Behaviors guided by chemical senses

Chemical guidance of predatory behavior has been extensively studied in snakes and lizards. Many species exhibit attack and ingestive behaviors on presentation of chemical cues derived from prey. In these experiments, the chemicals typically are presented on cotton-tipped applicators in the absence of any visual or tactile cues associated with prey. Therefore we can be certain that chemicals alone are responsible for triggering the predatory actions. This does not mean that garter snakes (Thamnophis sirtalis), for example, pay no attention to other stimuli arising from prey (fish, sala manders, frogs, worms) under natural conditions. We know that cues such as movement are vital for attracting the snakes' interest. Yet if investigatory behavior does not bring the snake into contact with appropriate chemical cues, the snake is unlikely to bite the target. Visual cues may command attention and lead to careful inspection, whereas chemical cues trigger the final consummatory acts. Of considerable interest is that neonatal garter snakes exhibit such responses to certain chemical cues before having any experience with prey. This finding leads to the conclusion that responses to these cues are innate. Furthermore, neonatal garter snakes born in different parts of the geographic range respond most strongly to different prey extracts. This finding indicates that subpopulations of snakes have experienced differential selection based on variation in prey abundance. In general, the chemical cues to which snakes respond most strongly are those associated with prey that happen to be available in the snakes' habitat. Changes in prey populations are followed by changes in prey recognition mechanisms within the snake population.

Because reptiles possess nasal and vomeronasal chemosen-sory systems, herpetologists have been keen to learn the relative contributions of these two systems to predatory behavior. Experimenters have developed various techniques for blocking one or both of the systems to study predatory behavior. Garter snakes have been subjects of most of these investigations, although a few studies have involved western rattlesnakes (Crotalus viridis). The common result has been that prey recognition remains undisturbed when the nasal system is blocked but that this behavior almost disappears when the vomeronasal system is blocked. In some particularly elegant experiments, the blocks were reversible, and restoration of the vomeronasal system was followed by a return of the abilities to recognize and respond appropriately to prey. This body of research leaves no doubt about the importance of the vomeronasal organs.

The role of the nasal system and its interaction with the vomeronasal system remains a matter of speculation. Most investigators believe that the nasal system is extremely sensitive but not particularly discriminatory, whereas the reverse is true of the vomeronasal system. The nasal system is thought to serve an alerting function. It informs the snake that something of interest is nearby and sets the vomeronasal system in motion in the form of tongue flicking. Vomeronasal exami

Reptilian visual displays: 1. Cottonmouth uses gaping mouth as a defensive warning; 2. Frilled lizard looks larger as a defensive display; 3. A ring-necked snake draws attention away from its head and shows its coloration as a defense; 4. The alligator snapping turtle uses a food lure to attract its prey; 5. and 6. Territorial or mating displays for conspecifics—green anole (5) and tuatara (6). (Illustration by Dan Erickson)

Reptilian visual displays: 1. Cottonmouth uses gaping mouth as a defensive warning; 2. Frilled lizard looks larger as a defensive display; 3. A ring-necked snake draws attention away from its head and shows its coloration as a defense; 4. The alligator snapping turtle uses a food lure to attract its prey; 5. and 6. Territorial or mating displays for conspecifics—green anole (5) and tuatara (6). (Illustration by Dan Erickson)

nation can identify the molecules in question and activate the appropriate behaviors. In this view, nasal olfaction functions as vision or detection of vibrations does in that ambiguous stimulation of any of these senses can activate tongue flicking so that the animal can conduct a definitive examination. Unambiguous stimulation of any sense, usually by an approaching predator, typically activates immediate escape without the need for cross-modal verification. Detection of potential prey is frequently an ambiguous matter because of prey crypsis (camouflage) and consequent reduction in intensity of stimuli and requires the synergistic action of multiple sensory systems. In the latter context, the vomeronasal system performs the "gold standard" test, and the other senses invite the vomeronasal system to conduct the test.

The vomeronasal system plays an equally important role in the reproductive behavior of garter snakes. In experiments, males with impaired vomeronasal systems did not follow the trails of estrous females and did not attempt to court or copulate with such females when they encountered them. When the vomeronasal systems of the males were restored, normal sexual behavior reappeared, including the ability of the males to follow trails deposited by females. Male garter snakes

Egg laying strategies. 1. Peninsula cooter turtle; 2. American alligator; 3. Python; 4. Copperhead. (Illustration by Dan Erickson)

differentiate the trails of conspecific and heterospecific females only through chemical cues. Males can differentiate large and small (but reproductively mature) conspecific females with chemical information alone. In experiments, males preferred chemical cues from large females over those of smaller females, a fact probably correlated with the greater number of eggs produced by large females. In other words, mating with larger females produces greater fitness benefits for males than does mating with smaller females, and males reflect the effect of this selective pressure in their chemosensory preferences. When the length of females was controlled but mass was varied, males preferred the heavier females and did so when only chemical information was available. Male garter snakes apparently can use chemical information to discriminate females of varying nutritional conditions. This capability probably is associated with an effect of nutritional condition on quantity and quality of eggs produced by females.

