Study site, sample collection, and morphology analysis

Halophila plants (Fig 1) were collected from the Tanjung Adang Laut shoal, Johor (1° 19’ 48.0" N 103° 33’ 59.8" E), in March 2022 and June 2022, during low tide on exposed seagrass bed shoal. The salinity and temperature of the seawater were recorded during high tide via the YSI Professional Plus Series Multiparameter. Depending on availability, one to three populations of each species were collected randomly at a distance of 10–30 m to avoid sampling the same clone. Three populations were collected for H. ovalis and H. major, two for H. nipponica, and one for H. spinulosa. Plants containing leaf blades, petioles, and rhizomes were selected and washed with seawater to remove all the sediment, then placed in sealed polythene plastic bags and brought to the laboratory for further analysis. In the laboratory, samples were divided for morphology analysis and herbarium; young leaves of the plant were preserved with silica gels in a paper bag and placed in a desiccator for DNA extraction. The plants were pressed for herbarium voucher samples [36]. Approximately 13–17 mature leaves were randomly selected for morphometric measurement. Measurements included blade length, blade width, the L:W ratio, the space between the intramarginal vein, the ratio of the distance between the intramarginal vein and lamina margin, the angle of the cross-vein, the distance between the cross veins, and the number of paired cross veins, following key identification by [16, 37]. These measurements were taken via a 5 MP Dino-Lite Edge 3.0 microscope connected to the computer. Statistical analysis of variance (ANOVA, p < 0.05) and post-hoc Duncan’s new multiple range test (DNMRT, p < 0.05) were performed via IBM SPSS statistic version 27 to compare the vegetative part dimension between populations of H. ovalis, H. major and H. nipponica from the Tanjung Adang Laut shoal.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Halophila species at Tanjung Adang Laut shoal, Johor.

(A) H. ovalis, (B) H. nipponica, (C) H. major, and (D) H. spinulosa.

https://doi.org/10.1371/journal.pone.0309143.g001

DNA extraction, ITS amplification, and sequencing process

The quantity of dried young leaves varied for each species to achieve a homogenized powder weighing 10–20 mg. Approximately 20–30 paired leaves were used for H. ovalis and H. nipponica, 5–10 paired leaves were used for H. major, and 10–15 paired leaflets were used for H. spinulosa. The plant tissue was homogenized via a mortar [38]. DNA extraction was carried out via the DNeasy® Plant Mini Kit (Qiagen, Inc., Valencia, CA, USA) following the manufacturer’s instructions.

Nine samples of Halophila species were collected and subjected to ITS amplification. The region nuclear ribosomal internal transcribed spacer (nrITS), including the 5.8S [28, 39] sequence, was selected for PCR amplification. The primers used were ITS1 (5′-TCCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′), which amplified a 700–720 bp sequence. The total volume of 30 μl for PCR components included 10X PCR buffer, 25 mM MgCl₂, 0.2 mm dNTPs, 5 U/μL Taq DNA Polymerase (Apical Scientific Sdn Bhd, Selangor, Malaysia), 10–30 ng of template DNA, and 1 pmol of each primer. PCR amplification was conducted via a PTC-100® Thermal Cycler (Biosystems, 850 Lincoln Centre Drive, Foster City, California 94404, USA) equipped with a heated lid. The reaction profile involved initial denaturation for 3 min at 95°C, followed by 35 cycles of denaturation for 30 s at 94°C, primer annealing for 30 s at 58°C and extension for 60 s at 72°C, and a final extension for 5 min at 72°C, which was terminated by a final hold at 12°C. The PCR products were observed via 1.2% agarose gel electrophoresis and stained with ethidium bromide. The direct sequencing of the PCR products was carried out via 1st BASE (Selangor, Malaysia) in both directions. PCR replications were conducted for each sample, and the highest-quality sequences were chosen for phylogenetic analysis. The quality of Sanger DNA sequences (Q>20) was determined via FinchTV 1.4 (Geospiza, Inc.; Seattle, WA, USA; http://www.geospiza.com). The consensus sequence was obtained from BioEdit 7.2 (Rutherford House, Manchester Science Parks, Pencroft Way, Manchester M15 6SE, United Kingdom).

