Novel contributions to the peritrich family Vaginicolidae (Protista: Ciliophora), with morphological and phylogenetic analyses of poorly known species of Pyxicola, Cothurnia and Vaginicola

Abstract

The classification of loricate peritrich ciliates is difficult because of an accumulation of several taxonomic problems. In the present work, three poorly described vaginicolids, Pyxicola pusilla, Cothurnia ceramicola and Vaginicola tincta, were isolated from the surface of two freshwater/marine algae in China. In our study, the ciliature of Pyxicola and Vaginicola is revealed for the first time, demonstrating the taxonomic value of infundibular polykineties. The small subunit rDNA, ITS1-5.8S rDNA-ITS2 region and large subunit rDNA of the above species were sequenced for the first time. Phylogenetic analyses based on these genes indicated that Pyxicola and Cothurnia are closely related. The present study suggested that the loricate species probably represent a distinct lineage in peritrich evolution and both genera Cothurnia and Thuricola are monophyletic. Pyxicola pusilla, Cothurnia ceramicola and Vaginicola tincta are recircumscribed.

ciliates, Loricate sessilids, rDNA

INTRODUCTION

One of the largest components of the global ciliate biota is the peritrich order Sessilida, with more than 800 species inhabiting various aquatic ecosystems (Kahl, 1935; Foissner et al., 1992; Lynn, 2008; Sun et al., 2009; Foissner et al., 2010). They commonly feed on small-sized particles and contribute significantly to the water quality (Pillai, 1952; Azam et al., 1983; Weber et al., 2007). Most are sedentary and attach to substrates via the scopula, stalk or lorica. Although the application of silver staining methods to reveal the infraciliature and silverline system has greatly advanced the understanding of the taxonomy of sessilids (Lom, 1964; Foissner & Schiffmann, 1974, 1975), most studies have focused on the aloricate forms, leaving loricate taxa relatively under-researched (Ji & Kusuoka, 2009; Utz et al., 2014; Ji et al., 2015; Jiang et al., 2016; Kühner et al., 2016; Liu et al., 2017;Shen et al., 2017; Sun et al., 2017; Zhuang et al., 2018).

The family Vaginicolidae is a species-rich group of loricate sessilids, with nearly 200 nominal species (Kahl, 1935; Corliss, 1979). However, among these names are many synonyms and misidentifications, because: (1) several genera have a similar appearance (e.g. Pyxicola Kent, 1882, was for a long time treated as Cothurnia Ehrenberg, 1831, because of their similar stalked lorica); (2) the congeners of each genus share an almost identical body shape with each other (e.g. thuricolas all exhibit a trumpet-shaped body when fully extended); (3) although the presence of a lorica is the key character for identifying vaginicolids, some species descriptions are incomplete, causing difficulties in distinguishing species (e.g. the narrow and wide side views of the lorica in some species are very different, but this has largely been ignored in previous reports; i.e. Kahl, 1935; Trueba, 1978, 1980; Warren & Paynter, 1991). It is now widely accepted that molecular data can help to elucidate some of the problems in the taxonomy and systematics of ciliates (Li et al., 2008; Sun et al., 2012, 2013; Li et al., 2015; Gao et al., 2017; Yan et al., 2017), but sequence data remain scarce for vaginicolids (Zhuang et al., 2016).

The genus Pyxicola was erected by Kent in 1882 for operculated ciliates that had previously been regarded as species of Cothurnia. Trueba (1978) revised the genus to include eight species and two forms. Thereafter, only one new species found in China was added to the genus: Pyxicola ovata Xu, 1987. Unfortunately, details of the ciliature and molecular information for the genus are still lacking.

Cothurnia, the largest genus of Vaginicolidae, includes approximately 100 species (Warren & Paynter, 1991), but ciliary data are available for only two of them (Song, 1992; Zhuang et al., 2016) and only three gene sequences have been deposited in GenBank (one of which is dubious). Furthermore, some species of Cothurnia with cursory descriptions are difficult to separate from each other. Overall, an investigation of this genus based on modern criteria is urgently needed.

Vaginicola Lamarck, 1816 is also a species-rich genus, containing nearly 50 nominal species (Corliss, 1979; Foissner et al., 1992). Compared with the two genera mentioned above, the structure of members of this genus is relatively simple, which renders species identification more difficult. This makes the ciliature more important for species separation but, so far, no ciliature data are available for this genus. Meanwhile, most of the literature is too old to contain photomicrographs to exhibit details of the lorica morphology (Kent 1882; Stiller, 1971; Küsters, 1974).

In this study we investigate and redescribe three species belonging to the three above mentioned genera, based on the observation of specimens in vivo and after protargol staining. The ciliature of Pyxicola and Vaginicola is thus described here for the first time. Phylogenetic analyses are performed based on the small subunit (SSU) rDNA, ITS1-5.8S rDNA-ITS2 regions and the large subunit (LSU) rDNA sequences.

MATERIAL AND METHODS

Sample collection

Pyxicola pusilla (Fig. 1E₁) was isolated from the freshwater alga Vaucheria sp., collected together with in situ water on 16 March 2017 from Hangzhou Bay, 30°22′ N; 121°11′ E (Fig. 1C, E), Ningbo (Fig. 1A), China. The water temperature was 9.5 °C and the salinity was 2 psu. Cothurnia ceramicola (Fig. 1D1) was collected from the surface of Enteromorpha sp. on 14 April 2017 at Tangdao Bay, 35°56′ N; 121°13′ E (Fig. 1B, D), Qingdao (Fig. 1A), China. The water temperature was 16 °C and the salinity was 32 psu. Vaginicola tincta (Fig. 1F1) was also obtained from Vaucheria sp. in Ri Lake, 29°54′ N; 121°34′ E (Fig. 1C, F), Ningbo (Fig. 1A), China, on 29 March 2017. The water temperature was 15 °C. In each case, the specimens were directly isolated from the subsamples for subsequent studies in the laboratory. The three species of Thuricola discussed in the present work were previously described by Lu et al. (2018).

Figure 1.

Map and sampling sites. A, map of the four seas of China, red asterisks indicate the locations of Qingdao and Ningbo. B, satellite map of southwest Qingdao, red asterisk indicates Tangdao Bay. C, satellite map of north-east Ningbo, red asterisks indicate the two sampling areas (wetland of Hangzhou Bay, Ri Lake). D, coast of Tangdao Bay. D1, Cothurnia ceramicola collected from site D. E, a pool in wetland of Hangzhou Bay. E1, Pyxicola pusilla collected from site E. F, Ri Lake, a freshwater lake in Ningbo. F1, Vaginicola tincta collected from site F.

Morphological study

Living cells were investigated using a compound microscope equipped with a high-power oil-immersion objective and differential interference contrast optics. The ciliature was revealed using the protargol staining method (Wilbert, 1975). The protargol powder was manually synthesized, following the method prescribed by Pan et al. (2013). Counts and measurements were performed at 400 and 1000× magnifications. Drawings of living organisms were based on in vivo observation and photomicrographs, whereas those of stained specimens were made with the help of a camera lucida. Terminology is according to Trueba (1978) and Warren & Paynter (1991).

DNA extraction, amplification and sequencing

Genomic DNA was extracted according to the methods described in Luo et al. (2017). The SSU rDNA was amplified using the primers 82F (5′ -GAA ACT GCG AAT GGC TC-3′; Jerome et al., 1996) and 18SR (5′-TGA TCC TTC TGC AGG TTC ACC TAC-3′; Medlin et al., 1988). A fragment of approximately 500 bp containing the ITS1, 5.8S ribosomal gene and ITS2 was amplified using primers 5.8SF (5′-GTA GGT GAA CCT GCG GAA GGA TC-3′) and 5.8SR (5′-CTG ATA TGC TTA AGT TCA GCG G-3′; Huang et al., 2018). The PCR amplifications of LSU rDNA were performed with the primers F3 (5′-ACG/C CGC TGA/G AT/CT TAA GCA T-3′) and R2 (5′-AAC CTT GGA GAC CTG AT-3′) from Moreira et al. (2007).Q5 Hot Start High-Fidelity DNA Polymerase (New England BioLabs, USA) was used to minimize the possibility of PCR amplification errors. Sequencing was performed bidirectionally by the Tsingke Biological Technology Company (Beijing, China).

