Rediscovery of Remarkably Rare Anaerobic Tentaculiferous Ciliate Genera Legendrea and Dactylochlamys (Ciliophora: Litostomatea)

Simple Summary Rare organisms represent a challenge for researchers in all fields of study in biology. In the realm of ciliatology, the genera Dactylochlamys and Legendrea are considered to be such cases. Very little information has accrued in over a century since their first descriptions; only a few published reports, online images or videos of rarely encountered individual specimens exist. Dactylochlamys and Legendrea are also morphologically remarkable for their likely independently evolved tentacle-like structures. Recently, the first molecular data were published for the species L. loyezae. In our study, we present more robust phylogenetic analysis based on three molecular markers of Dactylochlamys pisciformis and all of the three known Legendrea species, showing that they likely represent a new anaerobic lineage of ciliates. We first provide a detailed morphological characterization of both genera using modern microscopy and staining methods. We identify and discuss the bacterial (Syntrophaceae) and archaeal endosymbionts harbored by both genera based on 16S rRNA gene sequences. We also discovered that Legendrea preys on gastrotrichs, which is supported by molecular data and a unique video material documenting the feeding behavior of one Legendrea species. This study brings essential information needed to better understand the phylogeny, life strategies, and rarity of these organisms and emphasizes the importance of citizen science. Abstract Free-living anaerobic ciliates are of considerable interest from an ecological and an evolutionary standpoint. Extraordinary tentacle-bearing predatory lineages have evolved independently several times within the phylum Ciliophora, including two rarely encountered anaerobic litostomatean genera, Legendrea and Dactylochlamys. In this study, we significantly extend the morphological and phylogenetic characterization of these two poorly known groups of predatory ciliates. We provide the first phylogenetic analysis of the monotypic genus Dactylochlamys and the three valid species of Legendrea based on the 18S rRNA gene and ITS-28S rRNA gene sequences. Prior to this study, neither group had been studied using silver impregnation methods. We provide the first protargol-stained material and also a unique video material including documentation, for the first time, of the hunting and feeding behavior of a Legendrea species. We briefly discuss the identity of methanogenic archaeal and bacterial endosymbionts of both genera based on 16S rRNA gene sequences, and the importance of citizen science for ciliatology from a historical and contemporary perspective.


Introduction
The phylum Ciliophora comprises a bewildering diversity of ciliated unicellular eukaryotes inhabiting virtually every biotope on Earth as both free-living organisms or  48 32.1 E), and the GTUB population from a garden mesocosm (see [12]). The Czech population of L. pespelicani (VLKOV) was found in a small permanent concrete-lined freshwater reservoir in the village of Vlkov (49 • 09 04.0 N 14 • 43 22.4 E).

Morphological Characterization
Living cells were hand-picked with glass micropipettes and studied at magnifications of 100-1000× with brightfield and differential interference contrast illumination, using an Olympus BX51 microscope equipped with an Olympus DP70 camera (Olympus Corp., Tokyo, Japan). Cells of VB2A were fixed with 4% (v/v) formalin and impregnated with protargol (Polysciences, Warrington, PA, USA) according to a modified Dieckmann's protocol [13][14][15]. In vivo measurements were made from photomicrographs of uncompressed, freely swimming cells, using calibrated software. Counts and measurements from protargol preparations were made at 1000× magnification. For scanning electron microscopy (SEM), cells were fixed for at least 30 min in 2.5% glutaraldehyde or in a 1:1 (v/v) solution of 5% glutaraldehyde and 2% osmium tetroxide, processed according to [14], and examined JSM-IT200 scanning electron microscope (JEOL LV, Peabody, MA, USA). Except as noted, morphological terminology and size classification follows [1,2,9,14]. For clarity, we refer to the cortical pellicular projections of Dactylochlamys pisciformis and the three species of Legendrea as "tentacles" and the trichocyst-bearing oral bulge structures of L. pespelicani as "papillae".

