Diversity of natural chemical defenses against predation.  The use of chemical defenses among plants and animals that deter predation and the evolution of counter measures against these defenses by predators is one of the oldest and most common ecological interactions.  This interaction serves as a most intricate illustration of co-evolution and so has been discussed at length in introductory texts on the subject of evolution (7).  A partial list of animals that either manufacture or sequester defensive poisons includes cnidarians (8), mollusks (9), fishes (10), amphibians (11), and reptiles (12).  The diversity of insects with chemical defenses is so great that to date no one has constructed a satisfactory accounting but even a brief survey is impressive (13, 14).  The diversity of plants with toxic defenses far exceeds that of animals.  Indeed, many animals secondarily collect and sequester poisons manufactured by plants.  Two of the earliest surveys list over 2000 species of poisonous plants in North America alone (15, 16) and a more recent source on the web lists over 50,000 references to poisonous plants (17).  The diversity of plant poisons in lower latitudes of South America, Asia and Africa has yet to be characterized, but is almost certainly magnitudes greater than in North America and Eurasia where it has been studied more intensively.

The physiological effects of these substances when consumed by foragers range from subtle to spectacular depending upon amounts consumed, the nature of the toxin and the physiological defenses of the consumer.  Some effects may appear idiosyncratic but when the physiological effects are fundamental they are similar among a wide variety of consumers.  These range from transitory illness and incapacitation to protracted incapacitation, and death.  A few well-known examples of classes of toxic substances include neurotoxins (18), neurotransmitter and hormone mimics (19, 20, 21), and metabolic toxins (22, 23, 24, 25, 18). 

In addition, foods may deliver an even larger diversity of parasites and diseases that attack the gut of consumers or otherwise induce malaise after the food is consumed.  These are so diverse and wide spread that it beyond the scope of this report to enumerate them.  Unpredictability is increased by the fact that some foods that are safe to consume at some times and places may induce illness under other conditions.   In general, feeding in the natural world has become a maze of both predictable and unpredictable consequences that must be mitigated if foragers are to survive, grow and breed.

Defenses against food borne poisons.  Given the danger and extreme diversity of potential food poisons in most ecosystems is it not surprising that animal consumers appear to have evolved defenses enabling them to forage more freely (26, 27).  Indeed, it is partly the evolution of these consumer defenses that is thought to have driven the co-evolution of prey defenses.  Consumer defenses against intoxication include: 

1. Foraging strategies.  These behaviors guide consumers through their environment in a way that maximizes nutrition and reduces exposure to poisonous foods at times and in locations of peak poison concentrations (e.g. 28, 29).  The rarest and most extreme strategy may involve food specialization in which feeding is restricted to only a few different food items unlikely to contain toxins (e.g. nectivorous Honeycreepers of family Drepanidae (30).  Foraging may also include consumption of foods that counteract the effects of poisons in other foods or parts of foods.  Rats and some other animals have even been reported to consume clay or kaolin to remove poisons from the gut (31).
     
2. Physiological tolerance and/or detoxification of food borne poisons.  Most obvious among these mechanisms is the capacity to quickly and efficiently purge the contents of the upper gut by vomiting.  This occurs when the upper gut is irritated by food borne poisons.  Vomiting can also be evoked when the medulla is stimulated by poisons that have made their way into the blood.  In addition are mechanisms usually involving metabolic pathways in the liver that detoxify poisonous substances, and may include capabilities facilitating rapid excretion of toxins or their metabolites (32).
     
3. Instinctive behaviors: neophobia and unconditioned repellency.  Neophobia is an apparently unlearned behavioral routine animals engage in when confronted with a novel food.  It consists in repeated sampling of the taste of new food followed by delays of 30 minutes or more.  In this way, consumers may detect the effects of otherwise hidden poisons in foods without fully consuming a dangerous dose.  Unconditioned repellency is an instinctive response to the taste of foods that have contained dangerous poisons so reliably and for so long that it has become adaptive for consumers to automatically avoid these foods on the basis of their appearance (33), or the taste or smell (34, 18, 35) alone.  An example familiar to humans is the class of substances known as plant alkaloids.  Plant alkaloids are found as secondary plant substances in so many different species of plants all over the world that they are considered to be a class of substances with a very ancient history.  All of the plant alkaloids are variously poisonous to most vertebrates and they are all chemically similar enough to be detected by humans and other animals as having a characteristic “bitter” taste (35).   Without previous experience with any food taste or with any consequence of feeding on anything, neonatal mammals including human babies, reject the taste of bitter substances like quinine with a stereotypic suite of responses including squirming, gagging and head shaking.  Capsaicin is another broadly distributed secondary plant substance thought to be mainly an insect repellent, but repellent to many vertebrate consumers as well.
     
