Widespread Association of Ciliates Colonizing Gills of Shrimps Inhabiting Vents and Seeps Across the Pacific Ocean

ABSTRACT

Bacterial symbiosis is well‐documented in chemosynthesis‐based ecosystems, but associations with microeukaryotes remain overlooked. In this study, using scanning electron microscopy and 18S rDNA barcoding, we investigate the presence, diversity, and biogeographic patterns of ciliate epibionts associated with two deep‐sea caridean families: Alvinocarididae and Thoridae. We identified a widespread lineage of ciliates colonizing the gills of different alvinocaridid species, extending their previously known distribution in freshwater and coastal habitats to deep ocean down to 3388 m. These ciliates form a distinct clade related to coastal Chonotrichia, but show clear genetic divergence from the previously described species. Geographic divergence of these ciliate populations was observed across the Pacific Ocean, with no evident structure related to their host species. These chonotrichian ciliates exhibited variation in occurrence across host species, individuals, and regions, indicating a facultative association with their hosts. In contrast, the thorid shrimps harbored rare and phylogenetically diverse ciliates. More rarely, we found ciliates related to known parasitic lineages hosted by both shrimp families, with signs of immune response (black gills) in some individuals colonized by these ciliates. Our results reveal previously overlooked protist‐crustacean associations in chemosynthetic ecosystems and highlight the ecological and biogeographic importance of this group in the deep ocean.

1. Introduction

Deep‐sea hydrothermal vent and hydrocarbon seep ecosystems are characterized by dense biological communities, dominated by animal species that are often specific to these habitats and found nowhere else (Van Dover ²⁰⁰⁰). These ecosystems are sustained by microbial communities through chemosynthetic primary production, unlike most others that rely on photosynthesis. A primary focus of hydrothermal vent and hydrocarbon seep research has been on both the visually dominant metazoans (e.g., aggregations of siboglinid tubeworms, bathymodioline mussels, vesicomyid clams, and Rimicaris shrimps) and their chemosynthetic microbial symbionts (Dubilier et al. 2008; Sogin et al. ²⁰²¹). In contrast, unicellular eukaryotes (i.e., protists) and parasitic associations have received much less attention in these environments. Despite this, they are recognized as key contributors to ecosystem functioning, serving as bacterial grazers, nutrient cyclers, and, in some cases, parasitic regulators of host populations (Hu et al. 2021). Free‐living protists in hydrothermal vents are mostly dominated by the major clades of Alveolata and Radiolaria, as reported in most culture‐independent studies (Sauvadet et al. 2010; Murdock and Juniper ²⁰¹⁹; Hu et al. ²⁰²³). Early surveys of hydrothermal vent protists also revealed a high diversity of potential parasites in sediments, which are probably also hosted by the dense animal populations thriving around (López‐García et al. ²⁰⁰³; Moreira ²⁰⁰³; Takishita et al. ²⁰⁰⁵). Conversely, in hydrocarbon seeps, protistan assemblages are mostly dominated by Radiolaria, Dinophyceae, and Cercozoa, while proportions of ciliates are comparatively lower (Li et al. 2024). These patterns indicate that microhabitats and fluid chemistry may have strongly influenced the community composition and diversity of protists. Recent studies have also begun to investigate their functional role in chemosynthetic ecosystems and revealed that grazing activities from protists consume up to 28%–62% of daily bacterial and archaeal biomass production in vent fluids (Hu et al. 2021). Similarly, in hydrocarbon seeps, protist predation and parasitism appear central to food webs and nutrient cycling (Li et al. 2024).

However, unlike macrofaunal communities, it remains contentious whether protist communities inhabiting vents and seeps consist mostly of cosmopolitan species or specialized taxa unique to these ecosystems (Murdock and Juniper 2019). Several studies have uncovered a wide diversity of novel protist lineages and taxa specific to a given vent locality (Sauvadet et al. 2010; Coyne et al. ²⁰¹³; Hu et al. ²⁰²³), whereas others show that nearly 95% of protists from hydrothermal vents are also found in deep‐sea background communities outside vent or seep influence (Murdock and Juniper 2019). These contrasting observations may be due to differences in sampling strategies and sample types (e.g., water, fluids, sediments, animal tissues), emphasizing the importance of microhabitats in determining their distribution. While clear evidence for vent‐endemic protists remains rare, one notable exception is the colonial folliculinid ciliates forming dense “blue mats” on sulfide structures at hydrothermal vents (Kouris et al. 2007).

In addition, protist communities have been observed in host‐associated niches in chemosynthetic environments, dominated by ciliates (Ciliophora), followed by Cercozoa and other lineages. These communities are found in the pallial cavity of the bathymodioline mussel Bathymodiolus thermophilus and the vesicomyid clam Turneroconcha magnifica from the vent fields in the Eastern Pacific (Sauvadet et al. 2010), which may lead to host mortality (Moreira 2003). More broadly, this highlights the role of hydrothermal vent habitats as “oases” for protists. However, documented cases of protist‐metazoan associations in chemosynthetic systems remain rare, with only one known example of an ectosymbiotic Syndiniales associated with mussels, Bathymodiolus japonicus and Gigantidas platifrons, and the vesicomyid clams, Archivesica okutanii and A. soyoae, at the Off Hatsushima seep in Sagami Bay, Japan (Noguchi et al. 2013).

