Collection

A total of 120 Onchidella celtica specimens were collected during a year (October 2020-2021) at Praia de Ancoradoiro (42º 20' 1.520" N; 008º 49' 36.820" W), a moderately exposed rocky shore in southern Galicia (NW Iberian Peninsula). Sampling was carried out monthly via reef-walking at low tide. Animals were taken by hand when grazing over the rock surface and non-randomly selected to represent monthly variation in size and colour (GOULDING et al., 2018a). They were measured in situ and placed separately in seawater. Some specimens were photographed in their habitat while most of them were photographed in the laboratory. For each specimen, external anatomy was described in vivo avoiding the loss of information of those characters that may be altered during fixation (URGORRI et al., 2021).

Anaesthesia, fixation and preservation

Anaesthesia, fixation and preservation of specimens were carried out following URGORRI (1981) and URGORRI et al. (2021). First, animals were anaesthetised with 7% MgCl2 diluted in equal parts of freshwater and seawater. After anaesthesia, fixation was performed gradually by dripping Bouin’s liquid (fixative liquid) with a Pasteur pipette. The addition of the fixative liquid was observed under the stereomicroscope preventing any sudden reactions of the specimen. After fixation, animals were embedded in Bouin’s liquid for several days, and then preserved in 70% EtOH.

41 individuals were anaesthetized as described above and the remaining 79 specimens were directly fixed and preserved on 100% EtOH. Each specimen was labelled with a code including location (ANC, "Ancoradoiro"), site (1, 2, 3) and date of collection (dd/mm/yyyy) and specimen number (1 to 10). All this material was deposited on the malacological collection of the Marine Biological Station of A Graña (Ferrol, Spain).

Micro-CT procedures

Study of the anatomy of the specimens was done by means of micro-CT. Two mature adults (ANC110092103 and ANC110092104) were examined to ensure that the reproductive system was fully developed; both specimens were preserved in 70% EtOH, dehydrated in 80%, 90% and 96% EtOH baths each of 12 h.

Two different treatments were further applied before scanning depending on the specimen to determine the presence of subdermal spicules (see below). ANC110092104 was stained with 1% iodine in 96% EtOH for 3 days, then dehydrated with hexamethyldisilazane (HMDS) for 2 h and air dried overnight (ALBA-TERCEDOR & SÁNCHEZ-TOCINO, 2011; FAULWETTER et al., 2013; CANDÁS et al., 2014). Iodine staining was first dispensed with for ANC110092103. Thus, the animal was directly transferred from the 96% EtOH to HMDS for the same time as above. Once scanned, the specimen was stained with iodine, dehydrated like ANC110092104 and scanned again.

The specimens were scanned on a Skyscan 1172 microtomograph (Bruker, Belgium) using the following parameters: 55 kv, 165 μA, no filter and 360º sample rotation. Pixel size were 3.66 μm for ANC110092104 and 4.54 μm and 4.06 μm for ANC110092103 stained and no-stained, respectively.

X-rays projection images were reconstructed with the NRecon software (Bruker, Belgium) obtaining 2D sections that were posteriorly cleaned with the CTAn software (Bruker, Belgium). Frontal, sagittal and transverse sections were visualized using DataViewer software, while three dimensional models were obtained with the CTVox software (Bruker, Belgium). 3D images were reconstructed using the Amira® v.5.4.4 software (Visage Imaging, Richmond, Australia).

External Anatomy

Dorsal notum of living animals generally light or dark grey, occasionally olive green (Figs. 1A,B). Juveniles and immature adults (< 3 mm) whitish, sometimes translucent (Fig. 1C). Pedal sole and hyponotum white or creamy, always in contrast with body dorsal part (Figs. 1C,D). Body oval or round, oblong when crawling. Size between 1-25 mm, reaching largest values in late summer. A slightly reduction in length observed after fixation.

Fig. 1. Onchidella celtica. Habitus and external appearance in vivo. A: Dorsal view of a dark grey adult ANC207102102. B: Dorsal view of a green olive adult ANC202032101. C: Dorsal view of a whitish juvenile specimen ANC329032106. D: Ventral view of a light grey adult ANC217012102. E: Ventral view of a green olive adult ANC126022102. 

