Morphological Diversity of Gracilaria blodgettii Harvey 1853 (Gracilariaceae, Rhodophyta) from Sarawak, Malaysian Borneo

Gracilaria red algae are notable for their economic importance as agrophytes, sold as salad vegetable, and used as the base for selected food and nonalcoholic beverages. A wild population of Gracilaria exists in coastal areas of Sarawak, Malaysian Borneo, but there is only limited knowledge on species diversity and its abundance leaving the untapped economic potential of this resource. This study was carried out to determine diversity of wild Gracilaria populations in Lawas, Santubong, and Asajaya, Sarawak, using the combination of morphological character examination and 5′ region of the mitochondrial cytochrome c oxidase 1 (CO1-5P) gene analysis. Identification of the species using morphological characters revealed three species, namely, Gracilaria changii, G. blodgettii, and G. arcuata, had been collected from the sampling sites. However, based on 672 bp CO1-5P gene sequence analysis, all the three species were identified as G. blodgettii; besides, low genetic divergence values (0.17%–0.34%) were scored between samples in this study with the same species in GenBank. In the phylogenetic trees, all samples in this study group together with other G. blodgettii have high bootstrap values; thus, this species is monophyletic. This study implies that species identification of Gracilaria and other seagrass taxa which have a phenotypic plasticity problem should include the CO1-5P gene analysis as it is a reliable gene marker for species diversity assessment.


Introduction
Genus Gracilaria Greville consists of more than 170 species worldwide, distributed from tropical to temperate waters, covering from intertidal to subtidal zones [1][2][3]. It can be found from Arctic Ocean to tropical seas of the northern hemisphere and countries of Southeast Asian regions such as Malaysia, Indonesia, ailand, Vietnam, Singapore, and Philippines [4,5]. Gracilaria is important in production of agar in food industry [6] and culture medium in research industry [7], and it serves as a habitat for various aquatic organisms [8,9], besides becoming food for the local people.
Up to now, 20 species of Gracilaria had been identified in Malaysia [10], where half of them were found in Sarawak, namely, G. arcuata Zanardini, G. articulata Chang & Xia, G.
Molecular studies have shown positive results in solving the identification and taxonomy of seaweeds worldwide. Gracilaria is known as seaweeds with high plasticity characteristics and simple morphologies with very minor variations among them and sometimes have different structures throughout its life cycles [14,15]. Various gene markers had been used by researchers in identification of Gracilaria such as 5′ region of mitochondrial cytochrome c oxidase 1 (CO1-5P) [16], plastid-encoded large subunit of the ribulose-1,5biphosphate carboxylase (rbcL) [17], nuclear internal transcribed spacer (ITS) [18], and intergenic spacer between the cytochrome oxidase subunits 2 and 3 (cox2-3 spacer) [19]. Recently, CO1-5P gene marker has been used widely due to its ability to identify red seaweed at the species level, revealing cryptic species [15,20].
Ho et al. [21] had sequenced the nuclear and chloroplast genomes of Gracilaria changii. ey reported that the partial nuclear genome is 35.8 Mb with 10,912 predicted proteins, while the chloroplast genome is 183,855 bp with 201 ORFs, 29 tRNAs, and 3 rRNAs. Esa [22] had produced preliminary CO1-5P gene sequences for Caulerpa spp. inhabiting Sarawak, and Song et al. [10] reported on microsatellite markers from expressed sequence tags (ESTs) of seaweed usage in differentiating various Gracilaria species, some of which had been obtained from Sarawak. Since phenotypic plasticity occurs in seaweeds, identification of Gracilaria species becomes a very challenging task. In this study, identification of Gracilaria species collected from Lawas, Santubong, and Asajaya, Sarawak, had used two approaches: (i) CO1-5P gene marker and (ii) conventional approach of identification using morphological characters.
For molecular analysis, the samples were cleaned using distilled water to remove the epiphytes, aquatic organisms, and sand particles which may trap within the thalli. en, each thallus was kept separately in a labelled plastic bag and stored in the −20°C freezer until further analysis. e DNA extraction of Gracilaria samples was done following the standard cetyl-trimethyl ammonium bromide (CTAB) protocol by Doyle and Doyle [28], followed by 1% agarose gel electrophoresis (AGE). e amplification of CO1-5P gene was done using the polymerase chain reaction (PCR) technique with primers designed by Saunders [16]: forward primer GazF1 (5′-TCAACAAATCATAAAGATATTGG-3′) and reverse primer GazR1 (5′-ACTTCTGGATGTCCAAAAAAYCA-3′). e total volume of PCR reaction was 25 μl, comprising 10.5 μl ultrapure water, 4 μl 10X buffer, 4 μl MgCl 2 , 2.5 μl dNTPs (Promega), 1.0 μl forward primer GazF1, 1.0 μl reverse primer GazR1, 0.5 μl Taq DNA polymerase (Promega), and 1.5 μl DNA template. e PCR was carried out using a PCR thermocycler (Biometra TAdvanced) under following conditions: predenaturation at 94°C for 1 minute, followed by 35 cycles of denaturation at 94°C for 1 minute, annealing at 51.3°C for 1 minute 30 seconds, extension at 72°C for 1 minute, and final extension at 72°C for 5 minutes. e success of PCR was checked using 1% AGE, and successful PCR products were sent to First Base Sdn Bhd, Selangor, Malaysia, for single-pass DNA sequencing.
CHROMAS software was used to display the CO1-5P sequences, whereas the validation of species used the basic local alignment search tool (BLAST). CLUSTAL X program (version 1.81) was used to align the DNA sequences. Neighbour joining (NJ) trees were constructed using MEGA 6.0 [31], while a model of K81uf + I + G was used for the Bayesian inference implemented on MrBayes program. All phylogenetic trees were constructed together with other Gracilariaceae species obtained from GenBank where Janczewskia hawaiiana and Osmundea pinnatifida act as outgroups (Table 1). e genetic divergence values were obtained using Kimura's two-parameter model [32].

