Finding of Pebrilla paguri (Ciliata, Folliculinidae, Heterotricha) epibiont inside the shells of two hermit crab species, Calcinus tubularis and Clibanarius erythropus (Decapoda, Anomura, Diogenidae)

Abstract

Ever since its discovery, Pebrilla paguri was considered to be an exclusive epibiont of hermit crabs. In fact, up to now, it has never been recorded on substrata other than the hermit crab exoskeleton. The finding of its loricae on non‐consecutive exuviae, released by the same specimen of Calcinus tubularis, suggested that the motile stage may transform into the sessile form and attach itself to the shell occupied by the crab. All the shells occupied by the C. tubularis and Clibanarius erythropus specimens showing the ciliate on the exoskeleton were identified to genus level; for each shell genus a map of the inner morphology was drawn in order to record the position and number of the loricae. The present research showed that P. paguri is not an exclusive hermit crab epibiont: it also attaches itself to shells, and particularly to the three largest whorls, corresponding to the portion occupied by the crab body. The presence of P. paguri in the shell may be determined by the preference for substrata characterized by a particular water stream or light levels. The results obtained show that the P. paguri substrata choice is more similar to that of other folliculinids, and more varied than previously reported.

Keywords:

• Pebrilla paguri
• lorica
• shell
• generalistic

Introduction

Folliculinids (Faurè‐Fremiet 1936; Hadži 1951; Lynn & Small, 2002) are a group of generally marine, suspension‐feeding heterotrich ciliates (Hadži 1951), commonly observed in the littoral and sublittoral zones (Mulish & Patterson 1983). Their life‐cycle consists of a free swimming phase, alternating with a sessile one (Faurè‐Fremiet 1936). During the latter, the ciliate is protected by a lorica. Ciliates attach themselves to a great variety of inert or vegetal substrates (algae, mangrove roots, wood) (Martínez‐Murillo & Aladro‐Lubel 1994, 1996, 1999; Aladro‐Lubel & Martínez‐Murillo 1999), as well as to animals (molluscs, polychaetes, crustaceans) (Andrews 1944; Sprague & Couch 1971; Grimes 1978; Martínez‐Murillo & Aladro‐Lubel 1994; Baccarani & Pessani 1997; Fernández‐Leborans & Herrero Cordoba 1997; Mayén‐Estrada & Aladro‐Lubel 1998, 2000, 2001; Fernández‐Leborans & Sorbe 1999, 2003; Fernández‐Leborans & Tato‐Porto 2000, 2002a, 2002b; Pessani et al. 2001, Abstract in 62° Congr. Naz. U.Z.I.: 58; Basile et al. 2003; Culotta et al. 2003; Fernández‐Leborans 2003a, 2003b, 2004).

Pebrilla paguri (Giard, 1888) is a typical commensal on hermit crabs (Fernández‐Leborans & Herrero Córdoba 1997), observed, so far, only adhering to the exoskeleton of: Pagurus bernhardus (Faurè‐Fremiet 1936; Grimes 1978); Clibanarius erythropus (Faurè‐Fremiet 1936; Baccarani & Pessani 1997; Pessani et al. 2001, Abstract in 62° Congr. Naz. U.Z.I.: 58; Basile et al. 2004); P. prideaux and P. excavatus (Fernández‐Leborans & Herrero Córdoba 1997); P. anachoretus, P. cuanensis, Calcinus tubularis and Paguristes eremita (Basile et al. 2004).

The finding of P. paguri loricae on the first and third (and not on the second) exuviae, released in the laboratory by the same C. tubularis specimen, suggested that the swimming phase of the ciliate may transform into the sessile one, which attaches itself to the shell occupied by the crab and lives in it. Consequently, we analysed the inside of the shells occupied by C. tubularis and C. erythropus specimens, on whose exoskeleton we had previously observed the presence of P. paguri loricae.

Materials and methods

In order to detect the presence of P. paguri loricae, all the shells occupied by the crabs (C. tubularis and C. erythropus) on whose exoskeleton the ciliate had been observed, were analysed. These crabs, belonging to the Authors' collections, had never been separated from the occupied shells and, at their death, were fixed together with their shelter in a solution of formalin 4% and seawater. In spring, summer and autumn during the years 1997–1999, the crabs were manually collected on rocky bottoms of the Ligurian Sea between 0.5 and 7 m.

The occupied shells were identified at genus rank; their total length and maximum width were measured using a calliper. Further smashing of a few of them showed empty loricae of P. paguri attached to the internal walls. The remaining shells were sectioned into two perfectly equal parts (figure 1) using a circular saw 4 cm in diameter and 0.3 mm thick. Of each shell genus, we drew two maps to depict the inner morphology; with this purpose in view, we also used X‐ray films. The position and number of loricae were reported on the maps.

Figure 1 Cerithium vulgatum shell sectioned into two parts along the columella axis and showing P. paguri loricae inside (indicated by the arrow); magnification 4.1×.