Figure 2.Habitus and external appearance of the scanned specimens. A: Dorsal view of ANC110092104. B: Dorsal view of ANC110092103 stained. C: Dorsal view of ANC110092103 unstained. D: Ventral view of ANC110092104. E: Detail of the cephalic region of ANC110092104. F: Ventral view of ANC110092103 unstained. G: Detail of the caudal region of ANC110092103 stained 

Dorsal surface covered of hemispherical tubercles (=notal/dorsal papillae, pustules, warts) in variable density (Figs. 2A-C) Tubercles more abundant and globular on central notum than on perinotal border, but number fairly constant along dorsal surface. Juveniles and small adults with fewer tubercles showing a smoother notum. Central and anterior notal tubercles bigger than perinotal and posterior ones. Tubercles in living animals switching from hemispherical to conical or pointed when subjected to wind, direct sunlight or dryness. Tubercles in preserved specimens flattened sometimes. Dorsal gills and dorsal eyes not present.

Notum laterals uneven. Larger tubercles (=marginal papillae) surrounded by small ones covering perinotal glands. Marginal papillae lighter in colour than on dorsal surface, similar to hyponotum. Posterior papillae generally bigger than anterior ones. Marginal papillae rounded or pyramid-shaped (Fig. 2C), sometimes indistinguishable in juveniles.

Hyponotal border reduced in juveniles but wider than pedal sole in adults completely covering head and foot in disturbed animals. Hyponotal line dividing hyponotum into a granulose outer area (i), broader than a smooth inner one (ii) (close to the foot). Pedal sole (Figs. 2D,F) smooth and similar in colour to hyponotum. Foot ending posteriorly to mouth, close to pedal gland opening. Ciliated groove running along foot right side to caudal region, reaching female genital opening, the latter at right side of anus. Anus caudomedian. Pneumostome present in hyponotal line, always above anus (Fig. 2G).

Cephalic region well defined, bearing two optical tentacles with eyes at tip, retracted in preserved animals (Fig. 2E). Mouth surrounded by a pair of sensorial oral lobes of same size and not fused centrally. Male genital opening on right side of right oral lobe (Fig. 2E).

Internal Anatomy

Subdermal structures and glands

Perinotal glands (=marginal/repulsive glands) in body wall below notum, surrounding visceral cavity (Fig. 5C). 8 pairs of perinotal glands on each side; 9 pairs only in ANC129032101. Perinotal glands pyriform, circular in section. Glandular duct thin, opening distally and leading medially to marginal papillae tip (Fig. 5E). All glands arranged in a single row and almost all in same frontal plane. Posterior perinotal glands located slightly more ventrally. Central perinotal glands wider than anterior and posterior ones.

Figure 3.A-B: Internal anatomy of ANC110092104 from right-sagittal view. 

Figure 4.A: Internal anatomy of ANC110092104 from left-sagittal view. B: Internal anatomy of ANC110092103 unstained from left-sagittal view. C: Detail of the cephalic region of ANC110092103 unstained from right-sagittal view. D: Detail of the caudal region of ANC110092104 from left-sagittal view. 

Figure 5.Internal anatomy of ANC110092104. A-C: Front-dorsal view. D: Front-ventral view. E: Transverse section of a perinotal gland showing the outlet of the glandular duct at the tip of the marginal papilla. 

Figure 6.Internal anatomy of ANC110092104 reconstructed with Amira® v.5.4.4 software. A: Dorsal view. B: Ventral view. Central nervous ring containing cerebral ganglia (pink), pleural ganglia (blue), visceral ganglion (green) and pedal ganglia (yellow). (*) showing the end of the penial retractor muscle. C: Right lateral view (digestive and albumen glands transparent for better visualisation). D: Left lateral view (digestive and albumen glands transparent for better visualisation). 

Pedal gland located between mouth and anterior foot end (Figs. 5D, 6B-D). Gland one-chambered, rectangular in section. Frontal wall of pedal gland reduced and merging with pedal sole tissue. Secreting pore, subdermal spicules or concretions not observed.