4. Feeding traditions.  Some food preferences among adult animals appear to be based upon foods with which they had become familiar as juveniles.  Juveniles of some species are much less neophobic than adults (36, 37, 38, 39, 40) and so readily consume foods brought to them by adults or shared when foraging with adults.  In so doing, they learn not only improved “food handling” behavior but also apparently learn which foods are most likely “safe” to consume (41).  With the development of greater neophobic behavior as adults, the outcome of this previous juvenile familiarity is a feeding tradition guiding foraging away from new and potentially dangerous foods and toward common and safe foods.  When examined among humans, feeding traditions (cuisines) enforced by juvenile familiarity with foods may be powerful determinants of adult food preference and are among the last cultural features to change when humans immigrate into new cultures (41).  Research on feeding traditions suggests that juvenile wolves being raised in captivity with the view of being one day released into the wild should specifically not be fed foods with tastes and scents similar to domestic livestock they are likely to encounter (e.g. beef or lamb flavored dog foods or composed of meat byproducts from these sources).  If for convenience or economic reasons commercially available foods must compose the majority of the diet, then foods with tastes and scent of domestic animals unlikely to be encountered should be substituted and these might be altered with masking tastes and scents.  In addition, juvenile wolves should be familiarized with natural large game animals native to the habitat where they will be released together with wild-trapped rodents, rabbits, insects and other foods the remains of which are found in wild wolf scats.  If young wolves fit the pattern described for other species then only a few meals of these foods safely eaten may affect adult preferences and foraging behavior.  Together with a protocol producing aversion to the primary domestic livestock they are likely to encounter, reintroduction programs will engage a greater compliment of natural inclinations to reduce conflict between wolves and agriculturalists.
     
5. Conditioned taste aversion.  Although foraging behavior, physiological specializations, instinctive responses to foods and feeding traditions all contribute to the selection of foods that may be safely consumed by foragers, these mechanisms cannot insure broad foraging with complete safety.  The array of potential food borne poisons is still so great and unpredictable that there remain many to escape detection before consumption.  Thus, the common fate of virtually all consumers is to sooner or later encounter some food that induces a severe, time-delayed illness after consumption.  Even with advanced sanitation, refrigeration and preservation of foods in western societies, the food supply still contains substances that sicken consumers. This is much more the case in third world economies and more still when consumers forage in natural ecosystems.   Conditioned taste aversion is a form of learning that may be thought of as the last stand against lethal poisoning.  It is the learning that happens when all other defenses have failed, poison has been consumed and now physiological defenses designed to ward off death from intoxication are brought to bear.  If the animal recovers, it is conditioned taste aversion that ensures that the particular food consumed prior to the onset of the illness cannot be consumed again. 

THE EXPERIMENTAL ROOTS OF CONDITIONED TASTE AVERSION

The modern understanding of conditioned taste aversion began with research work first reported by Garcia (42).  Garcia (27) traced reports that predators quickly learned to avoid insect prey with chemical defenses as far back as Wallace (43), Poulton (44), and Bates (45).  But, because the principles of conditioned taste aversion described by Garcia seemed to violate the laws of learning through instrumental and Pavlovian conditioning as accepted in the mid-twentieth century, his work was either ignored or severely criticized, sometimes with overtones of personal animus.  Nevertheless, the empirical evidence supporting the unique features of conditioned taste aversion accumulated to the point that this form of learning became widely known and accepted.  The first comprehensive review of the literature on the subject was published by Milgram (46) and the first comprehensive bibliography listed some 300 research articles appeared in 1976 (47).  By 1985 Riley and Tuck (48) listed nearly 1400 research articles on the subject and today, Riley and Freeman list nearly 3000 (49) but this list is still very incomplete.

Since feeding is a fundamental process in the animal kingdom and conditioned taste aversion is fundamental to this process, the process of conditioned taste aversion has attracted the attention of researchers interested in food selection in humans and agricultural animals, hunger and satiety, the nature of drug toxicity, characterization of drugs of abuse, cancer anorexia, the basic biochemistry of learning, control of alcohol abuse, evaluation of the motivational effects of drugs and feeding, drug-drug interactions, the processes of drug withdrawal, the principles of taste psychophysics, and wildlife management.

