Population genetic differentiation of the hydrothermal vent crab Austinograea alayseae (Crustacea: Bythograeidae) in the Southwest Pacific Ocean

To understand the origin, migration, and distribution of organisms across disjunct deep-sea vent habitats, previous studies have documented the population genetic structures of widely distributed fauna, such as gastropods, bivalves, barnacles, and squat lobsters. However, a limited number of investigations has been conducted in the Southwest Pacific Ocean, and many questions remain. In this study, we determined the population structure of the bythograeid crab Austinograea alayseae from three adjacent vent systems (Manus Basin, North Fiji Basin, and Tonga Arc) in the Southwest Pacific Ocean using the sequences of two mitochondrial genes (COI and 16S rDNA) and one nuclear gene (28S rDNA). Populations were divided into a Manus clade and a North Fiji–Tonga clade, with sequence divergence values in the middle of the barcoding gap for bythograeids. We inferred that hydrographic and/or physical barriers act on the gene flow of A. alayseae between the Manus and North Fiji basins. Austinograea alayseae individuals interact freely between the North Fiji Basin and the Lau Basin (Tonga Arc). Although further studies of genetic differentiation over a geological time scale, life-history attributes, and genome-based population genetics are needed to improve our understanding of the evolutionary history of A. alayseae, our results contribute to elucidating the phylogeny, evolution, and biogeography of bythograeids.


Introduction
Hydrothermal vent environments are characterized by a lack of light (aside from that generated by high-temperature fluids [1]), lack of photosynthesis, high pressure, steep temperature gradients, and high levels of metals and dissolved gases [2][3][4]. Since the discovery of hydrothermal vents along the Galapagos Rift in 1977, the description of new species has progressed PLOS ONE | https://doi.org/10.1371/journal.pone.0215829 April 24, 2019 1 / 15 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 We also identified population genetic divergence and migration events of A. alayseae, and discuss the biogeographic connections among these three vent regions in the Southwest Pacific Ocean.

Vent crab sampling and identification
Specimens of bythograeid crabs were collected using suction samplers mounted on remotely operated vehicles from twelve vent sites in the Manus Basin, the North Fiji Basin, and the Tonga Arc of the Southwest Pacific Ocean (Fig 1; Table 2). On board, all specimens were immediately preserved in 95% ethanol or stored at -80˚C until genetic analysis. The specimens were identified as A. alayseae, A. hourdezi, and Gandalfus puia on the basis of the morphological characteristics and cytochrome oxidase subunit I (COI) DNA barcodes, following the methods used in previous studies [33,39,40]. Detailed information about the specimens is provided in S1 Table. DNA extraction, PCR amplification, and sequencing A microscopic section of muscle tissue was dissected from a pereopod of each specimen for DNA extraction. Total genomic DNA was extracted using the RED Extract-N-Amp PCR Kit (Sigma-Aldrich Co., Brooklyn, NY, USA) [41]. The partial sequences of two mitochondrial genes (COI and 16S rDNA) and one nuclear gene (28S rDNA) were determined using previously published primers (Table 3). PCR amplification was performed in a total volume of 50 μL containing 1 μL genomic DNA, 4 μL dNTP mixture (2.5 mM each), 1 μL (10 pmol) of each primer, 5 μL 10X Ex Taq Buffer (Mg2+ plus), and 1.25 U Takara Ex Taq DNA Polymerase

Phylogenetic analysis
The new sequences obtained in this study were aligned with those of other bythograeids and the Atlantic blue crab Callinectes sapidus, used as an outgroup (S2 Table), which were retrieved from GenBank using the Geneious alignment method implemented in Geneious Prime, and adjusted through visual inspection. Intra-and interspecific variations in individual gene alignment were calculated using MEGA X [45] based on the p-distance value.
To construct a phylogenetic tree of bythograeid crabs, individual COI, 16S rRNA, and 28S rRNA gene alignments were concatenated to form a single multiple-sequence alignment using Geneious Prime. The best-fitting model of nucleotide substitution was then determined using the Akaike information criterion (AIC) in JModelTest 2.1.7 [46], and the model GTR + I + G was selected as the best evolution model. Next, maximum likelihood (ML) and Bayesian inference (BI) tests were performed using RAxML version 8.2.11 [47] and MrBayes 3.2.6 [48], respectively, implemented in Geneious Prime with the gene partition option. Confidence in the resulting bythograeid relationships was assessed based on the bootstrap proportion (BP) with 100 replications for the ML model. For the BI analysis, four Markov chain Monte Carlo chains were run for 1,000,000 generations and sampled every 200 generations. Bayesian posterior probability (BPP) values were estimated after the initial 500 (10%) trees were discarded as burn-in.
To determine the genetic relationships among hydrothermal vent field populations and COI haplotypes, median-joining networks [52] were created using Arlequin v. 3.5 [50] and graphed with PopArt v. 1.7 [53]. Migration rates, population sizes, and relative numbers of migrants were estimated using MIGRATE v. 4.4.0 [54].