External morphology

The external morphology of Onchidella celtica is well known (DELLE CHIAJE, 1841-1844; PHILLIPPI, 1844; JOYEUX-LAFFUIE, 1882; MARCUS, 1979; CHEMELLO Y D’ANNA, 1986; TWEEN, 1987; WEISS & WÄGELE, 1998; BARILLÉ-BOYER et al., 2000; GROH, 2006); our observations fairly agree with the previous aforementioned work. The length of the studied specimens matches with those from Great Britain (TWEEN, 1987), France (MARCUS, 1979; BARILLÉ-BOYER et al., 2000) and Morocco (MARCUS, 1979). In addition, the maximum length of 25 mm recorded is twice that reported previously for this species in Galician waters (ROLÁN et al., 1989; TRONCOSO & URGORRI, 1990).

According to DAYRAT et al. (2011b), the maximum length could vary between species, but the rest of external characters are affected by fixation and preservation processes and show great intra-specifically variation. Therefore, detailed measurements or length and width ratios, as well as colour or appearance of notum, are not appropriate to differentiate species of Onchidella (MARCUS, 1979; DAYRAT et al., 2011b). This difficulty in distinguishing species by external features is widespread in most genera of the family, as almost all of them have externally cryptic species (DAYRAT et al., 2017; GOULDING et al., 2018a, 2018b; DAYRAT et al., 2020). However, the sharpness of the notal structures in the scanned specimens is remarkable when compared to light microscopy or SEM examination (BARILLÉ-BOYER et al., 2000; XU et al., 2018).

Subdermal structures and glands

Subepidermal silicate spicules were described by LABBÉ (1933, 1934) below the notum and hyponotum of several onchidiids thus resulting in the erection of the order Silicodermatae; these findings were, however, refuted by several authors (see MARCUS, 1979). Spicule networks have been successfully studied with micro-CT in nudibranchs (ALBA-TERCEDOR & SÁNCHEZ-TOCINO, 2011; PENNEY et al., 2018; PAZ-SEDANO et al., 2021; URGORRI et al., 2021). Due to the preference of iodine for polysaccharides and calcified structures, hard tissues show an excess in the brightness of micro-CT images. Consequently, spicule networks are more suitable for visualisation in unstained samples (ALBA-TERCEDOR & SÁNCHEZ-TOCINO, 2011; GIGNAC et al., 2016; CANDÁS et al., 2018). In the present work, no evidence of spicule or any other hardened subdermal deposits was found in ANC110092103. Thus, biosilification in Onchidiidae still remains unproven (DAYRAT, 2009; EHRLICH, 2019).

MARCUS (1979) emphasized the importance of the histochemistry and structure of perinotal glands to identify species of Onchidella. Histochemistry cannot be assessed by micro-CT, but this technique still allowed for a clear visualisation of its structure and disposition below the notum (see Fig. 5E). Our results are in line with those depicted in JOYEUX-LAFFUIE (1882), BINOT (1965) and MARCUS (1979) for O. celtica showing the outlet of the narrow glandular duct at the tip of perinotal papillae. Regarding its number, almost all authors agreed that this species possesses 10 pairs of glands in both Mediterranean (DELLE CHIAJE, 1841-1844; PHILLIPPI, 1844; CHEMELLO & D’ANNA, 1986) and Atlantic populations (JOYEUX-LAFFUIE, 1882; BINOT, 1965; MARCUS, 1979). However, WEISS & WÄGELE (1998) also described 8-9 pairs of glands in O. celtica thus highlighting intra-specific variability as it occurs in other Onchidella species (WATSON, 1925; MARCUS, 1978, 1979; DAYRAT et al., 2011b; GOULDING et al., 2022). On the other hand, histology and morphology of the pedal gland of O. celtica has been described (VAILLANT, 1872; JOYEUX-LAFFUIE, 1882; FRETTER, 1943; BINOT, 1965; WEISS & WÄGELE, 1998); our results mostly agree with previous work except for FRETTER (1943) who described a medial septum internally dividing the gland, not found in our examinations.

Digestive system

The presence of an inconspicuous jaw-like structure within the lips of the scanned specimens has been found as in other Onchidella species, including O. celtica (JOYEUX-LAFFUIE, 1882; VON WISSEL, 1904; WATSON, 1925; DAYRAT et al., 2011b). The radula has correctly been located within the buccal mass, but the shape and number of teeth could not be determined. Failure in distinguishing radular teeth by micro-CT has already been pointed out previously (CANDÁS et al., 2018; MARTÍNEZ-SANJUÁN et al., 2022). Nonetheless, ZIEGLER et al. (2018) noted that if the size of the radula is substantial, very high-quality images can be achieved with this technique. In Onchidiidae, while the radular formula seems to have a low reliability for species delimitation due to its high intra-specific variability (MARCUS, 1979; BRITTON, 1984), the shape of teeth usually helps at the genus level when observed with SEM (DAYRAT et al., 2011b, 2017, 2018, 2019a). Further studies will determine the effectiveness of Micro-CT against SEM for the visualisation of this trait.

