Being a soft-bodied, succulent, fleshy water bag in a dry environment poses problems, which range from simply drying out, through being stepped on, to being leapt on with greed by some heartless predator. On the face of it, a land mollusc seems a highly unlikely proposition, but the terrestrial molluscs have evolved satisfactory answers to the daunting array of potential disaster which confronts them, and have done it so successfully that they can be found in large numbers in an extraordinary variety of habitats all over the world. These habitats do not include only the moist areas; snails are common in many arid areas, and can be spectacularly successful in withstanding desiccation. A famous example is that of a specimen of the South African desert snail Eremina desertorum, glued to a card in the British Museum in 1846, which emerged from its shell and happily crawled off into the middle distance when exposed to moisture four years later. This ability to wait through dry conditions has often been noted — the West Australian camaenid snails are good at it, as are the South Australian helicellids, many African helicidae, and the prosobranch land snails of the Khasi Hills of Assam. Many deserts of the world have their populations of land snails, all waiting patiently for the rains to come.

The ability to retain water is the key factor for a snail living on land. Snails are inveterate hypochondriacs, constantly fretting about their water balance, and modifying their entire life style to keep that balance within allowable limits. Obviously, there are basic structural adaptations involved, but modifications to behaviour patterns, such as a nocturnal habit, the ability to seek out moist micro-habitats, and the ability to aestivate are also important (Hyman, 1967; Solem, 1978). This paper, then, will look at the hazards faced by the land molluscs, the modifications which allow them to deal with those hazards, the selective presures which have influenced the development of the slug form, and the evolution of one group of slugs in particular, the Athoracophoridae.

The Class Gastropoda is divided into three subclasses, the Opisthobranchia, Prosobranchia, and Pulmonata; the prosobranchs and pulmonates both have land representatives. The prosobranchs retain the open mantle cavity found in marine forms, and when they retreat into the shell during dry conditions they seal the opening with an operculum. As a water retention device, this is superb; in fact the seal is so effective that accessory notches, slits, grooves or separate tubes have evolved to allow air to circulate past the barrier of the operculum (Rees, 1964). However, the arrangement of the prosobranch mantle cavity allows high rates of water loss when the snail emerges, unless conditions are very humid. In the land prosobranch, the gills are lost and respiration takes place on the inner surface of the mantle. The kidney secretes a stream of almost pure water from the posterior of the pallial cavity, and this moistens the cavity, allows respiration, and stops the tissues drying out — at a cost. The arrangement is only marginally efficient. The snail can be active in periods of very high humidity, but when ground level humidity drops below about 95 per cent, the snail must retreat behind its operculum and wait for wet weather to return. This obviously cuts down the opportunities the snail has for feeding and mating.

The pulmonate snails lack an operculum to seal the shell aperture. Instead they secrete an epiphragm, a membrane of mucus or calcified page 49 mucus which may prevent water loss just as effectively as an operculum, but which still allows air to filter back and forth. Thus pulmonates can withstand dry conditions just as well as prosobranchs, but it is when humidity falls that pulmonates have the edge. They gain this edge through a number of modifications.

First, the pallial cavity, instead of being open, is enclosed, and the mantle collar is fused to the neck of the snail. This creates a large internal chamber communicating with the outside through a pneumostome, and this allows water loss over the breathing surface to be controlled. A bonus is the ability to retain reserve water inside the mantle cavity; this reserve may be as much as one-twelfth the total body weight (Blinn, 1964). Having a savings account of water allows a much more flexible approach to changing conditions. A pulmonate can remain active when a prosobranch is forced by falling humidity to withdraw into its shell, and this enables the pulmonates to exploit a wider variety of habitats.

Second, the pulmonates have evolved various methods of reducing water loss during excretion and respiration. A new organ, a ureteric groove or tube, develops in pulmonates, and there is direct evidence that in some species at least water resorption from kidney filtrate does take place (Vorwohl, 1961; Martin et. al. 1965). In many pulmonates, the kidney opens inside the pallial cavity some distance away from the pneumostome, and the excretion products must then be flushed out. In the order Sigmurethra there are various combinations of ciliated grooves and-or closed tubes that arise from the kidney, pass posteriorly to the hindgut, then reflect forward and continue partway to or actually reach the pneumostome. The evolution of this tube permits water conservation, and having it exit direct to the exterior through the pneumostome means that excretory products can be got rid of without having to use extra water to flush them out of the pallial cavity. (Solem, 1978).