Previous studies have revealed marine crustaceans from shallow‐water habitats establish associations with a wide variety of microeukaryotic symbionts, mainly ciliates belonging to three major ecological groups: apostomatids, chonotrichs, and suctorians (Fernandez‐Leborans et al. ²⁰¹⁶). These groups differ in their life cycles, attachment strategies, and degrees of host specificity (Fernandez‐Leborans et al. ²⁰¹⁶). Among them, ectosymbiotic ciliates are observed on many groups, including amphipods, isopods, decapods, and barnacles (Fernandez‐Leborans et al. ²⁰¹⁶). For instance, apostomatids commonly colonize the gills of coastal crustaceans, generally without inducing strong host immune responses, although pathogenic associations such as black gill syndrome have been reported in some penaeid shrimps (Fernandez‐Leborans et al. ²⁰¹³; Landers et al. ²⁰²⁰; Frischer et al. ²⁰²²). Chonotrichia ciliates are frequently found as gill epibionts of coastal and freshwater crustaceans, such as Chilodochona carcini on the mouthparts of the green crab Carcinus maenas (Lynn 2016; Fahrni et al. ²⁰²²). Currently, such associations remain largely unknown for deep‐sea crustaceans.

Here, we explore the potential presence of similar ciliate symbionts associated with crustaceans from hydrothermal vents and hydrocarbon seeps, using two groups of deep‐sea caridean shrimps as model cases. First is the family Alvinocarididae (Caridea: Bresilioidea), which are globally distributed and restricted to chemosynthetic ecosystems, with 17 species known to inhabit vents and/or seeps in the Western Pacific (Methou, Chen, and Komai ²⁰²⁴). The second is Thoridae (Caridea: Alpheoidea), specifically species of the genus Lebbeus. This taxon is broadly distributed across multiple oceanic regions, including the Pacific Ocean, North Atlantic, Southern Ocean, and Arctic regions, and is often found in both chemosynthetic habitats as well as the ambient deep‐sea environments (Komai and Chen 2024). Using a combination of scanning electron microscopy (SEM) and DNA barcoding (the nuclear 18S rDNA) analyses to screen the gills of 11 alvinocaridid and 3 thorid shrimp species from chemosynthesis‐based ecosystems across the Pacific, we aim to address the following questions: (1) Are the gills of caridean shrimp from deep‐sea chemosynthetic ecosystems colonized by epibiotic ciliates? (2) Are these ciliates also found in other marine environments, or are they preferentially associated with chemosynthetic habitats? (3) Is the phylogenetic diversity of these ciliates structured more by their geographic distribution or by the host shrimp lineages?

2. Materials and Methods

2.1. Sampling

We collected 14 shrimp species totaling 224 individuals belonging to the Alvinocarididae (4 Rimicaris species and 7 Alvinocaris species) and Thoridae (3 Lebbeus species) families, from 21 chemosynthesis‐based ecosystems across the Pacific Ocean, including 16 hydrothermal vent fields and five hydrocarbon seep sites at depths ranging from 628 m to 3388 m (Figure 1) (see Tables S1 and S2 for additional details). Shrimp specimens were sampled using suction samplers mounted on various submersible vehicles, depending on the cruises, including the human‐occupied vehicles (HOVs) Alvin, Nautile, and Shinkai 6500, or the remotely operated vehicles (ROVs) Haima2, KM‐ROV, ROPOS, SuBastian, and Victor 6500. Immediately after collection, they were fixed onboard in 2.5% glutaraldehyde for 16 h, preserved in 80% or 100% ethanol, or stored at −80°C. We ensured that no residual surface water was present on the shrimps at the time of collection to minimize the possibility of contamination. In the laboratory, they were examined morphologically under a stereomicroscope for species, developmental stage (juvenile or adult), and sex (male, ovigerous, or non‐ovigerous female) identification. The body size of the shrimp was measured with a caliper to the nearest 0.1 mm. Gills were subsequently dissected and used for two types of analyses, including microscopic observations and DNA extraction for molecular investigations of potential ciliate associations, with more details provided below.

FIGURE 1.

(A) Geographic map displaying hydrothermal vent (triangle) and hydrocarbon seep (circle) localities where shrimp specimens used in this study were collected. The map was generated in R (version 2024.04.2) with the ggOceanMaps package (v.2.2.0) (Vihtakari 2026). Colors are consistent across figures and indicate the region where the samples were collected from: red (Eastern Pacific), green (Southwestern Pacific), and blue (Northwestern Pacific). (B) Alvinocaridid shrimps on sulfide rocks at the NW Eifuku vent field (Credit: Verena Tunnicliffe, University of Victoria, TN167 expedition). (C) Lebbeus shinkaiae shrimps gathering near mussel bed at the Minami‐Ensei vent field (credit: Chong Chen, YK23‐16S expedition, JAMSTEC).

2.2. Scanning Electron Microscopy (SEM)

For SEM observations, we primarily used specimens fixed onboard directly in 2.5% glutaraldehyde and rinsed in sterile seawater (Table S2). When on‐board glutaraldehyde fixation was not available, we used specimens preserved in 80% or 100% ethanol or those that had been frozen at −80°C. In the latter case, the frozen specimens were post‐fixed in 2.5% glutaraldehyde at the shore‐based laboratory and subsequently rinsed three times with sterile seawater. In total, gills of 16 shrimp specimens were used for SEM observations, comprising 9 species of Alvinocarididae and 2 species of Thoridae. Dehydration was performed on the glutaraldehyde‐fixed specimens—either on board or in the laboratory—using an ascending ethanol series up to 100%. All samples, including those preserved in 80% ethanol, were then dehydrated up to 100% ethanol, followed by 5 h in a critical point dryer EM CPD 300 (Leica, Vienna) and coated with a thin layer of a 50/50 gold–palladium alloy using a Quorum Technologies SC7640 sputter coater (New Haven, USA). Observations were conducted using a Quanta 200 SEM (FEI‐Thermo Fisher, Hillsboro, OR, USA). The microscope operated at an electron beam voltage of 5 to 10 kV, with magnifications ranging from 50 to 5000 ×, depending on the structures examined.