Molecular phylogeny methodology

All new sequences are deposited in GenBank, with the accession numbers, lengths and GC contents of each summarized in Table 8. Besides these 15 newly obtained sequences (including three ITS1-5.8S-ITS2 region and three LSU rDNA of Thuricola), other sequences used in the present phylogenetic analyses were downloaded from GenBank. In order to compare morphologies, only species that had already been definitively identified were chosen, including: (1) SSU rDNA sequences of 45 sessilids, three mobilids and four hymenostomatians (Glaucoma chattoni X56533; Ichthyophthirius multifiliis U17354; Tetrahymena corlissi U17356; Tetrahymena pyriformis EF070254); (2) ITS1-5.8S-ITS2 region sequences of 29 sessilids and two hymenostomatians (Ichthyophthirius multifiliis DQ270016; Tetrahymena pyriformis KX832097); (3) LSU rDNA sequences of 14 sessilids and two hymenostomatians (Ichthyophthirius multifiliis EU185635; Tetrahymena pyriformis X54004). The hymenostomatians mentioned above were used as out-group taxa.

Sequences were aligned using the GUIDANCE2 algorithm online (http://guidance.tau.ac.il/ver2/) with its default parameters (Landan & Graur, 2008; Sela et al., 2015). The ends of the resulting alignments were trimmed manually using the program BioEdit v.7.0 (Hall, 1999). The final alignments of the SSU rDNA, ITS1-5.8S-ITS2 region and LSU rDNA sequences were 1767, 544 and 1795 positions, respectively.

Maximum likelihood (ML) analysis with 1000 bootstrap replicates was computed at the CIPRES Science Gateway (http://www.phylo.org), using the GTR + gamma model performed by RAxML-HPC2 v.8.2.10 on XSEDE (Stamatakis, 2014). Bayesian inference (BI) analysis was carried out using MrBayes v.3.2.6 on XSEDE (Ronquist et al., 2012) with the GTR+ I+ G model selected by MrModeltest v.2.2 (Nylander, 2004) via the Akaike Information Criterion. Markov chain Monte Carlo simulations were run for 6 000 000 generations with a sample frequency of 100 generations and a burn-in of 6000 trees (10%). All the remaining trees were used to calculate the posterior probability using a 50% majority rule consensus. Tree topologies were visualized using MEGA v.7.0 (Kumar et al., 2016). Systematic classification follows Gao et al. (2016) and Sun et al. (2012).

Topology testing

To assess the monophyly of the genus Vaginicola and the family Vaginicolidae, the approximately unbiased (AU) test (Shimodaira, 2002) was used. Four constrained ML trees were generated by PAUP (Swofford, 2002) with enforced constraints (Table 9) and then compared with the best (unconstrained) ML topologies implemented in CONSEL (Shimodaira & Hasegawa, 2001). Internal relationships in the constrained group and among the remaining taxa were unspecified.

RESULTS
Subclass Peritrichia Stein, 1859
Order Sessilida Kahl, 1933
Family Vaginicolidae Fromentel, 1874
Genus Pyxicola Kent, 1882

No species of Pyxicola have previously been investigated successfully using silver staining methods (Trueba, 1978). This genus is characterized by the operculum attached to the zooid just below the peristomial lip. The lorica is inhabited by a single zooid that is attached to a non-contractile stalk. The anterior portion of the lorica is usually oblique and neck-like.

Pyxicola pusilla (Wrześniowski, 1866) Kent, 1882
(Figs 2A–P, 3A–U; Tables 1–3)

Since its original description, Cothurnia pusilla were reported repeatedly (Wrześniowski, 1866, 1867, 1870). However, the ciliature of P. pusilla has never been revealed. An improved diagnosis based on previous reports and new data is supplied.

Table 1.

Morphometric characterization of Chinese populations of Pyxicola pussila (upper line), Cothurnia ceramicola (middle line) and Vaginicola tincta (lower line) based on specimens in vivo (measurements in μm)

Characters	Min	Max	Mean	SD	CV	N
Body length	64	74	67.86	3.34	4.7	7
	113	158	133.71	16.55	12.4	7
	157	243	190.20	26.04	13.7	10
Body width	12	21	18.07	2.83	15.7	7
	13	17	14.71	1.25	8.5	7
	23	35	26.70	4.06	15.2	10
Diameter of peristomial lip	21	28	23.14	2.41	10.4	7
	25	29	28.00	1.22	4.4	9
	48	60	53.10	3.73	7.0	10
Lorica length	60	67	62.00	2.50	4.0	7
	95	105	100.00	3.96	4.0	7
	130	175	148.86	15.35	10.3	7
Lorica width	31	34	32.43	0.98	3.0	7
	30	33	31.71	1.11	3.5	7
	47	61	55.00	4.16	7.6	7
Diameter of aperture	19	22	20.00	0.96	4.8	7
	30	34	32.14	1.35	4.2	7
	49	64	57.14	4.98	8.7	7
Lorica length-width ratio	1.82	2.03	1.91	0.07	3.7	7
	2.94	3.28	3.15	0.11	3.5	7
	2.33	3.13	2.71	0.27	10.0	7
Endostyle length	1.0	1.5	1.33	0.19	14.3	7
	2.0	3.0	2.30	0.35	15.2	6
	–	–	–	–	–	–
Mesostyle length	2.0	2.6	2.39	0.26	10.9	7
	1.5	2.5	2.12	0.37	17.5	6
	–	–	–	–	–	–
Stalk length	2.0	19.0	10.29	6.47	62.9	7
	3.0	5.0	4.08	0.74	18.1	6
	–	–	–	–	–	–
Diameter of operculum	16	18	17.14	0.69	4.0	7
	–	–	–	–	–	–
	–	–	–	–	–	–
No. of striations from peristome to TBa	–	–	–	–	–	–
	50	53	51.00	1.22	2.4	5
	74	83	79.00	3.74	4.7	4
No. of striations from TB to scopulaa	–	–	–	–	–	–
	56	73	64.75	5.68	8.8	8
	76	84	80.20	3.35	4.2	5
No. of total striationsa	62	62	62.00	0	0	1
	107	119	113.40	5.03	4.4	5
	150	163	158.25	5.91	3.7	4

^aRough values.

CV, coefficient of variation in %; Max, maximum; Mean, arithmetic mean; Min, minimum; N, number of specimens; TB, trochal band; –, data not available.

Table 2.

Morphometric comparison of different populations of Pyxicola pusilla (based on specimens in vivo, measurements in μm)

Populations	Body length	Body width	Diameter of PL	Body protrudes out	Lorica length	Lorica width	Diameter of aperture	Shape quotient	Feature of lorica	Colour of lorica	Neck of lorica	Stalk length	Habitat	Data source
Pyxicola pusilla	64–74	12–21	21–28	little	60–67	31–34	19–22	c. 1.9	annulated	colourless to brown	short	2–19	FW	original
Cothurnia pusilla	58	–	16	little	46	22	–	c. 2.1	annulated	colourless to brown	–	3	FW	Wrześniowski (1866, 1867, 1870)
Cothurnia furcifer	–	–	–	–	50	–	–	–	commonly annulated	colourless to brown	–	1/6 b	FW	Hutton (1878)
Pachytrocha cothurnoides	–	–	–	little	50	–	–	c. 2.0	–	colourless	–	short	FW	Kent (1882)
Pyxicola pusilla	–	–	–	–	50	–	–	c. 2.0	–	–	–	1/10 b	FW	Kent (1882)
Pyxicola affinis	–	–	–	little	80	–	–	c. 2.0	–	colourless to brown	–	1/3–1/2 b	FW	Kent (1882)
Caulicola valvata	–	–	–	1/4a	50	–	–	–	–	–	–	1/3 b	BW	Stokes (1895)
Pyxicola carteri	78–82 (n = 2)	14–17 (n = 2)	27–38 (n = 2)	little to more	60–77 (n = 2)	31–34 (n = 2)	17–20 (n = 2)	c. 2.0	annulated	brown	inconspicuous	short	FW	Sommer (1951)
Pyxicola eforiana	–	–	–	less than 1/2a	75–80	–	–	–	annulated	brown	–	short	BW	Tucolesco (1962)
Pyxicola carteri forma constricta	–	–	–	–	54–58	25–29	12–14	c. 2.o	–	–	–	–	FW	Nusch (1970)
Pyxicola pusilla	–	–	–	little	66–77 (n = 2)	31–38 (n = 2),	21–23 (n = 2)	c. 2.0	annulated	colourless to brown	short	4–60	FW	Trueba (1978)
Pyxicola pusilla	–	–	–	–	50–55	28–30	19–28	c. 1.6	–	–	short	4–12	FW	Shen & Gu (2016)

The bold indicates these data was based on original study.Abbreviations: FW, fresh water; BW, brackish water; PL, peristomial lip; –, data not available; ^aproportion of body length; ^bproportion of lorica length.

Table 3.