Third, the pulmonates are well equipped to make good use of the water which they have. The physiological mechanisms which allow this are revealed in the relative impermeability of the mantle collar to water (Machin, 1966), and in the ability to withstand considerable water loss over long periods of time. Helix can survive water loss of around 50 per cent body weight, Arion 60 to 66 per cent, and Limax 75 to 80 per cent. In such a desiccated state Helix can survive 10 to 11 months, and Helicella over a year (Hyman, 1967).

Fourth, an array of other modifications also comes into play in resisting desiccation. These include the development of retractable tentacles, and the use of slime to cut down evaporation from moist surfaces.

Superimposed on all these structural modifications are behaviour patterns which cumulatively reduce the effects of desiccation and predation. A generally nocturnal life style, retirement to crevices in dry conditions, aestivation, burrowing (in slugs), and withdrawal into the shell during periods of lowered humidity are all effective. Taken together, this battery of adaptations and behaviour patterns has given the molluscs major successes in conquering the land.

The Qualifications for Slugdom

The Path to Slugdom

Slug evolution is a gradual process, and there are plenty of living species to illustrate every stage. The transition from snail to slug involves a large number of inter-related changes. Reduction and loss of the shell is necessarily accompanied by changes in the mantle cavity, the excretory system, the free muscle system, the arrangement of the heart and respiratory surfaces, and by a reduction of the visceral hump with major reorganisation of the internal organs. Consider first the pallial cavity of a stylommatophoran snail. The roof is heavily vascularised for gas exchange, the hindgut runs along the parietal-palatal margin to the pneumostome, the heart and kidney abut the posterior margin, and in the Sigmurethra the ureter runs posteriorly and then recurves forward along the hindgut to the pneumostome. If this already complex arrangement becomes compacted, provision must be made for maintenance of the respiratory area. Heavier vascularisation can provide this up to a point, as in the amphibulimine Bulimulidae (Van Mol, 1972), but compaction may well go on to such a degree that the amount of pallial cavity roof available is inadequate, even with heavy vascularisation. In this case, vascularisation can invade mantle lobe or shell lap, it can be concentrated in pouch-like lobes that hang from the remaining roof margin (Solem, 1978), (figure 1), or, in the final stages of compaction the respiratory surface may expand into a closely packed feltwork of blind diverticula, as in the Athoracophoridae (Figs 4 and 5). The last case will be covered in more detail later. As compaction proceeds, the other organs abutting the mantle

Legend to Figures

a	atrium
a.gl.	albumen gland
an.	anus
b.m.	buccal mass
d.g.	digestive gland
e.p.	excretory pore
g	gonad
h.d.	hermaphrodite duct
i	intestine
l	lung
l.d.	lung diverticula
oes.	oesophagus
o.gl.	oviducal gland
ov.	oviduct
p	penis
p c.	pericardium
p.gl.	prostate gland
pn.	pneumostome
p.r.m.	penis retractor muscle
r	rectum
ren	kidney
s.d.	secretory diverticulum
s.o.	sense organ
sp.	spermatheca
sp.d.	spermathecal duct
s.r.	shell rudiment
st.	stomach
u.	ureter
u.i.t.	uretero-intestinal tubule
v.	ventricle
v.d.	vas deferens
♂	male duct
♀	female duct
Scale divisions in mm

page 52

Fig. 1 Respiratory surfaces in snails and slugs. (a-d redrawn from Solem, 1974).

cavity become restricted to less and less space, and are flattened and distorted. The relationships of the kidney, heart, and hindgut undergo major changes as the visceral hump diminishes. These changes, of course, vary according to the group being studied, but some clear sequences have been documented. An example is the sequence found in the Thailand Helicarionidae (Solem, 1966) shown in figure 2. Here transitional stages can be clearly seen as the shell becomes reduced in size. As the pallial cavity is reduced, the structures associated with it must be crammed into smaller and smaller volume, and this means that the kidney and ureter in particular undergo radical shape changes. Finally the shell becomes so reduced in size that it can be partly or completely enclosed by flaps derived from the mantle margin, as in Austenia and Muangnua. Other examples of these compaction sequences abound, all of them different — but all can be interpreted as different means of coping with the problems which result from a reduction in pallial space.