Results
Gracilaria arcuata was found attached to the net of cage culture in Lawas. It had reddish brown colour when fresh with discoid holdfast; the branches were cylindrical, irregular, and arcuate and could grow up to 120 mm tall ( Figure 2(a)). Constriction was observed at the base of every branching and the tip either pointed or divided to two to five short stubby spinose branchlets. e medulla was composed of 4-5 layers of parenchymatous cells surrounded by 2-3 layers of small cortical cells (Figure 3(a)).
Gracilaria blodgettii was found attached to the root of mangrove in Asajaya. e thallus was dark red in colour with discoid holdfast and could grow up to 200 mm tall ( Figure 2(b)). e branching occurs frequently, either secund or irregular with constriction at the base of each branch, and it had pointed tip at the end of branches. e medulla was composed of 3-4 layers of parenchymatous cells surrounded by 2-3 layers of small cortical cells (Figure 3(b)).
Gracilaria changii was found attached to the net of cage culture in Santubong. e colour of G. changii was dark red with discoid holdfast and could grow up to 180 mm-220 mm ( Figure 2(c)). e branching occurs occasionally, irregular with constriction at the base of branches. e tip either pointed or divided into two short branchlets. e medulla was composed of 3-4 layers of parenchymatous cells surrounded by 2-3 layers of small cortical cells (Figure 3(c)).
Based on genetic divergence analysis (Table 3), the three Gracilaria species in this study had genetic divergence values between 0% and 0.15%, when compared among each other. e genetic divergence values between the three Gracilaria species and the outgroups ranged from 22.60% to 24.03%. G. blodgettii (GA01-GA10), G. arcuata (GL01-GL03), and G. changii (GSA01-GSA10) had genetic similarity with G. blodgettii from China and the Philippines as mentioned in Table 2 with a variation of 0.17% to 0.35%. For comparison, high genetic variation was observed between G. arcuata in this study and G. arcuata from Japan and the Philippines within the range of 13.59%-14.90%. e comparison of intraspecific values of G. changii found in Santubong and Lawas with those of similar species from other countries could not be obtained due to unavailability of CO1-5P information in GenBank up to December 2018.
Morphological characteristic data and molecular data showed incongruent results in this study (Table 4). ree   species of Gracilaria, namely, G. changii, G. blodgettii, and G. arcuata, were observed in Lawas, Santubong, and Asajaya, Sarawak, using the morphological approach, whereas only one species, namely, G. blodgettii, was found using the molecular approach. Summary of morphological and molecular data on Gracilaria obtained in this study is shown in Table 5.