Phylogenetic analyses

Nine sequences from this study (Nos. 1–9) and 28 additional undeposited sequences of Halophila species (Nos. 10–37) were obtained from the coauthor Japar Sidik Bujang; these sequences were sourced from various locations and are detailed in Table 1. These sequences were aligned via CLUSTALW in MEGA11 [40] with 63 accessions retrieved from GenBank (http://www.ncbi.nlm.nih.gov/genbank/), as shown in S1 Table. This information is accessible in the public repository of the National Library of Medicine (NLM) at https://www.nlm.nih.gov/accessibility.html. Phylogenetic analyses were conducted via three methods, i.e., maximum likelihood (ML), neighbour-joining (NJ), and maximum parsimony (MP). The Tamura 3-parameter and gamma-distributed (T92+G) models were chosen on the basis of the lowest Akaike information criterion (AIC) and Bayesian information criterion (BIC) values, indicating the best fit for the 100-sequence dataset [40]. The number of sequence differences was determined via the estimation of the pairwise distance model in MEGA11. The phylogenetics were edited via the iTOL web-based tool (https://itol.embl.de/). The number of haplotypes (N), haplotype diversity (h), and nucleotide diversity (π) were measured within each clade via DnaSP version 6 [41]. Haplotype data were also used to construct a Network 10.2.0.0 (Fluxus Technology) to generate haplotype networks via median-joining [42] for the ITS1-5.8S-ITS2 sequences via their respective alignments. A nonparametric analysis of molecular variance (AMOVA) was performed via ARLEQUIN version 3.5 [43] to assess genetic differentiation among and within clades. The significance of variance components was tested via permutation tests, with 1000 permutations used to generate a null distribution of the test statistics. Fixation indices (F_ST) were calculated to quantify genetic differentiation, and their significance was also evaluated via permutation tests. A p-value of less than 0.05 was considered indicative of significant genetic differentiation.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Sample collections from this study and additional undeposited sequences from Japar Sidik Bujang (JSB) for phylogenetic analyses.

https://doi.org/10.1371/journal.pone.0309143.t001

Distribution and ecology

Halophila nipponica previously unrecorded in Southeast Asia in the Tropical Indo-Pacific bioregion, was first recorded in Malaysia (Fig 2). This discovery was made in March and June 2022 at the Tanjung Adang Laut shoal, Johor, where small patches were growing in muddy-sandy areas with small leaves H. ovalis at -2 to -3 m mean sea level (MSL) depth. Salinity and temperature were 28.60–28.92 ppt and 28.70–28.90°C in March; and 27.93–28.85 ppt and 30.20–32.50°C in June, respectively.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Global distribution of seagrass.

Reprinted from [blue]; data from UNEP-WCMC & Short [44] under a CC BY license, with permission from [UNEP-WCMC], original copyright [2021]. Seagrass bioregions adapted from Short et al. [45] include countries with the same Halophila species from Malaysia: Temperate North Pacific: 26-Japan, 27-South Korea; Tropical Indo-Pacific: 2-Mozambique, 3-Madagascar, 4-Tanzania, 5-Kenya, 6-South Saudi Arabia, 7-North Saudi Arabia, 8-Kuwait, 9-Bahrain, 10-Yemen, 11-West India, 12-East India, 13-Sri Lanka, 14-Bangladesh, 15-Myanmar, 16-Thailand, 17-West Malaysia, 18-East Malaysia, 19-Indonesia, 20-Papua New Guinea, 21-Philippines, 22-Cambodia, 23-Vietnam, 24-China, 25-South Japan, 28-Guam, 29-East Australia; and Temperate Southern Oceans: 1-South-Africa, 29-West Australia. # H. okinawensis; * H. gaudichaudii.

https://doi.org/10.1371/journal.pone.0309143.g002

Morphology descriptions

The morphology of H. major, H. ovalis, and H. nipponica collected from the Tanjung Adang Laut shoal is illustrated in Fig 3. The leaf morphologies of H. ovalis, H. nipponica and H. major are shown in Table 2. H. nipponica had the shortest blade leaves among all populations (p < 0.05), ranging from 8.93 ± 0.46 mm to 9.11 ± 0.59 mm, followed by H. ovalis, ranging from 11.25 ± 1.07 mm to 13.75 ± 1.20 mm (p < 0.05). H. major had the longest blade length, ranging from 30.83 ± 2.58 mm to 38.36 ± 2.22 mm. The blade width in H. nipponica population 2 was narrower, while that in H. nipponica population 1 was similar to that in H. ovalis population 2, followed by that in H. ovalis population 3. However, the blade width of H. major was significantly wider than that of both H. ovalis and H. nipponica (p < 0.05). H. nipponica was differentiated by having the lowest number of paired cross-veins, ranging from 4 to 7 pairs, followed by H. ovalis, with 7 to 14 pairs, and H. major, with 13 to 18 pairs (p < 0.05). This differentiation was further supported by the ratio of the space between the intramarginal vein and the blade margin (r:R), which was 1:4.76–6.13 mm for H. nipponica, 1:8.00–14.16 mm for H. ovalis, and 1:17.33–36.97 mm for H. major. The distance between the cross-veins of H. nipponica and H. ovalis showed no significant difference, ranging from 0.99 ± 0.52 mm to 1.55 ± 0.62 mm, while H. major had a slightly greater distance, ranging from 2.53 ± 1.38 mm to 3.25 ± 1.92 mm.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Illustration of Halophila species collected from Tanjung Adang Laut shoal, Johor.

