INTRODUCTION

Previously, the pulmonate family Arionidae was regarded as including the subfamilies Arioninae, Anadeninae and the American subfamily Ariolimacinae (Pilsbry 1948). According to this view, the family Arionidae is widely distributed in the Holarctic region. The genus Oopelta, which shows affinities to this family, was reported from South Africa during the middle of the nineteenth century (Gray 1855). At least five species in this endemic genus are currently recognized. The family Oopeltidae and subfamily Oopeltinae were established for them by Cockerell (1891). Van Regteren Altena (1966) recommended that the genus should be included in Arionidae within the subfamily Oopeltinae; and in this context it has for a long time been considered to be the only taxon of the family Arionidae in South Africa.

More recently, a second South African subfamily, Ariopeltinae, was described (Sirgel 1985). It was based on two new genera, each containing a single species, namely Ariopelta capensis and Ariostralis nebulosa, respectively. Both species occur at ca 1350 m above sea level in a very wet, cool and restricted habitat near the summit of the Hottentots Holland mountain range in the Western Cape Province of South Africa. This area is characterized by the presence of various palaeogenic invertebrates which are mostly relict faunal elements (Stuckenberg 1962). Features of these two slug species suggest that they are relatively primitive members of the superfamily Arionoidea and they should thus also be regarded as relicts. Being slugs, the main evolutionary process in respect of the phylogeny of the superfamily is that of limacization. In this regard Arion, the Holarctic type genus of Arionidae, exhibits relatively advanced character states with, amongst others, the shell being almost totally degenerated and the retractor muscles of the left and right sides shortened to such a degree that they originate independently from the dorsolateral body wall. The members of Ariopeltinae differ from other Arionidae in still retaining a coherent shell plate, while the tentacular and pharyngeal retractor muscles are longer and have a common point of origin (in some even a common stem) at the posterior edge of the pallium. These conditions are clearly more reminiscent of the ancestral snail condition, with a single columellar muscle dividing into the various retractor muscle elements some distance from its point of origin within a well-developed shell. As will be explained below, further features that appear to reflect a primitive condition are that the walls of both the vagina and penis carry structures that project into the lumen. In the penes of both species, each of these projections is supported by an internal calcareous, claw-shaped spine. In the case of the vagina, the supportive core in Ariopelta capensis also consists of calcareous material, whereas in Ariostralis nebulosa it is cartilaginous. A kind of endoskeleton is thus present in these structures.

More recent authors reject Pilsbry's (1948) views on the family Arionidae, mostly by promoting his subfamilies to family rank within the superfamily Arionoidea. Wiktor (2001) furthermore ejects the American genera Prophysaon and Anadenulus from Anadenidae. Bouchet et al. (2005) recognize the families Arionidae, Anadenidae, Ariolimacidae, Binneyidae, Philomycidae and Oopeltidae, with the last-mentioned including the subfamilies Oopeltinae and Ariopeltinae. Schileyko's (2007) classification differs from that of Bouchet et al. (2005) by retaining Ariolimacinae as a subfamily of Arionidae. Henceforth, the system suggested by Bouchet et al. (2005) will be followed in this paper.

The original description of Ariopeltinae (Sirgel 1985) indicates a closer morphological resemblance to the Indo-Chinese Anadenidae (genus Anadenus), described by Godwin-Austen (1882), than to its fellow South African subfamily Oopeltinae. The current description of a new Ariopelta species now permits a better understanding of the relationship between the two South African subfamilies.

In cases where, as mentioned above for Ariopeltinae, a new subfamily with new genera (each containing a single species) is to be described, the situation becomes problematic. As there are no additional species in each genus that can be contrasted with the first, it is rather difficult, if not impossible, to determine which of the given characters are diagnostic at the generic level and which are species specific. When an author is forced to define both the genus and species in such circumstances it is to be expected that once further species in such a genus are discovered, an update of the diagnosis of the genus will become necessary. It might, for instance, be found that certain features originally regarded as diagnostic of the genus must be relegated to the species level and described in more detail. Similarly, the discovery and description of additional species may require that a subfamily has to be redefined to allow for the inclusion of additional morphological characters.