Taken together, this body of empirical science has fleshed out the unique features of conditioned taste aversion and the underlying processes that produce it.  This research also provides undeniable support for the basic principles of conditioned taste aversion that may be used to mitigate conflicts in wildlife management. 
     

THE SPECIAL FEATURES OF CONDITIONED TASTE AVERSION

Conditioned taste aversion is a form of learning that differs from other kinds of learning in at least nine important ways.

1.      Animals placed under general anesthesia (50), effectively surgically decorticated, or rendered unconscious by hypothermia or by electro-convulsive shock can nevertheless learn the association between a food taste and illness that follows the taste (reviewed in: 51).

2.      The learned association between a food taste and illness is extremely powerful and, without additional reinforcement, is very long lasting (3, 6, 52, 53, 54, 55, 56, 57, 58, 59, 60).

3.      Only a single illness punishment is necessary to produce this learning (Reviewed in: 61).

4.      External details of space and time are not relevant to the formation or expression of conditioned taste aversion.  That is, foods previously associated with illness are avoided with disgust wherever the learning took place and wherever they may be subsequently encountered (62).

5.      The time delay between feeding and illness onset can be up to 1-2 hours and the association between the food taste/scent or taste/visual cues will still be made (63).

6.      As long as taste precedes illness, other non-taste cues including the scent of food (for mammals) and the visual appearance of food (for birds) can be conditioned by the illness also.  In this way, foragers in nature can identify poisonous foods at a distance without having to actually sample the foods by taste.  That is, when only scent or visual cues are presented to experimental animals with no taste cue and then animals are made ill, little or nothing is learned (27, 61).  The taste component has been characterized as operating like a gate allowing food scent and visual cues to have access to the taste-illness axis (29, 64, 65, 66).  The capacity of predators to avoid poisonous prey at a distance on the basis of olfactory or visual prey cues is much more than a mere convenience to them and is a feature of conditioned taste aversion essential to our efforts to use the process to manage wildlife problems.  Predators can ill afford to expose their location, expend valuable energy and incur risk of physical injury in the act of pursuing prey that in the end they cannot consume.  Any without the capacity to accurately identify prey at some distance would be at a severe competitive disadvantage against any with even the rudiments of this capability.  Furthermore, our efforts to use conditioned taste aversion to spare domestic or other prey from predation rely upon the capacity of predators to make food choices at a distance on the basis of non-gustatory food cues.

7.      Some aspects of conditioned taste aversion superficially seem counterintuitive.  For example, predators that consume an inanimate meat bait laced with an illness-producing dose of a hidden aversion agent will nevertheless avoid at a distance live animate prey with the same taste and scent or visual characteristics as the bait (52, 60, 67. 68).  Humans have acquired profound conditioned taste aversions to previously preferred foods because, for example, they consumed the food a short time before the onset of an illness induced from chemotherapy.  Even though these individuals may be intimately aware of the process of conditioned taste aversion and will readily admit that the food was not at all the cause of the illness, they can nevertheless not force themselves to swallow it and may vomit at the mere scent of the now “offending” food (69, 70, 71).

8.      Conditioned taste aversion demonstrates a fundamental violation of an early assumption about how learning takes place.  Not all signals can be conditioned by any punishment.  Specifically, food preference is altered by illness but not by external punishment such as electric shock or rubber bullets (61).

9.       Conditioned taste aversion is a kind of learning with remarkable similarities in how it is acquired and how it is expressed across a surprising diversity of animals.  In addition to birds and mammals, conditioned taste aversion has been established among fishes, reptiles and invertebrates including coelenterates, mollusks and insects (Reviewed in 51, 72).

The factors relevant to the formation and expression of a conditioned taste aversion are:                                    

1.      The severity of illness.  There is generally a dose-dependence in which the greater the dose of poison the more profound the food aversion.  Exceptions include substances that produce aversions at very small amounts and do not produce proportionately greater effects with proportionately greater doses.  Some substance like thiabendazole, however, produce reliable conditioned taste aversions but do not produce outward symptoms of illness at all (3).

2.      The delay between feeding and illness onset.  Although aversions can be demonstrated with delays of 1-2 hours or more, the ideal delay is on the order of 30 minutes (73).