As the pallial cavity becomes reduced, the stomach moves forward, and the oesophagus is shortened. In the Thailand helicarionid Muangnua it is almost non-existent, and in the Athoracophoridae it has migrated onto the top of the buccal mass and is extremely short (Fig. 3) (Burton, 1962, 1980). Other space adjustments can be seen in the reproductive system, where a major evolutionary trend involves the progressive fusion of the male and female ducts, with the male duct becoming a ciliated cleft in the wall of the female duct. This allows the female duct to be relatively narrower, as it can distend during the passage of eggs until the male duct in the wall of the female duct is virtually obliterated. This topic will be further explored later. page 53

Fig. 2 Pallial area changes in Thailand Helicarionidae. Redrawn from Solem, 1966.

Finally, the free muscle system, so essential in the snail for retraction of the animal into the shell, undergoes radical reduction. The elaborate tail fan needed to retract head and foot in the snail disappears, and only a small buccal retractor, tentacle retractors, and a penis retractor muscle remain (Solem, 1974). page 54

Fig. 3 Athoracophorus bitentaculatus, internal anatomy in situ, with the dorsum reflected to show the position and arrangement of the pallial complex.

Sluggishness in the Athoracophoridae

The Athoracophoridae are a family of slugs found in New Guinea, eastern Australia, the Bismarck Archipelago, the Admiralty Islands, the New Hebrides, New Caledonia, New Zealand, and the subantarctic page 55 islands. They have always been considered aberrant, and regarded with deep suspicion by a number of malacologists, but recent work on the group (Burton, 1977, 1978, 1980, 1981a, 1981b in press; Barker 1978) has led to a reversal of many previous ideas, and it is now possible to assess the group in the light of the processes involved in slug evolution outlined above.

The family comprises two subfamilies, the northern Aneiteinae and the New Zealand and subantarctic Athoracophorinae, and contains only slugs — no semi-slugs or snails need apply. The shell is reduced to a number of calcareous deposits embedded in the remains of the mantle collar, and in many specimens even these are missing. There is no sign of a visceral hump, and the pallial complex has become flattened and compacted to a high degree, with a highly convoluted and elongated ureter and a compact, high-surface-area lung of a type found in no other mollusc. There is no question that the Athoracophoridae as a group have advanced far down the road to slugdom. Obviously, there can be no fossil evidence to show the details of the path they have taken, and all the clues must be gleaned from a study of comparative anatomy. Nevertheless, the Athoracophoridae are a varied group, and a close study reveals that the clues are plentiful. The New Zealand and subantarctic species in particular show a rich web of inter-relationships and a gradation of answers to specific problems. The conceptual tool which allows fruitful exploration of the problems posed by the Athoracophoridae is Solem's work on the effects of compaction (Solem, 1966, 1972, 1974, 1978).

The Athoracophorid structures that first arouse loud shouts of “aberrant!” among malacologists are the lung in particular and the pallial complex in general. The lung consists of a dorsal cavity opening to the exterior through a pneumostome; a feltwork of thin-walled diverticula radiate ventrally and laterally from the floor of the cavity. A blood sinus encloses these diverticula, and blood drains from the sinus directly into the atrium of the heart (Figs 4a, 4b, and 5). The efficiency of this system is not known, but there is no question that it provides a large respiratory surface crammed into a small, flattened space. Thus the athoracophorid slug is able to do without the complex mantle lobes and folds seen as respiratory surfaces in other slug families (Fig. 1). Further, the pallial complex as a whole shows a high degree of modification and compaction (Burton, 1981a in press). The complexly-folded ureter is a case in point. The Athoracophoridae lack a mantle cavity as such — the lumen of the lung could be regarded as a remnant of the mantle cavity, but it is reduced in size, and has lost all the functions of a mantle cavity apart from respiration. As the lung does not store pallial water, the respiratory surface is on the floor of the cavity rather than the roof. The function of pallial water storage has apparently been taken over by the ureter, which, particularly in the New Zealand and subantarctic slugs, has become complexly folded and much expanded. (Figs 4a, 4b, and 5). It seems possible that the water resorption properties of the ureter are not highly developed in the Athoracophorinae. Significantly, these slugs possess a tubule connecting the ureter with the intestine, and this tubule may allow water stored in the ureter to pass back to the gut and be reabsorbed in times of water-stress. The tubule enters the intestine just before it joins the rectum, and presumably the water can be drawn out of the rectum as the intestinal contents pass through. The northern representatives, the Aneiteinae, lack this tubule, and the arrangement of the pallial complex shows other primitive characters. In these slugs the anus, renal aperture, and pneumostome are closely grouped on the dorsolateral aspect of the back. page 56