Such adjustments are now deemed necessary in relation to the description and correct placement into the higher categories of the new Ariopelta species to be described here. Although it does have intrusions of the penis and vagina, calcareous or cartilaginous cores supporting them are lacking. As such cores were considered to be of diagnostic value at the subfamily level in the original description (Sirgel 1985), changes to the definition of the subfamily are now required. The same applies where a penis consisting of four successive regions was regarded as diagnostic at the subfamily level but which in the new species consists of only three regions. These changes also affect the diagnoses of the genera and species. New information gained now also enables more detailed descriptions to be made at the species level. Clearly, a revision of the subfamily Ariopeltinae is required.

MATERIAL AND METHODS

The new species was found within the Kogelberg Nature Reserve in a relatively small forest known as Wynandlouwsbos (34°17′13.89″S 18°52′58.48″E), situated at an altitude of 305 m near the base of the south-facing slope of the Kogelberg Mountains. Although some specimens were collected from the leaf litter, most were in wet tree trunks that were in an advanced state of decomposition. Examination of the contents of the alimentary canal revealed it to largely consist of the decaying wood in which the slugs occurred. Pollen of Asteraceae and Proteaceae was also found to be part of the diet.

All type specimens have been preserved in 70% ethanol and deposited to the collection of the KwaZulu-Natal Museum (NMSA).

TAXONOMY

Fig. 1.

Ariostralis nebulosa, genital system. Abbreviations: ALB GL - albumen gland, GA - genital aperture, GATR - genital atrium, OV - free oviduct, OVOT - ovotestis, PR - penis retractor muscle, SPOV - spermoviduct, SPT - spermatheca, Vag - vagina, VAG B - vaginal bulb, Vd - vas deferens, 1–4 indicate the successive parts of the penis. (From Sirgel 1985)

Fig. 2.

Ariostralis nebulosa, inner perforated wall of the vaginal bulb removed and photographed from its outer side. Abbreviations: PRF - perforations, SP STR - tissue strands suspending the inner wall of the vaginal bulb from its outer wall.

Figs 3, 4.

Ariopelta capensis, genital system: (3) details of the genital system (from Sirgel 1985); (4) distal part of genital system with vaginal bulb cut open to show its interior. Abbreviations: ALB GL - albumen gland, G A - genital aperture, G ATR - genital atrium, LAM - calcareous ring-shaped lamellae intruding into lumen of vaginal bulb, O SPD - opening of spermathecal duct, OV - free oviduct, OVOT - ovotestis, PR - penis retractor muscle, SPD - spermathecal duct, SPOV - spermoviduct, SPT - spermatheca, VAG - vagina, VAG B - vaginal bulb, Vd - vas deferens, 1–4 indicate the successive parts of the penis.

Fig. 5.

Ariopelta capensis, second and third parts of penis cut open to show calcareous spines intruding into the lumen: (2) second part with smaller spines, (3) third part with larger spines.

Fig. 6.

Ariopelta variegata sp. n., holotype in extended and in contracted conditions (composite image).

Figs 7–9.

Ariopelta variegata sp. n.: (7) dorsal view to show arrangement of retractor muscles; (8) genital system. Abbreviations: ALB GL - albumen gland, AO - position where aorta penetrates the diaphragm, GA - genital aperture, G ATR - genital atrium, L+R TR - left and right tentacle retractor muscles, L P - large papillae, OV - free oviduct, OVOT - ovotestis, PH R - pharynx retractor muscle, PR penis retractor muscle, RECT - position where rectum penetrates the diaphragm, SP- small papillae, SPOV - spermoviduct, SPT - spermatheca, VAG B - vaginal bulb, Vd - vas deferens, VSP - very small papillae, 1–3 indicate the successive parts of the penis; (9) penis with second and third parts opened to show interior.

TABLE 1

Characters separating Ariopelta variegata sp. n. from Ariopelta capensis.