Although few other reptile species have been studied in as much detail as have garter snakes, especially T. sirtalis, we know that male skinks can do some of the things that male garter snakes can do. In particular, male skinks can differentiate conspecific and heterospecific females with only chemical cues. Likewise, female skinks can differentiate chemical cues of conspecific and heterospecific males. It seems probable that male skinks also are able to differentiate conspecific females of varying size and condition, but these phenomena have not yet been tested. Nor has anyone tested whether fe male skinks or garter snakes can differentiate chemical cues derived from males of varying size or condition. We assume the vomeronasal systems of garter snakes and skinks are responsible for discrimination, but this has not yet been demonstrated in experiments that block the vomeronasal system while the olfactory system is unimpaired.

Chemical cues associated with predators are detected by various reptiles. The result is avoidance or other self-protection reactions. We cannot yet be certain that the vomeronasal system mediates these reactions because appropriate blocking experiments have not been conducted, but this is a reasonable hypothesis.

Some reptiles have been shown to discriminate between their own chemical cues and those of conspecifics. Some male reptiles can differentiate male conspecifics on the basis of chemical cues. Experiments along these lines have been done with rattlesnakes, desert iguanas, and sand swimming skinks. It is likely the abilities will be found to be widespread. A number of benefits can be derived from individual recognition, including the interesting social dynamic known as the "dear enemy" phenomenon. Common among territorial birds, the phenomenon consists of individual recognition by neighboring territorial males. These animals respect the boundaries between their territories and do not intrude on each other; thus each animal is allowed to relax the level of vigilance and aggressive behavior that would otherwise be devoted to boundary patrolling and defense. If one of the males is re placed with a new individual, the remaining neighbor exhibits an immediate elevation in vigilance and defense, revealing that he notices that his former dear enemy is no longer present. Eventually the original territory owner and the new neighbor enter a dear enemy relationship, so that each can save energy and avoid injury from fighting. The dear enemies enjoy mutual benefits by respecting each other's property rights. Because individual recognition is clearly an important component of this behavior, the dear enemy phenomenon is generally thought to be an advanced form of social interaction. This does not mean that it is found only in birds and mammals; the behavior has been shown to occur in salamanders and lizards. Whether it occurs in any species of snake, turtle, or crocodilian is conjectural at this time, although a chemo-sensory basis for the phenomenon clearly exists in at least some species.

Perhaps the most intriguing chemosensory behaviors of reptiles involve venom. Venom immobilizes and kills prey and contains powerful enzymes that greatly facilitate digestion. In experiments, rattlesnakes given a choice between enveno-mated and nonenvenomated mice otherwise equal in size, age, and sex selected the envenomated prey more frequently than they did the nonenvenomated prey. This remained true when the mice were wrapped in dark nylon mesh that blocked visual or tactile cues; thus we can be certain that chemical cues mediated the selection. When rattlesnakes were given a choice between a trail deposited by an envenomated mouse and one deposited by a nonenvenomated mouse, the snakes reliably selected the former trail, a finding that again indicates chemical cues mediate the behavior.

To understand how sensory bias might contribute to a snake's fitness, it is necessary to understand that rattlesnakes and many other venomous species are ambushers that strike and release adult rodent prey. This style of predation avoids injuries that could occur if struggling rodents were held in the snake's jaws after the strike. Although the venom kills the prey, the process takes as long as several minutes, during which the rodent's teeth, claws, and guard hairs can inflict serious damage. Releasing the rodent after the envenomating strike minimizes the snake's risk of injury, but the snake risks losing the prey, which can travel several meters from the site of attack while the venom takes effect. Recovering the rodent carcass becomes difficult, partly because the immobilized rodent no longer emits motion cues to attract the snake's attention and partly because the rodent may be behind objects or in a burrow, so thermal information (i.e., infrared radiation detectable by pit organs) is obscured or blocked entirely. Recovery of the rodent carcass therefore is mediated by chemical cues detected by the vomeronasal system. A rattlesnake that has delivered a successful predatory strike is capable of following the rodent's chemical trail with exactitude, even though the same snake usually does not follow such a trail in the absence of a predatory strike. The strike is necessary to trigger chemosensory searching and trail-following behavior. This is true for rattlesnakes; cottonmouths (Agkistrodon piscivorus), however, follow trails effectively whether or not a predatory strike has been delivered before the snake encounters the trail. Gila monsters (Heloderma suspectum) and Australian sand goannas (or monitors) Varanus gouldii, behave as cottonmouths do. For these predators, chemical cues on the substrate are sufficient to induce trail following, and this behavior may be associated with a propensity to search for carrion. For rattlesnakes and various other vipers, chemical cues on the substrate usually are not sufficient, and a successful predatory strike is critical.

In the case of rattlesnakes, two trails always are present in the poststrike environment, one deposited as the prey wanders into striking range (the preenvenomation trail segment) and one deposited as the prey moves away from the site of attack (the postenvenomation trail segment). Following the wrong, or preenvenomation, segment could result in losing the prey, wasted energy, and vulnerability to predators. It is not surprising that rattlesnakes discriminate the two trails and select the postenvenomation segment. What is the chemical basis for this fascinating discrimination?

