Proc. Nat. Acad. Sci. USA Vol. 71, No. 3, pp. 966-970, March 1974 Olfactory Basis of Homing Behavior in the Giant Garden Slug, Limax maximus (digitate ganglion/locomotion/orientation/terrestrial pulmonate) ALAN GELPERIN Department of Biology, Princeton University, PrincetQn, New Jersey 08540 Communicated by V. G. Dethier, November 14, 1973 ABSTRACT Time lapse photography of slugs living in in the soil and forage over an area extending at least 4.5 an experimental enclosure shows that these animals can meters from the home (14). Time lapse photography of grey return to a homesite from over 90 cm by a direct route. Slime trail following and vision are not involved in this field slugs, Agriolimax reticulatus, locomoting on an enclosed behavior. In the presence of a low velocity wind, homing soil surface shows that the animals often return to the same occurs upwind. Surgical disconnection of the presumptive hole in the soil from which they emerged earlier in the night olfactory apparatus (digitate ganglion) from the central (15). The present work documents homing behavior in Limax nervous system eliminates homing. Neurophysiological and initial physiological investigation of recordings from the receptor surface associated with the maximus presents digitate ganglion and the olfactory nerve demonstrate its sensory basis. the olfactory function of the digitate ganglion. The olfac- tory acuity and capacity for directed locomotion via olfac- MATERIALS AND METHODS tory cues are also relevant to studies of slug feeding be- havior, ecology, and learning ability. The behavioral experiments were done on slugs confined to a 1.5 by 1.7-meter area of moist filter paper bounded by a 2- "Simple" animals often reveal their possession of sophisticated inch wide border of crystalline NaCl. An inverted clay flower behavioral machinery when experimental questions are asked pot with four semicircular notches cut in its lip, was centrally in the proper context. This is nowhere better documented than located and served as the animal's daytime resting site. The in studies of orientation. The sun-compass orientation mecha- filter paper was kept moist by inverted water reservoirs. Food nism of bees and ants (1), the ability of noctuid moths to items such as carrot (Daucus carota), potato (Solanum tubero- steer their flight path away from bats using two sense cells sum). dog food (Ken-L-Ration), or mushroom (Boletus edulis) (2), and the apparent use of hydrodynamic cues by migrating were supplied in a petri dish at one corner of the arena. Fluo- lobsters (3) are examples of complex neural mechanisms which rescent room lights provided illumination and were automati- became apparent when physiological experiments were done cally controlled to produce a cycle of 12 hr of light and 12 hr in an ethological context. The present experiments on homing of darkness. In some experiments, a plastic covering was used in Limax maximus were undertaken to probe the complexity to shield the arena from air currents. of behavior possible in a preparation amenable to cellular A 16-mm camera modified for time lapse operation and neurophysiological analysis (4). equipped with a wide-angle lens was mounted vertically 2.7 Homing behavior has been documented in a wide variety of meters above the experimental arena. A xenon bulb flash unit molluscan species. Aristotle described the homing behavior with a flash duration of 2 msec was triggered synchronously of limpets and experiments to date are still searching for the with the camera shutter. The charging capacitor in the flash sensory basis of this behavior (5, 6). Octopus can return to its unit was changed from 300 uf to 40 Mf to produce the least nest after forays covering considerable distances (7). Small intense flash which would give a distinct image oln plus-X colonies of the intertidal pulmonate Onchidium nest in rock negative film with the lens wide open. A framing rate of 4 per crevasses and after a period of feeding away from the nest, minute was used. Typically the camera was activated from all members of a particular colony simultaneously return di- 1700 hr to 0900 hr the following day. No behavioral response rectly to their nest (8). Capture and release experiments have could be detected to the light pulse emitted by the flash unit. shown that the garden snail, Helix pomatia, can return to More than 1600 hr of activity were filmed and analyzed using sites favorable for overwintering with an angular error of less an analytical projector. than 300 over distances up to 40 meters (9). The sea hare The 25 slugs used in these experiments were Limax maximus Aplysia is diurnally active and returns to a particular location and included both locally collected animals and individuals in its tank of seawater to sleep (10). reared from eggs in the