Morphometric comparison of Pyxicola pusilla with closely related congeners (based on specimens in vivo, measurements in μm)

Species	Body length	Body width	Diameter of PL	Body protrudes out	Lorica length	Lorica width	Diameter of aperture	Shape quotient	Feature of lorica	Colour of lorica	Neck of lorica	Stalk length	Substrate	Data source
P. pusilla	58–82	12–21	21–38	commonly just beyond PL	45–80	25–38	12–23	1.9–2.1	commonly clearly annulated	colourless to brown	short and conspicuous	2–60	aquatic plant	List 1
P. carteri	–	–	–	1/4–1/3a	65–96	20–33	–	2.2–3.0	slightly and gentle annulated	colourless to brown	conspicuous	1–15	aquatic plant	Kent (1882); Trueba (1978)
P. operculigera	–	–	–	1/6–1/3a	45–70	22–35	–	1.7–2.6	never clearly annulated	colourless to brown	conspicuous**	12–130, average 80	Cordylophora or aquatic plant	Kent (1869); Trueba (1978)
P. psammata*	–	–	–	–	75–85	23–35	–	2.4–3.3	slightly and gentle annulated	–	inconspicuous	15–40	Microlistra spinosissima	Hadži, 1940; Trueba (1978)

Abbreviations: PL, peristomial lip; –, data not available; ^aproportion of body length; *based on fixed specimens; ^** inferred from the drawing.

List 1: Original; Wrześniowski (1866, 1867, 1870); Hutton (1878); Kent (1882); Stokes (1895); Sommer (1951); Tucolesco (1962); Nusch (1970); Trueba (1978); Shen & Gu (2016).

Improved diagnosis

Lorica about 50–80 × 25–40 μm in size, urceolate with a short and oblique neck, aperture circular and 19–22 μm in diameter. Lorica stalk highly variable in length. Zooid about 55–85 × 10–25 μm in vivo, peristomial lip single-layered, protrudes just above the aperture. Contractile vacuole located in anterior quarter of zooid. Infundibular polykinety 3 composed of three equally long rows of kinetosomes. About 60 striations from the peristome to the scopula. Fresh or brackish water.

Description based on Ningbo population

Zooid is solitary, about 60–75 × 10–20 μm in vivo and usually convex at mid-body (Figs 2A, B, D, 3A–D, F) although sometimes trumpet-shaped (Figs 2C, 3E). The single-layered peristomial lip is 21–28 μm in diameter, thin but everted strongly, and is situated just beyond the aperture when the zooid is fully extended. An operculum adheres to the pellicle below the peristomial lip and is adjacent to the lower edge of the lorica aperture (Figs 2A–D, 3A–G). The operculum is 16–18 μm in diameter and concave in its central part (Figs 2K, 3H–L). The colour of the operculum varies with age. The operculum is colourless when the cell is young and yellow to dark brown when the cell is old (Fig. 3A–G). The operculum closes the aperture of the lorica when the cell is contracted (Figs 2E, 3H, I). The peristomial disc is obliquely elevated. The pellicle is very elastic, with about 62 distinct striations from the peristome to the scopula (counted from one live cell). The trochal band is unrecognizable in vivo.

Figure 2.

Pyxicola pusilla in vivo (A–N) and after protargol staining (O, P). A–D, lateral views of different individuals, showing the varieties of body shape, lorica transparency and stalk length. E, a contracted individual inside the lorica. F, longitudinally oriented macronucleus. G–J, varieties of lorica shape. K, lateral view of operculum. L, apical view of operculum. M, details of lorica base, mesostyle and endostyle. N, detail of zooid, showing pellicular striations. O, macronucleus, the trochal band and the scopula. P, oral ciliature. Abbreviations: AR, annular ridge; BD, basal disc; CV, contractile vacuole; En, endostyle; G, germinal kinety; H, haplokinety; LS, longitudinal striae; Ma, macronucleus; Me, mesostyle; Op, operculum; Po, polykinety; P1–3, infundibular polykineties 1–3; Sc, scopula; St, stalk; TB, trochal band; Tu, tube. Scale bars = 30 μm.

Figure 3.

Photomicrographs of Pyxicola pusilla in vivo (A–O) and after protargol staining (P–U). A–G, lateral views of different individuals, showing the varieties of body shape, lorica shape and colour, and stalk length; arrows indicate the contractile vacuole. H, I, two constricted individuals. J, apical view of a contracted individual, showing the aperture of lorica and frontal view of operculum. K, L. lateral view of operculum; arrows indicate the depression of operculum border. M, N. posterior portion of individuals; arrows indicate the mesostyle, arrowhead marks the basal disc. O, showing the striations of pellicle. P, Q. two specimens; arrows and arrowheads indicate macronucleus and the trochal band, respectively. R–T, oral ciliature. U, infundibular polykineties 1–3. Abbreviations: G, germinal kinety; H, haplokinety; P1–3, infundibular polykineties 1–3. Scale bars = 30 μm.

The cytoplasm is colourless, with numerous small globular granules and several food vacuoles. A contractile vacuole is located at the anterior quarter of the body, about 10 μm in diameter. The macronucleus is vermiform, extends longitudinally throughout almost the whole body in vivo (Fig. 2F) and becomes twisted in stained cells (Figs 2C, 3P, Q). The micronucleus was not observed.

The endostyle, mesostyle and stalk are contiguously connected with longitudinal striae. This combination penetrates the lorica wall via a tube-like structure (Figs 2A–D, M, 3M, N). The zooid sits on top of the endostyle and the tube. The endostyle is 1.0–1.5 μm long and the mesostyle is 2.0–2.6 μm long, knot-like and flared at the equator (Figs 2M, 3M, N). The lorica attaches to the substrate via a non-contractile stalk that is continuous with the mesostyle, attaching to the substrate via a basal disc (Figs 2A–E, 3N). The stalk itself varies significantly in length (between 2 and 19 μm), but this variance is not correlated with age. Like the lorica, the basal disc becomes more coloured with age (Fig. 3A–G).

The lorica is urceolate, about 60–70 × 30–35 μm in size and with a colour the same as the operculum (Fig. 3A–G). The anterior portion of the lorica is constricted and bent, forming a short neck (Figs 2G–J, 3A–G), which is about one-eighth of the lorica in length. The aperture is circular and oblique, about 19–22 μm across (Figs 2G–J, 3A–G, J). There are three distinct annular ridges at the main part of the lorica (Figs 2A–J, 3A–G) that reaches its greatest width at the lowest ridge. The bottom of the lorica is double-layered and holds the mesostyle and endostyle within a tube extension (Figs 2G–J, M, 3H, I, M, N).

The oral ciliature is of the usual sessilid type. A haplokinety is accompanied by a polykinety and circles about 1.75 turns around the peristomial disc. Then they enter the infundibulum and make a further turn on opposite walls. Each of the three infundibular polykineties (P1–3) is composed of three rows of kinetosomes (Figs 2P, 3R–U). The three rows of P1 are equal in length and extend to the cytostome (Figs 2P, 3R–U). P2 terminates at the convergence of P1 and P3, row 1 commences slightly before the other two rows and row 3 being detached from the inner two rows at the abstomal end (Figs 2P, 3R–U). P3 is shorter than P1 and P2 and terminates slightly behind P2, but clearly before P1. The rows in P3 lie parallel to each other and are equal in length (Figs 2P, 3R–U). The germinal kinety lies parallel to the haplokinety in the abstomal two-thirds of the infundibulum (Figs 2P, 3R–T). An epistomial membrane was not observed. Since the zooid is highly contractile and resides in the lorica, the silverline system was unsuccessfully stained by the silver nitrate method. Nevertheless, the pattern and number of silverlines could be inferred by the transverse striations of the pellicle when viewed in vivo.

Genus Cothurnia Ehrenberg, 1831

Species of Cothurnia usually have one or two zooids in a lorica that lacks any closure apparatus. Zooids attach to the lorica base directly or via an endostyle, with transverse striations on the pellicle. The shape of lorica in this genus is highly diverse. The incomplete data of most Cothurnia species necessitate a reinvestigation of this species-rich genus based on modern criteria.

Cothurnia ceramicola Kahl, 1933
(Figs 4A–L, 5A–R; Tables 1, 4, 5)

Cothurnia ceramicola was originally reported by Kahl in 1933 as an epibiont of the alga Ceramium from Kiel, Germany, and redescribed several times thereafter. However, its ciliature data is still unavailable. An improved diagnosis based on previous reports and new data is hereby supplied.

Table 4.

