Study of the morphology and genetics of the seagrass Halophila ovalis in the Wallace Line coastal waters on the Minahasa Peninsula, Indonesia

AACL Bioflux, 2025, Volume 18, Issue 6. http://www.bioflux.com.ro/aacl 2475 Study of the morphology and genetics of the seagrass Halophila ovalis in the Wallace Line coastal waters on the Minahasa Peninsula, Indonesia Billy T. Wagey, Carolus P. Paruntu, Frans F. Tilaar, Rene C. Kepel Department of Management Aquatic Resources, Sam Ratulangi University, Manado 95115, Indonesia. Corresponding author: B. T. Wagey, bill[email protected].id Abstract. This study aimed to quantify the morphological characters of Halophila ovalis and provide baseline information on the species' genetic characters (DNA chloroplast). The morphological data analyses employed the Excel 2016 Program combined with XL-STAT 2015 for the principal components analysis (PCA). The sequence data were analyzed using MEGA-X and version 6 DnaSP Program and compared with specimens of the Genbank. These were then used to construct the phylogenetic topology, nucleotide base composition, nucleotide base sequence, polymorphic sites, and haplotype. Seagrasses belonging to Halophila are typically small and narrow, less than 1 cm wide, and have a distinctive "straplike" shape with a pointed tip. The PCA identified four morphometric variables (leaf width, leaf length, petiole length, and rhizome diameter) related to PC1, which explains 51% of the observed variation. Species confirmation with the nBLAST technique found that the specimens were H. ovalis with the highest score and identity reaching 100%. The phylogenetic topology showed that H. ovalis from different coastal waters in Minahasa Peninsula occurred in a single clade with the GenBank metadata. Samples from various sites in Indonesia (including Bali) were identical to the recently studied H. ovalis reflecting low variability in nucleotide base sequences. Hence, H. ovalis specimens from the sites in this study showed high genetic connectivity. Key Words: DNA, genetics, morphology, phylogeny, seagrass. Introduction. Several terminologies are frequently used interchangeably with seagrass, including seagrass, sea grass, weeds, and hilamun (Wagey 2013). However, the name "seagrass" is used in this publication. Seagrass research has long been neglected in Indonesia, even though these plants may be found in nearly all of the country's marine water. They provide crucial habitats for numerous animals, including large commercial and recreational fisheries. They are also an important food source for mega-herbivores, such as green sea turtles, dugongs, and manatees. Seagrasses contribute to the stabilization of sediment and produce much organic carbon. However, anthropogenic effects directly threaten these plants and their environments, and there is still a long way to go before answering the major research issues and putting them into management plans (Short et al 2016; Thangaradjou & Bhatt 2018). As a megadiverse country, Indonesia hosts extraordinary biological richness, encompassing a wide array of terrestrial and marine life (Myers et al 2000). In order to document and monitor this diversity and to help offset the accelerating loss of species and habitats, ongoing efforts in data collection on organismal variety are essential both for research imperatives and practical conservation needs (Myers et al 2000). Historically, seagrass species identification has relied primarily on morphological characteristics (Wagey 2017). However, in tropical regions with high species richness and morphological complexity, identification based only on vegetative traits can be challenging due to significant phenotypic plasticity and overlapping morphological features among species (Wagey 2017). Most tropical seagrass research has focused on ecological parameters such as species distribution, shoot density, and meadow extent rather than detailed taxonomic resolution (Orth & Heck 