The stomach of Onchidiidae exhibits a marked regionalisation which is better appreciated when it is full (FRETTER, 1943). Here, we have clearly observed the four chambers described in most Onchidella species (WATSON, 1925; FRETTER, 1943; TWEEN, 1987; WEISS & WÄGELE, 1998; DAYRAT et al., 2011). The smaller one (=stomach IV) has been clearly seen with micro-CT but is usually omitted in some descriptions and merely termed as “intestinal caecum” (JOYEUX-LAFFUIE, 1882; CHEMELLO & D’ANNA, 1986). The sizes of the chambers, the whole stomach and the digestive gland strongly vary between individuals (DAYRAT et al., 2011b). Thus, providing detailed measurements of each organ seems useless for taxonomic purposes (MARCUS, 1979).

According to the dorsal pattern of the intestine, PLATE (1893) distinguished four types of intestinal loops (types I-IV) in Onchidiidae, to which a fifth one (type V) was later added by LABBÉ (1934). Almost all onchidiid genera possess one or two types of intestinal loops and, despite some intra-specific variation within a particular genus, nearly all species are characterised by a single intestinal type (DAYRAT et al., 2019b). The taxonomic value of this character for genera delimitation has recently been emphasized in the phylogeny of the family (GOULDING et al., 2022). The intestinal U-shaped loop of the scanned specimens (see Fig. 6C) fits with type IV, which is the only one present in the genus Onchidella (BRITTON, 1984; DAYRAT et al., 2011b; GOULDING et al., 2022).

Nervous system

The structure of the nervous system of Onchidiidae is essentially identical in all species (WATSON et al., 1925). The morphology and location of all ganglia has been clearly observed in the scanned specimens. The position of the visceral ganglia shows a high variability within O. celtica and O. borealis, i.e. displaced to the right (JOYEUX & LAFFUIE, 1882; WEISS & WÄGELE, 1998; this work) whereas other work reported them near to the left pleural ganglion (PLATE, 1893; VON WISSEL, 1898; STANTSCHINSKY, 1907); therefore, this character seems lacking taxonomic relevance (WEISS & WÄGELE, 1998). Regarding the nerves, WEISS & WÄGELE (1998) described 14 pedal nerves arising from pedal ganglia, while only 9 were observed in the scanned animals. The applied resolution was probably not sufficient to distinguish pairs of overlapping nerves (CANDÁS et al., 2018; MARTÍNEZ-SANJUÁN et al., 2022).

However, lower numbers of pedal nerves have been found in other Onchidella species (WATSON, 1925).

Circulatory system and renal-pulmonary comple

Detailed descriptions of the circulatory system were provided by VAILLANT (1872) and JOYEUX-LAFFUIE (1882). Our results did not show any relevant differences in comparison to previous work. Within the genus Onchidella, the position of pericardium constrains the right lung and kidney branches causing an asymmetrical course of both organs (WATSON, 1925). In O. celtica, the right branch is always longer than the left one (JOYEUX-LAFFUIE, 1882; FRETTER, 1943; WEISS & WÄGELE, 1998), whereas in other genera it may be the opposite or simply have a symmetrical development (BRITTON, 1984). As for the inner epithelium of the kidney, three different types were recognized by PLATE (1893) in Onchidella: lamellate, folded and smooth. The last one is present in O. celtica (WEISS & WÄGELE, 1998) and was also found in the scanned specimens. The observation of the renal-pulmonary complex by using micro-CT is noteworthy, since these organs are often overlooked due to its thin and inconspicuous epithelium (XU et al., 2018).