2.3. DNA Extraction, PCR, Sequencing

DNA extractions were performed on 208 shrimp specimens using one or more dissected gill filaments following the manufacturer's instructions of the DNeasy PowerSoil Kit (Macherey‐Nagel, Düren, Germany). Partial sequences of the 18S ribosomal gene were amplified by PCR using the ciliate‐specific primers F384dT (5′‐AAGGGCACCACCAGGAGTGG‐3′) and R1147dT (5′‐CGCTTTCGTGATYCTTGGTTCTC‐3′) (Dopheide et al. 2008). PCR reactions were prepared in a 25 μL volume, comprising 15.25 μL of ultrapure water, 5 μL of Green GoTaq buffer, 2 μL of MgCl₂, 1 μL of dNTPs, 0.75 μL of each primer, 0.25 μL of GoTaq Polymerase (Promega, Corporation, Madison, WI, USA), and 1 μL of gill‐derived DNA. Amplification conditions were as follows: initial denaturation at 94°C for 5 min, followed by 40 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 2 min, with a final extension at 72°C for 7 min. The PCR efficiency was checked via 0.8% agarose gel electrophoresis. Samples with weak or faint amplification were subjected to nested PCR, using 1 μL of the first‐round product as template under identical reagent and cycling conditions, except that the cycle number was reduced to 10 or 15 (vs. 40 cycles initially). For some samples showing double bands on the agarose gel, gel extraction was performed on the band corresponding to the expected size (about 750 bp), matching the one observed in samples with a single clear band. Extraction was carried out using the NucleoSpin Gel and PCR Clean‐up Kit (Macherey‐Nagel, Düren, Germany).

Sequencing was performed using Sanger sequencing by Eurofins Genomics (Cologne, Germany). Raw chromatograms were cleaned, assembled, and edited using Geneious v10.0.9 (https://www.geneious.com), removing artifacts, and assembling fragments obtained with both primers for each individual. The resulting sequences were compared against the NCBI GenBank database using BLASTn (Basic Local Alignment Search Tool) for taxonomic identification, with the database version available as of December 2025.

2.4. Phylogenetic Reconstruction and Haplotype Network

Using BLAST results for a rough check of the taxonomic affinity, cleaned sequences were grouped by ciliate classes, resulting in two different alignments of sequences belonging to the Phyllopharyngea class (subclass Chonotrichia) and the class Oligohymenophorea (subclasses Apostomatia and Peritrichia). Alignments were performed with the MUSCLE algorithm (Edgar 2004) in Geneious. The sequences obtained in this study were aligned separately against reference ciliate sequences retrieved from the NCBI GenBank, including those reported in (Lynn 2016) and (Fahrni et al. 2022) for the Phyllopharyngea alignment and (Metz and Hechinger 2021) for the Oligohymenophorea alignment. Phylogenetic reconstructions were performed using IQ‐TREE v1.6.12 (Minh et al. 2020) with 1000 ultrafast bootstrap replicates (Hoang et al. 2018). The best‐fitting nucleotide substitution model was selected according to the Bayesian Information Criterion (BIC) using the ModelFinder module integrated in IQ‐TREE. Resulting phylogenetic trees were visualized and edited using TreeViewer v2.0.2 (Bianchini and Sánchez‐Baracaldo ²⁰²⁴). The 18S rDNA sequences of Tokophrya huangmeiensis (KJ567607) were used as outgroups to root the trees.

Given that the majority of ciliate 18S rDNA sequences obtained from the gills of the alvinocaridid shrimps belong to a specific group within Chonotrichia, we conducted haplotype network analysis specifically on this group to examine their genetic structure. The alignment of the targeted sequences was performed using the previously mentioned methods, and the resulting alignments were utilized to construct a haplotype network using a median‐joining method implemented in PopArt v1.7 (Leigh and Bryant 2015).

2.5. Prevalence of Ciliate Epibionts

Shrimp individuals were grouped by subregion and by host species to consider both the influence of geography and host identity on the prevalence index. Given the recurrence of similar sequences belonging to Chonotrichia in our dataset, we calculated a prevalence index to estimate the frequency of gill colonization by Chonotrichae. This index was calculated as follows:

$$\begin{array}{c}\text{Prevalence}=\left(\text{Number of shrimp gills with presence of Chonotrichia}/\right.\hfill \\ \left.\text{Total number of shrimp gills examined}\right)\times 100\end{array}$$

This yields the percentage of shrimp specimens colonized by Chonotrichia within each group. Specimens whose 18S rDNA sequences affiliated with Chonotrichia were considered colonized by these epibionts. Conversely, the absence of colonization on individuals without 18S rDNA sequences or with sequences affiliated with another ciliate group was double confirmed visually. This was done using gills of the opposite side observed either by SEM or with a stereomicroscope using floxin for enhancing the contrast between the ciliate epibionts and the shrimp tissue. For a few samples where microscopic observation was no longer possible, prevalence was assessed based solely on the barcoding results (Table S1). All Lebbeus specimens (e.g., 3 species with a total sample size of 17) were removed from the prevalence analysis, as they were not colonized by Chonotrichia ciliates, except for one individual, which would have biased the colonization patterns observed.

3. Results

3.1. SEM Observations of Shrimp Gill Epibionts

Observations of gills by SEM confirmed the presence or absence of ciliates in several species of shrimps inhabiting chemosynthesis‐based ecosystems. Nevertheless, colonization by these ciliates varied greatly in density. For instance, some species exhibit marked or even very dense colonization for most individuals, such as Alvinocaris longirostris (Figure 2A,D), Alvinocaris lusca (Figures 2B and S2E), Rimicaris vandoverae (Figure 2C), and Rimicaris variabilis (Figures 2E and S2B). Other species like Rimicaris cambonae, however, showed limited colonization or no colonization at all (Figure 2F). There is also considerable variation within host species, as illustrated by the presence of both uncolonized and heavily colonized host shrimps with up to 650 ciliates on a single branchial filament (Figure 2A) to just a few or even none in different individuals of A. longirostris from the same locality (Off Hatsushima, Sagami Bay; Figures 2A, S2C and S2D). Similarly, for R. variabilis , specimens from one locality (Pacmanus, Manus Basin) were colonized (Figure S2B) while those from the other locality (La Scala, Woodlark Basin) were not (Figure S2A). For one specimen, the colonization was also heterogeneous across branchial filaments, with filaments 2 and 3 generally showing higher ciliate densities than filaments 1, 4, and 5. In some specimens, such as R. variabilis (Figure 2E), filamentous bacteria were also observed on the gills and around the ciliates. No colonization by ciliates could be observed visually with our SEM observations on the gills of thorid shrimp species, neither Lebbeus parvirostris (Figure S1A) nor Lebbeus shinkaiae (Figure S1B).

