2.1 Sample Collection

A total of 12 mature samples of Periophthalmus, three samples for each species, were collected from four different sites (Table 1) in the Mekong Delta (Figure 1) from April 2020 to March 2021 consisting of Tra Vinh (9°40′29.5″N 106°34′49.5″ E); Soc Trang (9°26′19.7″N 105°10′48.1″ E); Bac Lieu (9°05′50.5″N 105°29′54.7″ E) and Ca Mau (8°58′10.4″N 105°22′58.9″ E). These sites were selected for sampling to represent the natural mudflat areas in the Mekong Delta, which were heavily influenced by the Mekong River flow. Furthermore, these sites were reported to have abundant mudskipper species (Dinh et al. 2022). The specimens were identified based on external morphological characteristics described in Table 2 and are consistent with previous descriptions by Jaafar and Murdy (2017) and Tran et al. (2013).

The mudskipper fish was sampled monthly by hand-catching at the mudflat in each sampling site. The mudskipper samples were then cleaned, and preliminary species were classified based on the outward feature (Table 2). Mudskipper fishes were anesthetized with tricaine methanesulfonate solution (MS222) and kept in absolute cold alcohol (Animal Welfare Assessment No. BQ2020-03/KSP), then transported to the Lab and reserved at −20°C till use for DNA isolation.

2.2 Molecular Procedures

2.3 Dataset Preparation

2.4 Genetic Diversity and Phylogenetic Analyses

3.1 PCR Amplification and Species Identification

The amplification of the Cytb mitochondrial gene from 12 mudskipper specimens from the mudflat areas in the Mekong Delta produced a fragment of about 1300 bp in length. Sequence lengths after chromatographic-based editing ranged from 1135 to 1140 bp.

Sequences in GenBank are considered to be most similar when the maximum and total scores are the same, query coverage is close to 100%, and identity is close to 100%. Based on these criteria, all sequences had low similarity to three reference species in GenBank: P. minutus, P. magnuspinnatus, and P. barbarus. As there is not much data on the three mudskippers kept in GenBank, the identity percentage is low.

At a glance at Table 4, all 12 samples of mudskipper in the present study were identified as belonging to only one genus, Periophthalmus. The similarity of the samples in the study with the data in GenBank is relatively low, about 86.20%–87.20%. This result suggested that they belong to the same genus, Periophthalmus, but are different species. All of the Cytb gene sequences of mudskipper samples have been submitted in GenBank with accession numbers OP777919–OP777922 for P. chrysospilos, OP777923–OP777926 for P. variabilis, and OP777927–OP777930 for P. gracilis.

3.2 Nucleotide Composition

In general, three mudskipper species had different nucleotide compositions. Specifically, in P. chrysospilos, the nucleotide compositions of C were highest, while in P. gracilis, %T was highest, and in P. variabilis, %C and %T were equivalent (Table 5). All three species had %AT content higher than %GC; however, %AT was highest in P. gracilis, followed by P. variabilis, and lowest in P. chrysospilos. The variation in nucleotide percentage indicated variation between the three Periophthalmus species.

Three Periophthalmus species were rich in C (30%–32%) but low in G (14%–15%). Similar nucleotide percentages were also recorded in studies of Cantatore et al. (1994) and Almodóvar et al. (2000). Symonová and Suh (2019) noted that the ratio of AT/GC in fish genomes was similar, and %GC was inherently low. In this study, AT/GC was about 1.2 in all sequences, and %AT was consistently greater than %GC. In the research of Çiftci et al. (2013) about the sturgeon species in the Black Sea, the %AT of the Cytb gene was also higher and consistently more remarkable than the %GC, similar to the report of Apostolidis et al. (1997) on Salmo trutta L. (Greek brown trout). The percentage of AT was about 53%–55%, suggesting that the genetic diversity of these mudskippers was moderate. The high AT was common in fish and could be found in easily hypervariable locations (Simon 1991). High AT nucleotide content is of mitochondrial DNA (Min and Hickey 2007). High AT content may result from natural selection favoring the accumulation of A and T nucleotides over G and C nucleotides in the mitochondrial genome. The exact mechanism of this AT bias is not yet fully understood (Arisuryanti et al. 2024).

3.3 Genetic Distance

The intraspecific and interspecific genetic differences are essential information for species classification by DNA barcode (Hebert et al. 2003; Chakraborty and Ghosh 2014). Zemlak et al. (2009) stated that fish species are considered to be the same species if the genetic distance is less than or equal to 3.5%; otherwise, they are considered different species. In the study, the conspecific genetic distances of three mudskipper species were 0% (Table 6), proving that there were no genetic variations between the individuals of each species sequenced in this study. Thus, the Cytb gene helped to identify three mudskipper species of the genus Periophthalmus accurately.