(A) H. nipponica, (B) H. ovalis, and (C) H. major.

https://doi.org/10.1371/journal.pone.0309143.g003

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Vegetative comparison of Halophila species at Tanjung Adang Laut shoal, Johor.

https://doi.org/10.1371/journal.pone.0309143.t002

Halophila nipponica (Fig 4) was identified as a dioecious plant with a fragile narrow rhizome, 0.40–0.60 mm thick and rhizome internodes 10.00–20.00 mm long. Each node had one unbranched root with numerous root hairs. The leaf scales (Fig 4A) were transparent, folded, and convex, with entire margins of approximately 2.35–2.75 × 2.86–3.60 mm. The petioles were 10.00–12.50 mm long and 0.40–0.45 mm thick. The leaves were light green, elliptical, and orbicular in shape (Fig 4B), with entire margins of 8.21–10.35 mm in length and 5.18–7.26 mm in width, an L:W ratio of 1.29–1.64:1 mm and a half lamina width of 2.61–3.75 mm (Table 2). The apex of the leaf was rounded, whereas the leaf base was rounded or oblique. The space between the intramarginal vein and the blade margin was 0.52–0.70 mm. The ratio of the distance between the intramarginal vein and the lamina margin was 1:4.76–6.13 mm. The number of paired cross-veins ranged from 4 to 7, with a distance between cross-veins of 0.43–2.50 mm and an angle of 50.25°–82.86°. The young mature male flower (Fig 4D) consisted of a transparent spathe, that was 1.27–2.18 mm long and 0.99–1.14 mm wide, three whitish tepals that were 1 mm long and 0.52 mm wide, and a 1 mm long pedicel. The female flower (Fig 4E) had an ovary of 0.87–1.20 × 0.67–1.00 mm, a hypanthium 0.36–5.08 mm long, and three styles of 4.60–7.20 mm long.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Morphology features of H. nipponica from Tanjung Adang Laut shoal, Johor.

(A) The habit of H. nipponica, (B) Leaf blade, (C) Male flower, and (D) Female flower. Us, upper scale; Ls, lower scale; Sp, spathe; T, tepal; P, pedicel; St, style; H, hypanthium; O, ovary.

https://doi.org/10.1371/journal.pone.0309143.g004

Halophila ovalis (Fig 5) is a dioecious plant with a creeping, fragile, narrow rhizome that is 0.50–1.40 mm thick and rhizome internodes 17.00–25.00 mm long. Each node had a single unbranched root with numerous root hairs. The leaf scales were transparent, folded, and convex, with entire margins of approximately 3.20–3.29 × 2.62–3.64 mm. The petioles were 8.00–17.00 mm long and 0.40–0.45 mm thick. The leaf blades were light green, elliptical (Fig 5B), and ovate (Fig 5C) in shape, with entire margins of 9.55–15.32 mm in length and 5.31–10.15 mm in width, with an L:W ratio of 1.39–1.96:1 mm and a half lamina width of 2.84–5.42 mm (Table 2). The leaf apex was obtuse and rounded, whereas the base was rounded or oblique. The space between the intramarginal vein and the blade margin was 0.25–0.49 mm, with a distance between the intramarginal vein and lamina margin of 1:8.00–14.16 mm. The number of paired cross-veins ranged from 7 to 14, with single- and double-branch veins. The distance between cross veins was 0.13–1.98 mm, at an angle of 50.55°–86.89°. Male and female flowers, as well as fruits, were absent from the collected samples.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Morphology features of H. ovalis from Tanjung Adang Laut shoal, Johor.

(A) The habit of H. ovalis, (B) Elliptic leaf, and (C) Ovate leaf.

https://doi.org/10.1371/journal.pone.0309143.g005

Halophila major (Fig 6) is a dioecious plant with a thick whitish rhizome that is 1.00–1.80 mm in diameter and rhizome internodes that are 11.30–23.00 mm long. Each node had one unbranched root with numerous root hairs. The leaf scales were transparent, folded, and convex, with 3.00–4.00 × 2.00–2.50 mm. The petioles were 30.00–38.00 mm long and 0.80–1.00 mm thick. The leaf blades (Fig 6B) were dark green, elliptical, ovate, and oblong with entire margins of 26.81–42.30 mm in length and 15.10–22.03 mm in width, with an L:W ratio of 1.50–2.36:1 mm and a half lamina width of 7.99–11.30 mm (Table 2). The leaf apex was obtuse and rounded, whereas the base was rounded or oblique. The space between the intramarginal vein and blade margin was 0.30–0.59 mm, with a ratio of the distance between the intramarginal vein and lamina margin of 1:17.33–36.97 mm. The number of paired cross-veins ranged from 13 to 18, with 2 to 5 branched veins. The distance between cross veins was 0.46–5.73 mm at an angle of 48.22°–72.05°. The male flower was absent. The female flower (Fig 6C) had an ovary of 2.00–2.15 × 1.00–1.20 mm, a hypanthium 4.00–7.00 mm long, and styles 14.00–20.00 mm long. The fruit (Fig 6D) was 2 mm in diameter.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Morphology features of H. major from Tanjung Adang Laut shoal, Johor.