FIGURE 2.

Scanning electron microscopy (SEM) observations of gills from multiple shrimp specimens collected from chemosynthesis‐based ecosystems, either colonized or not by ciliates. (A) Overview of an Alvinocaris longirostris gill filament colonized by several ciliates (500 μm), (C) Close lateral view of ciliate epibionts colonizing the gills of Alvinocaris lusca (50 μm), (B) Close apical view of ciliate epibionts colonizing gills of Rimicaris vandoverae (25 μm), (D) View of a gill filament of Alvinocaris longirostris colonized by several ciliate epibionts (500 μm), (E) View of several ciliate epibionts and the surrounding bacteria on a gill filament of Rimicaris variabilis (50 μm), (F) Overview of a Rimicaris cambonae gill filament showing the absence of colonization by ciliate epibionts (500 μm).

The observed ciliates were approximately 50 μm in size, although variations in size ranging from 25 to 100 μm were observed (Figures 2 and S1), even within the same gill filament (Figure S1H). They display a characteristic central invagination, surrounded by a well‐defined ciliary arrangement forming a funnel‐shaped collar composed of a coiled structure housing the cilia (Figure 2B). The body is elongate, and the ciliates are mainly localized at the base of the gill filaments, where they attach to the surface of the lamellae closest to the gill arch by means of a stalk (Figure 2A,C). This structure appears to serve as an anchoring support without penetrating the gill tissues (Figure S1C).

3.2. Phylogeny, Haplotype Networks and Prevalence of Phyllopharyngea Ciliates

We obtained 120 new 18S rDNA sequences (length: 551 to 644 bp) of ciliates from the gills of alvinocaridid shrimps. Phylogenetic reconstruction revealed that most sequences (93 out of 120 sequences representing 18 unique sequences) were classified in Chonotrichia, a subclass within the class Phyllopharyngea (Figure 3). These sequences form a distinct monophyletic clade (bootstrap = 93) in our phylogenetic reconstruction, most closely related to Chilodochona carcini (GenBank Accession No.: KU588417) with a sequence similarity ranging from 93.9% to 94.1%, or to Spirochona gemmipara (KY349104) with a sequence similarity ranging from 94.3% to 94.6%. In contrast, a more limited number of 18S rDNA sequences (4 out of 120 sequences representing three unique sequences) from thorid Lebbeus shrimps were classified in Phyllopharyngea. These sequences appear more scattered within our phylogenetic reconstruction of Phyllopharyngea, with one sequence (Phyllo_01 PX752129) isolated from L. shinkaiae most closely related to Isochona sp. (AY242116; 99.2% similarity), and another sequence (Phyllo_03 PX752131) also isolated from L. shinkaiae being most closely related to Acineta flava (HM140400; 94.9% similarity). One sequence (Phyllo_02 PX752130) isolated from a gill of L. unguiculatus was affiliated with a group of sequences related to Dysteria parasemilunaris voucher (PP905720; 99.5% similarity). Only one sequence isolated from the gill of a L. unguiculatus specimen (Chono_01 PX752096: see Figure 4B and Table S1) was affiliated with the group of sequences related to Chilodochona carcini (KU588417; 93.63% similarity).

FIGURE 3.

Phylogenetic tree of the Chonotrichae and related taxa (Phyllopharyngea class) on a partial fragment of the 18S ribosomal gene (Alignment length: 551 bp). The colors boxes indicate the family to which the ciliate host belongs, red: Thoridae, purple: Alvinocarididae. The asterisks indicate symbiotic ciliates associated with other metazoan hosts. The tree was inferred from SSU rDNA sequences using IQ‐TREE v2.2.5 under the TIM2 + F + I + R6 model selected by ModelFinder. NCBI accession numbers are shown to the right of each species name. The new sequences obtained in the present study are indicated in bold. Ultrafast bootstrap support values (10,000 replicates) are shown at the nodes, the ones with bootstrap of 100 are represented by black circles. The scale bar represents 0.05 substitutions per nucleotide position.

FIGURE 4.

(A) Haplotype network based on a partial fragment of the 18S ribosomal gene (Alignment length 551 bp) of Chonotrichia ciliates according to geographic regions (86 individuals). Colors indicate the region of the specimens collected: red (Eastern Pacific), green (Southwestern Pacific), and blue (Northwestern Pacific). Numbers on the network indicate some of the genetic distance among haplotypes (see Table S5 for full detailed genetic distance among each haplotypes). (B) Haplotype network based on the 18S ribosomal gene fragment of Chonotrichia ciliates according to phylogenetic host species of the North West Pacific (55 Individuals). Colors indicate the shrimp host species of the Chonotrichia ciliate. For both networks, each circle represents a distinct haplotype, with its size proportional to the number of sequences belonging to that haplotype. Lines connecting the circles represent nucleotide substitutions between haplotypes (see Tables S3 and S4 for a detailed list of COI haplotypes and their corresponding GenBank ID).