Morphometric comparison of different populations of Cothurnia ceramicola (based on specimens in vivo, measurements in μm)

Populations	Body length	Body width	Diameter of PL	Body protrudes out	Lorica length	Lorica width	Diameter of aperture	Shape quotient	Shape	Feature of lorica	Endostyle	Mesostyle	Stalk	Habitat	Data source
C. ceramicola	113–158	13–17	25–29	1/6–1/3a	95–105	30–35	30–34	2.9–3.3	cylindroid	annulated	2–3, striated	1.5–2.5	3–5	MW	original
C. ceramicola	–	–	–	1/4–1/3a	60	c. 24*	c. 21*	c. 2.5*	cylindroid	slightly annulated*	short, striated*	short, striated*	short	MW	Kahl (1933)
C. ceramicola	–	–	–	more than 1/3a	60	c. 24*	c. 21*	c. 2.5*	cylindroid	slightly annulated*	short, striated*	short, striated*	short	MW	Khal (1935)
C. ceramicola	–	–	–	much	74–77	–	–	–	cylindroid	–	short	short	short	MW	Precht (1935)
C. ceramicola (?)	–	–	–	c. 1/5a*	78–82	–	–	–	cylindroid	slightly annulated*	short, striated*	short, striated*	short, striated	MW	Precht (1935)
C. ceramicola	–	–	–	less than 1/4a	65–75	–	–	–	–	annulated**	short	short, striated	short	MW	Felinska (1965)
C. ceramicola	100–150	15–17	26–28	1/3–2/3a	67–102	29–35	23–31	–	cylindroid to elliptical	annulated sometimes	short, striated	short, striated	7–9, striated	MW	Küsters (1974)
C. ceramicola	–	–	–	1/3a	50–65	24–33	–	–	elliptical	smooth*	short, striated*	short, striated*	10, striated*	MW	Song (1992)
C. ceramicola	140–220	–	–	less than 1/2a	110–130	20–45	–	–	cylindroid	annulated	short, striated*	short, striated*	short, striated*	MW	Sun et al. (2009)
C. ceramicola	71–76	14	22–24	1/5a	69–76	29–32	29–32	c. 2.4	cylindroid	smooth*	short, striated*	short, striated*	short, striated*	FW	Shen & Gu (2016)

Abbreviations: FW, fresh water; MW, marine water; PL, peristomial lip; –, data not available; ^aproportion of body length; *inferred from the drawing; ^** the span of annular ridges is very large.

Table 5.

Morphometric comparison of Cothurnia ceramicola with closely related congeners (based on specimens in vivo, measurements in μm)

Species	Body length	Body width	Diameter of PL	Body protrudes out	Striations	Lorica length	Lorica width	Diameter of aperture	Shape quotient	Shape	Feature of lorica	Endostyle	Mesostyle	Stalk	Data source
C. ceramicola	71–220	13–17	22–29	1/6–1/2a	conspicuous	60– 130	20–45	21–34	2.5–3.3	cylindroid	3–5 annular ridges	2–3, striated	1.5–2.5, striated	3–9, striated	List 1
C. curvula	60	8–10	–	just beyond	inconspicuous	60	30	15	c. 2.0	aperture narrowed	3 annular ridges in middle	short	short	short	Entz (1884); Warren & Paynter (1991)
C. harpactici	105	17–20	20	1/6a	conspicuous	75–88	28–35	30	c. 2.3*	cylindroid	3–4 annular ridges	short, striated	short, striated	15, striated	Kahl (1933); Warren & Paynter (1991)
C. fibripes	90	10	15–20	1/3a	conspicuous	60	30	15	c. 2.0	aperture narrowed	annular ridges in middle	short	short, striated	short	Kahl (1933); Warren & Paynter (1991)

Abbreviations: PL, peristomial lip; –, data not available; ^aproportion of body length; *inferred from drawing.

List 1: original; Kahl (1933, 1935); Precht (1935); Küsters (1974); Sun et al. (2009).

Improved diagnosis

Lorica about 60–130 × 20–45 μm in size, wall transparent and colourless, cross-section circular. Aperture about 20–45 μm in diameter, circular and distinctly everted. Zooid about 70–220 × 13–17 μm in vivo, when fully extended, protruding a quarter to a half out the lorica, with a single-layered peristomial lip. Contractile vacuole dorsally located. Endostyle and mesostyle knot-like and stalk with conspicuous longitudinal striae. Three equally long kinetosomal rows in infundibular polykinety 3, terminating before infundibular polykinety 1. About 50–55 striations from the peristome to the trochal band, 55–75 striations from the trochal band to the scopula.

Description based on Qingdao population

One or two zooids reside in the lorica. The solitary zooid is relatively smaller and the two co-existing zooids are abreast and anisometric (Figs 4A, C, 5A–F). When fully extended, the zooid is 110–160 × 13–17 μm in size and trumpet-like. The peristomial lip is single-layered, about 25–30 μm across and conspicuously everted; the peristomial disc projects slightly from the peristomial lip and forms a convex contour (Figs 4A, C, 5A, H). Pellicular striations can be recognized under a magnification of 200× (Figs 4H, 5A, J). The numbers of striations from the peristome to the trochal band and the trochal band to the scopula are 50–53 (N = 5) and 56–73 (N = 8), respectively. The trochal band is located above mid-body and can be recognized as a high crest on the body surface (Figs 4A, C, 5A–F).

Figure 4.

Cothurnia ceramicola in vivo (A–J) and after protargol staining (K, L). A, lateral view of an individual with single zooid. B, showing macronucleus in vivo. C, lateral view of an individual with two zooids. D–G, varieties of lorica shape. H, striations of pellicle. I, apical view of zooid, showing the location of contractile vacuole. J, posterior portion of individual, showing lorica base, endostyle, mesostyle, stalk and basal disc. K, showing macronucleus and the trochal band. L, oral ciliature. Abbreviations: AR, annular ridge; BD, basal disc; CV, contractile vacuole; En, endostyle; EM1–2, epistomial membrane 1–2; G, germinal kinety; H, haplokinety; In, infundibular; Ma, macronucleus; Me, mesostyle; Po, polykinety; P1–3, infundibular polykineties 1–3; St, stalk; TB, trochal band; Tu, tube. Scale bars = 50 μm.

Figure 5.

Photomicrographs of Cothurnia ceramicola in vivo (A–L) and after protargol staining (M–R). A–F, lateral views of different individuals with single zooid (A, C, E, F) and two zooids (B, D); arrow indicates the contractile vacuole. G, the single-layered peristomial lip. H, arrow indicates the contractile vacuole. I, apical view of zooid; arrow indicates the contractile vacuole, arrowhead indicates the infundibulum. J, lateral view of lorica; arrows indicate the annular ridges on the lorica, arrowheads indicate the everted aperture. K, apical view of aperture. L, posterior portion of individual; arrow indicates the mesostyle, arrowhead indicates the lorica base, double-arrowheads indicate the basal disc. M, N, oral ciliature. O, P, two stained specimens. Q, R, oral ciliature; arrow indicates the epistomial membrane 2. Abbreviations: EM1–2, epistomial membrane 1–2; G, germinal kinety; H, haplokinety; P1–3, infundibular polykineties 1–3. Scale bars = 50 μm.

The cytoplasm is hyaline without special features, with numerous reflective granules usually less than 1 μm in diameter. Several spherical food vacuoles are scattered throughout the cytoplasm in the upper half of the body. The freshly detached food vacuoles are generally fusiform. The contractile vacuole is approximately 10 μm across when fully expanded, situated at the dorsal wall of the infundibulum and its centre is just lower than the peristomial lip (Figs 4A, C, I, 5A, H, I). A vermiform macronucleus stretches almost along the entire body length and is relatively straight in vivo (Fig. 4B). Due to significant shrinkage in the fixation process, the macronucleus is twisted in protargol stained specimens (Figs 4K, 5O, P). The micronucleus was not observed.

The attachment apparatus of the zooids and the lorica consists of four parts that are continuous with each other, namely endostyle, mesostyle, stalk and basal disc. The endostyle is nearly cylindrical, 2–3 μm long, while the mesostyle is widened, knot-like and 1.5–2.5 μm long (Figs 4J, 5L). Endo- and mesostyle pass through the lorica base via a tube extension around them. The non-contractile stalk is 3–5 μm long with an expanded base (Figs 4J, 5L). Striae run lengthways throughout from mesostyle to stalk. The basal disc is colourless, considerably expanded and adheres to the substrate (Figs 4J, 5L).

The lorica is about 95–105 × 30–35 μm in size, with a ratio of length to width of 2.9–3.3, quasi-cylindrical, the anterior portion slightly narrower than the basal portion and circular in cross-section (Figs 4A–G, 5A–F, J). The aperture is distinctly everted and nearly as wide as the lorica (Fig. 5K). The wall of the lorica is transparent, with three to five continuous annular ridges (Figs 4D–G, 5A–F, J), mainly situated in the upper half. The lorica itself tapers to the base in a curve, with a bottom that is double-layered and a tube that surrounds and holds the endostyle and mesostyle (Figs 4J, 5J, L).