(A) The habit of H. major, (B) Leaf blade, (C) Female flower, and (D) Fruits. Ls, lower scale; St, style; H, hypanthium; O, ovary; F, fruit.

https://doi.org/10.1371/journal.pone.0309143.g006

Halophila spinulosa (Fig 7) is a dioecious plant with woody rhizomes that are 0.85–1.90 mm thick and rhizome internodes that are 15.70–35.10 mm long. Each rhizome node had an unbranched root with numerous root hairs and a nonflowering or flowering erect shoot bearing compound leaves (Fig 7A) with leaflets. The leaflets were arranged distichously and tristichously (Fig 7B). The leaflet blades were dark green, oblong-linear, 7.00–16.00 mm long and 2.30–4.00 mm wide, with 4–7 pairs of cross-veins. The leaflet margin was folded near the base one side (Fig 7D and 7E). The leaf tip was rounded and serrated. The midrib was conspicuous and united at the top with intramarginal veins. Male and female flowers, as well as fruit, were absent from the collected sample.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Morphology features of H. spinulosa from Tanjung Adang Laut shoal, Johor.

(A) Vegetation habit of H. spinulosa, (B) Compound leaf, (C) Serrate leaflet, and (D) Lower leaflet with leaf folded. Tr, tristichous; Di, distichous.

https://doi.org/10.1371/journal.pone.0309143.g007

Phylogenetic analysis

Phylogenetic analyses were conducted via three methods, i.e., maximum-likelihood (ML), neighbor-joining (NJ), and maximum parsimony (MP), employing Tamura 3-parameter and gamma-distributed (T92+G) models with a dataset of 100 sequences. The MP method yielded the most suitable sequence alignment, resulting in six well-supported clades (Clades I-VI) among the Halophila species, as presented in Fig 8.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. The phylogenetic tree of Halophila species is inferred from maximum parsimony analysis.

1000 replication bootstrap using 651 base pairs of nrDNA, including ITS1, 5.8S rDNA, and ITS2. Bootstrap support values above 50% are shown on branches. Over 70% of the bootstrap value is threshold confidence. (*), sample of this study; (#), undeposited sequences from Japar Sidik Bujang.

https://doi.org/10.1371/journal.pone.0309143.g008

Clade I included H. ovalis, which clustered with a 90% bootstrap value alongside H. minor, H. hawaiiana, and H. johnsonii. Within Clade I, the small-leaved H. ovalis from this study clustered with all H. ovalis samples from the Tanjung Adang Laut shoal, and H. ovalis (SL) and H. ovalis (RL) from the Merambong shoal undeposited sequences from Japar Sidik Bujang. This is also supported by an 82% bootstrap value and a subgroup of Singapore samples with an 81% bootstrap value. H. ovalis (MY121200PMB# and MY121200MG#) from Sabah were clustered with H. ovalis from Malaysia (KF620343 and KF620344). In Clade II, H. nipponica from this study clustered with a strong 99% supported bootstrap value and was separated from the subgroup comprising H. nipponica, H. okinawensis, H. gaudichaudii, and H. minor.

H. major in this study were clustered in Clade III with 99% bootstrap supported value escorted samples from Malaysia, Indonesia, Thailand, Vietnam, Myanmar, Sri Lanka, Japan, and H. australis from Australia. H. major of this study was subgrouped with Indonesia and separated with H. major from Sabah, Malaysia with a 61% bootstrap value. Clade IV clustered H. decipiens from Malaysia, Vietnam, Japan, Australia, and the USA according to a 99% bootstrap value. Clade V received 97% bootstrap values for H. beccarii from Malaysia and Vietnam and H. engelmannii from the USA. In Clade VI, H. spinulosa from Malaysia was clustered with Australia by 88%, supporting the bootstrap value.