Haplotype network of 551‐bp 18S rRNA sequences of Chonotrichia indet. revealed a strong geographic structuring of haplotype distribution (Figure 4A). Specifically, two dominant haplotypes were shared among specimens from the Northwestern Pacific (Okinawa Trough, Sagami Bay, Izu‐Bonin‐Mariana Arc), whereas a single dominant haplotype was prevalent in the Southwestern Pacific (Lau Basin, Futuna Arc, Manus Basin, Woodlark Basin). By contrast, specimens from the two Eastern Pacific vent sites each possessed a unique haplotype, which may point to site‐specific endemism. Nevertheless, this interpretation remains tentative owing to the limited sample sizes available for these localities. Moreover, some specimens differed from this general trend, like those inhabiting the Hikurangi Margin in the Southwestern Pacific, that possess the same haplotype as specimens from the Northwestern Pacific. Similarly, specimens from the South China Sea shared haplotypes with both specimens from the Northwestern Pacific and the Southwestern Pacific. Finally, although we examined host identity as a potential structuring factor, we found no clear haplotype partitioning by host species (Figure 4B), indicating that host species does not appear to drive the geographic structuring observed here. This analysis was restricted to the Northwestern Pacific, as it was the only region in our dataset where multiple shrimp species coexist at the same location.

Chonotrichia ciliates were detected in every subregion across all 11 study regions, indicating their ubiquitous distribution despite variations in abundance (Figure 1 and Table 1). High prevalence values (≥ 50%) were found in 5 of the 11 regions studied, notably in Manus Basin vents (85%; n = 20), Woodlark Basin vents (75%; n = 4), the Sagami Bay seep (71%; n = 14), Galápagos Rift vents (67%; n = 3), and Costa Rica Margin seeps (50%; n = 6). Moderate prevalence values were found in the Okinawa Trough vents (46%; n = 78) and South China Sea seeps (41%; n = 17), whereas the Futuna Arc vents (27%; n = 11), the Hikurangi Margin vents (25%; n = 12), the Lau Basin vents (20%; n = 10), and most notably the Izu‐Bonin‐Mariana Arc vents (10.5%; n = 19) displayed low prevalence values. All shrimp species examined were colonized by Chonotrichia ciliates, except Rimicaris cambonae, for which no epibionts were observed.

TABLE 1.

Prevalence of alvinocaridids colonized by Chonotrichae ciliates.

Subregion	Site	Depth range (m)	Species	Total number of specimens	Number colonized by Chonotrichia	Prevalence index by site (%)	Prevalence index by subregion (%)
Okinawa Trough	Amami Rift	628	Alvinocaris dissimilis	10	9	70.00	55.38
			Rimicaris leurokolos	10	5
	Minami Ensei Knoll	706–710	Alvinocaris dissimilis	8	3	47.37
			Rimicaris leurokolos	11	6
	Iheya North Knoll	979–1056	Alvinocaris longirostris	5	1	75.00
			Rimicaris leurokolos	11	11
	Noho Site	1546–1555	Alvinocaris longirostris	5	0	0.00
	Tarama Knoll	1684	Alvinocaris longirostris	5	1	20.00
Izu‐Bonin Marianes Arc	NW Eifuku	1619	Alvinocaris marimonte	4	0	0.00	10.5
			Rimicaris cambonae	3	0
	Myojin Knoll	1228	Alvinocaris marimonte	9	0	0.00
	Snail	2850	Rimicaris vandoverae	3	2	66.67
Sagami Bay	Off Hatsushima	956–1070	Alvinocaris longirostris	10	9	90.00	90.00
South China Sea	Haima	1446	Alvinocaris kexueae	3	1	36.36	41.00
			Alvinocaris longirostris	8	3
	Site F	1125	Alvinocaris longirostris	6	3	50.00
Lau Basin	ABE	2153	Rimicaris variabilis	10	2	20.00	20.00
Hikurangi Margin	Glendhu Ridge	1986–2300	Alvinocaris webberi	12	3	25.00	25.00
Futuna Arc	Kulo Lasi	1406	Rimicaris variabilis	4	1	25.00	39.28
	Fatu Kapa	1570	Rimicaris variabilis	7	2	28.57
Manus Basin	Pacmanus	1738	Rimicaris variabilis	16	15	93.75	25.00
	Susu Knolls	1210–1516	Rimicaris variabilis	4	2	50.00
Woodlarck Basin	La Scala	3388	Rimicaris variabilis	4	3	75.00	75.00
Galapagos Ridge	Tempus Fugit	2602	Alvinocaris lusca	3	2	67.00	67.00
Costa Rica Margin	Mound 12	997–1000	Alvinocaris costaricensis	6	3	50.00	50.00

3.3. Phylogeny of Oligohymenophorea Ciliates

Although most 18S rDNA sequences from shrimp gills were identified as Phyllopharyngea ciliates, BLASTn search and phylogenetic reconstruction revealed that some 18S rDNA sequences (23 out of 120 sequences representing 18 unique sequences) were affiliated with the class Oligohymenophorea, within the Apostomatida and Peritrichia groups (Figure 5). Most Apostomatida sequences were unique to a single shrimp host species, and identical ones came from individuals collected within the same region and sometimes from the same site. For instance, three sequences were shared among three R. variabilis specimens from the Futuna Arc vents (Rva20, Rva29 & Rva30), two between A. marimonte from the Izu‐Ogasawara Arc vents (Amt25 & Amt29), and three among Lebbeus spp. from the Okinawa Trough vents (Lpa1 & Lpa2) and the Sagami Bay seep (Lun14). Nearly all sequences from the subclass Apostomatia were positioned within three well‐supported clades (bootstrap $\ge$ 90): one forming a sister branch to the group formed by the genera Hyalophysa, Synophrya, and Vampyrophrya (Oligoh_04 PX752114, Oligoh_03 PX752113 & Oligoh_05 PX752115), one basal node near genus Metacollinia (Oligoh_07 PX752117 & Oligoh_08 PX752118), and one basal branch sister to all Apostomatida comprising the rest of the sequences. One additional Apostomatida sequence (Oligoh_06 PX752116) showed the highest similarity to Lynnia grapsolytica (MT906161; 96.58% similarity). Three additional sequences were positioned within the subclass Colliniidae, with two sequences most closely related to Pseudovorticella spathulata (MH513617; 97.16% and 97.79% similarity) and one sequence (Oligoh_16 PX752126) most closely to Rhabdostyla commensalis (MN543651; 97.95% similarity).

