Phylogeography of the marine benthic alga Gracilaria salicornia (Gracilariales, Rhodophyta) in Southeast Asia

1 Introduction

Southeast Asia is one of the most important centers of marine biodiversity and species distribution (Ludt and Rocha 2015). Complex geography and oceanography have contributed to the various patterns of species diversity and distributions (Gallagher et al. 2024). Historical phylogeography assumes a central role in assessing current distribution patterns by analyzing the genetic relationships between different populations (Grant 2015). Surface seawater currents have influenced genetic connectivity between distant population, while geographical barriers, sea-level fluctuations, and other factors have resulted in genetic fragmentation of population (Ludt and Rocha 2015).

Previous studies on Southeast Asian seaweeds have highlighted various patterns of population structure and phylogeography, as mentioned below. Their distribution patterns have been influenced by various driving factors; sea-level rise (Kantachumpoo et al. 2014), geographical barrier (Bulan et al. 2022; Muangmai et al. 2023; Wichachucherd et al. 2014), and regional sea circulation or ocean currents (Bulan et al. 2022; Fontana et al. 2024; Hu et al. 2017; Muangmai et al. 2023). Historical phylogeographic studies revealed post-glacial dispersals or mid-Pliocene divergence (Chan et al. 2013, 2014; Fontana et al. 2024; Hu et al. 2017).

The intertidal red alga Gracilaria salicornia (C.Agardh) E.Y.Dawson (Gracilariales) is a good example for the historical phylogeography because it’s common in Southeast Asia and mitochondrial COI-5P data were accumulated in previous studies (Ferrer et al. 2019; Muangmai et al. 2023; Ng et al. 2015; Yang et al. 2013). Lim et al. (2001), using random amplified polymorphic DNA method, reported some genetic difference between the articulated and the non-articulated morphotypes within G. salicornia. However, the genetic difference between the two morphotypes has not been supported by subsequent molecular studies (Ferrer et al. 2019; Ng et al. 2015; Yang et al. 2013). Genetic diversity has been documented for populations from Malaysia, the Philippines, and Thailand (Ferrer et al. 2019; Muangmai et al. 2023; Ng et al. 2015; Yang et al. 2013). Muangmai et al. (2023) showed contrasting patterns of genetic diversity and population discontinuity in G. salicornia by geographical barrier between the Andaman Sea and the Gulf of Thailand. The aim of the present study was to analyze COI-5P sequence from seven populations of G. salicornia in Vietnam and to infer the phylogeography of Southeast Asian populations. This is the first historical phylogeographic study of G. salicornia utilizing both mismatch distribution and Bayesian skyline plot analyses.

2 Materials and methods

2.1 Sampling

Gracilaria salicornia specimens were collected at seven locations, ranged from Nui Thanh (15.5113 N, 108.6724 E), Quang Nam in the north to Ha Tien (10.3886 N, 104.4452 E), Kien Giang in the south (Figure 1; Supplementary Table S1). Depending on their frequency, 5–16 specimens were collected about 50 cm apart. Individuals were mostly green in shaded locations, however, plants in the exposed intertidal flat were yellow (Figure 1). Both articulated and non-articulated morphotypes were collected, but the articulated morphotype was most frequent. Branch tips from each individual were cleaned of epiphytes and kept with silica gel for DNA extraction. Vouchers were deposited in the Marine Ecology Laboratory of the Institute of Tropical Biology, Ho Chi Minh City, Vietnam.

Figure 1:

Habitat of Gracilaria salicornia. (A) Green plants with articulated axes (Ninh Hoa, Vietnam; 7 December 2023). (B) Yellow plants with articulated axes (Ly Son, Vietnam; 22 June 2022). (C) Dark green clump with compressed to cylindrical axes (Ha Tien, Vietnam; 1 April 2022). (D) Intertidal flat where G. salicornia commonly occurs (Nui Thanh, Vietnam; 16 June 2022).