Phylogenetic position of A. alayseae within the bythograeid lineage
Austinograea alayseae is distributed widely across hydrothermal vent areas in the Southwest Pacific Ocean. In this study, we obtained the sequences of three genes, two mitochondrial genes (COI and 16S rDNA) and one nuclear gene (28S rDNA), from 38 specimens of A. alayseae, which included 10 individuals from the Manus Basin (Manus population), 11 individuals from the North Fiji Basin (North Fiji population), and 17 individuals from the Tonga Arc (Tonga population). Then, phylogenetic trees of bythograeids were constructed using the concatenated sequences of the three genes (Fig 2; (Table 4). The proportions of intraspecific variation in 16S rDNA and COI were 0.00-1.69% and 0.00-4.08%, respectively. We observed the maximum variation between the Manus and North Fiji-Tonga clades (Table 5). In particular, the maximum value of COI variation was found in the middle of the barcoding gap, which is the gap between inter-and intraspecific variation, for bythograeids [34,40].

Population genetic divergence of A. alayseae
We examined whether the differences between these two A. alayseae clades represented genetic divergence at the population level. Alignment of the COI sequences of A. alayseae allowed detection of 32 variable nucleotide sites at the third position and two sites at the first position in the codons. The degree of intrapopulation variation ranged from 0.00% to 1.19%, and the maximum divergence was found in the North Fiji population ( Table 5). The degree of interpopulation variation between the Manus population and the other two regions was 3.23-4.08%, and the maximum value for divergence between the North Fiji and Tonga populations was only 1.02%, within the range of intrapopulation variation among the three populations. In addition, based on pairwise comparison of F ST , the Manus population differed significantly from the North Fiji and Tonga populations, whereas no difference was found between the North Fiji and Tonga populations.
Based on the variable sites in the COI sequences of A. alayseae, we examined 20 haplotypes, which showed an overall haplotype diversity of 0.94 (± 0.022). The Manus, North Fiji, and Tonga populations consisted of seven, eight, and nine haplotypes, respectively. All three populations showed relatively high (>0.8) degrees of haplotype diversity. The highest diversity value (0.93 ± 0.067) was obtained for the North Fiji population. Four haplotypes were identified in both the North Fiji and Tonga populations, whereas none were shared between the Manus and North Fiji-Tonga populations. According to the AMOVA, haplotype variation between the Manus and North Fiji-Tonga populations was greater than that between North Fiji and Tonga populations and those among individuals within each population (90.88% vs. 0.00% and 9.12%, respectively), suggesting genetic isolation of A. alayseae in the Manus and North Fiji-Tonga hydrothermal vent field regions. On the other hand, based on migration rates and population sizes, the estimated number of migrants between North Fiji and Tonga indicated bidirectional flow of A. alayseae from North Fiji to Tonga (4289 migrants/generation) and from Tonga to North Fiji (4467). In the haplotype network, the COI haplotypes of A. alayseae were divided distinctly into two clades, the Manus clade and the North Fiji-Tonga clade, with 14 nucleotide substitutions (Fig 3), which well reflects the phylogenetic relationship inferred using three genes (Fig 2). This result indicates that at some time in the past, A. alayseae living in the Southwest Pacific Ocean might have experienced a strong population bottleneck, which influenced separation of the two clades (Fig 2; Table 5). Since then, both clades have undergone independent population expansion (Tajima's D and Fu's F S < 0).

Population genetic structure of invertebrates in the Southwest Pacific Ocean
Among vent organisms distributed evenly around hydrothermal vent fields in the Southwest Pacific Ocean, the vent crab A. alayseae, vent shrimp Chorocaris sp. 2, and black snail Ifremeria Table 4. Intra-and interspecies variations in COI, 16S rDNA, and 28S rDNA sequences of Bythograeidae. Sequence variations were calculated from the nucleotide sequences using the p-distance method in MEGA X.  (Fig 3). Based on this difference among the three species, we inferred that the A. alayseae clades were generated through a stepwise mutation process over a long period of time, with the possibility that intermediate haplotypes became extinct.