Fig. 4 Pallial complexes in the Athoracophoridae. (a) Triboniophorus graeffei (Australia) (b) Pseudaneitea dendyi (Canterbury, N.Z.) Both drawn from serial sections.

This close association is recognised as a basic gastropod condition (Hyman, 1967; Purchon, 1968). In the New Zealand and Subantarctic Athoracophorinae this association is broken to various degrees (Figs 6 and 7). The anus migrates away from the pneumostome (Palliopodex verrucosus) then leaves the mantle area altogether (Pseudaneitea gigantea), then migrates down to the margin of the back (P. aspera, Athoracophorus page 57

Fig. 5 Pallial complex of Athoracophorus bitentaculatus, exploted view. Drawn from serial sections. Not to scale.

bitentaculatus). The pneumostome, which is no longer associated with the anus, is free to move medially, and the renal orifice migrates anteromedially. Once the close association seen in the Aneiteinae is broken, each orifice is free to move to the most advantageous position (Burton, 1980). The shift in anal position may be associated with the desirability of keeping the dorsum free of faecal material. Even in the Aneiteinae, which have the anus inside the mantle area, the breadth of the mantle area makes it possible for the slug to extrude faeces on to the side of the dorsum rather page 58

Fig. 6 Compaction and flattening in slugs. (a) Limax maximus; (b) Triboniophorus graeffei; (c) Pseudaneitea gigantea; (d) Palliopodex verrucosus; (e) Pseudaneitea dendyi; (f) Athoracophorus bitentaculatus.

than the top, since the slugs are rounded in cross section (Figure 6). The New Zealand slugs, notably Athoracophorus bitentaculatus, are dorsoventrally flattened, and in all these species the anus has migrated to a position at the margin of the back, so that faecal material never soils the dorsum. Clearly, in this respect the Aneiteinae show the primitive condition, and that in the Athoracophorinae is derived. The process of flattening and compaction has obviously been at work in the group.

Another example of the results of compaction is seen in the reproductive system. Examination of athoracophorid reproductive systems sheds some light on the evolution of pulmonate and opisthobranch reproductive systems in general. Since the late 1880's it has been traditional to state that a monaulic (having combined male and female ducts as a single gonoduct) pallial “spermoviduct” is a primitive stage, and that the advanced condition is shown by the separation of the male and female ducts. Plate page 59

Fig. 7 Changes associated with flattening in the Athoracophoridae. The representative chosen as the starting point is Triboniophorus graeffei.

(1898), Pelseneer (1935), Morton (1955), Duncan (1960), Rigby (1963, 1965), Harry (1964), and Ghiselin (1965) all subscribe directly or by inference to the theory that the initial stages of hermaphroditism superimposed the male duct on the female, resulting in a single gonoduct passing from posterior to the pallial cavity to the external gonopore. Solem (1972) takes precisely the opposite viewpoint, arguing that the separation of the pallial gonoducts is widely established in groups that are considered to be more primitive than and-or ancestral to the Stylommatophora. Here, the situation in the Athoracophoridae is instructive. They are the only family in which both the fused and separate conditions of the gonoducts are found (Burton, 1980), with three species (Palliopodex verrucosus from the Auckland Islands, Pseudaneitea ramsayi from Three Kings Islands and Athoracophorus bitentaculatus on the mainland) being monaulic, and all the others showing the diaulic conditions (Figs 8, 9, and 10). In the light of the controversy outlined above, which came first? On balance, it seems most likely that in the Athoracophoridae the diaulic condition is primitive, the monaulic condition is derived, and the three monaulic species arrived at the condition independently. First, the northern Aneiteinae, demonstrably more primitive than the Athoracophorinae in the arrangement of the pallial complex, all show the diaulic condition (Solem 1959, Oberzeller 1870, Burton 1980). Second, the New Zealand and subantarctic species show a lot of variation in their size and in their body cross section, i.e. in their state of compaction. They range from large rounded slugs such as Pseudaneitea gigantea and P. papillata to small, highly compacted, very page 60