Observations were made using a stereomicroscope (Wild M5A; eye‐piece 10 or 20×, objective 6–12–25–50×) with micrometer eye‐piece.

Data were statistically analysed using the Kruskal–Wallis (1952) test, the Friedman (1937) test, and linear regressions.

Results

Out of 583 analysed crabs, only 81 (13.9%) (45/177 C. tubularis and 36/406 C. erythropus specimens) showed P. paguri on the exoskeleton. The crabs were the same used for the research by Basile et al. (2004). The shells of all these 81 crabs were examined and 36 (44.4%) of them had P. paguri loricae on their inside walls (Tables I and II): the loricae had the typical specific morphology, already described by Faurè‐Fremiet (1936), Fernández‐Leborans & Herrero Córdoba (1997) and Basile et al. (2004). All the loricae observed on the shells showed no live ciliate inside but had the same morphology and biometry as the one attached to the crab and showing the ciliate, identified at species level by Basile et al. (2004).

Table I. Number of P. paguri loricae on crab body, CZ (Crab Zone), NCZ (No Crab Zone), and CO (Columella) for each C. erythropus specimen.

Crab body	Shell species	CZ	NCZ	CO
8	M. turbinata	4	0	0
5	M. turbinata	6	0	0
45	M. turbinata	4	0	1
2	M. turbinata	4	0	0
11	M. turbinata	2	1	0
1	M. turbinata	1	1	0
5	M. turbinata	18	0	0
3	M. turbinata	11	0	0
6	C. vulgatum	0	8	0
3	C. vulgatum	10	0	8
2	C. vulgatum	8	0	0
5	M. turbinata	2	0	1
18	M. turbinata	16	0	2
21	M. turbinata	19	0	3
29	C. vulgatum	11	11	4
21	H. trunculus	15	3	0
7	M. turbinata	3	0	0

Table II. Number of P. paguri loricae on crab body, CZ (Crab Zone), NCZ (No Crab Zone), and CO (Columella) for each C. tubularis specimen.

Crab body	Shell species	CZ	NCZ	CO
14	C. vulgatum	94	20	18
3	C. vulgatum	12	0	15
2	C. vulgatum	4	7	3
1	M. turbinata	441	0	60
3	C. vulgatum	8	0	0
4	C. vulgatum	3	44	17
5	H. trunculus	14	12	8
11	B. corneum	25	13	8
2	C. vulgatum	1	1	0
1	C. vulgatum	38	1	0
1	M. turbinata	3	0	3
11	C. vulgatum	18	21	0
29	C. vulgatum	49	0	3
7	C. vulgatum	10	0	0
2	C. vulgatum	245	62	93
3	C. vulgatum	81	13	5
1	H. trunculus	1	1	1
3	C. vulgatum	1	0	4
6	C. vulgatum	1	8	6

Loricae were not evenly distributed. In each shell species, the majority of them (around 70%) were attached to the three largest whorls of the shell. The internal morphology of the shell was related to the position occupied by the crab. Three zones were recognized: the three largest whorls, corresponding to the zone occupied by the crab (Crab Zone, CZ), the remaining whorls (No Crab Zone, NCZ), and the Columella (CO). To verify the possible preferential distribution of the ciliate, P. paguri loricae adhering to each zone were counted. The number of ciliates was also correlated to the shell species and therefore to a shelter shape (Tables I, II and III).

Table III. Number of the shells occupied by the two hermit crab species, and mean number of P. paguri loricae (Pp) attached to the different shell zones (CZ, Crab Zone; NCZ, No Crab Zone; CO, Columella). B, Buccinulum corneum; C, Cerithium vulgatum; H, Hexaplex trunculus; M, Monodonta turbinata.

Shells	Pp loricae				Shell zone (Pp)
	C. erythropus		C. tubularis		CZ	NCZ	CO	Tot.
	No.	Mean (SD)	No.	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
B	0	0	1	46	25	13	8	46
C	4	15.0 (8.7)	14	64.7 (103.7)	33 (59.7)	10.9 (17.1)	9.8 (21.6)	53.7 (93.2)
H	1	18.0	2	18.5 (21.9)	10 (7.8)	5.3 (5.9)	3 (4.4)	18.3 (15.5)
M	12	8.2 (7.1)	2	253.5 (350.0)	38.1 (115.1)	0.1 (0.4)	5 (15.9)	43.3 (131.9)
Total	17		19

No statistically significant difference was found in the number of loricae according to the shell species (Kruskal–Wallis test: H = 7.284; df = 3, NS), whereas a different distribution was recorded according to the shell zone: the number of loricae counted on CZ was significantly higher than those on NCZ and CO (Friedman test: F = 6.82, df = 2, P<0.05; F = 18.62, df = 2, P<0.05), independently of the morphological feature of the shell (trochoid shape, such as Monodonta turbinata, or conical shape with more or less wide base, such as Buccinulum corneum, Hexaplex trunculus and Cerithium vulgatum). Moreover, the total number of loricae was significantly higher inside the shells occupied by C. tubularis specimens than in those by C. erythropus (Kruskal–Wallis test: H = 7.027, df = 1, P<0.05).