2023). AACL Bioflux, 2025, Volume 18, Issue 6. http://www.bioflux.com.ro/aacl 2476 Several studies on seagrass genetic diversity and genetic population have been conducted in spatially widespread coastal seascapes in the South China Sea (Hernawan 2018), the Western Indian Ocean (Jahnke et al 2019), English archipelago (Alotaibi et al 2019), and Western Australia (Sinclair et al 2020). The investigations included both tropical and temperate seagrass. Southeast Asia has the greatest diversity of seagrass with 12-15 species (Fortes et al 2018; Thangaradjou & Bhatt 2018). Molecular investigations like genetic diversity assessment, genetic population analysis, and seagrass phylogenetics are critical to understanding seagrass distribution, evolution, and conservation (Jahnke et al 2019; Lamit & Tanaka 2019; Clarito et al 2020). DNA barcodes have practical applications such as ecological surveys and sample verification. Translating these applications into usable DNA barcodes requires selecting one or more loci that can be routinely and accurately sequenced to identify large, taxonomically diverse samples and to facilitate interspecific comparisons; this selection is guided by three main criteria: primer universality, discriminatory power between species, and ease of amplification or recoverability of the sequence (Yang et al 2018). Considering these criteria, this study investigated the genetic diversity of the seagrass Halophila ovalis in the coastal waters of the Minahasa Peninsula. Limited taxonomic knowledge often leads to nomenclatural errors and species misidentifications, which have been documented among Halophila species (Guiry & Guiry 2023). Indonesia, as one of the world’s megadiverse countries, urgently needs to validate and document species within its jurisdiction. Members of the genus Halophila have a relatively wide distribution except at high latitudes (Hemminga & Duarte 2000). Therefore, this study focuses on the morphological and genetic characterization of H. ovalis in North Sulawesi, Indonesia. Material and Method Study sites. Five study sites representing different coastal zones of the Minahasa Peninsula, North Sulawesi, Indonesia, were selected: 1) Arakan/Rap-rap (ARA) – northern part; 2) Malalayang (MAL) – western part, Sulawesi Sea; 3) Bahoi–Likupang (BAH) – eastern part, Molucca Sea; 4) Tanjung Merah (TjM) – southern part; and 5) Ratatotok–Belang (RAT) – southeastern part, Molucca Sea (Figure 1). Figure 1. The location of location of the sampling sites in North Sulawesi, Indonesia. AACL Bioflux, 2025, Volume 18, Issue 6. http://www.bioflux.com.ro/aacl 2477 Sampling was conducted between March and June 2022 at depths of 3 m and 5 m under bright weather conditions. Each sampling site was georeferenced using a Global Positioning System (GPS) device. Site coordinates, sampling dates, and measured environmental parameters are presented in Table 1. The data presented in Table 1 should be considered as descriptive only due to limited data collection of in situ parameters. Although there are differences in the environmental parameters, for example, Arakan has warmer water temperature in the shallow (3 meters) but also lowest temperature at 5 meter depth. The rest of the variables, however, such as salinity were similar across sites. Table 1 Environmental condition of the study sites Parameters Site ARA MAL BAH TjM RAT Date 19-3-2022 19-3-2022 19-3-022 21-3-2022 28-6-2022 Latitude (N) 1°23'28.50" 1°27'46.22" 1°43'13.06" 1°24'14.41" 0°51'05.57" Longitude (E) 124°32'26" 124°47'08.49" 125°01'12.53" 125°07'14.05" 124°42'31.66" Weather Bright Bright Bright Bright Bright Water temp., 3 m (°C) 30.80 31.90 29.80 30.90 29.50 Water temp., 5 m (°C) 24.10 28.10 26.10 27.01 27.00 Salinity, 3 m (‰) 25.00 25.00 25.00 25.00 25.00 Salinity, 5 m (‰) 23.00 23.00 23.00 23.00 23.00 pH, 3 m 6.15 5.88 6.30 6.55 No data pH, 5 m 5.20 5.50 6.05 6.23 No data Substrate type (3 m) Sandy Sandy Sand mixed with coral shards