3 Results

We generated 629–857 bp of COI-5P from 55 samples of G. salicornia from Vietnam (Supplementary Table S1). A total of 399 sequences, including published sequences from GenBank, were used for genetic diversity and phylogeographic analyses. A total of 27 haplotypes were identified, including five haplotypes found in Vietnam, three of which were private. Pairwise divergence ranged from 0.18–1.8 % between populations from different countries.

Gracilaria salicornia had high haplotype diversity (Hd = 0.765 ± 0.018) and low nucleotide diversity (π = 0.00456 ± 0.00023) (Table 1). Vietnamese populations (Hd = 0.352 ± 0.079, π = 0.00074 ± 0.00019) and the Chinese populations (Hd = 0.331 ± 0.143, π = 0.00082 ± 0.00039) had similar haplotype and nucleotide diversities. Genetic diversity values were twice as high in the Philippines and Thailand populations (Hd = 0.615 ± 0.031, π = 0.00553 ± 0.00034, and Hd = 0.670 ± 0.024, π = 0.00172 ± 0.00012, respectively).

The ML tree of G. salicornia haplotypes was monophyletic and well resolved (100 % ML bootstrap value, 1.0 for BPP) (Supplementary Figure S1). G. salicornia was divided into three haplogroups (I–III), following the criteria of Ferrer et al. (2019); haplogroup I comprised 18 haplotypes from China, Hawaii, Indonesia, Malaysia, the Philippines, Singapore, Thailand, and Vietnam with statistical support. Haplogroup II comprised eight haplotypes from Japan and the Philippines, but it was not statistically supported. A single haplotype (H19) from the Philippines was placed in haplogroup III.

Of 27 haplotypes found in G. salicornia, two (H1, H2, and H4) were shared with other countries and 15 were private, each found in a single population (Figure 2). The haplotype network consisted of the same three haplogroups as the ML tree. Haplotype H1 (n = 175) was the most dominant and widely distributed haplotype in China, Indonesia, Malaysia, the Philippines, Singapore, Thailand, and Vietnam, and we designated it as the ancestral haplotype. Haplotype 2 was shared between Hawaii and the Philippines, and haplotype 4 occurred in China, the Philippines, and Vietnam.

Figure 2:

Haplotype network and geographic distribution of COI-5P haplotypes in Gracilaria salicornia. (A) Median-joining network. Each circle denotes a single haplotype with size proportional to frequency. Small cross lines represent single mutation steps and black circles represent extinct or unsampled haplotypes. Haplotypes are colored by country, as shown in the key. (B) A map showing haplotype distribution in Vietnam. Pie charts denote the proportion of haplotypes present in each site. Haplotypes are colored as shown in the key.

Non-hierarchical AMOVA showed that 55.92 % (p < 0.001) of the genetic variation was found among countries (Table 2). The hierarchical AMOVA showed that 65.6 % (p < 0.01) of variation occurred between three haplogroups, followed by within-haplogroup variation (25.9 %, p < 0.001).

Under the sudden expansion model of the mismatch distribution analysis, the population expansion of G. salicornia was estimated to start about 251,976 years ago (95 % CI: 105,522–787,488) with low and non-significant sum of squared deviation (SSD = 0.02090) and Harpending’s raggedness (H_Rag = 0.03970) values (Table 3). The expansion time of haplogroup I was estimated to be 78,742 years ago (95 % CI: 68,900–96,890). Both the Chinese and Vietnamese populations were estimated to have expanded about 236,000 years ago.

The pairwise mismatch distribution was bimodal for the overall G. salicornia populations (Figure 3A). Haplogroup I was unimodal (Figure 3B), being closely aligned with the expected pattern under the sudden expansion model. Both Tajima’s D and Fu’s Fs values were statistically significant for haplogroup I (D = −1.72366, p < 0.05; Fs = −11.841, p < 0.01) (Table 1), indicating a recent demographic or range expansion. In the remaining two haplogroups (II and III), the Tajima’s D and Fu’s Fs values were not statistically significant; therefore, a mismatch distribution analysis was not performed.