When enough shells were available, i.e. respectively for C. erythropus specimens in M. turbinata and C. tubularis in C. vulgatum (Tables I and II), no statistically significant linear correlation was found between the number of loricae attached to CZ and to the crab body (C. erythropus: n = 12, R ² = 0.023, NS; C. tubularis: n = 14, R ² = 0.001, NS).

Discussion

The morphology of the loricae adhering to the inside of shells here analysed is referable to that observed for P. paguri by Faurè‐Fremiet (1936), Fernández‐Leborans & Herrero Córdoba (1997) and Basile et al. (2004).

In each group of protozoan epibionts, there are generalistic species and specific species. Among the latter, for example, Vorticella incostans on the cladoceran Macrothrix hirsuticornis (Fernández‐Leborans & Tato‐Porto 2000a) and Acineta karamani on the decapod Atyaephyra desmarestii (Fernández‐Leborans & Tato‐Porto 2000b) are known. Regarding P. paguri, which was never before reported attached to substrata other than the hermit crab, this research shows for the first time that the species is more generalistic than was noted by Faurè‐Fremiet (1936), Fernández‐Leborans & Herrero Córdoba (1997) and Basile et al. (2004). In fact, loricae also adhere to the inside walls of the shells occupied by the crabs. The use of substrata such as gastropod shells, bivalves and polychaete tubes is, instead, a well‐known phenomenon for other folliculinids (Faurè‐Fremiet 1936; Andrews 1944; Mulisch et al. 1986).

A low percentage (13.9%) of crab specimens showed P. paguri loricae on the exoskeleton and less than half of them, corresponding to 6.1% of the original sample (36/583), had the ciliate inside the shell. Hence, to have a larger number of cases, allowing an exhaustive statistical analysis, it would be necessary to collect thousands of crabs, which we did not consider a positive factor from the point of view of the Ligurian Sea populations. Therefore, the absence of published data regarding the use of gastropod shells as a substratum seems to be due to:

	the very low percentage of crabs with P. paguri attached either to the exoskeleton or the inside of the shell; and
	the fact that P. paguri adheres to shells preferably when the ciliate density on the crab exoskeleton is high.

On the other hand, as 44% of the 81 considered shells showed the epibiont, it may be assumed that the shell seems to be quite a good substratum to which P. paguri attaches.

The presence of P. paguri in the shell may be determined by the preference for substrata characterized by particular water stream or phototropism conditions, which, among the abiotic factors, play the main role in ciliate substrata choice (Mulisch et al. 1986; Fernández‐Leborans & Herrero Córdoba 1997). Inside the shell, there is lack of light and a weak water‐stream. In particular, the latter phenomenon promotes bacterial proliferation and concentration of pieces of food used by P. paguri (Andrews & Reinhard 1943).

Moreover, as there is no statistically significant difference in the number of ciliates attaching to different shell genera, and therefore to conical or trochoid shells, we can assume that, if the occupying crab shows P. paguri on its body, the ciliate may equally well adhere to the inside of the shell regardless of its shape.

Regarding the preferential loricae localization:

	the low number of ciliates on the NCZ can be explained by the concentration of debris, frequent in this shell zone. Deposit accumulation is known to reduce folliculinid presence (Mulisch et al. 1986);
	the scarcity of loricae on the columella is referable to the fact that the crab body rubs against this portion of the shell, so that the settled loricae are probably removed;
	the large number of folliculinids on the CZ can be due to the fact that it corresponds to the portion of the shell occupied by the body when the crab is withdrawn, therefore showing the light and water stream conditions which P. paguri is already used to on the crab body.

In agreement with what was reported for the folliculinids attached to the crab body (Basile et al. 2004), the number of ciliates is higher on the shells occupied by C. tubularis than on those occupied by C. erythropus. This phenomenon is explained by the ethology of C. tubularis, which is characterized by a sedentary nature and scarce tendency to shell exchange (Damiano et al. 1998). As a consequence, a greater number of successive generations of ciliates can settle on the shells occupied by C. tubularis than on those by C. erythropus, a species which performs numerous shell exchanges.

In conclusion, as opposed to what was previously known, P. paguri is not an exclusive hermit crab epibiont, but is more generalistic than what was known, as it also attaches itself to shells occupied by crabs. The presence of P. paguri in the shell may be determined by the preference for substrata characterized by a particular water stream or light level. The results obtained here allow us to enlarge the knowledge of P. paguri biology and substratum choice, which seems to be very similar to that of other folliculinids (commonly attached to non‐living secreted surfaces) than previously reported.

Acknowledgements

The authors wish to thank Dr Massimo Meregalli for his valuable advice and for improving the English version.