Sandy Sandy mud Substrate type (5 m) Sandy Sandy Sand mixed with coral shards Sandy Sandy mud Associated biota (3 m) No data Anemones, Chaetodon Echinothrix & Diadema setosum No data Fish, molluscs, echinoderms Associated biota (5 m) No data Anemones, Chaetodon Diadema setosum No data Fish, molluscs, echinoderms Habitat type (3 m) Sand-coral reef Sand-coral reef Sand-coral reef No data Seagrass meadow, coral reef, Anemone, soft coral Note: ARA = Arakan; MALA = Malalayang; BAH = Bahoi; TjM = Tanjung Merah; RAT = Ratatotok. Environmental parameters. Water parameters were measured in situ at each site. Temperature, pH, and dissolved oxygen were recorded using a HORIBA LAQUA PD-220K handheld pH/ORP/DO/Temperature meter, calibrated with standard buffer solutions (pH 4.0, 7.0, and 10.0) before use. Water temperature ranged from 24.1 to 31.9°C, and pH varied between 5.20 and 6.55. Salinity was determined using a handheld refractometer, ranging from 23 to 25‰. Substrate type was determined by direct underwater observation and categorized as sandy, sandy mud, or sand mixed with coral fragments. The habitats were primarily sand-coral reef environments, with Ratatotok–Belang also characterized by seagrass meadows, soft corals, and anemones. Associated benthic organisms included sea anemones (Anthozoa), butterflyfish (Chaetodon spp.), sea urchins (Echinothrix and Diadema setosum), molluscs, and reef fishes. Sample handling and morphological identification. Seagrass samples of H. ovalis were collected by SCUBA divers from each study site (ARA, MAL, BAH, TjM, and RAT). Samples were rinsed with freshwater to remove sediment and epiphytes, then transported to the Integrated Laboratory, Faculty of Fisheries and Marine Sciences, Sam Ratulangi University, for analysis. A total of 30 samples (n = 30) were analyzed for each morphological variable: leaf width (LW), leaf length (LL), petiole length (PL), internode (IN), rhizome diameter (DoR), number of left cross veins (NCVL), number of right cross veins (NCVR), and number of AACL Bioflux, 2025, Volume 18, Issue 6. http://www.bioflux.com.ro/aacl 2478 leaves (NL). Measurements were taken using a digital caliper for quantitative traits. The presence of reproductive structures such as fruits, flowers, and scales (scutes) was also recorded. Species identification followed diagnostic features described by Waycott et al (2004), El Shaffai (2011), and Wagey (2013). Molecular work. Observations and measurements of morphological characters were conducted both in the field and at the Integrated Laboratory of the Faculty of Fisheries and Marine Sciences, Sam Ratulangi University. DNA extraction was carried out at the Biotechnology Service Unit, Faculty of Mathematics and Natural Sciences, Sam Ratulangi University, Manado. Polymerase chain reaction (PCR) amplification was performed using 5× FIREPol PCR Master Mix in a total reaction volume of 50 µL, containing 20 pmol of each primer and template DNA. The primer pairs used were: rbcLaF (forward): 5′–ATG TCA CCA CAA ACA GAG ACT AAA GC–3′ and rbcLaR (reverse): 5′–CTT CTG CTA CAA ATA AGA ATC GAT CTC–3′. PCR thermal cycling conditions were as follows: initial denaturation at 95°C for 2 min, followed by 40 cycles of 95°C for 40 s, 55°C for 40 s, and 72°C for 50 s, with a final extension at 72°C for 10 min. The PCR products were separated using 1% agarose gel electrophoresis in 1× TBE buffer and visualized under UV illumination. Amplified products were sent, along with the primers, to 1st BASE Malaysia for sequencing. Data analysis. Morphological data of H. ovalis (DoR, IN, LL, LW, NCVL, NCVR, NL, and PL) obtained from five sampling sites at two depths (3 m and 5 m) were first organized and tabulated in Microsoft Excel. These data were then subjected to Principal Component Analysis (PCA) using XLSTAT 2015 to examine patterns of morphological variation among samples and to determine which traits contributed most to overall variability. Prior