FIGURE 5.

Phylogenetic tree of ciliates from the class Oligohymenophorea based on a partial fragment of the 18S ribosomal gene (length: 551–644 bp). The colored boxes indicate the family to which the ciliate hosts belong: red, Thoridae; purple, Alvinocarididae. The asterisks indicate ciliates associated with other metazoan. The tree was inferred from 18S rDNA sequences using IQ‐TREE v2.2.5 under the TIM2 + F + I + R3 model selected by ModelFinder. NCBI accession numbers are shown to the right of each species name. The sequences obtained in the present study are indicated in bold. Ultrafast bootstrap support values are shown at the nodes, with values of 100 represented by black circles. The scale bar represents 0.05 substitutions per nucleotide position.

Although systematic observations were not carried out for all these shrimp specimens, blackened gills were observed for some of the individuals from which an 18S rDNA sequence affiliated to Apostomatia was obtained, such as the R. variabilis specimen Rva20 (Oligoh_01; PX752111) from the Futuna Arc (Figure 6).

FIGURE 6.

Photograph of a gill exhibiting black gill syndrome from a specimen of Rimicaris variabilis (Rva20).

4. Discussion

4.1. A Widespread Lineage of Chonotrichae Ciliates on Gills of Alvinocaridid Shrimps

Epibionts of crustaceans belonging to Chonotrichia were previously reported in coastal or freshwater habitats (Lynn 2016; Fahrni et al. ²⁰²²). Our combined results from SEM (Figure 2) and molecular barcoding (Figures 3, 4, 5) demonstrate that Chonotrichia ciliate epibionts are widely distributed on the gills of alvinocaridid shrimps from hydrothermal vents and hydrocarbon seeps throughout the Pacific Ocean (Figure 1), thus occupying deep and remote habitats down to 3388 m depth.

Our phylogenetic reconstruction places these ciliates as a sister group to Chilodochona carcini (Figure 3), a group of ciliates also known to live as an epibiont colonizing the green crab Carcinus maenas in coastal environments (Lynn 2016). Yet their relatively low similarity (93.9%–94.1%) and distinct morphological features to the closest known Chonotrichia imply the specificity of our newly detected lineage to deep‐sea environments. The distinctiveness of this previously unknown Chonotrichia lineage must be viewed in the context of a current under‐representation of available reference sequences for this group, as highlighted by (Lynn 2016; Fahrni et al. ²⁰²²), with about 112 reported species but only 5 with associated molecular data. However, the specificity of Chonotrichia epibionts colonizing alvinocaridid shrimps is also supported by their host fidelity. Our examination of 208 host shrimp specimens shows that these ciliates were found almost exclusively on alvinocaridids (190 specimens totaling 14 species from two genera), and not on the thorid shrimp genus Lebbeus (17 specimens totaling 3 species), apart from one L. unguiculatus individual (Lun13), which was detected by 18S barcoding but could not be confirmed by SEM. Alvinocaridid shrimps are known only to inhabit chemosynthetic‐based environments (Lunina and Vereshchaka 2014)—and even be restricted to hydrothermal vents for those belonging to the genus Rimicaris (Methou, Chen, and Komai ²⁰²⁴) likely constraining their ciliate epibionts to the same habitat distribution. Unlike macrofaunal species, most studies on protist diversity have not been able to determine whether they are primarily specific to chemosynthetic environments or also cosmopolitan in deep‐sea surroundings (Sauvadet et al. 2010; Coyne et al. ²⁰¹³; Murdock and Juniper ²⁰¹⁹; Hu et al. ²⁰²³). Even today, the colonial folliculinid ciliates forming dense “blue mats” in the Juan de Fuca Ridge are one of the rare cases of a vent‐endemic protist (Kouris et al. 2007). Our study reports another likely case of ciliates specific to chemosynthetic habitats: the association between Chonotrichia ciliates and alvinocaridid shrimps.

Given the observed variation in occurrence across host species, individuals, and regions, Chonotrichia ciliates appear to have a facultative association rather than an obligate relationship with their alvinocaridid hosts. Although the precise biological functions of these epibionts remain to be fully elucidated, their distribution patterns imply that they may play ecological roles comparable to those of closely related epibiotic ciliates. For instance, the size variations of ciliates observed on the same gill suggest continuous or progressive colonization, a process that may influence microbial turnover and interactions at the host surface (Bickel et al. 2012). Besides, host characteristics, such as cuticle structure, mucus composition, and immune responses, might be crucial in colonization success, in agreement with patterns observed in other epibiotic ciliates (Fokin 2004; Lynn ²⁰¹⁰; Nowack and Melkonian ²⁰¹⁰).

Unlike Alvinocarididae, colonization of thorid shrimp in the genus Lebbeus by ciliates differs considerably, with only some individuals colonized (6 out of 17 specimens) each time by a very distinct lineage of Phyllopharyngea affiliated to Acineta flava (Suctoria), Isochona sp. sensu Snoeyenbos‐West et al. 2004 (AY242116) (Chonotrichia), or Dysteria paraselluminaris (PP9005721). Only one symbiont affiliated to Acineta sp. sensu Gregori et al. 2016 (LK934647.1) was retrieved from one A. marimonte shrimp, an alvinocaridid species that does not appear to be colonized by the specific Chonotrichae lineage. All the ciliates identified associated with Lebbeus appear to belong to potentially new or existing lineages with 94.9% to 99.5% similarity to their closest known relative. However, unlike the epibionts specific to alvinocarids, it is difficult to determine whether they are specific to chemosynthetic environments, the deep ocean, or even cosmopolitan marine environments. Indeed, even though L. shinkaiae (Komai et al. 2012) and L. parvirostris (Komai and Chen 2024) have only been found in hydrothermal vents, other species within this genus, like L. unguiculatus , occupy other non‐chemosynthetic deep‐sea habitats (Chan et al. 2010) and can be considered transient fauna dwelling in vents and seeps. Future studies on the associations of ciliates and other protists with a wider range of crustacean and invertebrate hosts will help to clarify these patterns of specificity to their hosts or their environments. Analyzing seawater and sediment from the same sites, which were unfortunately not available for the present study, would help determine whether these ciliates are also present in the surroundings to support the hypothesis that the ciliates are acquired from the environment.