Figure 3:

Mismatch distribution for Gracilaria salicornia (A) and G. salicornia haplogroup I (B); and Bayesian skyline plots for G. salicornia (C) and G. salicornia haplogroup I (D). For mismatch distributions, the black bar chart represents observed distributions, whereas the grey line represents simulated data under a spatial expansion model. For Bayesian skyline plots, x-axis indicates time since the present in years and y-axis indicates estimated effective population size. Bold line is the median estimate, and thinner lines indicate the 95 % highest posterior density (HPD) interval.

The generalized skyline plot showed that the divergence time from its recent common ancestor was about 0.5 Ma, and indicated population growth up to about 0.2 Ma, followed by stable population sizes until the present (Figure 3C). Haplogroup I showed a steady population growth over about 0.175 Ma (Figure 3D). Both haplogroups II and III were not able to analyze, as mentioned in mismatch analysis.

4 Discussion

Our sampling of 55 specimens across seven localities represented a diverse range of habitats for G. salicornia in Vietnam, which comprised five haplotypes, three unique (H6–H8) and two (H1, H4) shared with neighboring countries. Both shared and unique haplotypes are equally important in phylogeographic analysis; shared haplotypes indicate genetic connectivity between Vietnam and other countries, while unique haplotypes indicate mutation accumulation specific to Vietnam. The genetic diversity of Vietnamese populations is lower, than that of the Philippines and Thailand populations, due to the high frequency (44/55 specimens examined) of the ancestral H1. Our results indicate a recent expansion of the Vietnamese population that is supported by both Tajima’s D and Fu’s Fs. This demographic or spatial expansion was estimated to start in the middle Pleistocene (about 236,000 years ago). The Vietnamese population is an important piece in the phylogeographic puzzle of G. salicornia from Southeast Asia.

The presence of three haplogroups (I–III) within G. salicornia is supported by our phylogeny, haplotype network, and distribution of haplotypes, as was also reported in a previous study (Ferrer et al. 2019; Muangmai et al. 2023). Haplogroup I is distributed mostly in Indonesia, Malaysia, Thailand, Vietnam, and China, being on the west side of the marginal sea in the South China Sea. Haplogroup II occurs in the Philippines and Japan, being on the east side of the marginal sea, and likely reflecting the influence of the Kuroshio current. Despite comprising a relatively high number of mutation steps in the haplotype network, haplogroup II was broadly defined due to its restricted geographic distribution in the Philippines and Japan (Okinawa). Further study is required to decipher the genetic structuring of haplogroup II, which was not statistically supported in the present study, despite detailed collections in many locations in the Philippines (Ferrer et al. 2019; Yang et al. 2013). Haplogroup III includes a single specimen from the Philippines (Ferrer et al. 2019). Despite extensive collections across multiple localities in China, Malaysia, Thailand, the Philippines (Ferrer et al. 2019; Muangmai et al. 2023; Ng et al. 2015; Wang et al. 2023; Yang et al. 2013), and Vietnam, in the present study, no additional specimens of haplogroup III were found. We infer that the intervals of low sea level in the South China Sea, which became an inland sea during the Pleistocene (Wang et al. 1995), may have introduced west-east distribution pattern of G. salicornia. Demographic or range expansion related to sea-level fluctuations in the Quaternary are not uncommonly reported in tropical marine animals and the brown algal genus Sargassum (Chan et al. 2013, 2014; Grant 2015; Ludt and Rocha 2015; Wang et al. 2016; Zhang et al. 2014).

The recent chronogram of Gracilaria showed that the most recent common ancestor (MRCA) of G. salicornia began to diverge in a period of time between the Eocene and the Quaternary (approximated from Figure 3 in Lyra et al. 2021). This phylogeny was constructed using protein-coding genes of plastid and mitochondrial genomes of specimens from Hawaii and Singapore (Lyra et al. 2021). Haplotype network in the previous study (Ferrer et al. 2019) as well as the present study revealed that a single mutation event occurred between H2 (KF824534 from Hawaii, and KT357381, KY7774623-5 from the Philippines) and H3 (MZ336093 from Singapore) (Supplementary Tables S1 and S2), both being included in the haplogroup I (Figure 2). Evolutionary estimation of the MRCA of G. salicornia will highlight its phylogeography, however, it is beyond the focus of the present study because of the lack of organellar genomes from representatives of haplogroups II and III.