DISCUSSION

Collinge (1900) described a new species of slug from Cape Town as Amalia ponsonbyi (Milacidae). More recently, it was suggested that it should be transferred to the genus Ariopelta as an additional species, differing from Ariopelta capensis in minor details of the genital system (Sirgel 1985). Collinge's indication that the body colour is “Dorsally sepia blue, gradually becoming yellowish laterally and towards the foot fringe” also differs from the evenly black body wall of A. capensis. Collinge based his description on a single preserved specimen, now lost. This, together with the fact that no further specimens of such a species have since been found, appeared to be sound reasons for regarding A. ponsonbyi as a nomen dubium. The discovery ofAriopelta variegata sp. n. reported here, however, provides supporting evidence for the existence of A. ponsonbyi. The colour and other external characters, as described by Collinge (1900), seem to be identical to those of the new species. Secondly, the structure of the genital systems shows a high degree of resemblance. However, Collinge's figure and description of his species as having a slender distal region of the penis with two small successive dilations near its junction with the genital atrium, as well as the reported presence of an atrial gland (Fig. 10), make A. ponsonbyi seem sufficiently different from A. variegata sp. n. to rule out the possibility of it being the same species. It could be argued that the genital system Collinge described was possibly underdeveloped and that the atrial gland he mentioned might be an erroneous interpretation of tissue surrounding the genital atrium and connecting it to the body wall. Even then, though, his species is still disqualified as being the same as A. variegata sp. n.

The concave ventral side of the shell of both Ariopelta capensis and Ariostralis nebulosa is often found to be filled with a calcareous deposit. No such thickening deposit was detected in any of the shells of the specimens of Ariopelta variegata sp. n. that were examined. In all specimens of A. variegata which were dissected, whether young or apparently fully-grown, the albumen gland was always seen to be very small, even in those individuals where the penile and vaginal structures were well developed and the ovotestis very large. These differences could be a consequence of collection of the examined specimens prior to having reached complete development, or shortly after egg-laying. It must, however, be pointed out that in numerous dissected specimens of A. capensis and A. nebulosa, the albumen gland was always relatively smaller than the very large one usually found in Oopelta species.

As far as the retractor muscles are concerned, Ariopelta capensis shows a relatively primitive condition, with the tentacular and pharyngeal retractors having a common origin at the posterior border of the pallial organs (Sirgel 1985), from where they extend anteriorly as a common trunk to opposite the centre of the pallium where the tentacle retractors of the left and right sides diverge. The pharyngeal retractor only diverges from the left tentacle retractor further anteriorly, more or less opposite the anterior border of the pallium, and only splits into a left and right muscle close to their insertion on to the buccal mass. Also, the penis retractor shares a common origin (but no common trunk) with these muscles. In Ariostralis nebulosa, the condition is more advanced in that the tentacle retractors on the left and right sides have a contiguous origin but no common trunk. Moreover, the pharyngeal retractors of the left and right sides are separate over most of their length. The penis retractor has a separate origin a short distance to the left of the other muscles. The condition in Ariopelta variegata sp. n. resembles that in A. capensis, except that the origin of the penis retractor is also separated towards the left; and the pharyngeal retractors are separated for about half their length (Fig. 7). These muscles thus show a tendency to separate and shorten. This is carried much further in the advanced Oopeltinae, where the tentacle retractors are so drastically shortened that even the branches to the upper and lower tentacles each have a separate origin located far anteriorly on the lateral body wall of their respective sides.

In both Ariopelta species, the ovotestis consists of tightly packed acini of which the basal parts are often divided into a few short lobes. Apically, the acini each taper to pass over into a small duct which joins others to constitute the main hermaphrodite duct. In Ariostralis nebulosa, the condition is different. The acini are unlobulated, loosely packed, large, and cylindrical (Sirgel 1985).

In contrast to the condition in Ariopelta capensis and Ariostralis nebulosa, where the globular vaginal bulb is separated from the genital atrium by a more slender duct, in Ariopelta variegata sp. n. it extends all the way to the genital atrium in fully developed specimens (Fig. 8). A slender distal section is, however, found in immature specimens (Fig. 11). In such cases, it clearly only represents the underdeveloped distal part of the vaginal bulb. Collinge (1900) did not indicate any globular section in the vagina of Ariopelta ponsonbyi but he drew both the free oviduct and the vagina as a relatively wide duct (Fig. 10).

As far as the penis is concerned, a fourth portion extends as a slender tube between the third portion and the genital atrium in Ariopelta capensis as well as in Ariostralis nebulosa, and Collinge (1900) indicated the existence of a similar slender duct for Ariopelta ponsonbyi. In fully developed specimens of Ariopelta variegata sp. n., the condition differs because the third, dilated portion of the penis extends all the way to the genital atrium. In immature specimens, though, a slender duct extends between the middle, dilated part of the penis and the genital atrium (Fig. 11). At first sight, this latter duct may seem to represent a fourth part of the penis as found in the other three species mentioned here. In those species, however, the middle part consists of only one dilation, indicating that the slender distal duct in these young specimens of the new species is merely the underdeveloped third part of the penis that has yet to develop a dilation. In both A. capensis and A nebulosa, the relatively short second portion of the penis carries small calcareous spines, while much larger spines are present in the more dilated and longer third portion. In A. variegata sp. n., the situation is the opposite, with the second portion of the penis being the larger one and carrying the larger papillae, although devoid of calcareous spines. The papillae in the smaller, third portion are very small.