Fig. 8 Pseudaneitea dendyi, reproductive system.

flattened slugs like P. ramsayi and Athoracophorus bitentaculatus. P. gigantea in particular shows a number of primitive characteristics, such as a dorso-lateral anal position close to the pneumostome, a smooth skin, and a well-developed central radula tooth, a condition recognised as primitive in other groups (Solem, 1959); its reproductive system shows the diaulic condition, as do those of all the other large, rounded species (Burton, 1980). The three monaulic species all show the effects of compaction; they are small and flattened, and can on this account be regarded as advanced. However, they are not closely related, and the land masses on which they occur have never been in contact (Fleming, 1979). Third, the fusion of gonoducts as part of a general process of compaction seems more likely than their separation, as fusion effectively gives two ducts in the space of one.

In the three monaulic athoracophorid species, the oviducal gland forms the wall of the oviduct, rather than being a discrete gland attached to one side, and the prostate gland is a diffuse structure emptying into the male duct at intervals along its length. This arrangement of the prostate gland is presumably necessary because prostatic fluid can escape from the male duct along its length and must be replenished at intervals. The monaulic arrangement is most similar to that seen in the Aneiteinae, which have an elongate, lobular prostate gland feeding into the vas deferens at intervals, page 61

Fig. 9 Athoracophorus bitentaculus, reproductive system, exploded view. Drawn from serial sections.

page 62

Fig. 10 Suggested evolutionary trends in athoracophorid reproductive systems.

page 63 and an oviducal gland incorporated into the wall of the oviduct, and it is tempting to speculate that the monaulic arrangement is derived from such a condition. However, as all our information is derived from living species there is no real evidence to support this conjecture, and it is very difficult at this stage to arrive at any firm conclusions. It would appear that Solem's hypothesis explains the unique situation seen in the Athoracophoridae more satisfactorily than does the traditional view, and that the three monaulic species arrived at the condition independently, but these conclusions must be treated with reserve.

Similarly, taking the compaction hypothesis into account, some evolutionary trends could be inferred in the reproductive systems of the diaulic Athoracophorinae. Such trends include the development of a discrete prostate and a distinct oviducal gland, and a general shortening of the reproductive tract (Fig. 10). It must be stressed that this figure does not imply an evolutionary relationship between the species named; each species simply illustrates a different stage in a possible evolutionary sequence.

Thus the process of compaction has been carried almost to an extreme in the Athoracophoridae, and in Athoracophorus bitentaculatus in particular. In this species it has resulted in a slug so flattened that it can easily fit into small crevices and spaces between leaves denied to fatter slugs, and thus gain protection from desiccation and predation. Second, compaction and flattening confers on the slug a much higher foot area-weight ratio, which improves its ability to climb up vertical surfaces and allows it to browse on trunks and leaves at night.

The Athoracophoridae as a group are highly advanced in some respects. Solem (1974) has described slugs in general as representing the acme of land snail evolution. As Athoracophorus bitentaculatus in particular has carried the process of compaction to its logical extreme, it can be regarded as a supreme example of the current state of the art in the land mollusca.

References

BARKER, G. M. 1978: A reappraisal of Athoracophorus bitentaculatus, with comments on the validity of genus Reflectopallium (Gastropoda: Athoracophoridae). N.Z. Journal of Zoology 5 (2):281-8.

BLINN, W. C. 1964: Water in the mantle cavity of land snails. Physiological Zoology, 37: 329-37.

BURTON, D. W. 1962: New Zealand land slugs — Part 1. Tuatara 9 (8):87-97.

— 1977: Two new species of Pseudaneitea Cockerell (Athoracophoridae: Gastropoda) from New Zealand. Journal of the Royal Society of N.Z. 7 (1): 93-8.

— 1978: Anatomy, histology, and function of the reproductive system of the tracheopulmonate slug Athoracophorus bitentaculatus (Quoy and Gaimard). Zoology Publications from Victoria University of Wellington 68: 1-16.