4.2. Prevalence and Biogeographic Patterns of Chonotrichae Ciliates

Although Chonotrichia ciliates frequently colonize the gill surfaces of alvinocaridid shrimp species throughout the Pacific, their prevalence varied considerably across geographical locations from 93% of the individuals in certain populations to undetectable levels in others (Table 1). This prevalence reaches or exceeds 50% for nearly half the sites, even though some values should be taken with caution due to a low number of specimens analyzed in some areas, particularly for the Eastern Pacific. In addition, some species, such as A. marimonte (n = 14) and R. cambonae (n = 9), had no Chonotrichia ciliates identified on their gills (Figure 2). Thus, despite their widespread distribution, the occurrence of these ciliates varies not only among different shrimp species but also among individuals of the same species, which indicates a facultative rather than obligate association.

However, the drivers of inter‐ and intra‐specific variation in chonotrichian colonization among shrimp host species and localities remain incompletely understood. Part of the inter‐individual variability could be related to the molting cycle of alvinocaridids throughout their adult life, renewing their cuticle and probably the associated ciliates (Corbari et al. 2008). While the complete absence of ciliates in certain areas (e.g., NW Eifuku or Myojin Knoll; see Table 1) may be linked to local conditions of vent and seep geofluids, observations from the Okinawa Trough present a contrasting case. There, Chonotrichia epibionts colonize all examined alvinocaridid species across the geochemical spectrum, from hosts like R. leurokolos living near the hottest fluids to those like A. dissimilis inhabiting mussel beds located away from the venting fluids (Yahagi et al. 2015). This pattern implies that colonization may depend more on host‐specific factors than on local chemistry.

Our data also suggest a global geographic structuration of Chonotrichia, with a slight differentiation between large regional assemblages from the Northwestern, the Southwestern, and the Eastern Pacific (Figure 4A). This geographic structuration corresponds approximately to the biogeographic provinces defined for macrofaunal species at hydrothermal vents (Moalic et al. 2012; Tunnicliffe et al. ²⁰²⁴) and their connections with hydrocarbon seeps (He et al. 2023). Nevertheless, some exceptions to this geographic pattern are notable. For instance, ciliate epibionts from shrimp hosts off New Zealand show stronger genetic similarities with those from the Northwestern Pacific despite more than 7000 km of distance. This genetic similarity between distant regions mirrors the one observed for their alvinocaridid hosts, like A. dissimilis , whose distribution range extends from the Okinawa Trough near Japan to the Hikurangi Margin in New Zealand (Methou, Ogawa, et al. ²⁰²⁴). Additionally, hydrocarbon seeps in the South China Sea appear to function as a corridor that links the Northwestern and the Southwestern Pacific regions (He et al. 2023). Indeed, distant populations of A. longirostris , which are present in both the South China Sea and Sagami Bay seeps, the Okinawa Trough vents, and as far south as the Manus Basin vents, have also been shown to exhibit similar patterns of gene flow and population structure (Dai et al. 2026). The wide distribution of host populations across large geographic regions could thus facilitate long‐distance transport of their epibionts (Goffredi and Orphan 2010). Nevertheless, it should be noted that the usage of the relatively conserved 18S rRNA gene marker may be limited in its ability to resolve species identity and fine‐scale geographic structure among Chonotrichia ciliate populations.

4.3. Other Infrequent and Potentially Parasitic Ciliates Belonging to Oligohymenophorea

In addition to the dominant Chonotrichia epibionts, ciliates belonging to Oligohymenophorea could also be found, although more sporadically, notably with some Peritrichia (Vorticella or Pseudovorticella) or some Apostomatia (Metacollinia, Hyalophysa, Pseudocollinia, or Lynnia among others) (Figure 5). The overall lower proportion of Oligohymenophorea observed in chemosynthetic ecosystems compared to nearshore or aquaculture environments (Frischer et al. 2022) may reflect habitat‐specific differences in community composition. Compared to Chonotrichia, their presence is more geographically restricted, most often within a single region, and for some of them, it could correspond to a parasitic relationship. This is supported by the immune response observed in an individual of the alvinocaridid shrimp R. variabilis (Rva20) displaying blackened gills (Figure 6). This specimen is colonized by an apostome ciliate most closely related to Hyalophysa, a genus known to parasitize crustaceans in coastal environments (Frischer et al. 2022). Some species of this genus, such as Hyalophysa lynni, have been reported to induce physiological imbalances and trigger an immune response characteristic of black gill syndrome (Bickel et al. 2012; Fernandez‐Leborans et al. ²⁰¹³; Landers et al. ²⁰²⁰; Frischer et al. ²⁰²²). Although no direct tissue damage was observed in the individual Rva20, the presence of Hyalophysa and the associated blackened gill coloration typical of an immune response suggests a potential physiological impact that has not been observed in individuals colonized by Chonotrichia, either in coastal shrimps (Frischer et al. 2022) or in deep‐sea alvinocaridids examined in this study.

