High genetic diversity and mixing of coastal horseshoe crabs (Tachypleus gigas) across major habitats in Sundaland, Indonesia

Introduction

High rates of gene flow are common in marine organisms that are spread across large geographic ranges (Palumbi, 1994; Crandall et al., 2019). Several marine organisms also exhibit low levels of genetic differentiation across large geographic scales (Avise, 2000). Population structures are affected by genetic drift, strong post-settlement selection (Hedgecock, 1986), and spatial-landscape patterns (Johnson & Black, 1998; Watts & Johnson, 2004). Species with limited dispersal abilities are often composed of highly genetically structured populations with small geographic ranges (Collin, 2001). This creates opportunities to compare the depths and positions of intraspecific genetic differentiation when using location as an extrinsic factor (Bernardi & Talley, 2000).

Horseshoe crabs, an interesting group of marine organisms considered “living fossils” (Eldredge & Stanley, 1984), have been extant for almost 500 million years. There are four extant species of horseshoe crabs: the American horseshoe crab (Limulus polyphemus) found along the eastern coast of North America from Maine to Mexico (Walls, Berkson & Smith, 2002; Rutecki, Carmichael & Valiela, 2004), and three Asian horseshoe crabs species (the mangrove horseshoe crab (Carcinoscorpius rotundicauda), the coastal horseshoe crab (Tachypleus gigas), and the tri-spined horseshoe crab (Tachypleus tridentatus)) (John et al., 2018; Vestbo et al., 2018) that are sporadically distributed across Southeast Asia and Japan. They are ancient marine arthropods that exhibit life-histories and habitat preferences that suggest a restricted dispersal ability (Sekiguchi, 1988). The Asian species are found in Indonesian coastal waters, dispersed around Sumatra, Java, Kalimantan, and Sulawesi (Rubiyanto, 2012; Mashar et al., 2017; Meilana et al., 2016).

Throughout their life cycle, horseshoe crabs are highly dependent on environmental conditions in coastal habitats. Most research suggests that they are declining both locally and regionally due to over-harvesting for food and biomedicine, and coastal development (Itow, 1993; Botton, 2001; Chen, Yeh & Lin, 2004) and the loss of suitable spawning grounds. T. gigas was once relatively common along the northern Java Sea. However, coastal and mangrove horseshoe crab populations have an undetermined conservation status due to insufficient data (John et al., 2021). Furthermore, most population genetic studies on horseshoe crabs have focused on the American horseshoe crab, with little attention paid to the Asian horseshoe crab (Pierce, Tan & Gaffney, 2000; King & Eackles, 2004; King et al., 2005; Yang et al., 2007; Rozihan & Ismail, 2011; King et al., 2015). Therefore, this study examined the genetic diversity, connectivity, and population structure of coastal horseshoe crabs by screening an AT-rich region of mitochondrial DNA, an established genetic marker for arthropods (Brehm et al., 2001). Our aim was to use genetic evidence to facilitate horseshoe crab conservation efforts in Indonesia.

Materials & Methods

Study area and sample collection

With the help of a local fisherman, adult and juvenile T. gigas specimens were collected from shallow waters in six locations around Indonesia: Bintan, Balikpapan, Demak, Madura, Subang, and Ujung Kulon (Fig. 1). We collected the hemolymph from a total of 91 T. gigas specimens between April 2019 and August 2020. There were eight, 14, 16, 13, 20, and 20 samples from Bintan Island (BT), Balikpapan (BP), Demak (DK), Madura (MD), Subang (SB), and Ujung Kulon (UK), respectively. The hemolymph was collected from each individual and immediately preserved in absolute ethanol. Field experiments were approved by the Research Council of the Study Program from IPB University (letter number 1426/IT3.F3.2/KP.03.03.2019).

Results

Population connectivity

The relationship of the 34 haplotypes was illustrated using a median-joining network (Fig. 2). The haplotype network showed that there were shared haplotypes (H1, H3, H5, H6, H8, H9, and H18) across the geographic sites. H3 was the most common, and was identified in all populations except UK and including 15 individuals. H5 was found in 12 individuals from the BT, BP, DK, SB, and UK populations. However, specific haplotypes were only found in certain locations. The UK population had the highest number of specific haplotypes (seven). Meanwhile, BT had the lowest number of haplotypes (two) (Fig. 3).

We assessed historical demography based on mtDNA AT-rich region haplotype frequencies. There were shared haplotypes in all locations (Fig. 2). Furthermore, the Tajima’s D test values (Table 5) were negative across all populations, with the exception of DK, MD, and SB. They showed no significant p-values, indicating that there was no evidence of selection. Similarly, the Fu’s F s test results (Table 5) were negative (except in DK), with no significant p-values across all six populations. This indicated an excess number of haplotypes, as expected due to a recent population expansion.

Discussion

In this study, there was high haplotype diversity in six coastal horseshoe crab populations in the northern Java Sea, Bintan, and Balikpapan waters of Indonesia. There was also a high number of polymorphic sites (43, with 34 defined haplotypes) in Indonesian coastal horseshoe crab populations. The mean haplotype diversity (h = 0.935) was quite high, while nucleotide diversity (π = 0.006) was low across all populations. Similarly high haplotype diversity values were reported in T. gigas (h = 0.797 ± 0.129 and π = 0.058 ± 0.001; Rozihan & Ismail, 2011) in Malaysia and tri-spined horseshoe crab (T. tridentatus) in Taiwan (h = 0.626 ± 0.075 and π = 0.003 ± 0.005; Yang et al., 2007).