Sequence characteristics and genetic divergence

The final alignment of nine sequences from this study, represented in Table 3, consisted of 651 position datasets (including gaps), resulting in a total number of nucleotides ranging from 609 to 621 position datasets. The percentage of guanine-cytosine content (%G~C content) of H. ovalis at the Tanjung Adang Laut shoal, Johor, was identical among the three collection points (63.28%), whereas respectively, while H. nipponica presented a lower percentage of G~C content (62.91%). H. major contained 64.9% and 65.06% G~C content, whereas H. spinulosa presented the lowest content, at only 58.30%.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Nucleotide composition (%).

https://doi.org/10.1371/journal.pone.0309143.t003

The analysis of sequence divergences involved 100 nucleotide sequences, as shown in S2 Table, via the Tamura-3 parameter model over 1000 replications to locate pairwise distances. In the present study, H. nipponica and H. ovalis presented a significant difference of 0.040 in sequence distance. The ITS sequences of H. ovalis were not different in terms of genetic distance between their populations and were identical to those of H. ovalis from Singapore (OQ597146 and OQ597093) (Table A in S3 Table). A comparison with other H. ovalis sequences revealed genetic differences between 0.002 and 0.027 (1–17 bp differences), with the highest divergences between H. ovalis (red leaf) from the Merambong shoal. The observed sequence divergence of H. nipponica was 0.015 (10 bp difference) compared with that of its temperate Japanese counterparts and 0.015–0.017 (10–11 bp difference) compared with that of its South Korean counterparts (Table B in S3 Table).

A comparison of the pairwise distance of the H. major ITS from the Tanjung Adang Laut shoal, Johor, revealed minimal sequence differences of 0.002–0.003 (1–2 bp differences) between populations (Table C in S3 Table), in comparison with other H. major sequences, which presented variations of 0.002–0.011 (1–7 bp differences). H. spinulosa at the Tanjung Adang Laut shoal demonstrated low sequence divergence towards species from Malaysia (0.000–0.002; 0–1 bp differences) but notably greater differences than H. spinulosa from Australia (0.017; 11 bp differences) and H. tricostata (0.031; 20 bp differences), as shown in Table D in S3 Table. Interspecies pairwise distance comparisons between Clades I (H. ovalis) and II (H. nipponica) revealed an average sequence divergence of 0.032 (96.8% similarity; 21 bp differences), as shown in Table 4. The interspecies pairwise distance between Clades III (H. major) and I was lower than that between Clades II, with values of 0.044 (95.6% similarity; 29 bp difference) and 0.051 (94.9% similarity; 33 bp difference), respectively. This trend continued with Clades IV, V, and VI. Clade VI, containing H. spinulosa, showed considerable sequence distance from all other clades, with values ranging from 0.144 to 0.165 (83.5–85.6% similarity; 94–107 bp differences).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Genetic divergence between clades (interspecies pairwise distance).

https://doi.org/10.1371/journal.pone.0309143.t004

Haplotype diversity of Halophila species

Further analyses were conducted to increase the amount of supporting data on the determination of haplotype diversity and its differences between other Halophila species. A total of 100 sequences were estimated from six different clades (Table 5). The total nucleotide diversity (π) for all clades was 0.04950. The total haplotype diversity (Hd) was 0.968 from the 47 haplotype numbers (H). Clade V presented the highest haplotype diversity (1.000), where H. engelmannii was grouped with H. beccarii, followed by Clades III, I, II, and IV (0.914, 0.908, 0.859, and 0.524, respectively), and the lowest was in Clade VI (0.464). The number of segregating sites (S) of Clade IV was the lowest, with six sequence site differences, while it was the highest for Clade V (55 sites).

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Summarize of Halophila species by clade.

https://doi.org/10.1371/journal.pone.0309143.t005

The sequences of all individuals were used to construct a haplotype network (Fig 9), resulting in a total of 47 haplotypes (H) from 14 species in different regions. Clade I included 16 haplotypes (N = 48), H1–H15, including H. ovalis, H. minor, H. johnsonii, and H. hawaiiana. The haplotype network of Clade II formed eight haplotypes (N = 13), H17–H24, including H. nipponica, H. okinawensis, H. gaudichaudii, and H. minor. Clade III included 14 haplotypes (N = 21), H25–H38, involving H. major and H. australis. Clades IV (H. decipiens) and V (H. beccarii and H. engelmannii) both presented three haplotypes each, H39–H41 and H42–H44, whereas Clade VI presented three haplotypes (N = 8), H45–H46 and H47, including H. spinulosa and H. tricostata.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 9. Haplotype network of Halophila species and related species.

The circle size shows the number of samples. Mutations were designed as bars between network lines. Black dots represent root haplotype or ancestral haplotype. A-G are expected root/ancestral haplotypes for each clade. The haplotype network design was done using Network 10.2.0.0 software, and the colours were edited in Microsoft PowerPoint. (*), Samples of this study.

https://doi.org/10.1371/journal.pone.0309143.g009

In Clade I, the H. ovalis of this study was located at H7 along with the Singapore sample, attached from H9 (KF620346 H. ovalis Malaysia) with a single mutation/segregation site and eight mutation sites towards H2 (H. ovalis (RL) and H. ovalis (SL) from the Merambong shoal, Malaysia). H. minor from Thailand shares the same haplotype with H. ovalis on H8. H. nipponica in this study was referred to as H24, which is linked to a presumed ancestor with 24 mutation sites, connecting H. okinawensis and H. nipponica from Japan and Korea (H20–H23). Additionally, there was a genetic distance from another ancestor towards H. gaudichaudii (H19) and H. minor (H17–H18).

