A New Branchipolynoe (Aphroditiformia: Polynoidae) Scale worm from Deep-Sea Hydrothermal Vents in The Onnuri Vent Field on The Northern Central Indian Ridge

Background: Deep-sea hydrothermal vents are dynamic environments with exotic faunas. In this study, we found a new species of Branchipolynoe (Aphroditiformia: Polynoidae) polynoid scale worm living in the recently discovered mussel Gigantidas vrijenhoeki in deep-sea hydrothermal vents and methane seeps at depths of 2,014 - 2,023 m. Associations between scale worms and giant mussels are common in hydrothermal ecosystems. Results: We analyzed the morphology of Branchipolynoe from the Onnuri vent eld (OVF) on the northern Central Indian Ridge and sequenced the full mitochondrial genome. Based on its morphological traits and mitochondrial genes, we designated the specimens as Branchipolynoe onnuriensis n. sp., described herein. This species resembles B. longqiensis and B. tjiasmantoi, but is distinguished by its notopodial acicular lobe form and the tips of the subacicular neurochaetae. The identity of the new species was well supported by genetic distance and phylogenetic analyses of the mitochondrial c oxidase subunit I (COI) gene. Genetically, the new species is closest to the Western Pacic species B. tjiasmantoi; phylogenetic analyses support the correlation between Indian Ocean and Western Pacic hydrothermal polychaetes. This study provides a foundation for exploring the evolutionary relationship between scale worms and giant mussels.


Introduction
The rst hydrothermal vents were discovered on a sea ridge crest off the Galapagos Islands in 1977 [1]. Many vents continue to be discovered in global mid-ocean ridge systems, back-arc diffusion centers, and off-axis submarine volcanoes [2]. Deep-sea hydrothermal vents are dynamic environments with steep nutrient gradients and physicochemical conditions caused by volcanic and tectonic phenomena [3]. Investigations of the diversity and distribution of the deep-sea hydrothermal vent community are critical due to their unique nature [4]. Vents are commonly associated with dense communities and large biomass of organisms that are patchily distributed on the deep-sea bottom [5,6]. These communities contribute signi cantly to the production of chemically synthesized biomass and demonstrate the remarkable adaptability of life in hydrothermal ecosystems [7]. These organisms are maintained in hydrothermal habitats by chemically synthesized energy sources such as hydrogen sul de and methane from hydrothermal mineral deposits [8].
Since the discovery of a hydrothermal vent community at the Kairei vent eld near the Rodriguez Triple Junction in the Indian Ocean in 2000 [9], ve hydrothermal vent ecosystems have been found along the northern Central India Ridge (CIR) and Southwest Indian Ridge (SWIR), including the Dodo, Edmond, Longqu, Solitaire, and Onnuri vent elds [10][11][12]. Therefore, the biogeographical connectivity of vent fauna in the Indian Ocean has become a focus of research [2]. Mid-ocean ridges of the Indian Ocean have been hypothesized as a biogeographical pathway linking vent fauna in the Western Paci c and Atlantic Ridge systems [11]. Thus, the evolutionary a nity of vent invertebrates in the Western Paci c and Indian Ocean vent communities has been proposed to assemble asymmetrically, with a positive bias in the Western Paci c [7].
Polychaetes, in suborder Aphroditiformia (Annelida), are commonly referred to as scale worms due to their dorsal scales (elytra) [12]. Scale worms are classi ed into 242 genera within seven families (Acoetidae, Aphroditidae, Eulepethidae, Iphionidae, Pholoidae, Polynoidae, and Sigalionidae). Scale worms are found throughout the tropics, polar regions, intertidal zones, and deep seas [13][14][15][16]. Branchiate scale worms in the genus Branchipolynoe Pettibone, 1984 live within the mantle cavities of bathymodiolin mussels in hydrothermal vents and methane seeps [17]. Notably, these scale worms have very well developed, arborescent branchiae, which are much larger than those of other branchiate polynoid species [18]. Members of genus Branchipolynoe are also unusual for their small scales, as most scale worms have scales that completely cover the dorsum [19]. In the Indian Ocean, Branchipolynoe scale worms have been reported from deep-sea hydrothermal vents on the SWIR [20].
Branchipolynoe species have been recorded as symbionts of host mussels; however, recent studies have suggested that they may be parasitic [21]. The polynoid Branchipolynoe seepensis has been found in mussels that inhabit hydrothermal vents and methane seeps, causing gill lament displacement and minor gill tissue damage [22].
These scale worms eat particulate organic matter ltered by the mussels, and accidentally consume mussel tissues during the feeding process; thus, they are not considered true parasites or commensals but rather kleptoparasites [23]. According to a biogeographical study of Branchipolynoe phylogenetics, the most recent common ancestor of eastern Paci c Branchipolynoe species may have inhabited methane seeps, and then migrated west to hydrothermal vents and methane seeps of the western Paci c and Indian Oceans [19].
In this study, we examined Branchipolynoe specimens collected on a survey of the hydrothermal vents of the Onnuri vent eld (OVF) in 2019. We identi ed and described a new scale worm species and sequenced the full mitochondrial genome. Finally, we described the anatomy of the new species and evaluated its phylogenetic associations within Branchipolynoe.