Discussion
Based on morphological descriptions, samples GSA01-GSA10, obtained from Santubong, were G. changii; samples GA01-GA10, from Asajaya, were G. blodgettii; and samples GL01-GL03, from Lawas, were G. arcuata as they matched descriptions by Dhargalkar and Kavlekar [23], Ismail [24], Lin [25], Nurridan [12], Nurridan [11], Ohmi [26], and Yamamoto [27]. e presence of these three species in Sarawak is noted in the checklist of seaweed by Nurridan [12] and in her updated checklist in 2013. In addition, Othman et al. [9] had also reported the presence of G. arcuata in Lawas, Sarawak. In another study, based on morphological characters' examinations alone, Othman et al. [13] claimed that there were three species, namely, G. changii, G. blodgettii, and G. coronopifolia, found attached to the nest of cage culture in Santubong and the root of mangrove trees in Asajaya, Sarawak. Previous researchers, for example, Phang et al. [14] and Saunders [15], reported that Gracilaria possesses high plasticity characteristics and simple morphologies with minor variations among species and may have different structures throughout its life cycle. us, the morphological data of Gracilaria in this study should be critically examined.
In this study, CO1-5P sequences obtained were between 660 bp and 672 bp in length. Similarly, Kim et al. [20] reported that the Korean Gracilariaceae sequences had a length of 670 bp to 685 bp, while Gracilariaceae in Qingdao, China, had a length of 664 bp for CO1-5P genes [33]. All 23 sequences obtained in this study matched G. blodgettii from China and the Philippines with the expected value (E value) equal to zero. E value was used to determine the level of significance between DNA sequences obtained with the mitochondrial DNA voucher deposited in GenBank, where    6 Scientifica the closer the E value to zero, the higher the similarity of the match [34]. erefore, all sequences in this study had significant match with the database, confirming all samples were G. blodgettii. According to Kim et al. [20], Le Gall and Saunders [35], and Saunders [15], the intraspecific divergence value more than 2.0% is considered as different species. e low value of genetic divergence (0.17% to 0.35%) among these three species in Sarawak with relevant vouchers from GenBank suggested that they belong to one species, namely, G. blodgettii. In the phylogenetic trees, it is noted that G. blodgettii, G. arcuata, and G. changii in this study with G. blodgettii from GenBank (Clade I, the first subclade) had recorded a high bootstrap support with values of 99% (NJ) and 1.00 (BPP). e strong value of bootstrap support given to this clade means it is very likely that the relationship is true, further supporting that all specimens belonged to the same species.
According to Saunders [15], misidentification of Gracilaria species may likely to happen due to its simple and high plasticity characters. Zhao et al. [33] reported that not all red seaweeds could be identified based on the morphological approach, especially Gracilariaceae due to the following reasons: (i) they have high varieties of morphologies within the species; (ii) they have simple morphological characteristics and highly convergent morphology which make them look similar to others leading to confusion during identification; (iii) lack of distinct parts or features to differentiate among species; and (iv) their reproduction cycle is very complex due to heteromorphic alternation of generation. Similarly, Md Sah et al. [14] also claimed that the identification of Gracilaria could be problematic because of limitations of distinct morphological and reproductive characteristics. For example, in Virginia, omsen et al. [36] had corrected the identification of G. vermiculophylla which at first has been referred to as G. verrucosa and G. tikvahiae.   [15]. ere are also cases where the same species have different characteristics because of environmental factors that lead to confusion during identification [37,38]. In this study, identification of Gracilaria species (G. blodgettii, G. arcuata, and G. changii) using the morphological approach most likely resulted in misidentification because of either a variety of morphologies, complexity of the life cycle, or the environments where they grow. In addition, to avoid further confusion, omsen et al. [36] suggested that each specimen should be kept properly using air-drying or herbarium-pressing methods so that future studies could refer to them as reference materials. Guiry and Guiry [3] also reported in their taxonomic note that the samples identified as G. blodgettii should be compared with the topotype material from Key West, Florida, USA, for confirmation and avoiding further misidentification. For now, the sequences of Gracilaria samples in Sarawak, Malaysian Borneo, should be identified as Gracilaria aff. blodgettii until we could compare the samples with G. blodgettii from Key West, Florida, USA. us, the study on seaweed identification should be continued to resolve the taxonomy and create clear understanding on their morphology including plasticity characteristics that respond to environmental changes.

Conclusion
is study suggests that three Gracilaria species obtained in Lawas, Santubong, and Asajaya, Sarawak, which were initially identified as G. blodgettii, G. arcuata, and G. changii, should be renamed into one species, namely, Gracilaria blodgettii Harvey 1853. Minor differences between specimens in terms of morphology may be due to environmental influence. For future seaweed research, CO1-5P gene markers should be sequenced and analysed, besides the conventional methods of using species identification keys provided by previous researchers.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.