3.      The novelty of the food consumed prior to illness onset.  Generally, predators form aversion more easily to unfamiliar foods than they do to familiar foods that have safely been consumed before.  This may require more than one pairing of the food with illness before the predator finally forms the aversion.

4.      The complexity of the food cues prior to the onset of illness.  If many different foods are consumed either sequentially or simultaneously before illness, then there might not be a simple, single association between signal and consequences. Subsequent behavior of consumers tends to be a perfectly reasonable routine likely to disclose which foods are safe and which are not.  This routine often includes pronounced avoidance of novel foods and may include approach to familiar foods with behavior similar to neophobia.  The learning task may require two or more illnesses before the animal correctly associates the poisoned food with the illness.  If, however, the illness is severe and the source of the illness remains unclear, then the consumer may abandon the effort entirely (6, 54).

5.       In the special case in which mammalian predators are offered meat baits treated with a dose of an aversion agent designed to produce conditioned taste aversion that suppresses attack and consumption of live untreated prey, the meat bait must have the same taste and scent cues as that of live prey.  Predators that consume poorly constructed baits that do not have the same taste and scent cues as live prey will form an aversion to the baits but continue to attack and consume untreated live prey (3, 74).  Put simply, one should not be surprised if animals made sick after eating beef do not avoid chicken.  Mammalian predators that do not discriminate between the taste and scent cues of treated baits and those of untreated live prey avoid both on the basis of taste and scent alone.  That is, what is relevant to the formation and expression of conditioned taste aversion is the taste and scent of the prey, not details of location or movement or sound or behavior (3).

CTA Appendix 4

THE NEUROANATOMICAL AND NEUROPHYSIOLOGICAL REASONS FOR CONDITIONED TASTE AVERSION’S UNIQUENESS. 

(Please note:  Many of the details concerning how taste and scent information are processed centrally have yet to be clarified and so what follows is an oversimplified summary.  For a more authoritative discussion of this content refer to Bures (51), Garcia (61) and Grill (75).)

The anatomy.  In all reptiles, birds, and mammals three cranial nerves bring taste information from mouth to brain, but this information does not first or foremost go to the conscious part of the brain, the cerebrum.   Instead, these three cranial nerves: the Facial, the Glossopharyngeal, and the Vagus nerves all converge at the same place in the medulla of the brain.  This place is called the nucleus of the solitary tract.  Lying immediately near the nucleus of the solitary tract is another medullary center known as the area postrema.  An additional branch of the Vagus nerve brings information from the lining of the upper gut.  Together with a small portion of the upper brain, the gustatory cortex, these are the principal parts of the nervous system necessary for conditioned taste aversion learning and with the exception of some information that is shared with the upper brain, this anatomy is involved with exclusively unconscious regulation of interior body physiology.  It is the medulla, for example that regulates respiration and the vascular system.  That is, the medulla acts reflexively and has very powerful executive control over physiological functions of the body. 

The nucleus of the solitary tract is the “emesis” center of the brain.  Stimulation of this area of the medulla produces vomiting.  Emesis can be achieved experimentally by electrical stimulation or it can be achieved naturally when the branch of the Vagus nerve arriving from the upper gut conveys information indicating that a poison in the gut is irritating the gut lining.  The area postrema of the medulla has direct contact with the circulatory system.  Blood borne poisons that escaped detection by the Vagus nerve may nevertheless stimulate the area postrema, which in turn initiates vomiting as well.

The sequence of events producing learned conditioned taste aversion.  When food that has never before been associated with illness enters the mouth, the nucleus of the solitary tract allows the food to be swallowed.  If the food contains a hidden poison that irritates the lining of the upper gut then stimulation of the solitary nucleus by the vagus nerve produces emesis, which usually purges the poison.  Recovery from the illness soon follows.  If some of the poison is absorbed without irritating the lining of the gut then the area postrema may detect blood borne poisons and initiates emesis, which also purges poison remaining in the upper gut.  Additional physiological mechanisms of the liver and the kidneys may eliminate residual poison, followed by full recovery.  But, the medulla retains a permanent trace of the taste(s) that preceded illness and upon subsequent contact with that same food taste(s) refuses to allow the consumer to swallow the food.