The oral ciliature is broadly similar to that of most sessilids: the haplokinety and polykinety spiral through a 1.5 turn around the peristomial disc before plunging into the infundibulum and diverging within the infundibulum to lie on opposite walls (Figs 4L, 5M–R). Each of the three infundibular polykineties is composed of three rows of kinetosomes. The three rows in P1 are equally long and their adstomal ends terminate at the same level (Figs 4L, 5Q, R). P2 terminates adstomally above the adstomal end of P1 and P3 and the abstomal ends of each row are diverged and staggered (Figs 4L, 5Q, R). The rows of P3 are parallel and equal in length, terminating adstomally just beyond the adstomal ends of P2. Since the outer two rows of P3 are very proximal, they sometimes overlapped in stained specimens (Figs 4L, 5N, Q, R). The germinal kinety lies parallel to the haplokinety in the abstomal half of the infundibulum (Figs 4L, 5Q, R). Epistomial membrane 1 is located near the entrance of the infundibulum (Figs 4L, 5M, Q). Epistomial membrane 2 is shorter than epistomial membrane 1 and located slightly ahead of the distal end of the haplo- and polykinety (Figs 4L, 5M). The silverline system is in the Vorticella pattern and the number of silverlines can be determined by counting pellicular striations in cells observed in vivo.

Genus Vaginicola Lamarck, 1816

Species of this genus have a comparatively simple formation. Commonly, one or two zooids attach to the lorica floor via the scopula, with transverse striations on the pellicle. The lorica attaches to the substrate directly without stalk. Although species of Vaginicola are commonly found in a wide variety of habitats and some have been investigated many times, their species identification is exclusively based on live specimens.

Vaginicola tincta Ehrenberg, 1830
(Figs 6A–I, 7A–U; Tables 1, 6, 7)

Vaginicola tincta was originally described by Ehrenberg (1830) and was subsequently redescribed several times (Ehrenberg 1838; Kahl, 1935; Sommer, 1951; Stiller, 1971; Vucetich & Escalante, 1979; Foissner et al., 1992; Kreutz & Foissner, 2006; Shen & Gu, 2016). However, its ciliature is still unknown. We here provide an improved diagnosis based on present and previous observations.

Improved diagnosis

Lorica about 75–175 × 30–65 μm, cross-section circular. Aperture about 45–65 μm in diameter, circular and everted. Fully extended zooids about 100–250 × 20–35 μm in vivo, peristomial lip single-layered. Contractile vacuole located dorsally at the same level as peristomial lip. Rows of P2 shortened progressively from row 1 to row 3, row 1 merging with P1 abstomally and terminates above other two rows, row 3 divergent from other two rows abstomally. P3 composed of three rows, commencing at the same level and terminating slightly above adstomal end of P1; inner row shorter than other two rows and terminates earliest. About 70–85 striations from the peristome to the trochal band and 75–85 striations from the trochal band to the scopula.

Description based on Ningbo population

Usually with a single zooid, sometimes coupled with a second anisometric zooid (Figs 6A, B, 7A–F). When solitary, less than one-third of the zooid protrudes through the aperture. When two zooids are present, more than one-third of the larger zooid protrudes through the aperture. The zooid is up to about 155–245 μm long and 20–35 μm wide and trumpet-like in vivo. The peristomial lip is single-layered and thin and is significantly wider (48–60 μm in diameter) than the body (Figs 6A, B, 7A–F). The peristomial disc is elevated above the peristomial lip and has a slightly convex surface (Figs 6A, B, 7I). Pellicular striations with convex ribbing are conspicuous at magnifications of 400× and higher (Figs 6E, 7A, N). There are about 74–83 (N = 4) striations from the peristome to the trochal band and 76–84 striations (N = 5) from the trochal band to the scopula. The trochal band can be recognized as a high crest on the body surface in the mid-body region (Figs 6E, 7N).

Figure 6.

Vaginicola tincta in vivo (A–G) and after protargol staining (H, I). A, B, two individuals with single zooid (A) and two zooids (B). C, D, two contracted individuals. E, striations of pellicle. F, apical view of zooid, showing the location of contractile vacuole. G, macronucleus in vivo. H, showing macronucleus and the trochal band. I, oral ciliature. Abbreviations: CV, contractile vacuole; EM1–2, epistomial membrane 1–2; G, germinal kinety; H, haplokinety; In, infundibulum; Ma, macronucleus; Po, polykinety; P1–3, infundibular polykineties 1–3; TB, trochal band. Scale bars = 80 μm (A–G), 40 μm (H).

Figure 7.

Photomicrographs of Vaginicola tincta in vivo (A–N) and after protargol staining (O–U). A–F, lateral views of different individuals with single zooid (A, C–E) and two zooids (B, F), showing varieties of body shape, lorica shape and colour. G, H, two contracted individuals. I, arrow indicates the contractile vacuole. J, apical view of zooid; arrow indicates the contractile vacuole, arrowhead indicates the infundibular. K, apical view of lorica. L, anterior portion of lorica; arrow indicates the everted aperture. M, posterior portion of lorica; arrow indicates the fluctuation of lorica. N, striations of pellicle; arrow indicates the trochal band. O–Q, three stained specimens; arrows indicate the trochal band. R, T, oral ciliature; asterisks indicate the beginning of haplokinety and polykinety. S, U, infundibular polykineties 1–3; asterisks indicate the ending of row1 within P2 and P3. Abbreviations: EM1–2, epistomial membrane 1–2; G, germinal kinety; H, haplokinety; Ma, macronucleus; Po, polykinety; P1–3, infundibular polykineties 1–3. Scale bars = 80 μm (A–H), 40 μm (O–Q).

The endoplasm is transparent and comparatively featureless, but full of reflective granules (less than 1 μm across). Several comparatively large granules (2–3 μm across) are distributed at the posterior end of the body (Figs 6A, 7A). A few fusiform or globular food vacuoles are scattered throughout the cytoplasm. The contractile vacuole is approximately 7 μm in diameter and is situated at the dorsal wall of the infundibulum, at the same level as the peristomial lip (Figs 6A, B, F, 7I, J). The macronucleus is vermiform and extends the entire length of the body, with several inconspicuous curves in vivo (Fig. 6G). The zooid contracts strongly on fixation causing the macronucleus to become more twisted in protargol-stained specimens (Figs 6H, 7O, P). The micronucleus was not observed.

The lorica is about 130–175 × 45–65 μm in size, nearly cylindrical in shape, with the middle portion slightly convex (Figs 6A–D, 7A–H) and a circular cross-section (Fig. 7K). The diameter of the flared aperture is nearly the same as the widest part of the lorica. Due to the gradual accumulation of ferric deposits on its surface (Kreutz & Foissner, 2006), the lorica changes colour from colourless to dark brownish with age (Fig. 7A–H). Sometimes, the ferric deposits are unevenly distributed, forming a relatively transparent area at the posterior and/or a comparatively light-coloured area at the anterior (Fig. 7A, B). An inconspicuous annular depression exists at the rear end (Figs 6A–D, 7M). The bottom is moderately flat with a slight curve.

Table 6.

Morphometric comparison of different populations of Vaginicola tincta (based on specimens in vivo, measurements in μm)

Populations	Body length	Body width	Diameter of PL	Body protrudes out	Striations	Lorica length	Lorica width	Diameter of aperture	Shape quotient	Shape	Colour of lorica	Data source
V. tincta	157–243	23–35	48–60	1/7–1/3a	fine	130–175	47–61	49–64	2.3–3.2	cylindroid	colourless to brown	original
V. tincta	–	–	–	–	–	–	–	–	–	–	–	Ehrenberg (1830)
V. tincta	–	–	–	–	–	110	–	–	–	cylindroid*	brown*	Ehrenberg (1838)
V. tincta	–	–	–	1/3a	–	85	–	–	–	cylindroid	brown	Kahl (1935)
V. tincta	–	–	–	–	–	–	–	–	–	–	brown	Sommer (1951)
V. tincta	128	–	–	–	–	85	48	–	–	cylindroid	brown	Stiller (1971)
V. tincta	–	–	–	–	–	87–105	28–30	20	–	cylindroid	brown	Vucetich & Escalante (1979)
V. tincta	130	–	–	1/3a	fine	85	40	–	–	cylindroid	brown	Foissner, Berger & Kohmann (1992)
V. tincta	100	–	–	–	–	–	–	–	–	cylindroid	brown	Kreutz & Foissner (2006)
V. tincta	155–178	–	22–28	1/4–1/2a	fine	75–133	30–53	–	2.0–2.7	cylindroid	colourless	Shen & Gu (2016)

Abbreviations: PL, peristomial lip; –, data not available; ^aproportion of body length; ^*inferred from drawing.