to analysis, the data were standardized to remove the effect of differing measurement scales among parameters. The PCA produced eight principal components (F1-F8) that summarized the relationships among morphological traits. The first two components (F1 and F2) explained the largest proportion of the total variance and were used to visualize grouping patterns among sites and depths. The loading values of each variable on the PCA axes were interpreted to identify the traits that contributed most strongly to each component. High positive loadings indicated parameters that increased together, while negative loadings suggested inverse relationships. In this study, F1 typically represented the overall leaf dimension traits (LL, LW, PL, and NL), capturing variation related to leaf size and growth form, whereas F2 was mainly associated with structural traits such as internode length (IN) and rhizome diameter (DoR), reflecting differences in horizontal growth and substrate adaptation. The remaining components (F3-F8) explained minor variations and were considered less influential in discriminating site and depth effects. The PCA biplot was used to visualize the clustering of sampling sites and to interpret how morphological traits contributed to variation among shallow and deep populations of H. ovalis. Following the PCA, a Pearson correlation matrix of the morphological variables was computed to evaluate the strength and direction of linear relationships among traits, providing additional insights into interdependence between morphological parameters. Molecular analysis. The DNA sequences obtained were edited and aligned using the MEGA-X program (Kumar et al 2018). Sequence similarity was verified by comparison with reference sequences in the GenBank database using BLAST and the Barcode of Life Data System (BOLD). Phylogenetic trees were reconstructed in MEGA-X using default parameters with 1,000 bootstrap replications. Polymorphism analyses - including nucleotide composition, polymorphic sites, transitions, transversions, indels (insertions/deletions), and haplotypes - were conducted using DnaSP version 6.12.03 (Rozas et al 2017). AACL Bioflux, 2025, Volume 18, Issue 6. http://www.bioflux.com.ro/aacl 2479 Results Morphological variations of seagrass H. ovalis. The results of measurements (mean) and observations obtained showed that H. ovalis at all locations had the characteristics of oval/or-shaped leaves, smooth and slippery surfaces; LW 10.59-14.18 mm (3 m depth) and 7.82-14.38 mm (5 m depth); LL 19.14-29.92 mm (3 m depth) and 14.10-28.46 mm (5 m depth); 15-26 pairs of leaf veins, and this species is found in sandy substrate mixed with coral rubble. The representative samples of H. ovalis are shown in Figure 2. Figure 2. Leaf shape patterns of H. ovalis from the sampling sites. A, Arakan/Rap-rap; B, Bahoi; C, Tanjung Merah, D, Malalayang; E, Ratatok; and F, sample photomicrograph (40x) of the edge of H. ovalis leaf from Ratatok. Scale bars = 10 mm. Variations in each morphological trait are presented as barplots in Figure 3. At both depths, H. ovalis exhibited significant variation in morphological parameters across the five sampling sites. In the shallow sites (A), leaf-related traits such as LL, LW, and PL showed the highest values, particularly at Ratatotok and Tanjung Merah, indicating favorable growth conditions. DoR and IN remained relatively consistent across sites, with only slight increases in deeper sediments. At greater depth (B), a general reduction in most parameters was observed except for PL and NL, which remained comparatively high at Ratatok. The number of cross veins (NCVL and NCVR) exhibited moderate variability among sites and depths, suggesting morphological plasticity in response to environmental gradients such as light availability and substrate type. Overall, H. ovalis displayed site-dependent morphological differences rather than depth-dependent differences, reflecting its