— 1980: Anatomical studies on Australian, New Zealand, and subantarctic Athoracophoridae (Gastropoda: Pulmonata). N.Z. Journal of Zoology 7:173-98.

BURTON, D. W. 1981a: Pallial systems in the Athoracophoridae (Gastropoda: Pulmonata). N.Z. Journal of Zoology 8: 391-402.

BURTON, D. W. 1981b: The status of the genus Reflectopallium Burton (Athoracophoridae: Gastropoda). Journal of the Royal Society of N.Z., in press.

DUNCAN, C. J. 1960: The evolution of the pulmonate genital system. Proceedings of the Zoological Society of London, 134(4):601-9.

FLEMING, C. A. 1979: The Geological History of New Zealand and its Life. Auckland University Press.

GHISELIN, M. 1965: Reproductive function and the phylogeny of opisthobranch gastropods. Malacologia 3(3):327-78.

HARRY, H. H. 1964: The anatomy of Chilina fluctuosa Gray re-examined, with prologomena on the phylogeny of the higher limnic Basommatophora (Gastropoda: Pulmonata). Malacologia 1(13):355-85.

HYMAN, L. H. 1967: The Invertebrates: Mollusca I. McGraw-Hill, New York.

MACHIN, J. 1966: The evaporation of water from Helix aspersa. IV. Loss from the mantle of the inactive snail. Journal of experimental Biology 45:269-78.

MARTIN, A. W., STEWART, D.M. & HARRISON, F. M. 1965: Urine formation in a pulmonate land snail Achatina fulica. Journal of experimental Biology, 42:99-123.

MORTON, J. E. 1955: The evolution of the Ellobiidae with a discussion on the origin of the Pulmonata. Proceedings of the Zoological Society of London, 125:127-68.

OBERZELLER, E. 1970: Ergebnisse der Osterreichischen neukaledonien-Expedition 1965. Terrestriche Gastropoda, II: Veronicellidae und Athoracophoridae. Annalen des naturhistorischen Museums in Wien 74:325-41.

PELSENEER, P. 1935: Essai d'ethologie zoologique d'apres l'etude des mollusques. Academie Royale de Belgique, Classe des Sciences, Publications de la Fondation Agathon de Potter, 1: 1-662.

PLATE, L. H. 1898: Beitrage zur Anatomie und Systematik der Janelliden. Zoologische Jahrbucher. Abtheilung fur Anatomie und Ontogenie der Thiere (Jena) XI: 193-269.

REES, W. J. 1964: A review of breathing devices in land operculate snails. Proceedings of the Malacological Society of London, 36:55-67.

RIGBY, J. E. 1963: Alimentary and reproductive systems of Oxychilus cellarius (Muller) (Stylommatophora). Proceedings of the Zoological Society of London, 141(2):311-59.

RIGBY, J. E. 1965: Succinea putris: a terrestrial opisthobranch mollusc. Proceedings of the Zoological Society of London, 144(4):445-86.

RUNHAM, N. W. & HUNTER, P. J. 1970: Terrestrial Slugs. Hutchinson, London.

SOLEM, A. 1959: Mollusca of the New Hebrides. Fieldiana, Zoology 43(1-2).

— 1966: Some non-marine molluscs from Thailand, with notes on classification of the Helicarionidae. Spolia Zoologica Musei hauniensis 24:1-108.

— 1972: Tekoulina, a new viviparous Tornatellinid land snail from Rarotonga, Cook Islands. Proceedings of the Malacological Society of London 40:93-114.

— 1974: The Shell Makers. John Wiley and Sons, Canada.

— 1978: Classification of the land Mollusca. Pp. 49-98 in V. Fretter & J. Peake (Eds), Pulmonates, vol. 24, Systematics, Evolution and Ecology. Academic Press, London.

VAN MOL, J. J. 1972: Notes anatomiques sur les Bulimulidae (Mollusques, Gastropodes, Pulmones). Annales de la Societe Royale Zoologique de Belgique 101(3):183-225.

VORWOHL, G. 1961: Zur Funktion der Exkretionsorgane von Helix pomatia L. und Archachatina ventricosa Gould. Zeitschrift fuer vergleichende Physiologie 45:12-49.