Specimen collection and preservation
Specimens of G. vrijenhoeki containing Branchipolynoe were collected from deep-sea hydrothermal vents in the OVF on the northern Central Ridge of India (Additional le 1). The OVF is located along the CIR at 11°24.880' S, 66°25.420' E. All specimens were obtained at depths of 2,014 and 2,023 m using a video-guided hydraulic grab

Morphological examination
We used Leica DMC 4500 cameras mounted on a Leica M205C stereomicroscope for micrography. We used the Helicon Focus v6 software (Helicon Soft Ltd., Kharkiv, Ukraine) to merge images into a stack of pictures, which was then stained with methyl green to observe the parapodial lobes and chaetae in detail. For scanning electron microscope (SEM) observations, several parapodia were isolated from the specimens, rinsed with absolute ethanol, dehydrated, coated with gold, inspected, and photographed using a Hitachi S-4300 SEM. All known Branchipolynoe species were tabulated, and several key characteristics were compared including the elytra, laments, branchiae, dorsal cirri, notochaetae, and neurochaetae.

Genome sequencing and Trimming
Two libraries (insert size, 550 bp) were constructed using the TruSeq DNA Nano 550bp kit. The libraries were sequenced using the Illumina Novaseq 6000 platform. Low-quality reads (less than Q20) were trimmed using Trim Galore! (ver. 0.6.6) and the reads having shorter than 120 bp or with unknown nucleotides ("N") were discarded. We obtained a total of 131,865,089 reads.

Mitochondria genome assembly and annotation
After the ltering process, de novo assembly was performed using the MITOZ [25] and SPAdes (ver. 3.14.0) [26]. Putative mitochondrial contigs generated by both programs were identi ed and annotated on the MITOS web server [27]. The circular mitogenome was visualized using Circos (ver. 0.69-8).

Phylogenetic analyses
To reconstruct phylogenetic relationships, we constructed a dataset including the COI genes from 12 Branchipolynoe species and 1 outgroup species, Austrolaenilla antarctica (Additional le 3), which were downloaded from National Center for Biotechnology Information (NCBI) database. Multiple alignment of the coding genes was performed using MAFFT (ver. 7.475) with the default options. COI distance matrices were implemented using the MEGA X software and the Kimura two-parameter model [28] ( Table 1).
We used the Bayesian information criterion (BIC) to select the best model for phylogenetic tree reconstruction obtained using the IQ-TREE software [29]. The TPM2+F+G4 evolutionary model was the best t for the dataset sequences. Maximum likelihood (ML) phylogeny was reconstructed using the RAxML tool [30]. ML node support was determined from a majority consensus tree constructed using 1,000 bootstrap replicates. Bayesian inference (BI) analysis was performed using MrBayes (ver. 3.2.7a) [31]. Markov chain Monte Carlo (MCMC) searches were run twice for 10 6 generations with four chains, sampling every 500 generations for each analysis. The phylogenetic tree was visualized using FigTree (ver. 1.4.4).