Members of Oopeltinae usually have a pigmented parietal peritoneum, and at least part of the covering of the gonad is pigmented. In Ariostralis nebulosa, the parietal peritoneum and the surface of the gonad are slightly pigmented, imparting a light greyish colour to these parts. In both species of Ariopelta, the internal body parts are not pigmented. These differences in internal pigmentation might not, however, be genetically determined, but could represent a secondary feature that is dependent on the lifestyle of the animals and on the amount of insolation received. Members of Oopeltinae occur in relatively open areas where they are exposed to solar radiation. In Oopelta polypunctata and O.flavescens, the bodies of which are often lemon-coloured externally, that part of the gonad exposed to the body wall is intensively pigmented, whereas the area covered by other organs is less pigmented. Although sharing the same locality, the sparse internal pigmentation in Ariostralis nebulosa, in contrast to its total absence in Ariopelta capensis, may result from the fact that A. nebulosa often dwells on the surface of dense vegetation (especially in misty conditions), while A. capensis normally remains in the deepest layers of the dense vegetation. During the day, Ariopelta variegata sp. n. is found within rotting logs or buried in the leaf litter of the forest.

In the past, the position of the South African genus Oopelta within the family Arionidae was regarded as problematic. In fact, Cockerell (1891) and Simroth (1910) proposed long ago that it should be given independent family rank. This point of view can be understood, considering that at the time this genus was perceived as a solitary group, the area of distribution of which was widely separated from the Holarctic region where other members of Arionoidea were known to occur. Furthermore, although certain characteristics of Oopelta correspond to those described for Arionidae, others seem to deviate significantly. For example, Oopelta differs by having an oxygnathous jaw; no caudal pore; the genital pore situated anteriorly near the tentacles; a penile structure; and the retractor muscles of the four tentacles being extremely shortened, each with its own origin. Taking all these facts into account, Van Regteren Altena (1966) came to the conclusion that Oopelta is a group of great antiquity and long isolation and should be given subfamily rank, but he nevertheless retained them within the family Arionidae sensu Pilsbry (1948). Clearly, more information was needed to finally understand the position of Oopeltinae in relation to the various families within the superfamily Arionoidea.

The discovery and description of Ariopeltinae as a new subfamily in South Africa showed it to have certain features that correlate well with those of some members of the Asian Anadenidae. In both groups, the genital aperture is situated anteriorly close to the tentacles and intrusions including calcareous structures occur in the distal parts of the genital system (Schileyko 2007). The pattern of arrangement of the main retractor muscles shows a high degree of resemblance. A caudal pore is absent and the foot fringe is narrow in Anadenidae. Thus there are similarities in these respects to, at least, the genus Ariopelta as well as all the members of Oopeltinae. The jaw in both South African subfamilies is oxygnathous. In Anadenidae, it is believed to conform to that of the European Arioninae in merely consisting of a number of ribs. However, Heynemann's (1863) description of Anadenus schlagintweiti indicates signs of an oxygnathous type of jaw. He describes it as having 16 ribs of which the laterals are fused, while the central one projects like a beak (“die mittleren treten etwas schnabelartig vor”; Heynemann 1863: 141). These correlations seem to be significant enough to regard Ariopeltinae as a subfamily within Arionoidea, and they suggest affinities with Anadenidae. The widely separate geographical distribution of these two subfamilies, South Africa and Indo-China, however, suggests divergence in the distant past from a common ancestor. The shared features must therefore be regarded as very primitive characters which have been preserved in both groups since antiquity. This is in line with Stuckenberg's (1962) observation that various palaeogenic relict faunal elements occur in the mountainous area, where Ariopeltinae are now found. The same phenomenon seems to be applicable to several Anadenidae species, according to the descriptions of Schileyko (2007). A comparison between these taxa that show primitive features and the advanced Arionidae of Europe accordingly seems to suggest a Gondwana origin for the superfamily Arionoidea.