adaptability to varying coastal conditions in North Sulawesi. Based on the PCA in Figure 4, the resulting eigenvalues showed that the variance explained by the first two principal components (axes 1 and 2) accounted for 66.21% of the total variability in the dataset (Figure 5; Table 2). This indicates that most of the variation in the morphological data can be effectively represented along these two axes. The remaining 33.79% of the variance was explained by components 3 to 8. Overall, the analysis accounted for 78.69% of the total variation, suggesting that the PCA provided a reliable summary of the morphological characteristics of H. ovalis. To interpret the contribution of each variable to the model, the position of the variables along axis 1 and axis 2 was examined (Figures 4 and 5). Among the eight AACL Bioflux, 2025, Volume 18, Issue 6. http://www.bioflux.com.ro/aacl 2480 variables, LL and LD were positioned closer to axis 1, indicating a strong influence on the first principal component. In contrast, NL was positioned nearer to axis 2, suggesting a smaller effect on the first axis. Variables that point in the same direction - such as IN, DoR, and PL - were positively correlated, and their proximity to the correlation circle indicates a significant contribution to overall morphological variation. Figure 3. Morphological parameters of the seagrass H. ovalis from the sampling locations at two depths: A, shallow (3 m) and B, deep (5 m). Error bars are the standard error of the means (n = 30). LW = leaf width (mm), LL = leaf length (mm), PL = petiole length (mm), IN = internode (mm), DoR = rhizome diameter (mm), NCVL = number of left cross vein, NCVR = number of right cross vein, NL = number of leaves. Figure 4. PCA biplot of seagrass morphometric variables. AACL Bioflux, 2025, Volume 18, Issue 6. http://www.bioflux.com.ro/aacl 2481 Figure 5. Plot of cumulative variable value of root characteristics. PC 1 and 2 explain > 70% of the variability. Note: F1-8 represents PC components. Table 2 Summary of principal component analysis (PCA) results for the eight morphological variables of Halophila ovalis, showing eigenvalues, percentage of variance, and cumulative variance explained by each principal component (F1–F8). PCA summary F1 F2 F3 F4 F5 F6 F7 F8 Eigenvalue 4.08 1.22 0.99 0.76 0.51 0.29 0.08 0.06 Variability (%) 51.00 15.21 12.48 9.52 6.36 3.61 1.04 0.79 Cumulative % 51.00 66.21 78.69 88.20 94.56 98.17 99.21 100.00 Note: F1-8 represents PC components The Pearson correlation matrix (Table 3) reveals strong positive correlations among several morphological traits of H. ovalis. LW and LL are highly correlated (r = 0.903), indicating that larger leaves tend to be both longer and wider. PL also shows moderate to high correlations with LW (r = 0.616) and LL (r = 0.659), suggesting that leaf size and petiole elongation increase together. DoR is moderately correlated with LW (r = 0.573) and LL (r = 0.513), implying that thicker rhizomes are associated with larger leaf structures. The NCVL and NCVR sides exhibit a very strong correlation (r = 0.929), reflecting bilateral vein symmetry in the leaf morphology. In contrast, the NL shows very weak or negligible correlations with all other traits, suggesting that leaf count per shoot is largely independent of individual leaf dimensions or rhizome thickness. Overall, these correlations highlight coordinated growth patterns among vegetative structures but limited association with leaf number in H. ovalis. Table 3 Pearson correlation matrix of the morphological variables Variables LW LL PL IN DoR NCVL NCVR NL LW 1 0.903 0.616 0.493 0.573 0.666 0.695 -0.008 LL 0.903 1 0.659 0.470 0.513 0.547 0.564 -0.019 PL 0.616 0.659 1 0.341 0.321 0.282 0.297 0.084 IN 0.493 0.470 0.341 1 0.461 0.194 0.205 0.004 DoR 0.573 0.513 0.321 0.461 1 0.358 0.340 0.070 NCVL 0.666 0.547 0.282 0.194 0.358 1 0.929 -0.004 NCVR 0.695 0.564 0.297 0.205 0.340 0.929 1 -0.023 NL -0.008 -0.019 0.084 0.004 