Etymology and Host
Named in honor of the discoverer of the OVF, the host is Gigantidas vrijenhoeki Jang & Won, 2020.
Prostomium ellipsoidal, bilobed with almost rounded anterior lobes. Short, conical median antenna inserted between two anterior lobes and pair of short conical palps (Fig. 1C). Median antenna and palps smooth and tapering to the slender tip. Palps extending beyond the prostomium. Prostomium lacking frontal laments, eyes, and lateral antennae. First segment fused to prostomium with two pairs of short anterior cirri. Anterior cirri smooth and slightly slender, not exceeding the prostomium length (Fig. 1C). Thick, extended muscular pharynx with ve pairs of dorsal and ventral small, sac-like terminal papillae surrounding the mouth (Fig. 2A).
Branchiae on segments 3-21 dense and arborescent, with short terminal laments (Fig. 2B), not extending beyond the elytral border. Branchiae separated into two types, showing dorsal and ventral emergence, respectively. No discernible dorsal tubercles. Branchiae gradually decreasing in size anteriorly and posteriorly. Parapodia subbiramous. Notopodia smaller than neuropodia, with few notochaetae projecting beyond notopodia (Additional le 4). Neuropodia large, rounded, enclosing numerous neurochaetae with rounded lobes. Notochaetae smooth, stouter, and shorter than neurochaetae (Fig. 3A). Notochaetae few, more abundant on the middle and posterior than anterior segments, slightly tapered with serrated distal part; tip rounded, shaft with inconspicuous rows (Fig. 3B-D). Neurochaetae numerous, more abundant in the middle than anterior and posterior segments, arranged as a vertical fan; tapered, with subdistal swelling and small spines along edge, serrations starting at the midpoint on only one side and extending distally.
Ventral cirri small, smooth, without papillae, and attached to the middle regions of neuropodia; projecting anteriorly ( Fig. 1D). Elongated ventral papillae of female on segments 11 and 12; projecting posteriorly, reaching the subsequent segment. Elongated ventral papillae of males on segment 12, not extending beyond half the length of the subsequent segment (Fig. 1D). Pygidium small, with pair of short, thick, tapered conical anal cirri.

Morphological variation
Page 6/19 The holotype 28 mm long and 13 mm wide, including parapodia. Paratypes vary in size, 23-31 mm long and 9-15 mm wide. All specimens with elongated ventral papillae on segments 11 and 12, suggesting that males were not found.

Remarks
Nine species have been described in the genus Branchipolynoe [4,13,19,20,32,33]. The diagnostic characteristics of the genus were determined from Branchipolynoe symmytilida and amended by Pettibone for B. seepensis [13,33]. This latter revision included the rst position of the branchiae, the presence of the frontal lament, and form of the parapodium. Subsequently, Zhou (2017) published a description of B. longqiensis in the Indian Ocean and Lindgern (2019) published descriptions of ve new species from the Paci c Ocean (Additional le 3). Members of the genus have 21 segments and 10 elytra; the elytra partially cover the dorsal region. Only B. symmytilida has frontal laments, and all have a bilobed prostomium, except B. kajsae. In Branchipolynoe n. sp, the branchiae start at the third segment, and the parapodium is subbiramous, forming a very similar shape to B. longqiensis and B. tjiasmantoi, but with a short, rounded notopodial acicular lobe, inconspicuous pharynx papillae, and different shapes of the tips of sub-acicular neurochaetae (Additional le 5).

Phylogenetic analyses
Among Branchipolynoe species, the mitochondrial COI genetic distance ranged from 0.056 to 0.237, with an average of 0.175 (Table 1). The species with the closest COI genetic distance to that of B. onnuriensis n. sp. was B. tjiasmantoi from the western Paci c, with a value of 0.056. This value is similar to those of B. halliseyae and B. kajsae (0.059), which are referred to as sister species in previous studies due to their similar morphology [19]. The second closest COI genetic distance belonged to B. longqiensis (0.099) found in the Indian Ocean.
The ML tree and BI analyses inferred from Branchipolynoe mitochondrial COI sequences produced a single topology for each region (Fig. 5). The phylogenetic trees isolated the Branchipolynoe species and supported a clade comprising B. onnuriensis n. sp. and B. tjiasmantoi (ML: 78%, BI: 1).

Base composition
To assess the mitochondrial genome, we calculated its nucleotide composition (A%, C%, G%, T%, A+T%, C+G%), AT skew, and GC skew. AT and GC skew were calculated as follows: AT skew = (A -T%) / (A% + T%) and GC skew = (G -C%) / (G% + C%). The overall nucleotide composition of the complete mitochondrial genome was 28.45% A, 24.44% C, 8.99% G, and 38.12% T. The proportion of AT content (66.57%) was ~1.99 times higher than that of GC content (33.43%) (Additional le 7). Most genes showed positive AT skew, except rns and rnl. All genes also showed negative GC skew, indicating that PCGs in B. onnuriensis n. sp. contained a higher percentage of T and C than A and G, except rns and rnl.