Table 7.

Morphometric comparison of Vaginicola tincta with closely related congeners (based on specimens in vivo, measurements in μm)

Species	Body length	Body width	Diameter of PL	Body protrudes out	Feature of PD	Lorica length	Lorica width	Diameter of aperture	Shape quotient	Shape	Aperture	Below aperture	Posterior	Colour of lorica	Data source
V. tincta	100–243	23–35	48–60	1/7–1/2a	–	75–175	30–61	49–64	2.0–3.2	cylindroid	flared	weakly narrowed	annular constriction	colourless to brown	List 1
V. ceratophylli	235–260	–	–	1/3a	–	115–145	65–80	–	1.8–2.0	cylindroid	flared	weakly narrowed	plate	brown	Penard (1922); Shen & Gu (2016)
V. plicata	150–185	–	–	1/4a	–	120–150	35–42	–	2.5–3.5	cylindroid	flared	weakly narrowed	annular constriction	brown	Shen (1980)
V. festivus	182	20–23	29	1/2a	fingerlike projection	100–106	34–46	–	c. 2.5	truncated cone	not flared	widened*	smooth*	colourless	Shen & Gu (2016)

Abbreviations: PD, peristomial disc; –, data not available; ^aproportion of body length; *inferred from the drawing.

List 1: original; Ehrenberg (1830, 1838); Kahl (1935); Stiller (1971); Foissner, Berger & Kohmann (1992); Kreutz & Foissner (2006); Shen & Gu (2016).

Table 8.

List of species, for which SSU rDNA, ITS1-5.8S-ITS2 region and LSU rDNA sequences were newly obtained

Species	SSU rDNA			ITS1-5.8S rDNA-ITS2				LSU rDNA
	Acc. No.	Length of seq.	GC%	Acc. No.	Length of seq.	GC%	Acc. No.	Length of seq.	GC%
P. pusilla	MK184555	1571bp	42.71	MK182738	456 bp	31.58	MK182732	1767 bp	41.88
C. ceramicola	MK184556	1615 bp	44.22	MK182739	445 bp	40.00	MK182733	1766 bp	43.60
V. tincta	MK184557	2083 bp*	41.28	MK182740	473 bp	29.81	MK182734	1766 bp	41.62
T. obconica	–	–	–	MK182741	459 bp	33.33	MK182735	1765 bp	42.04
T. folliculata	–	–	–	MK182742	459 bp	33.33	MK182736	1759 bp	42.07
T. kellicottiana	–	–	–	MK182743	459 bp	33.12	MK182737	1765 bp	41.93

*containing an intron at 3’ end.

Table 9.

Approximately unbiased (AU) test results

Topology constraints	AU value (P)
	Concatenated tree	SSU rDNA tree	ITS1-5.8S rDNA-ITS2 tree
Vaginicola	2.00e–04	2.00e–04	0.012
Vaginicolidae	–	–	0.777

P < 0.05 refute the possibility of monophyly; P > 0.05 do not refute the possibility of monophyly.

The oral ciliature basically accords with that of other sessilids. The haplokinety and polykinety make a 1.75 turn around the peristomial disc and an additional turn on the opposite walls in the infundibulum (Figs 6I, 7R–U). Each infundibular polykinety (P1–3) is three-rowed. The rows in P1 are equally long and diverge slightly at the adstomal ends (Figs 6I, 7R–U). The arrangement of the rows in P2 is distinctive in that they become progressively shorter from inside to outside. Row 1 joins in P1 abstomally and terminates above the other two rows, while row 3 detaches from the other two rows abstomally and terminates at the same level as row 2 (Figs 6I, 7R–U). The abstomal end of P3 is nearly flush with row 1, ending adstomally above the adstomal ends of row 2 and row 3, which are approximately at the same level (Figs 6I, 7R–U). Occasionally, row 2 terminates slightly ahead of row 3. The adstomal end of P3 is slightly above that of P1. The germinal kinety accompanies the haplokinety in the abstomal half of the infundibulum (Figs 6I, 7R, T). There are two epistomial membranes. Epistomial membrane 1 is near the opening of the infundibulum (Figs 6I, 7R, T). Epistomial membrane 2 is ahead of the distal end of the haplokinety (Figs 6I, 7T).

Phylogenetic trees

Trees based on concatenated genes

Both phylogenetic trees generated with ML and BI methods are congruent with respect to major nodes, so only the ML tree is shown here, with nodal support provided from both algorithms (Fig. 8). The loricate vaginicolids are divided into two major clades, representing a distinct branch separated from the aloricate sessilids with relatively strong support (84% ML, 0.96 BI). Cothurnia ceramicola robustly groups with two cothurnids (92% ML, 0.99 BI). Then these cluster together with P. pusilla (69% ML, 1.00 BI) and these then form a clade with V. tincta. The other major clade is fully supported (100% ML, 1.00 BI) and comprises three species of Thuricola and Vaginicola crystallina.

Figure 8.

Phylogenetic tree inferred from concatenated genes (SSU rDNA, ITS1-5.8S rDNA-ITS2, LSU rDNA), revealing the phylogenetic relationship within the family Vaginicolidae. Numbers near nodes denote maximum likelihood (ML) bootstrap values and Bayesian inference (BI) posterior probability, respectively. Asterisks indicate disagreements between ML and BI. The scale bar indicates five substitutions per 100 nucleotide positions. The out group includes Glaucoma chattoni, Ichthyophthirius multifiliis, Tetrahymena corlissi and Tetrahymena pyriformis.

Trees based on SSU rDNA

The entire ML tree and the portion of the BI tree that has a different topology are shown in Fig. 9. The topology of the ML tree is broadly consistent with the concatenated genes tree. Members of Vaginicolidae form an independent branch, albeit with low support (42% ML). Cothurnia and Thuricola are both monophyletic. However, Vaginicola tincta occupies the basal position in the Vaginicolidae clade, although with low support (37% ML). In the BI tree, the vaginicolids form a polytomy.

Figure 9.

Phylogenetic tree inferred from SSU rDNA sequences, revealing the phylogenetic relationship within the family Vaginicolidae. The topology of Vaginicolidae with the BI analyses is shown here because it different from the ML analyses. Numbers near nodes denote maximum likelihood (ML) bootstrap values and Bayesian inference (BI) posterior probability, respectively. Asterisks indicate disagreements between ML and BI. The scale bar indicates five substitutions per 100 nucleotide positions. The out group includes Glaucoma chattoni (X56533), Ichthyophthirius multifiliis (U17354), Tetrahymena corlissi (U17356) and Tetrahymena pyriformis (EF070254).

Trees based on ITS1-5.8S-ITS2 region

The topologies of the ML and BI trees are almost identical, therefore, only the ML tree is shown (Fig. 10). Unlike the concatenated and SSU rDNA trees, Vaginicolidae is not monophyletic. The AU test could not statistically significantly reject (P = 0.777 > 0.05) the monophyly of Vaginicolidae. Vaginicola tincta clusters with Pyxicola pusilla (68% ML, 0.99 BI) that together group with a branch uniting Astylozoidae, Vorticellidae and Ophrydiidae. Cothurnia ceramicola clusters with a group that includes three thuricolas and V. crystallina (51% ML, 0.91 BI).

Figure 10.

Phylogenetic tree inferred from ITS1-5.8S-ITS2 region sequences, revealing the phylogenetic relationship within the family Vaginicolidae. Numbers near nodes denote maximum likelihood (ML) bootstrap values and Bayesian inference (BI) posterior probability, respectively. Asterisks indicate disagreements between ML and BI. The scale bar indicates ten substitutions per 100 nucleotide positions. The out group includes Ichthyophthirius multifiliis (DQ270016) and Tetrahymena pyriformis (KX832097).

Trees based on LSU rDNA

The topologies of the ML and BI trees are almost identical (Fig. 11). All six species of Vaginicolidae sequenced in this work form a branch (75% ML, 1.00 BI) and are separated from the aloricate species, which is consistent with the concatenated tree. Pyxicola pusilla, V. pusilla and C.ceramicola group together, forming a sister clade to the Thuricola clade (77% ML, 1.00 BI). However, P. pusilla clusters with V. tincta with low support (53%) in the ML tree, whereas it clusters with C. ceramicola with moderate support (0.77) in the BI tree.