DNA Extraction, Amplification, and Sequencing
In total, 3 cells from the VB2A population and single cells from the VLKOV, MOKOTL, MOKOTP1, and MOKOTP1Q populations were hand-picked with glass microcapillary pipettes, washed 5 times with distilled water, and added into 30 µL of DNA/RNA shield (Zymo Research, Irvine, CA, USA). The total DNA was isolated using the MasterPure TM Complete DNA and Purification Kit following the manufacturer's instructions. Three individuals of SAGEGLEN were picked for DNA extraction with a modified Chelex method [16].

Phylogenetic Analyses
An alignment of 18S rRNA gene sequences included 167 taxa, comprising 135 litostomateans and 32 outgroup taxa representing Armophorea and CONThreeP [29]. Five litostomatean sequences are newly reported in this study and all other sequences were obtained from GenBank. The dataset of concatenated 18S-ITS-28S rRNA region sequences consisted of 168 taxa, including 6 newly determined sequences and 162 sequences (135 litostomateans and 32 outgroup taxa) obtained from GenBank. The sequences were aligned using the G-INS-i algorithm [30] on the MAFFT server (https://mafft.cbrc.jp/alignment/software/, accessed on 15 March 2023). The alignment was manually edited using AliView [31] to trim primer sequences. The length of the final 18S rRNA gene data set was 1825 positions and the concatenated data set comprised 3163 positions. Phylogenetic trees were constructed using the maximum likelihood (ML) and Bayesian inference (BI) methods. The best fit substitution model (GTR + I + Γ) was selected using Modeltest-NG [32,33]. ML analysis was performed in RAxML-NG [34] on the CIPRES portal (http://www.phylo.org/sub_sections/portal/, accessed on 15 March 2023) under the GTR + I + Γ substitution model. Statistical support was assessed using 1000 bootstrap pseudoreplicates and convergence assessed using weighted Robinson-Foulds distance. Bayesian analysis was performed using MrBayes 3.2.7 [35] on the CIPRES portal using the GTR + I + Γ substitution model. For the 18S rRNA gene alignment, Markov chain Monte Carlo (MCMC) analyses of four chains (three hot, one cold, temperature 0.2) and for the concatenated alignment six chains (five hot, one cold, temperature 0.1) were each run for 10 million generations, with a sampling frequency of 1000 generations. The first 25% of sampled trees were removed as burn-in. Convergence was assessed using RWTY [36]. Uncorrected p-distances were calculated using MEGA v10.2.6.
Oral bulge truncate conical, inconspicuous in vivo, about 2 µm high, 4 µm in diameter after protargol impregnation. Oral bulge extrusomes and cytopharyngeal basket not observed ( Figure 2). Based on our own observations, we accept Kahl's synonymization of D. hystrix Wetzel, 1928 with D. pisciformis [10]).  About 10-12 ordinarily spaced somatic ciliary rows on slightly wavy left-hand spiraling ridges, short inclined ciliated pectinelle between anterior end of each somatic kinety and base of oral bulge, each pectinelle composed of three or four basal bodies. Kineties composed of monokinetids bearing long fine cilia (12-20 µm long), interspersed with tentacles. Tentacles very slender, tubular, retractile, about 15 µm long when extended, inconspicuously capitate when retracted, and each encloses an argyrophilic extrusome. Dorsal brush not observed. extending into the cytoplasm (white arrowhead), the contractile vacuole in diastole (white arrow), the macronucleus (black arrowhead), and the oral bulge (black arrow). (H) Surface view showing the long, fine somatic cilia (black arrowheads) and the tentacles (white arrowheads). (I) Lateral view of suctorian swarmer form (similar to Parapopdophrya solaris) from the same sample as VB2A, showing an extended capitate tentacle (white arrowhead), the apical part of the cell superficially resembling the oral bulge of Dactylochlamys (black arrow), and the somatic cilia resembling the pectinelles of Dactylochlamys (black arrowheads); otherwise somatic cilia are absent unlike in Dactylochlamys. Scale bars: 20 µm (A-F,I), 10 µm (G,H).  Oral bulge truncate conical, inconspicuous in vivo, about 2 µm high, 4 µm in diameter after protargol impregnation. Oral bulge extrusomes and cytopharyngeal basket not observed ( Figure 2).