Discussion
Branchipolynoe onnuriensis n. sp. is the second Branchipolynoe species to be found in deep-sea hydrothermal vents in the Indian Ocean. Morphologically, B. onnuriensis is very similar to B. longqiensis from the Indian Ocean and B. tjiasmantoi from the western Paci c. To date, the genus Branchipolynoe has been described from only 10 species discovered in hydrothermal vents and methane seeps worldwide, including 5 in the eastern Paci c, 2 in the western Paci c, 1 in the Atlantic, and 2 in India; among these, B. eliseae, B. halliseyae, B. kajsae, B. meridae, and B. tjiasmantoi were recently described from the eastern and western Paci c [19]. Lindgren (2019) suggested that the most recent common ancestor of Branchipolynoe species may have lived in methane seeps in the eastern Paci c. Hydrothermal vents and methane seeps comprise similar basic ecological systems; therefore, it appears that Branchipolynoe moved west to colonize vents and seeps of the western Paci c and Indian Oceans. Future studies should elucidate Branchipolynoe migration patterns to other regions in the Indian Ocean and their evolution to adapt to these harsh environments.
The species discovered in this study was di cult to distinguish from previously discovered several species using external morphological characteristics, but showed su cient genetic distance to classify it as a new species. It is increasingly common to describe animal species without obvious morphological differences based on molecular data alone [34]. This technique allows the reliable isolation of polynoid species, which is important for understanding the distribution and geographic patterns of this taxon [35]. Polynoids are the most diverse and widely distributed polychaete group found in hydrothermal vents and seeps, and are a good model for evaluating the biogeographic distribution of deep-sea environmental faunas [4]. Further studies of polynoids are needed to enhance our understanding of deep-sea ecosystems.
Branchipolynoe live in hydrothermal vents or methane seeps, and B. pettiboneae has been found in both. Two undescribed species have been reported from hydrothermal elds on the CIR and SWIR [36]. Branchipolynoe sp. "Dragon," a commensal scale worm found in bivalve mussels in the Longqi vent eld on the SWIR, shows genetic similarity to an undescribed species (Branchipolynoe sp. "VG-2002") that was recorded and sequenced from the Kairei vent eld on the CIR. Further research on Branchipolynoe species in the Indian Ocean is needed to elucidate the adaptations of animals living in vents of the CIR.
There is no consensus on the minimum COI distance required to classify a species, and each taxon has a different minimum COI distance [37]. Therefore, a minimum interspecies distance greater than the maximum interspecies distance is often applied to identify new species [38]. In polychaetes, a COI distance of ≥ 2% is considered to indicate a different species, and a ≥ 10% difference is used to indicate a different genus. For the newly discovered species, the minimum genetic distance among species was 0.056, representing a difference of > 5% from its most closely related species, B. tjiasmantoi. Thus, even without morphological data, the COI distances indicate that B. onnuriensis n. sp. is a new species.
The newly discovered polynoid is symbiotic to the recently discovered mussel G. vrijenhoeki. Unlike other free-living vent polynoids (e.g., Branchinotogluma), Branchipolynoe species typically inhabit the inner mantle space of Bathymodiolus mussels. Therefore, the lack of a record of symbiotic polynoids in the OVF may be due to the apparent rarity of Bathymodiolus in this region [39]. B. pettiboneae was rst discovered in Gigantidas platifrons; however, most specimens have been found in Bathymodiolus mussels. Symbiotic species subsist on the byproducts of chemical synthesis by the host, and obtain protection from the external environment from the host shell [40]. Rindren (2019) found no relationship between haplotype and host mussel or depth in a new species found in Costa Rica, and no signi cant within-species genetic distances at different depths or in different hosts. Branchipolynoe species are found in a variety of hosts and at a range of depths; thus, it appears that any mussel is capable of inhabiting hydrothermal vents or methane seeps can act as a suitable host. Different host bivalves adapt to different hydrothermal vents or methane seeps, and Branchipolynoe has evolved to coexist with these shell sh.
Most members of Branchipolynoe live individually inside the mussel mantle pallial cavity [41]. Sometimes two are found, but these are very rare occurrences. Among species found in hydrothermal vents, the polynoid B. symmytilida was found in the mussel B. thermophilus, and the polynoid B. pettiboneae was found in the mussel Bathymodiolus brevior. The polynoids B. longqiensis and B. tjiasmantoi were found in the mussels Bathymodiolus marisindicus and B. brevior, respectively. Branchipolynoe species have also been newly discovered in cold seeps. Branchipolynoe larvae can travel long distances between vents or among numerous ventilation systems, allowing gene ow [19]. Further studies are needed to determine whether Branchipolynoe species have host mussel preferences, how they coexist and evolve, and the role that they play in hydrothermal environments.  Maximum likelihood (ML) tree inferred from mitochondrial cytochrome c oxidase I (COI) sequences of genus Branchipolynoe. The COI sequence of Austrolaenilla antarctica was used as an outgroup. Numbers at nodes represent ML and Bayesian inference (BI) support values