0.070 -0.004 -0.023 1 Note: LW = leaf width, LL = leaf length, PL = petiole length, IN = internode, DoR = rhizome diameter, NCVL = number of left cross vein, NCVR = number of right cross vein, NL = number of leaves, n = 30 for each variable. AACL Bioflux, 2025, Volume 18, Issue 6. http://www.bioflux.com.ro/aacl 2482 Genetic confirmation of seagrass H. ovalis. After obtaining a DNA band that matches the length of the target gene, then sequencing was carried out to confirm the amplification of the rbcL gene from each sample (Lin et al 2021). The sequencing results were confirmed with Basic Local Alignment Search Tools (BLAST) to determine whether the gene sequence obtained was the target gene of seagrass H. ovalis. This result can be proven. Confirmation was achieved by calculating the percent homology between the obtained rbcL sequence and reference rbcL sequences. Query from Genbank showed a high match rate of 100%. The high match indicates that the target gene obtained is the true H. ovalis rbcL gene (Stevanus & Pharmawati 2021). Seagrass H. ovalis was identified at Arakan/Raprap, Malalayang–Manado, Bahoi– Likupang, Tanjung Merah, Ratatotok–Belang, and Bali (Wagey et al 2016) based on the results of the alignment of the rbcL gene sequences that was 564 bp. Alignment was conducted to determine the nucleotide base-sequence homology level of the rbcL gene sequence obtained between samples from different locations. The composition of the nucleotide base frequency of the seagrass H. ovalis based on 7 nucleotide sequences is A = 28.01%, T = 28.27%, C = 21.02% and G = 22.70% with the dominant bases being T and A. Table 4 shows only one nucleotide base substitution at position 564 of the nucleotide sequence. Table 4 H. ovalis nucleotide base sequence showing substitution at position 564 from different sampling locations Position Nucleotide base variation by location Bali Arakan Malalayang Bahoi Tanjung Merah Ratatatok-1 Ratatok-2 541 G - - - - - - 542 T - - - - - - 543 T - - - - - - 544 T - - - - - - 545 A - - - - - - 546 T - - - - - - 547 G - - - - - - 548 A - - - - - - 549 A - - - - - - 550 T - - - - - - 551 G - - - - - - 552 T - - - - - - 553 C - - - - - - 554 T - - - - - - 555 A - - - - - - 556 C - - - - - - 557 G - - - - - - 558 T - - - - - - 559 G - - - - - - 560 G - - - - - - 561 T - - - - - - 562 G - - - - - - 563 G - - - - - - 564 A - - C - C C Only position 564 was polymorphic (T ↔ C); all other alignment positions (1-563) were identical across samples; no indels detected. Haplotype assignment: Hap_1 = T (Bali, Arakan/Rap-rap, Tanjung Merah); Hap_2 = C (Malalayang, Bahoi, Ratatotok-1, Ratatotok-2); h = 2; Hd = 0.5714. The nucleotide base substitution is a transition, not a transversion. In addition, there was no indel (insert = insertion and deletion = deletion) of nucleotide bases. The transition point is the substitution of nucleotide bases from bases A and G (purines) or between AACL Bioflux, 2025, Volume 18, Issue 6. http://www.bioflux.com.ro/aacl 2483 bases T and C (pyrimidines), while transversion is the substitution between a purine base and a pyrimidine base (Lyons & Lauring 2017). The results of the phylogenetic reconstruction of H. ovalis sequences from different locations in the Minahasa peninsula are presented in Figure 6. Wagey et al (2016) also reported genetically identical H. ovalis in Arakan, Bali and Tanjung Merah. In comparison, the results of DNA polymorphism analysis did not have a clear haplotype. That indicates the 7 H. ovalis sequences from these locations have the same ancestor. Figure 6. Phylogenetic reconstruction (Neighbor-Joining method – MEGA X). Haplotype distribution can be described as follows: Number of haplotypes, h: 2 Haplotype diversity, Hd: 0.5714 Hap_1: 3 [1-2 5] Hap_2: 4 [3-4 6-7] Hap_1: 3 [H. ovalis (Bali)-H. ovalis (Arakan/Raprap) H. ovalis (Tanjung Merah)] Hap_2: 4 [H. ovalis (Malalayang)-H. ovalis (Bahoi) H. ovalis (Ratatotok-1)-H. ovalis (Ratatotok-2)]