Description based on in vivo observations of MOKOTL (n = 3; Figure 3; Video S2): Size 85-102 × 55-102 in vivo, contractile. Shape broadly obovate to nearly globular, anterior end obliquely truncate; up to 25 flexible, trailing tentacles restricted to posterior one half of cell. Horseshoe-shaped macronucleus, micronuclei not observed. Contractile vacuole eccentrically located at posterior end, sometimes indented by median pellicular cleft when tentacles retracted, occupies entire posterior end in diastole ( Figure 3C,D). Extrusomes in oral bulge, morphology not determined, tentacles with two types of extrusomes; one type curved, filiform, closely packed in central bundles at distal ends of tentacles (about 4.5 µm long); second type comprises layer beneath distal ends of tentacles, small (about 1.5 µm long), rod-shaped ( Figure 3F,H,I). Ejected extrusomes not observed. Tentacles extensible, covered with granular layer near junction with cell body, distinct cortical ruffles (collar) not observed. Cytoplasm colorless but cells appear dark under low magnification. Cells packed with refractive globules, food vacuoles with ingested prey. Cells swim steadily at moderate pace and rotate rightward about long axis (Video S2).  Ciliature holotrichous, somatic cilia about 6 µm long in closely spaced longitudinal kineties; number of kineties not determined. Each tentacle with single circumferential subterminal kinety. Dorsal brush inconspicuous with approximately 3 µm long clavate cilia; number and morphology of brush rows not determined.
Remark: We were unsuccessful in obtaining protargol preparations of L. loyezae.
• Legendrea pespelicani Penard, 1922 Amended diagnosis based on [9] and current study: With characters of genus. Size 180-210 × 60-90 µm in vivo. Broadly spatulate in lateral view, laterally compressed. Oral bulge oblique, invariably contains one trichocyst-bearing papilla at either end, usually one or two papillae in between; thus, oral bulge outline moniliform or dumbbell-shaped when viewed anteriorly. Macronucleus filiform, very long, tortuous strand. About six stout, mushroom-shaped, ciliated, extrusome-bearing tentacles commence at or anterior to equator, distributed at irregular intervals on dorsal and ventral margins of cell, always one posterior polar tentacle. Tentacular extrusomes longer (about 25 µm) than those in tentacles of L. ornata or L. loyezae. Distinctly spathidiid somatic ciliary pattern. Circumtentacular kineties composed of many more dikinetids than in L. ornata. Ciliary rows interrupted by tentacles while deviating around tentacles in L. ornata. Tentacle extension characteristic for L. loyezae and L. ornata not observed.
Oral bulge oblique (about 40 • to long axis of cell), convex anteriorly, long, narrow (about 80 × 5 µm). Circumoral kinety about 80 µm long, 5 µm wide between papillae, invariably encloses one extrusome-bearing papilla at dorsal and ventral end, usually one or two additional papillae between (two of nine individuals had only two oral bulge papillae: one dorsal, one ventral). Oral bulge papillae about 3.5 µm high × 10 µm across. Outline of circumoral kinety moniliform or dumbbell-shaped when viewed anteriorly due to deviation around large papillae (Figure 5C,D,G). Large broadly conical oral basket extending nearly entire length of cell ( Figure 5H).
ably one tentacle at posterior pole ( Figure 6A). Active tentacle extension not o Total of 3 dorsal brush rows composed of dikinetids bearing 3.5 µm long clav rows commence to left of dorsal oral bulge papilla; B1 and B2 extend posterior 20% of cell length; B3 highly unusual, i.e., proximal part similar in morphology an to B1 and B2 but posterior part composed of patchily distributed monokine groups of dikinetids extending about 50% of cell length ( Figure 6C,D,F).