H. major in this study consisted of two haplotypes, H26 and H27, which differed by a single mutation site. Variant H26 was observed in the H. major samples of this study, as well as in Pulau Mabul, Pulau Salakan, and Pulau Gusungan. A sample of H. major from this study diverged as H27 from H26, indicating a potentially expanding haplotype within the population. In Clade VI, the H. spinulosa sample from this study shared H45 with Malaysian samples from Merambong shoal and Seluyong shoal in Johor. Moreover, this haplotype was found to have a shared ancestor with H. spinulosa samples from Australia, with six mutation sites differentiating them. The network haplotype constructions were a general estimation used to determine the separation between intraspecies division and mutation graphically. Specific individual comparisons of the same species across different bioregions will result in better phylogeographical determination.

A haplotype network using a universal ITS genomic region revealed closely and distantly related species. As shown in Fig 9, the haplotypes from different species formed six clusters as phylogenetic trees (Fig 8) divided by the expected root node or ancestral node. Nodes A, B, and C have one mutation/segregation from each node. The H. ovalis clade has the closest distance to the H. nipponica clade, with six mutations/segregations from node A to node G. The H. major clade has ten mutations/segregations from node B to node E. The H. decipiens clade has a greater number of mutations/segregations from node B, with 21 mutations/segregations from node B. H. spinulosa clade has 30 mutations/segregations from node C. H. beccarii clade tends to have the highest number of mutations/segregations and nodes from nodes A, B, and C.

The analysis of molecular variance (AMOVA) results shown in Table 6 revealed that 76.40% of the genetic variation was among clades, whereas 23.60% was within clades. The variance of components were 23.799 among the clades and 7.353 within the clades. The variance components were 23.799 among the clades and 7.353 within the clades. Permutation tests confirmed that both sources of variation were statistically significant (p < 0.05). The fixation index (F_ST) was 0.76395, indicating substantial genetic differentiation among clades. The F_ST value significantly differed from the null distribution generated by permutation tests (p < 0.05), supporting the conclusion of high genetic differentiation.

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. AMOVA (analysis of molecular variance) results from six clades.

https://doi.org/10.1371/journal.pone.0309143.t006

Morphometric characteristics of Halophila species

The morphometrics of Halophila species tend to be variable across different global geographic [21, 24, 46]. The morphological data for the Halophila section were compared with those for other species from other locations, as shown in S4 Table. The measurements of H. major in this study revealed longer blade lengths of up to 42.30 mm, in contrast to H. major populations from Mabul Island, East Coast of Malaysia, which typically exhibit lengths ranging between 18–22 mm [8]. Leaf lengths in other regions also varied, with Vietnam reporting values of 10–18 mm [34], Thailand reporting values of 25.9–29.4 mm [25], Indonesia reporting values of 25.04 ± 4.88 mm [29], and Ly Son Islands, Vietnam, reporting values of up to 30.66 ± 1.09 mm [26]. Compared with the samples from the present study, H. major and H. ovalis from Singapore are closer in length and width but have a slightly greater number of cross veins, with 15–21 pairs for H. major and 12–17 pairs for H. ovalis [33]. The morphological variation of Halophila leaves is influenced by environmental factors such as substrate, salinity, depth, and associated habitat in Malaysia [23, 24]. Accordingly, three categories were divided into big-leaved, intermediate-leaved, and small-leaved variants based on leaf size dimensions and pairs of cross veins. H. ovalis in Sarawak, found at similar depths presented small leaves on muddy substrates, intermediate leaves on sandy substrates, and big leaves on muddy–sandy substrates [23]. H. ovalis from the Merambong shoal had both small-leaved and big-leaved plants on the same substrate but were exposed to different conditions [24].

The key identification in this study lies in the ratio of the distance between the intramarginal vein and the blade margin with a half-width lamina (r:R), which allows differentiation between H. major, H. ovalis, and H. nipponica (S4 Table). H. major also resulted in a higher ratio of r:R from other regions because they have wider leaves. H. major from Southeast Asia recorded a range of 1:16.6–27.1 mm from Thailand, Malaysia, Indonesia, and Vietnam [8, 25, 29, 47], and H. major from Japan showed a similar range of 1: 20–25 mm [16]. In this study, H. ovalis was distinguished from H. major by the ratio of r:R from 1:8–14 mm, a finding that is consistent with similar observations reported in other studies [8, 16, 25]. Additionally, this study identifies the tristichous arrangement of leaflets as a distinctive characteristic of H. spinulosa (Spinulosae section), as this characteristic was initially discovered at the Tanjung Adang shoal [4].