It is also plausible that some hydrothermal vent areas, or some periods, could be more affected than others by these potential parasites. In coastal penaeids, severe outbreaks of black gill disease have particularly affected areas off the coasts of South Carolina and Georgia, compared to other regions (Landers et al. 2020; Frischer et al. ²⁰²²). Seasonality also appears to play an important role in these epidemics, with the presence of black gills and Hyalophysa lynni parasites mostly in late summer when water temperatures reach their maximum and oxygen levels are at their lowest (Frischer et al. 2022). Although seasonal dynamics and biological cycles are structured differently in the absence of light in deep‐sea chemosynthetic ecosystems (Mat et al. 2020; Methou et al. ²⁰²²; Zhang et al. ²⁰²⁵), black gill syndrome in alvinocaridid shrimps could also follow periodic bursts throughout the year. However, our dataset is not designed to specifically resolve biogeographical patterns within this group, due to the limited and uneven sampling across regions and host species. We therefore cannot assess biogeographical or seasonal structuring of these ciliates based on the presently available data. Future work with broader and more balanced temporal sampling across habitats would be required to properly test for potential biogeographical or seasonal patterns in this group.

Our results align with the conclusions of Dykman et al. (2023), who proposed that the ecological isolation and unstable nature of hydrothermal vent environments do not significantly limit the diversity of metazoan parasites, particularly trematodes, but also copepods or nematodes. Contrary to previous expectations, the high density of animals inhabiting these ecosystems could instead promote the transmission and maintenance of epibionts (Moreira 2003). This diversity also includes parasites with indirect life cycles that involve multiple hosts (Dykman et al. 2025). Our work expands this diversity of potential parasites to microeukaryotes with the example of alvinocaridid shrimps displaying immune blackened gill responses potentially in relation to infection by apostome ciliates related to Hyalophysa. Our study has only scratched the surface of the many potential relationships between ciliates and animals living side by side in deep‐sea chemosynthetic ecosystems, and we anticipate that many more remain to be discovered.

Funding

Lison Hey and Pierre Methou were supported by ISblue project, Interdisciplinary Graduate School for the Blue Planet (ANR‐17‐EURE‐0015) and co‐funded by a grant from the French government under the program “Investissements d'Avenir” embedded in France 2030. Analyses conducted during this study also benefited from French State aid managed by the National Research Agency under France 2030 LIFEDEEPER: ANR‐22‐POCE‐0007. Dewi Langlet was supported by JSPS KAKENHI grant 23K05942. This study was also supported by two General Research Funds (GRFs) and one Collaborative Research Fund (CRF) of the Hong Kong SAR government (16309324, 16100425, and C2013‐22G), a project from the Otto Poon Center for Climate Resilience and Sustainability of The Hong Kong University of Science and Technology (CCRS25SC01), and two projects from the Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou) (2021HJ01 and SMSEGL24SC01). The R/V Yokosuka research cruise (YK23‐16S) was supported by the Cooperative Research Program of Atmosphere and Ocean Research Institute, The University of Tokyo (R/V Yokosuka, cruise YK23‐16S). The R/V Kairei research cruise KR15‐16 was supported by Council for Science, Technology, and Innovation (CSTI) as the Cross Ministerial Strategic Innovation Promotion Program (SIP), Next‐generation Technology for Ocean Resource Exploration. The R/V Falkor (too) cruise FKt231024 (Project Zombie: Bringing dead vents to life—Ultra fine‐scale seafloor mapping) was funded by the Schmidt Ocean Institute. The Schmidt Ocean Institute also kindly covered the Open Access costs of this paper. The AT42‐03 expedition off Costa Rica was funded by the National Science Foundation (NSF) grant NSF OCE 1635219.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Scanning electron microscopy (SEM) observations of gills from multiple shrimp specimens collected from chemosynthesis‐based ecosystems, either colonized or not by ciliates. (A) Overview of a gill filament of Lebbeus parvirostris showing no colonization by epibiotic ciliates (200 μm), (B) Overview of a gill filament of Lebbeus shinkaiae showing no colonization by epibiotic ciliates (1.5 mm), (C) Close‐up of an epibiotic ciliate that colonized a gill filament of Alvinocaris longirostris (100 μm), (D) Close‐up of an epibiotic ciliate colonizing a gill filament of Alvinocaris longirostris (50 μm), (E) Epibiotic ciliates colonizing the gills of Rimicaris leurokolos (200 μm), (F) Close‐up of an epibiotic ciliate colonizing the gills of Alvinocaris longirostris, showing bacteria inside (15 μm), (G) Overview of a gill filament of Alvinocaris longirostris colonized by epibiotic ciliates (1.5 mm), (H) Overview of a gill filament of Rimicaris variabilis colonized by epibiotic ciliates (200 μm), (I) View of an epibiotic ciliate on the gill lamellae of Alvinocaris costaricensis (50 μm), (J) Overview of an epibiotic ciliate colonizing the gills of Alvinocaris dissimilis (250 μm), (K) View of epibiotic ciliates colonizing the gills of Rimicaris variabilis along with surrounding bacteria (200 μm).

Figure S2: Scanning electron microscopy (SEM) observations of gills from multiple shrimp specimens collected from chemosynthesis‐based ecosystems, either colonized or not by ciliates. (A) Overview of a gill filament of Rimicaris variabilis showing absence of colonization by ciliates (500 μm), (B) Overview of a gill filament of Rimicaris variabilis showing presence of colonization by ciliates and bacteria (500 μm), (C) Overview of a gill filament of Alvinocaris longirostris showing absence of colonization by epibiotic ciliates (500 μm), (D) Overview of a gill filament of Alvinocaris longirostris showing presence of colonization by epibiotic ciliates (500 μm), (E) Overview of a gill filament of Alvinocaris lusca colonized by epibiotic ciliates (500 μm).

Table S1: List of samples analyzed for molecular barcoding.

Table S2: List of samples analyzed for SEM.

Table S3: List of Chonotrichae epibiont haplotypes by region.

Table S4: List of Chonotrichae epibiont haplotypes by host species.

Table S5: Genetic distance among Chonotrichae epibionts haplotypes.

Data Availability Statement

The data that support the findings of this study are openly available in NCBI GenBank at https://www.ncbi.nlm.nih.gov/, reference number PX752096–PX752132.