Figure 11.

Phylogenetic trees inferred from LSU rDNA sequences, revealing the phylogenetic relationship within the family Vaginicolidae. Numbers near nodes denote bootstrap values. The scale bar indicates five substitutions per 100 nucleotide positions. The out group includes Ichthyophthirius multifiliis (EU185635) and Tetrahymena pyriformis (X54004).

DISCUSSION

Comments on Pyxicola pusilla
(Fig. 12; Tables 2, 3)

This species was first described as Cothurnia pusilla (Wrześniowski, 1866). Kent (1882) established the genus Pyxicola for the operculated loricate peritrichs and reassigned this species as P. pusilla. This species has been described multiple times under different names and the descriptions for this form vary, especially in respect to the annular ridges of the lorica (inconspicuous to conspicuous) and the length of the stalk (from 3–60 μm; Kent, 1882; Stokes, 1895; Finley & Bacon, 1965; Nusch, 1970; Stiller, 1971; Trueba, 1978; Shen & Gu, 2016). Even so, there are some common features in these reports. For example, the lorica shape quotient (i.e. the length divided by the width) is about 2, the aperture is obliquely truncated, the colour of the lorica changes with age, but is often brown, and the cell protrudes just beyond the aperture. Our population corresponds with all these common features. Accordingly, we believe that the Ningbo isolate is a population of P. pusilla.

Before the genus Pyxicola was created, these operculated loricate peritrichs were treated as cothurnids. Thus, some populations were placed in the genus Cothurnia at one time or other. The isolate reported by Hutton (1878), namely Cothurnia furcifer, matches perfectly with Wrześniowski’s form (Fig. 12B) and thus can be considered as P. pusilla. Pyxicola affinis was first described by Kent (1882) and subsequently cited by Blochmann (1886) and Hickson (1903) as Cothurnia affinis and shows a larger lorica (80 × 40 μm vs. 46 × 22 μm). It has a longer stalk than Wrześniowski’s form (27–40 μm vs. 3 μm) (Fig. 12C). Pachytrocha cothurnoides, reported by Kent (1882) and also as Cothurnia cothurnoides by Blockmann (1886), corresponds in all points with Pyxicola pusilla except that the indurate operculum of Pyxicola is replaced by a fleshy pad for this isolate (Blochmann, 1886) (Fig. 12D). We agree with Kahl (1935) and Trueba (1978) who suggested that this isolate was very likely a mutilated Pyxicola pusilla with a fallen operculum. Stokes (1895) reported a new genus and species, Caulicola valvata, largely because the operculum adheres to the aperture rather than the zooid, but this was probably an optical illusion (which in fact also occurred in our observations; Fig. 12E). So, we agree with Trueba’s view in treating C. valvata as a junior synonym of P. pusilla. Trueba (1978) considered that P. carteri sensu Sommer, 1951 is a population of P. pusilla. We follow the proposal, because the shape quotient of its lorica is 2 and the contractile vacuole is situated at the same position as in our population (from the drawing; Fig. 12F). Pyxicola eforiana reported by Tucolesco (1962) from brackish water is also a synonym of P. pusilla. The lorica of this form is larger than in the original population (75–80 vs. 46) and our population (75–80 vs. 60–67), and a larger proportion of the body projects out of the lorica (one-third vs. significantly less than one-third in the original and our population). Furthermore, it corresponds well with both the original description and present population (Fig. 12G). Nusch (1970) described a form under the name of Pyxicola carteri forma constricta and considered that Pyxicola nolandi Finley & Bacon, 1965 is a synonym. However, the lorica of this form is plumper than in the original population with a shape quotient about 2. Thus, we agree with Trueba (1978) in regarding this as a population of P. pusilla. In addition to the above forms, three other populations of P. pusilla were described (Kent, 1882; Trueba, 1978; Shen & Gu, 2016). These are similar to each other and correspond with the original description. Unfortunately, some previous descriptions are cursory and the photomicrographs of P. pusilla are only available from Trueba’s paper. It is possible, therefore, that there are some mistakes in the above synonymy list. Pyxicola annulata described by Leidy (1882) was divided into two forms by Trueba (1978) and considered as synonyms of P. pusilla and P. carteri, respectively. However, in this case Trueba’s opinion is contestable: although the form with comparatively conspicuous annular ridges was treated as P. pusilla, one-fifth of the body protrudes outside the lorica, with a small contractile vacuole just below the peristomial lip (from Fig. 8 from pl. 2 in Leidy, 1882) and its lorica shape quotient is 2.5–3.0; all these characters are divergent with P. pusilla. Thus, we consider P. annulata to be a synonym of P. carteri.

Figure 12.

Different isolates of Pyxicola pusilla (A–H) and its similar species (I–K), different isolates under the names of Cothurnia ceramicola (L–R) and its similar species (S–U), different isolates under the names of Vaginicola tincta (V–Z) and its similar species (Z1–Z4). A, Cothurnia pusilla sensu Wrześniowski, 1866. B, Cothurnia furcifer sensu Hutton, 1878. C, Pyxicola affinis sensu Kent, 1882. D, Pachytrocha cothurnoides sensu Kent, 1882. E, Caulicola valvata sensu Stokes, 1895. F, Pyxicola carteri sensu Sommer, 1951. G, Pyxicola eforiana sensu Tucolesco, 1962. H, after Shen & Gu, 2016. I, Pyxicola carteri (from Kent, 1882). J, Pyxicola operculigera (from Kent, 1869). K, Pyxicola psammata (from Hadži, 1940). L, after Kahl, 1933. M, after Precht, 1935. N, after Felinska, 1965. O, after Küsters, 1974. P, after Song, 1992. Q, after Sun et al., 2009. R, after Shen & Gu, 2016. S, Cothurnia curvula (from Entz, 1884). T, Cothurnia harpactici (from Kahl, 1933). U, Cothurnia fibripes (from Kahl, 1933). V, Vaginicola tincta after Ehrenberg, 1838. W, after Kahl, 1935. X, after Stiller, 1971. Y, after Vucetich & Escalante, 1979. Z, after Shen & Gu, 2016. Z1, Vaginicola ceratophylli (from Penard, 1922). Z2, Vaginicola lagena (from Vucetich & Escalante, 1979). Z3, Vaginicola plicata (from Shen, 1980). Z4, Vaginicola festivus (from Shen & Gu, 2016). Scale bars = 30 μm.

The other three species in the P. pusilla complex, namely Pyxicola carteri Kent, 1882, Pyxicola operculigera (Kent, 1869) Kent, 1882 and Pyxicola psammata Hadži, 1940, should be compared with P. pusilla, considering, in particular, that P. pusilla possesses an urceolate lorica smaller than 100 μm with an annulated wall and a habitat in freshwater or brackish water. Pyxicola carteri differs from P. pusilla in having a slenderer lorica (shape quotient about 2.5 vs. 2.0) with a longer neck and a body that protrudes up to one-third of its length outside the lorica (vs. just the peristomial lip beyond aperture; Fig. 12I) (Kent, 1882; Trueba, 1978). Pyxicola operculigera possesses a long stalk, usually longer than the lorica and even up to 150 μm, which is much longer than the 2–60 μm long stalk of P. pusilla. Moreover, its lorica is never clearly annulated, whereas it is commonly and clearly annulated in P. pusilla (Fig. 12J) (Kent, 1882; Trueba, 1978). The lorica of P. psammata is much slenderer than that of P. pusilla (shape quotient from 2.4–3.3, average 2.9 vs. about 2.0) and the operculum of the former is smaller than that of the latter (7 μm across vs. 12–18 μm across; Fig. 12K) (Hadži, 1940).

Comments on Cothurnia ceramicola
(Fig. 12; Tables 4, 5)

The lorica of our population is much larger than Kahl’s (95–105 vs. 60) and with relatively distinct annular ridges compared to the slight ridges in Kahl’s form (from the drawing). We think these are population-dependent differences, because unstriated, partly striated and completely striated loricas were found in a single species in another vaginicolid genus, Platycola (Warren & Carey, 1983). Our population matches Kahl’s form with respect to the marine habitat, having a lorica that is cylindroid in shape, slightly narrower towards the top and 1/4 to 1/3 of the body projects outside the lorica (Kahl, 1933). It also shares an obviously striated endostyle, as well as a clearly striated and truncated-cone shaped mesostyle (Fig. 12L). Therefore, we identify it as Cothurnia ceramicola.