Genetic and haplotype diversity

Genetic and AMOVAs of multiple Halophila species, which are divided into six clades, revealed significant genetic differences (F_ST: 0.76395) and varying levels of interspecies divergence, ranging from 0.032 to 0.165. H. ovalis demonstrated highly significant differences (F_ST: 0.679) among 14 populations from different regions in the Western Pacific and Eastern Indian Oceans [8]. The use of the universal ITS marker in this study proves valuable for studying closely related species of the same genus, owing to the high degree of sequence variation and genetic differences it detects.

For example, H. nipponica exhibited a shorter distance from H. ovalis than H. major did, indicating that H. ovalis may act as an intermediate species between the two. Additionally, owing to significant morphological differences, H. spinulosa and H. beccarii presented the greatest genetic divergence within the Halophila section. Our results revealed sequence divergence, with an average of 0.044, between H. ovalis and H. major clades. H. major in Thailand was reported to have an average genetic diversity of 2.60 ± 0.40% and 4.70 ± 0.90% from H. ovalis [29]. In Vietnam, the evolutionary divergence between H. ovalis and a new record, H. major, was reported to be 0.043–0.051 [47]. In the following year, they reported that the H. major clade differed from H. ovalis by 0.033–0.038 [8]. Evolutionary divergence refers to the broader process of populations or species diverging from a common ancestor, whereas genetic divergence is a specific aspect of this process, with a focus on the accumulation of genetic differences between these divergent lineages [28, 48, 49]. Extensive genetic diversity in dominant seagrass populations can improve growth performance and stability, even when disruptions occur. Therefore, it is crucial to prioritize the preservation of high genetic diversity within seagrass populations to maintain the normal structure and functioning of the ecosystems they contribute [50]. A genetic study in Thalassia testudinum demonstrated high genetic diversity and strong connectivity among morphologically diverged meadows across acute physiochemical gradients, suggesting strong phenotypic plasticity as a general-purpose trait under natural selection in this species [51].

The study also revealed high haplotype diversity (Hd) for H. nipponica, H. ovalis, and H. major, ranging from 0.859 to 0.914. This result is influenced by the large sample size and wide geographical distribution, encompassing diverse environments and habitats, providing ample opportunities for genetic differentiation and the accumulation of distinct haplotypes [47]. Conversely, H. decipiens, and H. spinulosa with low sample sizes, reached 1.00 in the case of H. beccarii because of their grouping with H. engelmanii, which contains many mutations. In a study of a specific species of H. major, distinct levels of haplotype diversity (Hd) were reported, with region IV (Sri Lanka) displaying a low range of Hd (0.202), region V (Bay of Bengal and Coast of Japan) showing a mid-range of haplotype diversity (0.710–0.713), and the highest Hd values found in region II (Wallacea site: Philippines and Indonesia) and region III (Sahul Shelf sites: Australia), ranging from 0.933 to 1 [26]. In another study, H. ovalis from Southeast Asia demonstrated greater diversity than H. ovalis from the Red Sea [30]. This explains why intraspecies studies are more specific than interspecies studies, but there is a problem in yielding some clear phylogeographical separation among the regions [26].

Previous haplotype studies focused on one species, H. major [26] and H. ovalis [30], in different regions, whereas this study compared multiple species, H. ovalis, H. major, H. nipponica, and H. spinulosa, in different regions. Haplotype 8 (H8) might be the ancestral haplotype for H. ovalis and is shared by Malaysia, Singapore, Thailand, Vietnam, Taiwan, and Japan. Previous studies have shown that H. ovalis from Southeast Asia and East Asia share the same haplotype and present a high number of haplotype providers [30]. The genetic phylogeny, historical biogeography, and evolution of Hydrocharitaceae indicated that Halophila possibly originated in Southeast Asia 15.9–41.3 million years ago [52]. The studies of specific species revealed higher haplotype numbers than those of multiple species did, thus revealing a more detailed evolutionary history and identifying unique haplotypes. For example, H. major from 11 countries was reportedly divided into five regions with 22 haplotypes [26]. In contrast, our study identified only 14 haplotypes from 9 countries grouped into a single clade. H. ovalis from Japan were separated individually by haplotypes from Southeast Asia [30]. However, H. ovalis from East Malaysia shares the same haplotype as that from Thailand [8], which is similar to our results.