laboratory. No differences in homing The literature contains scattered suggestions of homing be- ability between these two categories of animals were observed. havior among slugs. The ability of Limax maximus to show Anatomical studies oln the optic tentacles were done using homing behavior has been referred to anecdotally by Taylor standard histological techniques to prepare 8-,Am serial sec- (11), Pilsbry (12), and Frdmming (13). Field observations of tions of whole tentacles stained with Mallory Heidenhain's the California banana slug, Ariolimax columbianus, suggest triple stain (16). Neural pathways in the digitate (= tentacu- that animals establish a homesite by excavating a depression lar) ganglion were stained using axonal iontophoresis (17) to introduce Co++ ions into axon.s. The tissue was then treated Abbreviation: EOG, electro-olfactogram. with ammonium sulfide, dehydrated with ethanol, and cleared 966 Downloaded by guest on October 1, 2021 Proc. Nat. Acad. Sci. USA 71 (1974) Olfactory Basis of Slug Homing 967 A B' Ce D FIG. 1. Track of single slug during one night. in basic methyl benzoate (18). Ganglia so treated were studied and photographed as whole mounts. Electrical recordings from the receptor surface of an optic FIG. 2. Homeward paths of several different slugs. Arrow tentacle were made with saline-agar filled electrodes of tip provides constant compass direction reference. diameter 50-100 /Am connected to a neutralized input capaci- tance dc amplifier. Polyethylene suction electrodes were at- Fig. 2 presents 10 return Ipaths selected to represent the tached to the olfactory nerve after severing its central con- variation in directness of homing observed in this study. The nections. Both types of signals were recorded relative to a maximum distance from which homing occurred was 93 cm, ground electrode in the saline bath. The signals were dis- the outer limit of the arena. The animals used varied in body played on a multichannel oscilloscope and either photo- length from 7.5 to 16.5 cm, and inter-optic tentacle distance graphed directly or recorded using an FM tape recorder. varied from 1.5 to 2.5 cm. Fig. 2 also illustrates that the same The saline used had the following composition in mM: Na slug can use different paths home on successive nights (Paths 70, K 2.5, Ca 3.4, Mg 0.8, Cl 81, glucose 0.6, Tris 50. C, F, J) and that two slugs living together can use different paths home on the same night (Paths A, I). RESULTS The nonrandom nature of homing was tested mathemati- The animals spend the daylight hours in the dark and humid cally in the following way. A set of linear path segments with environment provided by the homesite. With a latency vary- origins close to the periphery of the arena was selected; only ing from several minutes to several hours after lights off, they those paths which were linear because the animal moved emerge and move about the arena at speeds ranging between along the salt barrier were excluded. For the path whose ori- 0.069 cm/sec to 0.26 cm/sec. These travels bring them in con- gin was closest to the home, the flower pot subtended 150 of tact with the salt barrier, the food dish, other slugs, and ulti- the horizon. I assume, for purposes of this test, that the animal mately, the homesite. Periods of locomotor activity are inter- selects his direction of travel from a 1800 sector. This assump- slpersed with periods of sleep, sexual activity, or feeding. The tion yields a probability of contacting the home by chance of return to the homesite is often quite direct and over virgin 15°/180° or 8.3%. The sample of 41 linear path segments con- territory. tained 13 (32%) which were homing paths, a clearly non- A representative tracing of the travels of a slug about the random distribution. experimental arena is shown in Fig. 1. Three of the trips are Experiments were then directed to the question of the sen- short and in the immediate vicinity of the pot. These short sory cue providing direction to the homeward path. The ani- trips predominate during the first night in the apparatus mals are not following slime trails home, although they can when the pot and filter paper are clean. Two of the trips are follow slime trails and do so routinely to locate sexual part- more extensive and have terminal segments which clearly ners. The use of visual cues is possible but unlikely. The op- suggest a directed locomotion back to the homesite. All of the tical system and fine structure of the slug eye suggest poor activity occurred during the dark period; the animal made its visual acuity (19) and light was available for only 2 msec final return home 2.2 hr before light onset. We do not know every 15 sec. The use of vision in homing was tested in two when the animal actually decided to return home, but the ways.