Venom is a complex material, containing well over 30 identifiable fractions, some of which are lethal to rodents, some not. The latter fractions are thought to be synergizers or amplifiers, not necessarily harmful in themselves but capable of increasing the damaging effects of other fractions. Finding the particular fractions responsible for the snake's discrimination of pre- and postenvenomation trail segments is almost impossible. One reasonable hypothesis is that proteolytic enzymes, major constituents of venom, are responsible. The idea is that such enzymes break down rodent protein and contribute to immobilization, death, and digestion of prey and that the snakes have evolved a perceptual sensitivity to these effects. This sensitivity facilitates the important task of locating and following the postenvenomation trail segment. In other words, proteolytic enzymes have an initial or primary function, and they have acquired a secondary one because of the snake's ability to perceive some of these primary chemical effects. Another hypothesis is that specialized components have been added to venom, not necessarily because of their

A male panther chameleon (Furcifer pardalis) behaves defensively by widely gaping its mouth, extending its gular pouch with its tongue, rearing up on its hind legs, laterally compressing its body, and hissing. These displays are common in the genera Chamaeleo, Calumma, Bradypodion,and Furcifer.If this display fails, the chameleon must attempt to flee to avoid predation. (Photo by Ardith Abate. Reproduced by permission.)

A male panther chameleon (Furcifer pardalis) behaves defensively by widely gaping its mouth, extending its gular pouch with its tongue, rearing up on its hind legs, laterally compressing its body, and hissing. These displays are common in the genera Chamaeleo, Calumma, Bradypodion,and Furcifer.If this display fails, the chameleon must attempt to flee to avoid predation. (Photo by Ardith Abate. Reproduced by permission.)

Green seaturtles (Chelonia mydas) mating. (Photo by Animals Animals ©H. Hall. Reproduced by permission.)

lethal effects but because of their perceptual, trail-enhancing effects. The term "trail marker substance" can be used in recognition of the idea that the venom component has acquired its perceptual role not secondarily but as its primary (and perhaps only) function.

Differentiating these hypotheses hinges on the fact that rattlesnake venom passed through gel filtration columns separates into fractions according to molecular weight and that proteolytic enzymes sort into several fractions while other fractions contain little or none of this material. Therefore it is possible to use these various fractions as experimental injectants. That is, mice given injections of each of these fractions suspended in distilled water can be paired with control (nonenvenomated) mice and presented to rattlesnakes. If proteolytic enzymes are the critical elements, then snakes ought to be able to discriminate mice injected with these enzymes from control mice. If such mice cannot be differentiated from controls, one implication would be that some other venom fraction is the critical one. The result in experiments with western diamondback rattlesnakes (Crotalus atrox) has been that proteolytic enzymes do not cause a mouse to be discriminated from a control but that another fraction (containing no proteolytic enzymes) has this effect. The chemical composition of this fraction has not yet been identified, but the indications are that the second hypothesis is correct. This work underscores the subtlety of the chemical cues used by reptiles. Reptiles can detect gross cues, such as those associated with rotting carrion, and they can use remarkably subtle cues in remarkably low concentrations.

Ambush predators must first select an appropriate site, a place which prey are likely to visit or to pass through. Several reptiles have been shown to make this selection on the basis of chemical cues deposited by prey that have recently moved through the site. Prairie rattlesnakes making their ver nal migration would stop if they were to encounter fresh chemical cues derived from deer mice (Peromyscus manicula-tus). Once the rattlesnake occupies an ambushing site, however, it usually waits until a successful predatory strike is delivered before paying any further attention to chemical trails. There are exceptions, of course: Some snakes engage in active foraging for burrows containing the neonates of various small mammals, and this search is undoubtedly guided by chemical cues. Thus strikes are not always necessary to activate chemosensory searching. It seems likely that reptiles of many species can use chemical cues to discriminate large versus small populations of prey and to differentiate sites currently occupied by prey from sites previously occupied but now abandoned. Evidence along these lines has been collected, but a great deal of research remains to be done on the topics of habitat selection and chemical ecology.

Although heavy reliance on the nasal and vomeronasal senses is characteristic of many species of reptiles, particularly the most advanced species, numerous species are less reliant on chemical cues than on visual information. Chameleons are examples not only in their use of visual cues in locating and capturing prey but also in their use of visual cues in social and reproductive behavior. Interspecific variation exists among reptiles in the modalities that mediate important behaviors. Therefore statements about the role of vomeronasal stimulation should not be generalized to all reptiles. The relationships between type of food, behaviors involved in acquiring food, and the sensory modalities used are a major area of her-petological investigation. Herbivorous lizards have been found to behave in surprising ways. These animals not only detect edible plants through chemical cues but also detect and reject other plants on the basis of the presence of defensive compounds. Herpetologists have made theoretically important discoveries regarding the chemical senses of reptiles, but we have probably only scratched the surface.

0 0

Post a comment