sequence divergence (98.4-98.9% sequence similarity; 7-10-bp difference) from those of H. okinawensis from Japan and 2.1-2.5% sequence divergence (97.6-97.9% similarity; 13-15-bp difference) from that of H. gaudichaudii from Guam (Fig 3; S2 Table). H. nipponica also showed relatively high sequence similarity in the ITS region to Halophila species in Clade II (S2 Table). H. nipponica showed ITS sequence similarity of   Fig 3).

Relative divergence times of Halophila species
Clades IV and V, which each consisted of a single species, showed more recent relative divergence times than Clades I, II, and III, which each consisted of several Halophila species ( Fig  4A). Relative divergence times for Clade I, including H. nipponica, and Clade II, including H. ovalis, were 2.3 ± 0.6 and 3.5 ± 1.0 Mya, respectively (Fig 4A). H. okinawensis and H. gaudichaudii were the youngest species (1.5 ± 0.68 Mya) in Clade I (Fig 4B). Divergence times for H. nipponica and H. ovalis were 2.9 ± 1.08 and 8.7 ± 1.99 Mya, respectively ( Fig 4B).

Genetic variability in Halophila nipponica
Populations of H. nipponica have been reported only in the warm temperate coastal waters of Korea and Japan [10,[14][15][16]. When H. nipponica was described as a new species, this species was considered endemic to Japan [14]. Recently, many H. nipponica meadows have been observed in the coastal waters of Korea [16,17], but no studies on taxonomic or genetic similarities have been conducted between the populations of H. nipponica in Japan and Korea. In the present study, ITS sequences of H. nipponica plants from various locations across its geographic range in the coastal waters of Korea and Japan were identical or showed very low divergence (less than 3-bp difference). Divergence of ITS sequences within H. nipponica is much lower than the interspecific divergence (3.0-28.5%; 18-172-bp difference) found among other Halophila species [15,19,25]. Additionally, all H. nipponica collections from various locations in Korea and Japan were included in a group, and well separated from other Halophila species in the MP analysis. These phylogenetic results using ITS sequences suggest that H. nipponica is distinct from other Halophila species. Additionally, H. nipponica from Korea and Japan are confirmed to be the same species, and have nearly identical ITS sequences (0-3-bp difference).
In this study, Halophila species in the section Halophila were clearly separated from the species in other sections by molecular phylogenetic analysis of ITS sequences, and were grouped into five monophyletic clades. Historically, H. nipponica, H. okinawensis, and H. gaudichaudii were described morphologically as separate species on the basis of leaf dimensions [14]. Subsequently, H. nipponica grouped with H. okinawensis and H. gaudichaudii because of relatively high ITS sequence similarity and these three seagrasses were considered conspecific [15]. However, in our study of these species, including new samples collected throughout its current known range, H. nipponica has higher genetic similarity among all samples ( 3-bp) than it does to either H. okinawensis (7-10-bp) or H. gaudichaudii (13-15-bp) (S2 Table). ITS sequence divergence of 0-9-bp (0-1.45%) has been considered the level of intraspecific variation for Halophila species [15,25,33]. Additionally, H. nipponica, H. okinawensis, and H. gaudichaudii were well separated into 3 groups in the MP analysis. Thus, these 3 Halophila species should be considered distinct taxa at either the specific or subspecific level.
H. okinawensis and H. gaudichaudii occur in the subtropical region of the western Pacific, whereas H. nipponica occurs only in the warm temperate region of the northwestern Pacific  H. nipponica, H. okinawensis, and H. gaudichaudii) (Fig 3). Halophila species, which are more closely distributed geographically with H. nipponica, usually showed higher genetic similarity with this species. These four Halophila species from the tropical, subtropical, and temperate regions are quite similar morphologically [14,15], but we found H. nipponica to be genetically distinct from H. ovalis as well as from H. gaudichaudii and H. okinawensis according to the ITS sequence analysis.
Although H. nipponica is distributed in the Temperate North Pacific Bioregion in which the temperate seagrass genera Zostera, Phyllospadix, and Ruppia are dominant, it occurs primarily near the boundary of the Tropical Indo-Pacific Bioregion in which Halophila species are common [9]. Thus, propagules of Halophila species in the tropical/subtropical Indo-Pacific may travel easily to the temperate coastal waters of Korea and Japan via the warm Kuroshio Current. H. nipponica is quite similar morphologically to H. ovalis, and this species was treated as H. ovalis previously [14,19]. Among the Halophila species in the tropical Indo-Pacific, H. ovalis appears to be the most similar species to H. nipponica both morphologically and genetically [14,15] Recently, rbcL and matK sequences have been used for the identification of the common and widespread Halophila species such as H. ovalis and H. decipiens [35][36][37]. Analysis of these additional genetic sequences will provide invaluable information on the species identification and evolution of Halophila species.
In conclusion, H. nipponica plants from the various locations in temperate coastal waters of the northwestern Pacific were nearly genetically identical based on ITS sequences. H. nipponica from the temperate regions of Korea and Japan was grouped with H. okinawensis and H. gaudichaudii from the subtropical regions of the western Pacific in Clade I. These temperate and subtropical Halophila species in Clade I showed high ITS sequence similarity to the tropical H. ovalis in Clade II. Based on geographical distribution and similarities in genetics and morphology, H. nipponica is suggested to have diverged from a tropical Halophila species such as H. ovalis, which was transported from the tropical Indo-Pacific via Pacific Ocean circulation and then adapted to warm temperate environments. According to divergence time estimates, the temperate H. nipponica was considered an older species than the subtropical H. okinawensis and H. gaudichaudii and may have a different evolutionary history with the subtropical Halophila species.