Discussion
Previous studies have confirmed the monophyly of bythograeid crabs [17], which are distinguished from other Brachyura by reduction of the eyes at the adult stage and complete adaptation to hydrothermal vent environments [56]. Based on phylogenetic analysis, bythograeid crabs have been divided into two main groups, the Bythograea clade and a clade composed of Austinograea, Gandalfus, Allograea, Cyanagraea, and Segonzacia [17]. However, the established phylogenetic trees were constructed using specimens from a single vent area for each   [55,57].
In the Bythograea clade, B. laubieri and B. vrijenhoeki, which co-occur on the Southeast Pacific Rise, are the most recently diverged species [31]. These sympatric species have relatively little genetic difference (6.6%) in their COI sequences, which is the minimum interspecific variation observed in the family Bythograeidae (Table 4) [40]. In this study, populations of A. alayseae showed relatively large degrees of intraspecific variation (3.23-4.08%) between the Manus and North Fiji-Tonga clades. An analysis of bythograeids excluding A. alayseae revealed <1% intraspecific variation and 6.62-16.13% interspecific variation (Table 4). Although the proportion of variation between the A. alayseae clades is closer to the interspecific variation patterns of other bythograeids, we could not identify any morphological difference based on the original descriptions of Austinograea [29,33]. In many cases, genetic differentiation among related species in recently diverged lineages is not reflected in morphological traits, which occasionally leads to taxonomic ambiguity [58][59][60]. After experiencing a population bottleneck, the original A. alayseae population in the Southwest Pacific Ocean could have experienced independent population expansion and genetic differentiation of the two A. alayseae clades. These findings for A. alayseae raise the possibility of incipient speciation, despite the current absence of morphological traits distinguishing the clades. In addition, Guinot and Segonzac (2018) [33] noted the presence of the invalid Austinograea species, Austinograea sp. aff. A. alayseae, in the Manus Basin. Thus, we cannot rule out the possibility that our specimens from the Manus Basin may be Austinograea sp. aff. A. alayseae which currently does not have confirmed taxonomic status (a new species or subspecies) due to insufficient morphological information and unavailable DNA sequences in public databases.
Larval dispersal is a main driver of gene flow and distribution of vent invertebrates, and it is related to local-and regional-scale geological and hydrological barriers, species-specific development processes, dispersal depth, and water temperature [9,11,12]. According to the modeling of potential larval dispersal distances with a planktonic larval duration of 83 days in the Southwest Pacific Ocean, a bidirectional connection was suggested only between North Fiji and New Hebrides (500 km apart), with an event of dispersal between these regions occurring once every 5,000-12,000 years [15]. Even with an increased planktonic larval duration of 170 days, the model did not assure a bidirectional connection between the Manus and North Fiji basins. Moreover, based on previous studies of the bythograeid crab B. thermydron and other species [61][62][63], the estimated duration of A. alayseae larval development does not exceed 3 months, which is too short for dispersal over a sufficient distance.
To elucidate the genetically differentiated clades of A. alayseae, geological and hydrographic barriers in the Southwest Pacific Ocean were also considered. Previous studies of geological features have led researchers to propose that the Vanuatu and New Guinea archipelagos, Solomon Islands, and New Hebrides, which are relatively young and not fully developed back-arc basins, acted as barriers or stepping stones, forming a genetic connection between the Manus and North Fiji basins [7,15,16]. Although we did not observe intermediate haplotypes between the Manus and North Fiji-Tonga clades, considering the recent differentiation of these clades, such intermediates may exist in newly formed vent regions. On the other hand, to explain the connection between the North Fiji and Tonga populations, we considered the theory that these volcanic areas originated from the disruption of a single arc <10 Mya [18]. Based on the estimated divergence times of decapod species [64] and our results, the divergence of the North Fiji-Tonga clade correlates with this geological event. Thus, the wide area of vent fields ranging from North Fiji to the Tonga/Lau region functioned as a geographical boundary, leading to reproductive isolation within the region and restricted gene flow from outside. In addition, the water masses in the North Fiji and Lau/Tonga regions are well mixed by the South Equatorial Current system [15,65]. This mixing might have enabled bidirectional migration of A. alayseae between the North Fiji Basin and Tonga Arc. After larval dispersal, A. alayseae could have settled quickly and reliably in these regions, which have similar environmental features [12,18].
Many questions remain unanswered concerning the origin and evolution of vent fauna in short-lived patchy habitats that last no more than a few decades. To improve our understanding of evolutionary processes, further research should be considered, including biogeographic analysis of additional taxa, investigation of mutation rates and average generation times of species over a geological time scale, and examination of the roles of other environmental features and life-history attributes, as well as genome-based population studies. Furthermore, to elucidate the gene flow of hydrothermal vent fauna between basins, we should investigate additional specimens of known species from unstudied regions throughout the Southwest Pacific Ocean.