Figs 10,11.

(10) Amalia ponsonbyi, generative organs (after Collinge 1900; annotations changed to fit terminology in present paper); (11) Ariopelta variegata sp. n., genital system of juvenile. Abbreviations: AG - atrial gland, ALB GL - albumen gland, G ATR - genital atrium, OV - free oviduct, OVOT — ovotestis, PR — penis retractor, SPOV — spermoviduct, SPT — spermatheca, VAG B — vaginal bulb, Vd — vas deferens, 1–6 or 1–3 - successive parts of penis.

The morphology of Ariopeltinae thus more convincingly suggests affinities between a South African taxon and the Indo-Chinese Anadenidae than Oopeltinae did. The common features (e.g. coherent shell, similar pattern of muscular system, and the penis with calcareous inclusions) found in the Ariopeltinae and Anadenidae, and differentiating both from the Oopeltinae still remains a puzzling problem as far as correct interpretation of affinities between these three taxa is concerned. It should, however, be mentioned that Ariopeltinae corresponds better with its fellow South African taxon in respect of a few minor features such as a more pronounced oxygnathous jaw, a more differentiated penis, and a better differentiated genital atrium. The discovery of A. variegata sp. n. now permits a clearer understanding of affinities and the evolutionary transition from the primitive Ariopeltinae-type condition to that of the more advanced Oopeltinae. This new species exhibits morphological characteristics approximating those of A. capensis but shows a developmental tendency towards the Oopeltinae condition in having a thinner shell; calcareous structures not present in the distal parts of the genital system; retractor muscles slightly more separated; and a lighter body colour. Furthermore, A. variegata occupies a habitat in which environmental conditions are in-between those to which the primitive Ariopeltinae and the much advanced Oopeltinae are exposed, namely in relation to altitude, humidity (conditions are slightly drier) and temperature.

The more pronounced morphological similarities between Anadenidae and Ariopeltinae than between either of them and the much advanced Oopeltinae may at first sight be interpreted as an indication of Anadenidae and Ariopeltinae being sister groups, with Oopeltinae representing an outgroup. Such an interpretation of a closer relationship between two groups occurring on widely separated continents than between two occupying areas of close proximity would be highly anomalous, so a more logical explanation should be sought for this phenomenon. As mentioned above, the line leading to the geographically widely separated Anadenidae and Ariopeltinae probably diverged in ancient times, at or even before the break-up of Gondwana. Many features exhibited by slugs belonging to these two groups seem to have been preserved to a high degree. This could be why similarities are to be seen.

The major morphologic differences between the two South African taxa remains to be explained. Such an explanation should be sought in the belief that the Western Cape Province of South Africa, in which both of these morphologically divergent taxa presently occur, experienced a wet climate with lush vegetation during the early Cenozoic (Deacon 1983). In the later stages of the Cenozoic, the area underwent a series of climatic changes and became drier. During the Pliocene, a dry summer / wet winter rainfall regime also became established (Hendey 1983). These changes would have resulted in the original humid conditions becoming limited to small microhabitats in the mountains. Consequently, faunal elements that adapted to such a climate would become restricted to these microhabitats. Since this favoured climate was maintained in these places, not much selective pressure for evolutionary change would have been exerted on the inhabitants, and many of their bodily features would be preserved. The lower-lying areas which had undergone the climatic changes would now present new niches in which adaptive radiation could take place. Extensive selective pressure would be exerted on the inhabitants, thus resulting in rapid evolutionary change. As members of Oopeltinae presently occupy these drier and more sparsely vegetated lowlands, they seem to be the result of such accelerated evolutionary changes. Perhaps this explains why they and their more conservative fellow South African taxon Ariopeltinae are so very different morphologically. However, their divergence from the same ancestor must have taken place at a much later stage than the divergence of their common ancestor from the ancestor that led to the Indo-Chinese Anadenidae, which morphologically resemble Ariopeltinae more closely. According to the descriptions of Wiktor (2001) and Schileyko (2007), it seems as if extensive evolutionary developments also took place in the family Anadenidae in Asia. Developments like these fit perfectly the view of Bouchet et al. (2005), which is to regard both Oopeltinae and Ariopeltinae as subfamilies of the family Oopeltidae, with Anadenidae as an independent family within the superfamily Arionoidea.