Reproductive system

The female reproductive parts of Onchidiidae are not relevant for species identification owing to their great variation in size and shape, which depends on the maturity and sexual activity of animals (BRITTON, 1984; DAYRAT et al., 2011b). However, some features observed here deserve to be commented when compared to previous descriptions of O. celtica. The presence of a fertilization chamber at the end of the hermaphroditic duct had only been quoted twice (FRETTER, 1943; CHEMELLO & D’ANNA, 1986). As for the mucous and prostate glands, most authors refer to them as "female mass gland" because they are difficult to distinguish from each other (WATSON, 1925; DAYRAT et al., 2011b). Indeed, it is often identified as "mucous gland" (JOYEUX-LAFFUIE, 1882; WEISS & WÄGELE, 1998). In this work, clear visualisation of the inner epithelium of both glands was achieved by micro-CT. Our descriptions are consistent with those of VAILLANT (1872), FRETTER (1943) and CHEMELLO & D’ANNA (1986). Moreover, the vaginal gland was clearly observed and reconstructed. The presence of this structure in O. celtica has been reported by most authors apart from VAILLANT (1872) and WEISS & WÄGELE (1998). According to DAYRAT et al. (2011b), the vaginal gland is found only in Onchidella and Hoffmannola STRAND, 1932 and its opening at the end of the vaginal tube seems to vary between the two genera.

Regarding the male reproductive organs, BRITTON (1984) states that “these are the most widely used and trustworthy characteristics both at generic and specific level”. The position of the male opening has been proposed as a reliable feature for supra-specific classification (HOFFMANN, 1928). However, there are many different interpretations in published work that often lead to misunderstandings (DAYRAT et al., 2011b). In Onchidiidae, the absence or presence of the penial gland is usually a shared character within nearly all the species of the genus (see GOULDING et al., 2022). The same occurs with the penial hooks, that could vary inter-specifically (GOULDING et al., 2018b). Onchidella lacks both features as confirmed again in the scanned specimens. Hence, we cannot elucidate whether the use of micro-CT is appropriate for its observation in other onchidiids.

At the specific level, the length and coiling of the vas deferens has been proposed for species delineation (DAYRAT et al., 2011b). Our results are in line with previous descriptions in O. celtica (SEMPER, 1885; MARCUS, 1979; CHEMELLO & D’ANNA, 1986; WEISS & WÄGELE, 1998). A small penial papilla was clearly seen in the scanned specimens near the attachment of the vas deferens to the penial sheath. This structure has already been described in O. celtica (JOYEUX-LAFFUIE, 1882; SEMPER, 1885; HOFFMAN, 1928; MARCUS, 1979) and is often absent in some species of the genus (MARCUS, 1979; DAYRAT et al., 2011b). The presence of penial concretions has been mentioned in several species of Onchidella including O. celtica (JOYEUX-LAFFUIE, 1882; WATSON, 1925) but they were not observed in any of the scanned animals including the unstained specimen.

PLATE (1893) classified the attachment of the retractor muscle of onchidiids into three types: Type I (near the central nerve ring), Type II (close to the pericardium) and Type III (at the hind end of the visceral cavity). In the studied specimens, the end of the retractor muscle has been clearly observed at the level of the pericardium (see Fig. 6B) corresponding to Type II as previously reported for O. celtica (JOYEUX-LAFFUIE, 1882; HOFFMAN, 1928). Nonetheless, most previous work described it as a Type III (FRETTER, 1943; MARCUS, 1979; CHEMELLO & D’ANNA, 1986; WEISS & WÄGELE, 1998). Even though MARCUS (1979) and DAYRAT et al. (2011b) stressed the taxonomic importance of this character at the specific level, the variation of the attachment in individuals of the same species has also been found in O. indolens (Couthouy, 1852) (WEISS & WÄGELE, 1998). This trait has been well observed by means of micro-CT; therefore, its application to a larger number of specimens could assess whether shows variation and its potential taxonomic relevance.

Conclusions

The phylogenetic relationships of Onchidiidae have been reconstructed very recently considering molecular markers (GOULDING et al., 2022). Except for Onchidella, the systematics of all genera of the family have been revised in the last few years from an integrative approach (see above). Therefore, an anatomical revision of all species of Onchidella seems necessary to clarify the aforementioned results.

Here, despite limitations restricted to the radula, the internal anatomy of Onchidella celtica has been successfully observed and reconstructed by using micro-CT. Futhermore, our observations are strongly in line with previous descriptions. Therefore, micro-CT is a very useful technique for a relatively rapid and non-destructive study of the internal anatomy of Onchidiidae. As this first approach has focused on a single species, future studies including other onchidiid genera will provide more consistent data to support our observations.