The standard threshold of congeneric genetic distances in taxonomy has not yet been found (Bingpeng et al. 2018); for example, 22.2% for Indian goby species (Viswambharan et al. 2015) or 9.93% for Australian goby species (Ward et al. 2005). The K2P congeneric genetic distance in the present study fell in the 15%–18% range. Namely, the interspecific genetic distances of P. chrysospilos—P. variabilis, P. chrysospilos—P. gracilis, and P. variabilis—P. gracilis were 15%, 18%, and 17%, respectively. This result is similar to the genetic distance between Periophthalmus species in research and species on GenBank, such as P. waltoni, P. modestus, P. magnuspinnatus, P. minutus, P. barbarus, P. argentilineatus, and P. novemradiatus (13%–18%). Thus, the interspecific genetic distance of three Periophthalmus species in this study of about 15%–18% is reasonable. This result shows that the Cytb gene helped correctly identify three mudskipper species of the genus Periophthalmus, consistent with their morphological differences.

3.4 Phylogenetic Analysis

The phylogenetic tree in Figure 2 was separated into two main clades arising from the root of this phylogenetic tree: clade I included only one genus, Periophthalmus; all species of Glossogobius and Butis were grouped in clade II. The relationships between the two clades were well-supported (99%–100% bootstrap percentage). Clade I was divided into 10 sub-clades corresponding to three species, P. chrysospilos, P. variabilis, and P. gracilis in this study, and seven control species on GenBank included P. waltoni, P. modestus, P. magnuspinnatus, P. minutus, P. barbarus, P. argentilineatus, and P. novemradiatus. A bootstrap value of 100% supported the formation of 10 sub-clades in the Maximum Likelihood methods, showing that these sub-clades were sturdy and dependable. Within each sub-clade, bootstrap values were lower than 100% (ranging from 64%–99%), possibly due to low numbers of species represented in the phylogeny, thus “missing links” reducing support.

The divergence of these sub-clades was supported by morphological data in which P. chrysospilos had small orange spots, P. variabilis had black spots, while P. gracilis had horizontal bars on the body. In addition, pelvic fins in P. variabilis and P. gracilis were separated but conjugated from the disc in P. chrysospilos. They shared a common feature: the upper jaw had a single row of teeth. The complete morphological differences of the three species (described in Table 2) and their apparent separation in the phylogeny (Figure 2) reinforce that they are three distinct species and not due to phenotypic plasticity. This outcome was in agreement with the previous report of Jaafar and Murdy (2017) and Tran et al. (2013).

There was no geographical variation among the three mudskipper species of Periophthalmus genus, that is, samples from different locations belonged to the same species, suggesting a lack of geographical structure among them. Similarly, in the study of Febrianti et al. (2023), Periophthalmus kalolo samples collected from different locations in Indonesia belonged to the same species due to the intraspecific genetic distance of 3.5%. Another example, P. kalolo samples also did not show any significant genetic differentiation among populations in Indonesia (Arisuryanti et al. 2024). Interestingly, Cytb showed no variation along approximately 200 km of coastline in the Mekong Delta from Tra Vinh to Ca Mau. In other words, it seems that the individuals' exchange between far away populations occurs frequently enough to homogenize the genetic pool. The lack of geographical separation may be due to the small sample size. Future studies will require larger sample sizes to better understand these three species' genetic diversity patterns and geographical structure. In short, the sequences of the Cytb gene as a DNA barcode helped to identify the phylogeny among three species in the Periophthalmus genus collected from the estuaries of the Vietnamese Mekong Delta. An improved understanding of the genetic relationship of this genus is valuable in conserving genetic diversity and species management.

Mudskippers contribute significantly to the balance of the food chain in mangrove ecosystems (Budi et al. 2018). They ingest detritus and organic matter and then release nutrients into the environment (Shojaei et al. 2022). Their presence and activities in mangroves affect the ecological health and stability of these areas. Moreover, mudskippers can accumulate metal contaminants, so they are considered valuable biological indicators and monitoring agents for mangrove ecosystems (Arisuryanti et al. 2024). Accurate species identification and understanding of the genetic relationships of Periophthalmus mudskipper species will help conserve mudskipper species and ensure the overall ecological health of these areas.