Oral bulge oblique (about 40° to long axis of cell), convex anteriorly, long (about 80 × 5 µm). Circumoral kinety about 80 µm long, 5 µm wide between invariably encloses one extrusome-bearing papilla at dorsal and ventral end, usu or two additional papillae between (two of nine individuals had only two oral b pillae: one dorsal, one ventral). Oral bulge papillae about 3.5 µm high × 10 µm Outline of circumoral kinety moniliform or dumbbell-shaped when viewed anteri to deviation around large papillae (Figure 5C,D,G). Large broadly conical oral ba tending nearly entire length of cell ( Figure 5H). Stomatogenesis holotelokinetal; proliferation of basal bodies first occurs in kineties in line with proter dorsal brush rows. Developing anterior ends of opisthe curve ventrally ( Figure 5K).      Amended diagnosis based on [7,10,40] and current study: With characters of genus. Size 70-180 × 60-90 µm. Shape oblong ellipsoidal when swimming, broadly ovoidal when at rest with tentacles extended. Rows of tentacles at alternating angles on dorsal and ventral margins, commence in anterior one-fourth of cell, continuous around posterior end. Extrusome bundles about 5-7 × 5 µm in vivo. Tentacles highly extensible, up to >25 times retracted length. Ciliary pattern distinctly enchelyodontid, unlike spathidiid pattern of L. pespelicani, i.e., anterior ends of somatic kineties perpendicular to circumoral kinety. Ciliary rows not interrupted by tentacles as seen in L. pespelicani. Fewer (ten on average) circumtentacular kinety dikinetids than seen in L. pespelicani (35 on average). Three dorsal brush rows; B1 and B2 approximately equal in length, longer than B3.
Oral bulge inconspicuous, slightly oblique. Circumoral kinety narrow elliptical, composed of dikinetids. Oral bulge extrusomes not observed. Oral basket inconspicuous in vivo, obconical, extends about one-half length of cell in protargol preparations ( Figures 8D and 9A,D).
Stomatogenesis holotelokinetal, basal body proliferation commences first in somatic kineties in line with dorsal brush rows ( Figure 9H).

Molecular Data and Phylogenetic Analysis
Analyses of the concatenated data set revealed that the genera Legendrea and Dactylochlamys form a supported clade (bootstrap value 80, Bayesian posterior probability 1) together with Arcuospathidium sp. and Apertospathula oktemae; Legendrea sequences formed a fully supported clade (Figures 11 and S1).
As the 18S rRNA gene sequences of L. loyezae and L. pespelicani are identical and differ from that of L. ornata by only one nucleotide, the analysis based on the 18S rRNA gene did not resolve the relationships between Legendrea species (Figure S2), and neither did the analysis of the concatenated dataset resolve the relationships among the three Legendrea morphospecies. L. ornata MOKOTP1 and SAGEGLEN sequences cluster together. Sequences of L. loyezae OP352778 and L. pespelicani VLKOV are more closely related to each other than to L. loyezae MOKOTL (Figures 11 and S1). However, the published sequence of L. loyezae OP352778 represents only the V4 region of the 18S rRNA gene (1007 bp), while L. pespelicani VLKOV and L. loyezae MOKOTL are concatenated sequences (18S-ITS-28S). Genetic distances (uncorrected p-distances) between ITS-28S fragments (1183 bp) of L. ornata MOKOTP1, L. loyezae MOKOTL, and L. pespelicani VLKOV range between 0.008 and 0.009. Only amplification of the ITS-28S rRNA gene region of Dactylochlamys sp. MOKOTP1Q isolate was successful. Newly determined sequences are available in GenBank (accession numbers: OP985785-OP985794).
We also obtained two partial 18S rRNA gene sequences of prey gastrotrichs (family Chaetonotidae) from a cell of L. loyezae (MOKOTL; Figure 3B) and from the only cell of L. ornata from MOKOTP1 (GenBank accession numbers: OQ848030-OQ848031).

Morphological Comparison of Legendrea Species and Similar Species
Legendrea pespelicani can be easily distinguished from L. ornata and L. loyezae by size (180-210 vs. <180 µm), cell shape (broadly spatulate vs. oblong, resp. broadly obovate), and by possession of oral papillae (present vs. absent). L pespelicani also differs from L. ornata by number of tentacles (4-8 vs. , number of circumoral kineties (30 vs. 10 in average), number of somatic kineties (66-70 vs. 28-36), and the disposition of the ciliary rows (interrupted by tentacles vs. deviating around the tentacles). L. loyazae can be distinguished from L. ornata by the position of tentacles (bundle at the posterior vs. margin of the cell on 3 /4 of the cell length), and from both other species by the eccentric posterior contractile vacuole which appears as two vacuoles due to median cleft at posterior. All Legendrea species are different to any other spathidiid in terms of tentacles (present vs. absent), except Dactylochamys pisciformis, which are not similar to those of Legendrea (very slender, not ciliated vs. thick with circumtentacular kineties). L. pespelicani differs from Spathidium papilliferum in morphology of the oral papillae (circumtentacular kineties present vs. absent).