New distribution records of Halophila nipponica from Malaysia

The Tropical Indo-Pacific bioregion, which extends from East Africa, Southeast Asia, and tropical Australia to the eastern Pacific, is the largest and most diverse region for tropical seagrasses [45]. Eleven Halophila species have been documented in Southeast Asia, including H. ovalis, H. major, H. minor, H. ovata, H. gaudichaudii, H. decipiens, H. sulawesii, H. beccarii, H. spinulosa, H. tricostata (Halophila sp. 2), and Halophila sp. 1 [7]. Although H. nipponica has not yet been identified molecularly in Southeast Asia, formal molecular identification of H. minor from the Philippines [28] has shown clustering within the H. nipponica clade [18]. Similarly, H. gaudichaudii, which was genetically identified [18] from Guam, is distributed in the Tropical Western Pacific Ocean [53]. Temperate H. nipponica from Japan grow on sandy or muddy sediment associated with H. major, Zostera marina, Z. japonica and Z. caespitosa [29]. These plants are able to grow not only in temperate Japan and South Korea, where the temperature is less than 20°C, but also in tropical waters, where they grow in summer [54]. Subtropical Japanese waters ranged and remained constant at 26°C in the summer months for the whole region, and there was a 3°C difference in the winter months between the southern islands [55]. The water in Malaysia exceeds 28–32°C, as H. ovalis has a wide range of tolerances to different seagrass bioregions [3, 8, 28, 30]. H. nipponica shows habitat preferences, but it has been shown to be endemic to the temperate bioregion of Japan [16].

Morphologically, H. nipponica from the Tanjung Adang Laut shoal differed from H. ovalis (Table 2). Further analysis via key identification [16, 37] revealed that this sample ranged from H. nipponica to H. gaudichaudii where the distance between the intramarginal and blade margins (r) and the ratio r:½ lamina width were overlapped, with H. nipponica from Chiba Pref. Japan (0.5–1.0 mm; 1:1.5–6.5 mm), H. gaudichaudii from Okinawa Pref. Japan (0.4–0.6 mm; 1:4.0–8.3 mm) and H. gaudichaudii from Marshall Island (0.32–0.58 mm) [16, 53]. H. nipponica from Okinawa Pref. was characterized by distinct leaf morphology across different localities, including wide leaves (elliptic type) from Ishigaki Island, intermediate leaves (linear type) from Okinawan Island, and narrower leaves (up to 1 mm wide) from Ie Island [17, 18]. In the other region, the morphological and reproductive features of H. nipponica from the coast of South Korea closely resembled those of the species in Japan [56]. H. gaudichaudii from Marshall Island [53] displayed a close morphological resemblance, featuring obovate leaves identical to those of H. gaudichaudii from Okinawa, Japan, characterized by a broader leaf type [16]. This variation in H. gaudichaudii has been treated synonymously with H. nipponica [17, 18], resembling the characteristics observed in our species. Historically, the Halophila section of the so-called Halophila ovalis-minor complex, which has elliptic and ovate leaf shapes including H. ovata, H. ovalis, H. minor, H. nipponica, H. okinawensis, and H. gaudichaudii, has been misclassified because of overlapping morphologies [16–18]. H. nipponica, H. okinawensis and H. gaudichaudii may be conspecific [17], with similar numbers of cross-veins and a rarity or absence of branched cross-veins.

Phylogenetically, H. gaudichaudii forms a distinct subclade towards H. nipponica and H. okinawensis [17], where our sample forms an outgroup from both subclades with 99% bootstrap values (Fig 8). Genetic distance analysis revealed that our sample was closest to H. okinawensis (Okinawa Pref.) with 0.8% and 1.2% (5–8 bp) differences and 1.0% (7 bp) difference from H. gaudichaudii from Guam. H. gaudichaudii exhibited a divergence of 0.8–1.7% (5–11 bp) from both H. okinawensis and H. nipponica [17]. A comparison of our sample with H. nipponica temperate Japan and South Korea revealed of 1.5% and 1.7% (10–11 bp) difference. H. nipponica from Korea is identical and has fewer than three bp differences from temperate Japanese species [19]. The distance of genetic differences revealed that subtropical species presented differences of up to 1.7% from temperate species, which is why our samples from the tropical Pacific presented the same distance. Haplotype diversity and network analysis (Fig 9; Table 5) revealed H. nipponica (including H. okinawensis, H. gaudichaudii, H. minor) from eight different haplotypes, each of which was divided by segregation or mutation into 16 total sites. Our sample exhibited a clear distinction, with one haplotype closely linked to H. okinawensis and another displaying a distant ancestral connection to H. gaudichaudii. Notably, the same haplotype was shared between H. nipponica populations from temperate Japan and Korea. The Philippines and Guam populations share the same ancestor and have been proven to occupy the same latitudinal range in Pacific water. This result indicates that seagrass bioregion have a large effect on genetic distance. However, the conspecific nature of this clade, as mentioned previously, might be due to its shared characteristics, and this clade has wide morphological barriers to its locality [17, 18]. H. okinawensis became a native species to Okinawan Pref., whereas H. gaudichaudii was distributed in subtropical Japan, Micronesia (Guam and Marshal Island), and the Philippines [16, 17, 53]. We propose that our sample represents a distinct population in the Johor region, characterized by genetic uniqueness and sequence segregation compared with temperate and subtropical waters. In light of these findings, we recommend identifying our sample as H. nipponica, which aligns with the H. nipponica clade however, it morphologically resembles H. gaudichaudii.