Precht (1935) described several populations, also collected from Kiel, that match perfectly with Kahl’s population and they are clearly conspecific, except for one form that was found from Gonothyraea loveni (Allman, 1859) where Precht put a question mark after the specific name, because the double-layered part of the lorica was considerably larger and the body protruded less (Fig. 12M). We think these are population-dependent differences. Felinska (1965) reported a form under C. ceramicola, but this identification is suspect, because her population differs distinctly from Kahl’s form by the narrowed aperture, the larger double-layered proportion (more than 1/3 vs. less than 1/4) and the much larger span of the annular ridges (Fig. 12N). The lorica of Küsters’ population is longer than Kahl’s (67–102 μm vs. 60 μm) and our population is more similar to Küsters’ population than to that of Kahl in the lorica size (Fig. 12O) (Küsters, 1974). Song (1992) described a form of C. ceramicola isolated from the surface of Penaeus chinensis (Osbeck, 1765), but the identification is doubtful: its lorica is clearly elliptical (vs. invariably cylindroid), has a smooth pellicle (vs. clearly striated pellicle) and a stalk of about 10 μm long (vs. 3–9; Song, 1992). Sun et al. (2009) recorded a population of C. ceramicola in the Yellow Sea, which corresponds closely with our population. They are conspecific, the only difference being that the P3 is composed of two rows of kinetosomes, although this is probably an illusion caused by the overlapped row 2 and row 3 (Fig. 12Q). Shen & Gu (2016) depicted a species as C. ceramicola, but this was probably a different species, because of its occurrence in freshwater (Fig. 12R).

Cothurnia ceramicola is mostly characterized by its marine habitat, annulated wall and combination of endostyle–mesostyle-stalk. Based on these characters, three closely similar species should be compared, i.e. C. curvula Entz, 1884, C. harpactici Khal, 1933 and C. fibripes Kahl, 1933. Cothurnia curvula can be separated from C. ceramicola by the curved anterior portion of the lorica (vs. straight), the obviously narrowed aperture (vs. equal to the lorica width) and the small proportion of the body that projects through the aperture (just the peristomial lip vs. 1/6–1/3 of the body length; Fig. 12S) (Entz, 1884). Cothurnia harpactici is close in size to C. ceramicola, but its stalk is much longer (15 μm vs. 3–5 μm; Fig. 12T) (Kahl, 1933; Precht, 1935; Warren & Paynter, 1991). Cothurnia fibripes is different from C. ceramicola in its distinctly shorter body length (60 μm vs. 95–100 μm) (Kahl, 1935). Furthermore, its lorica is plumper than C. ceramicola (shape quotient 2.0 vs. 2.9–3.3) with a much narrower aperture (15 μm vs. 30–34 μm; Fig. 12U).

Comments on Vaginicola tincta
(Fig. 12; Tables 6, 7)

This species is commonly found in freshwater habitats and was originally reported by Ehrenberg (1830) and redescribed in 1838 by himself (Fig. 12V) (Ehrenberg, 1830, 1838). Kahl (1935) made a revision of Vaginicola with a fine description of V. tincta (Fig. 12W). Later, Sommer (1951) and Stiller (1971) (Fig. 12X) reported two populations of this form. Vucetich & Escalante (1979) described a form that has a much slimmer lorica than V. tincta (Fig. 12Y). However, the form they described as V. lagena Kahl, 1935 fits better with V. tincta (Fig. 12Z2). Foissner et al. (1992) made an overview of V. tincta and provided an elaborate diagnosis. Recently, Shen & Gu (2016) reported that this species has been found many times in China (Fig. 12Z). Our population corresponds well with the original description and Foissner’s diagnosis (Foissner et al., 1992), thus we think it is a population of V. tincta.

This species is characterized by its relatively large body and large, cylindrical lorica. There are three very similar congeners, namely Vaginicola ceratophylli (Penard, 1922) Kahl, 1935, Vaginicola plicataShen, 1980 and Vaginicola festivus Li, 2016. They all have a nearly cylindrical lorica with a flared aperture, weakly narrowed portion below the aperture and brown colour. Furthermore, all three species have a similar body size and a similar lorica size. Vaginicola ceratophylli differs from V. tincta by the plate on which the lorica rests (Fig. 12Z1), although it may be that they are conspecific if the plate is an optical illusion caused by an adhering layer (Penard, 1922; Kahl, 1935; Warren & Paynter, 1991). Vaginicola plicata is distinguished from V. tincta by the wrinkled posterior portion of the lorica and the flared aperture. These characters are present in our population and perhaps also in other populations of V. tincta, but were overlooked by the researchers (Fig. 12Z3) (Shen, 1980). So, V. ceratophylli and V. plicata are probably synonymous. Vaginicola festivus possesses a finger-like projection at the border of the peristomial disc that is absent in V. tincta (Fig. 12Z4) (Shen & Gu, 2016).

Phylogenetic analyses

Among the phylogenetic trees of Peritrichia, the loricate sessilids tend to form a separate clade from aloricate sessilids, although the family Vaginicolidae is polyphyletic in the BI tree based on the SSU rDNA and the trees based on the ITS1-5.8S-ITS2 region. Although they are divided into two clades in the ML tree based on ITS1-5.8S rDNA-ITS2 tree, the possibility that they cluster together is not rejected by the topology test (AU test, P = 0.777 > 0.05).

Focusing on the loricate species, the relationship between them is basically in accordance with their morphology (Fig. 13). These species can be divided into four groups based on morphology: Thuricola and Vaginicola attach to the substrate directly, although Vaginicola lacks a stalk and Thuricola possesses an internal stalk and a closure apparatus (valve); Cothurnia and Pyxicola attach to the substrate via a stalk outside the lorica, but Cothurnia lacks the operculum; while Pyxicola has an operculum attached to the border of the peristomial lip (Kahl, 1935; Trueba, 1978, 1980; Clamp, 1991; Warren & Paynter, 1991; Lu et al. 2018).

Figure 13.

Morphological differences and correlations of the six sequenced vaginicolids in this work.

Species of Cothurnia cluster in a clade with good support in SSU rDNA and concatenated trees (Figs 8, 9), indicating that this genus is probably monophyletic. Pyxicola pusilla is relatively closely related to Cothurnia in most trees. This is supported by their morphology in that they have a lorica that is attached to a substrate via a stalk (Fig. 13). In contrast, Thuricola lacks a stalk to attach to the substrate, a feature clearly separating it from Pyxicola and Cothurnia. Three thuricolas are grouped in a clade in all trees with low to full support, showing the monophyly of this genus. Although the length of Thuricola branches is short in all trees, their morphologies show consistent and distinct differences. For example, there is a single valve in the lorica of T. obconica, compared with two spineless valves in the lorica of T. folliculata and one bigger spiniferous valve together with a smaller valve in the lorica of T. kellicottiana (Fig. 13).

Vaginicola crystallina (AF401521) clusters with Thuricola in all trees with full support (Figs 8–11). The morphological feature that the lorica is directly attached to the substrate without a stalk supports this cluster (Fig. 13). However, V. tincta is separated from V. crystallina in all trees, exhibiting a 140 nucleotide difference in the SSU rDNA sequence. This indicates that the genus Vaginicola is polyphyletic. In addition, the topology testing results strongly reject the grouping of them (AU test, P = 2e-004 in concatenated tree, P = 2e-004 in SSU rDNA tree, P = 0.012 in ITS1-5.8S rDNA-ITS2 tree). We cannot find morphological support for the separation, because the morphological data of the sequenced V. crystallina is absent. Vaginicola tincta is placed at a basal position relative to Pyxicola and Cothurnia in the concatenated genes trees and to other genera in Vaginicolidae in the SSU rDNA tree with ML analyses (Figs 8, 9). This suggests that the stalkless lorica may represent the ancestral state and the stalked Pyxicola and Cothurnia are likely to have evolved from stalkless ancestors.

In conclusion, the loricate sessilids diverged from the aloricate sessilids and probably represent an independent lineage in the sessilids. The genera Cothurnia and Pyxicola are closely related, which is supported by both the molecular data and morphology. The genera Cothurnia and Thuricola are both monophyletic, whereas Vaginicola seems to be polyphyletic. The stalkless vaginicolids may be the ancestral group of Vaginicolidae. Additionally, morphological and molecular data for the loricate species are still seriously insufficient; therefore, the systematic relationships within the loricate group remain uncertain, pending a re-evaluation following the acquisition of additional data.

ACKNOWLEDGEMENTS

This work is supported by the National Natural Science Foundation of China (project No. 41576134, 31772431, 31772413), the Fundamental Research Funds for the Central Universities (project No. 201762017) and the Deanship of Scientific Research at King Saud University (Research Group Project No. RGP-083). We especially thank Dr Feng Gao (OUC) and Ms Tengteng Zhang (OUC) for their valuable help with the phylogenetic analysis.