Previous studies reported generally high genetic diversity in coastal horseshoe crab (Rozihan & Ismail, 2011; Aini et al., 2020). Our results showed not only high genetic diversity, but also low nucleotide diversity. The high number of haplotypes indicates that these populations were large enough to maintain a high level of genetic diversity. These small differences are the signature of rapid demographic expansion from a small effective population size (Avise, 2000). Nucleotide diversity is a sensitive index when analyzing population genetic diversity (Nei & Li, 1979), and is influenced by life-history characteristics, environmental heterogeneity, population size (Nei, 1987; Avise, 2000), fishing pressure (Madduppa, Timm & Kochzius, 2018), level of larval transport, and degree exchange with other populations (Timm et al., 2017). The rate of mitochondrial evolution and historical factors play an important role in determining genetic variability patterns (Grant, Spies & Canino, 2006; Xiao et al., 2009; Yamaguchi, Nakajima & Taniguchi, 2010).

We detected low differentiation across populations (insignificant F_ST values between 0 and 0.13), with exceptions between populations UK-MD and UK-SB. This result indicated that there was little subdivision across populations. Several studies suggested restricted dispersal abilities for horseshoe crabs regarding short-term tagging. However, some others explained that this crab has a wide dispersal abilities based on long-term studies. Individual distances up to 30 km have been observed in Malaysian crabs (Mohamad et al., 2019), while the movement abilities of tri-spined horseshoe crab did not exceed 150 km (Yang et al., 2007). Similarly, the American horseshoe crab in the Great Bay Estuary (USA) has a maximum mean annual linear distance ranging between 4.5 km and 9.2 km (Schaller, Chabot & Watson, 2010). Studies by Swan (2005) over multiple years found that Limulus moved from 104 to 265 km from their release sites. Ecological observations showed that their hatched larvae swim freely for approximately 6 days and then settle in the bottom of shallow waters around their natal beaches (Shuster Jr, 1982). However, larvae have a strong tendency to concentrate in inshore rather than offshore waters (100–200 km) (Botton & Loveland, 2003), suggesting a limited ability for long-range dispersal between estuaries. Additionally, low F_ST levels reflect inter-population movement over mutigenerational intervals that short-term tagging studies cannot document. Long-term tagging studies have found that horseshoe crabs can move from >5–500 predominated km 5–30 km (Beekey & Mattei, 2015), and up to 767 km over their long lifetimes (E. Hallerman, 2020, personal communication). Long-term tagging study similar study by Rozihan & Ismail (2011) reported that the crab’s F_ST value along the west coast of peninsular Malaysia ranges from 0.111–0.557, indicating moderate to high genetic differentiation (Wright, 1978; Hartl & Clark, 1997). Other reports in the area used microsatellite markers to find a F_ST value between 0.144 and 0.846.

There were only seven shared haplotypes among the 34 total haplotypes observed among all 91 samples. The median-joining network analysis indicated past population expansions with shared haplotypes among localities. Overall, relationship patterns at the mtDNA level showed little geographical structure. The haplotype network revealed recent demographic processes, but the small sample sizes also limited the possibility of observing the intermediate haplotypes inferred to exist in the network. Moreover, results of Tajima’s D and Fu’s F s tests indicated the occurrence of population expansion. Common haplotypes shared between localities also can be explained by the historical biogeography in this Southeast Asian region known as the Sunda Shelves, which includes Java, Sumatera, and Borneo. Historically, Sundaland experienced both dewatering and inundation during the Pleistocene period. Haplotype sharing in this study is attributed to breeding migration and dispersal of pelagic larvae, as well as the sharing of common ancestors (Frankham, 1996). The occurrence of many geographic site-specific haplotypes can be explained by the small sample size and perhaps historical isolation during the Last Glacial Maximum. Many species became isolated in refugia, and genetic differentiation and divergence occurred due to the retreat and dispersal of glacial ice sheets (Hewitt, 2000).

A proactive management approach regarding the Asian coastal horseshoe crab (T. gigas) in Indonesia should consider population genetics. High haplotype diversity that occurs with low nucleotide diversity has been associated with population growth or expansion after a period of low effective population growth (Grant & Bowen, 1998). Our findings indicate that T. gigas in Indonesia have low genetic differentiation but high population connectivity and expansion. Therefore, our results suggest that there is a single stock of Indonesia coastal horseshoe crab. The best conservation strategy would be one that combines both local and regional management. To expand our knowledge base, an advanced population genetic analysis based on male and female horseshoe crabs and the nuclear genome (e.g., microsatellites or SNPs) should be conducted. This should also include expanding the scope of geographic sampling around Indonesia.

Conclusion

High genetic diversity and low levels of differentiation across coastal horseshoe crab (T. gigas) populations in Indonesia indicated a single stock with high connectivity. A locally and regionally based conservation management method is suggested as a precautionary approach to conserving the Indonesian coastal horseshoe crab.

Supplemental Information