Remarks on the Rarity and Ecology of Dactylochlamys and Legendrea
Both Dactylochlamys and Legendrea spp. have been very rarely reported, although Lauterborn and Penard suggested that the populations could be quite abundant [9,37]. Our experience was similar: the VB2A sample was unusually rich in Dactylochlamys (approx. 4 cells/mL of water and sediment), but the abundance quickly declined in several days, and we were unable to maintain the ciliate in long-term culture. In other cases (MRATIN, MOKOTP1Q), only a few cells were found in the whole volume of a 50 mL sample. In the case of Dactylochlamys, Penard cautioned about possible misidentification, because some swarmers of suctorian ciliates (Phyllopharyngea: Suctoria) (e.g., Enchelyomorpha vermicularis Smith, 1899) appear similar in vivo (compare Figure 1F,I) [9]. Thus, mentions in the literature not accompanied by illustrations are questionable, e.g., [41,42]. In the case of Legendrea, we found two relatively rich localities in the Czech Republic and one in the USA; other known localities are, for example, in France, Germany, Belgium, and Poland. This might suggest that the rarity of Dactylochlamys and Legendrea is rather a matter of specific habitat conditions than endemicity.
Interestingly, gastrotrich rRNA gene partial sequences (family Chaetonotidae) were recovered from cells of L. loyezae and L. ornata, indicating ingestion of gastrotrichs as prey. The process of L. ornata from sediments of a German freshwater pond catching and digesting a gastrotrich (family Chaetonotidae) is the first documentation of the role played by the tentacles in prey capture (Video S7). Chaetonotidae is a species-rich and widely distributed gastrotrich family that commonly inhabits hypoxic lacustrine sediments [43]. Although chaetonotid gastrotrich 18S rRNA gene sequences were the only eukaryotic contaminants from Legendrea, the importance of gastrotrichs in the diet of Legendrea species remains unknown.

Phylogenetic Analysis and the Relationship of Tentaculiferous Ciliates
A recent report (8) included a 994 bp V4-V9 rRNA region sequence from L. loyezae. Phylogenetic analysis could not resolve a position of Legendrea within Haptorida. The problems with attempting to resolve relationships within Litostomatea based on analysis of the rRNA cistron have been emphasized previously [44,45]. Consistent with the studies mentioned above, we did not succeed in resolving the internal relationships among Legendrea species in spite of addition of two more markers, i.e., using three times more positions in comparison to [8].
Despite pronounced morphological differences, all three identified Legendrea species have identical (L. loyezae and L. pespelicani) or almost identical (L. ornata) 18S rRNA gene sequences. However, the existence of the three Legendrea morphospecies is further supported by the analysis based on additional molecular markers; namely, part of the ITS region and part of the 28S rRNA gene. Clustering of L. loyezae OP352778 with L. pespelicani was interpreted as an artifact due to the short length of the former (1007 nucleotides) and ambiguous sites in both of the sequences of L. loyezae. This view is also supported by the p-distances of the ITS-28S fragments of each of the three Legendrea species, which range from 0.008 to 0.009. On the other hand, Jankowski even erected a monotypic genus for each of the three Legendrea species solely on morphology, but we consider with respect to the morphological and molecular analysis presented above [46].
We also obtained a partial ITS-28S rRNA gene sequence of a second Dactylochlamys population (MOKOTP1Q), but the material was insufficient for morphometric analysis. Molecular data and in vivo observation suggest that it is possibly a new species (Video S9).