There are two secondary contact areas in northern Philippines, where two geographically different haplogroups (I and II) of G. salicornia are geographically reunited; one area is Batanes Islands, Bashi Channel, where H1 and H2 populations (Haplogroup I) may contact H25 population (Haplogroup II), and the other is Balaoan, La Union, where H4 population (Haplogroup I) likely contact H27 population (Haplogroup II). This result indicates the range expansion of Haplogroup I from west to east, suggesting the potential of reestablishment of gene flow between two groups (Bertl et al. 2018).

Our haplotype network revealed that most haplotypes especially in haplogroup I were linked by a single mutation to the ancestral haplotype. This might be a result of habitat preference or sea surface circulation by seasonal monsoons (Fumo and Sherwood 2023; Grant 2015; Hewitt 2000; Ludt and Rocha 2015; Ng et al. 2017). Muangmai et al. (2023) reported that genetic divergence within Thailand populations of G. salicornia was likely due to the geographic barrier of the Thai-Malay Peninsula and complex ocean currents.

Gracilaria salicornia was supposedly introduced from the Philippines to Hawaii (Abbott 1999; Smith et al. 2004). Ferrer et al. (2019) found an identical haplotype between the Philippines (four specimens from Batanes, Batangas, and Pangasinan) and Hawaii (eight specimens from Hawaii, Molokai, Oahu of the Hawaiian Islands, Sherwood et al. 2010), and supported the likelihood of anthropogenic transport from the Philippines to Hawaii. However, Acanthophora spicifera (M.Vahl) Børgesen (Rhodomelaceae) is a well-known case of a red algal introduction from Guam to Hawaii (Sherwood and Guiry 2023). To confirm whether the Hawaiian specimens originated in the Philippines or Guam, COI-5P analysis of Guam specimens of G. salicornia is required.

The occurrence of G. salicornia in India was demonstrated based on published COI-5P data (MT939890). However, although its occurrence in South Africa was reported in a previous study based on RuBisCO spacer and morphology (Iyer et al. 2005), the sequence in GenBank (AY241149) fell outside the monophyletic G. salicornia clade, as also shown by Muangmai et al. (2023). Our sequencing of three specimens from Lala Neck, Durban, South Africa failed due to contamination by marine bacteria. The occurrence of G. salicornia in South Africa to Tanzania and Red Sea (e.g., Einav et al. 2021; Iyer et al. 2005; Silva et al. 1996) requires molecular confirmation. Sequence data is lacking from Australia and Pacific Islands, although the morphology of the species was well documented (Meneses and Abbott 1987; Withell et al. 1994).

In conclusion, our data from Vietnamese G. salicornia is a significant addition to the phylogeography of the species in Southeast Asia, a tropical biodiversity hotspot (Boag et al. 2021) and brings together previous studies of other regions for an important contribution to our understanding of the genetic diversity and biogeography of global populations from outside Southeast Asia. This is the first study to examine the genetic variability and structure of G. salicornia in Southeast Asia, incorporating publicly available sequences from GenBank. Furthermore, to our knowledge, this represents the second report on the historical phylogeography of red algae in Southeast Asia using both mismatch distribution and Bayesian skyline plot analyses following Fontana et al. (2024) who studied Dichotomaria elegans (Nemaliales) from Taiwan, at the northeastern boundary of Southeast Asia. While our findings highlight genetic variability and haplogroup patterns across Southeast Asia waters, we advocate for additional sampling in surrounding regions to fill haplotype gaps and to further investigate population growth and range expansion scenarios in G. salicornia, an agar-yielding species.