Our analysis showed that Dactylochlamys and Legendrea are not closely related to Actinobolina, contradicting the traditional notion of the family Actinobolinidae, comprising the genera Actinobolina, Belonophrya, Legendrea, and Dactylochlamys, based on the presence of tentacle-like structures which have evolved independently [1,10,47]. As noted previously, homoplasies are rife throughout the Spathidiidae and, together with many plesiomorphies, have so far obscured the evolutionary relationships in the group [44,[48][49][50]. Interestingly, Dactylochlamys and Legendrea are members of the same well-supported clade despite their obvious morphological differences. The other members of the clade (i.e., Apertospathula oktemae and an undescribed Arcuospathidium sp.) belong to spathidiid genera which do not bear any tentacle-like structures [2]. Spathidium papilliferum Kahl, 1930, another only distantly related spathidiid, has oral bulge papillae superficially similar to those of L. pespelicani but lacking circumtentacular kineties [10,48]. Analysis also showed that a supported clade including S. papilliferum (DQ411857, DQ411858; morphologically uncharacterized population) and an Epispathidium species (KT246081, KT246094; undescribed) are not closely related to the other S. papilliferum (KY556645, KY556652), suggesting possible misidentification ( Figure 11). Thus, the diversity of tentacle-or papillae-bearing spathidiids could be even higher than currently indicated. The genus Belonophrya is highly similar to Actinobolina, but molecular data and ultrastructural details of its tentacles are still unavailable [51]. Another tentacle-bearing ciliate, Holophrya ornata is now synonymized with Legendrea bellerophon based on its morphological characteristics ( Figure 12) [40]. We also obtained a partial ITS-28S rRNA gene sequence of a second Dactylochlamys population (MOKOTP1Q), but the material was insufficient for morphometric analysis. Molecular data and in vivo observation suggest that it is possibly a new species (Video S9).
Our analysis showed that Dactylochlamys and Legendrea are not closely related to Actinobolina, contradicting the traditional notion of the family Actinobolinidae, comprising the genera Actinobolina, Belonophrya, Legendrea, and Dactylochlamys, based on the presence of tentacle-like structures which have evolved independently [1,10,47]. As noted previously, homoplasies are rife throughout the Spathidiidae and, together with many plesiomorphies, have so far obscured the evolutionary relationships in the group [44,[48][49][50]. Interestingly, Dactylochlamys and Legendrea are members of the same well-supported clade despite their obvious morphological differences. The other members of the clade (i.e., Apertospathula oktemae and an undescribed Arcuospathidium sp.) belong to spathidiid genera which do not bear any tentacle-like structures [2]. Spathidium papilliferum Kahl, 1930, another only distantly related spathidiid, has oral bulge papillae superficially similar to those of L. pespelicani but lacking circumtentacular kineties [10,48]. Analysis also showed that a supported clade including S. papilliferum (DQ411857, DQ411858; morphologically uncharacterized population) and an Epispathidium species (KT246081, KT246094; undescribed) are not closely related to the other S. papilliferum (KY556645, KY556652), suggesting possible misidentification ( Figure 11). Thus, the diversity of tentacle-or papillaebearing spathidiids could be even higher than currently indicated. The genus Belonophrya is highly similar to Actinobolina, but molecular data and ultrastructural details of its tentacles are still unavailable [51]. Another tentacle-bearing ciliate, Holophrya ornata is now synonymized with Legendrea bellerophon based on its morphological characteristics ( Figure  12) [40].  Given the complex relationships between tentaculiferous lineages in Spathidiidae or Haptoria, it is not surprising that tentacles of other litostomateans are also likely not homologous. Tentacles in the family Mesodiniidae, possibly a sister group to Litostomatea, are placed anteriorly, are reinforced by a cylindrical structure of 14 microtubules, and bear extrusomes [1,52,53]. Another tentacle-bearing species with unclear phylogenetic position is Enchelyomorpha vermicularis, once included in Actinobolinidae, tentacles of which are also reinforced by microtubules and do not bear extrusomes, which is now considered to be the swarmer of a globular suctorian on morphological grounds but has not yet been sequenced [1,54]. Tentacles of the members of the class Phyllopharyngea have a completely different morphology compared to both Legendrea and Dactylochlamys. Typical suctorian feeding tentacles are formed by two concentric cylinders of microtubules, the inner one being reinforced by microtubule fibrils, and the outer with various taxon-specific microtubular structures [1,55]. Although, as indicated by their name, suctorians have been assumed to "suck" cell contents from their prey, this mechanism has been cast into doubt by Rudzinska [56]. In the parasitic subclass Rhynchodia, only a single sucking tentacle-like structure, probably a transformed cytostome, is present [1].

Putative Prokaryotic Endosymbionts
Using Sanger sequencing of the partial 16S rRNA gene, we identified putative methanogenic endosymbionts of L. pespelicani as Methanobacterium sp. and Methanosaeta sp. in L. loyezae. Using Illumina sequencing, we corroborated the identity of the archaeal symbiont of L. pespelicani as Methanobacterium sp. and identified the archaeal symbiont of Dactylochlamys pisciformis as Methanocorpusculum sp. Autofluorescence typical of archaeal methanogens was also observed in Dactylochlamys (Video S8) [57]. Interestingly, it seems that there is some symbiont diversity among the studied species, where genera belonging to the same clade and even each of the Legendrea species harbor different methanogenic symbionts. Methanogenic Archaea commonly form syntrophic symbioses with anaerobic ciliates using the hydrogen and acetate produced by the host mitochondria for their metabolism, e.g., [21,[58][59][60]. This suggests that both Dactylochlamys and Legendrea are most likely facultative or obligate anaerobes, potentially having hydrogen-producing mitochondria. The role of these prokaryotes as endosymbionts requires confirmation by fluorescence in situ hybridization and/or transmission electron microscopy. Nevertheless, our results indicate that both Dactylochlamys and Legendrea represent a novel anaerobic lineage of ciliates. Interestingly, in addition to archaeal symbionts, we also identified a putative bacterial symbiont related to the family Syntrophaceae. The Syntrophaceae are strictly anaerobic and grow only in the presence of hydrogen-utilizing partners, such as archaeal methanogens [61]. Further investigations of symbiotic interactions between the host, archaeal, and bacterial endosymbionts in Legendrea are needed.

The Role of Citizen Science in Ciliatology
Taxonomic and biogeographic studies of a wide variety of groups, including protists, have always been plagued by the problem of undersampling. This fact, in part, underlies the uncertainty and contentiousness associated with such topics as protist endemicity [62][63][64]. Since the number of academic professionals dedicated to and funded for mainly taxonomic work has rapidly declined, the importance of the much larger population of "expert amateurs" (interested individuals without formal academic credentials in the particular specialty) in species identification, specimen collection, ecology, and biogeographical data collection has come into clearer focus [65][66][67][68][69][70]. The role of the expert amateur was central to science world-wide prior to the professionalization of science in the early 20th century, after which contributions by non-academics were increasingly ignored or criticized as inferior. Alfred Kahl, one of the most notable ciliatologists of the last century, a high school teacher by training, was probably an unfortunate object of this developing attitude [38,71]. Although a formal definition of "citizen-science" infers that it occurs mainly as an activity in collaboration with, or under the direction of, academic professionals usually as part of structured projects [72], we consider the data generated by expert amateurs working independently to be of considerable value, especially with regard to rare taxa such as Legendrea. Most of the information that has accrued in the century since the first description of the genus Legendrea has been gathered almost exclusively by expert amateurs and made publicly available in online forums and social media. The study of Weiss et al. and this study are perfect examples of the fruitfulness of such cooperation [8].

Conclusions
Although both genera, Legendrea and Dactylochlamys, presented in this study are considered to be remarkably rare, we were able to collect enough data to not only corroborate their phylogenetic position close to or within Spathidiidae but also revealed that both genera are closely related to each other despite apparent differences in their tentacle-like structures. We studied their morphology with modern methods and found out that these two ciliate genera are in fact anaerobes harboring prokaryotic endosymbionts. Our work demonstrates that for studying rare organisms, both modern single-cell methods and the contribution of citizen science are essential. Thus, we encourage the increasing collaboration of academic professionals with their expert amateur counterparts, inclusion of their data in peer-reviewed research publications, and acknowledgement of their contributions.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/biology12050707/s1, Figure S1: Maximum likelihood tree of the class Litostomatea based on 18S-ITS-28S rRNA region sequences, Figure S2: Maximum likelihood tree of the class Litostomatea based on 18S rRNA gene sequences;  Table S4: Illumina consensus nucleotide sequences of putative prokaryotic symbionts; List of Videos S1−S9.