The neuroanatomy of this area of the brain, head and upper gut together with the inevitable features of the feeding sequence explain the unique features of conditioned taste aversion:

1.      Animals rendered unconscious by general anesthesia or by surgical removal of the cerebral cortex can still learn illness-induced conditioned taste aversion because these procedures affect conscious voluntary activity and do not affect the unconscious vegetative events necessary for the acquisition of conditioned taste aversion.

2.       Conditioned taste aversions are powerful and long lasting because the medulla is well placed to powerfully regulate internal body function and among these functions is the decision as to what food substances are acceptable and what are not on the basis of food taste and of past physiological consequences of consumption.

3.       Only a single taste-illness episode is sufficient to produce the effect presumably because it is not necessary for repeated pairings of signal (taste) and punishment (illness) for animals to correctly associate the actual “cause” and “effect”.  In other kinds of learning, the conscious brain will be aware of many different kinds of events happening simultaneously and each might be a signal predicting some punishment.  Only after repeated pairings of signal and punishment do animals resolve the learning task and form an accurate association.  Since only taste (signal) and physiological change (consequences) information converge at the medulla the association between these two things cannot be easily confused by other inputs.

4.      External non-food/non-taste stimuli are not relevant to the taste-illness axis because the medulla does not receive this information.  Instead, what is tracked is the most relevant and accurate measure of the identity of food (its chemistry as reflected by taste/scent) and food value (physiological change following swallowing).

5.      The time delay between signal (taste) and consequences (illness) can be much longer than in other forms of learning because the medulla and gustatory cortex are insulated from other sources of confusing intervening stimuli and so “interference theory” of forgetting may not apply.

6.      That non-gustatory senses, like scent and visual cues can be used by animals to make accurate feeding choices without the need to actually taste prey, is undisputed.  Exactly how this process works is still a matter of investigation.

7.      That conditioned taste aversion seems counterintuitive to us is not surprising, considering where what we refer to as intuition lies in our brain.  Once it is recognized that some events, including food preferences, are not open to conscious “negotiations”, this aspect of conditioned taste aversion, including the prediction that predators will avoid live prey after having been treated with inanimate meat baits, becomes much more readily understood.   It is possible that no vertebrates, including humans, ever made a fully conscious decision concerning the foods they prefer to eat.

8.   That external punishment such as electric shock does not affect food preference is

      also understandable: external pain information does not converge at the point

in the medulla where it can become associable with taste information.

9.  That such a wide diversity of vertebrates share the same predisposition to

      associate food taste with a single time-delayed illness is predicted by the fact that

      all of the vertebrates share similar anatomy of cranial nerves and the brain stem.

      Invertebrates with dissimilar nervous systems nonetheless form illness-induced

      conditioned taste aversion with properties to vertebrates.  This is likely the result

      of a remarkable case of convergent evolution enforced by similar demands to

      avoid toxic food.
 

SUMMARY

Conditioned taste aversion is clearly a form of learning specifically adapted to protecting

animals from repeated consumption of toxic foods.  The neuroanatomy and physiology responsible for the unique features of this learning are exclusively devoted to this task,

are common among all vertebrates so far studied and contain redundancy in the form of two medullary centers that can act to produce the effect: one in case food toxins act locally upon upper gut lining and the other in case toxins pass through the lining and become systemic.  The process of conditioned taste aversion wields unopposed executive control over what animals can and cannot eat with consequent effects upon their foraging behavior. 

Thus, when the process of conditioned taste aversion is used to mitigate conflicts between agriculturalists and efforts to reintroduce wolves, it is not a case of applying some human invention with the foibles and limitations inherent to all human contrivances.  Instead, it is a case of applying the discovery that vertebrates are powerfully predisposed to associate the taste and scent/appearance of foods with time-delayed illness in such a way as to regard once highly preferred foods with disgust wherever and whenever they are subsequently encountered.   All attempts to apply conditioned taste aversion are, of course, subject to empirical verification in every detail but their ultimate success or failure will depend upon the extent to which the natural requirements of the process are satisfied in the particular application.  When these requirements are met then subsequent events have all of the reliability that millions of years of trial and success have produced.  When humans fail to meet these requirements or attempt to apply conditioned taste aversion to situations in which it is inapplicable, then this “failure” can hardly be laid at the feet of the process.  In other words, it is simply inappropriate to ask if conditioned taste aversion “works”.  It is only appropriate to ask how it may be made to apply under clearly specified conditions. 

THE APPLICATION OF CONDITIONED TASTE AVERSION TO WILDLIFE MANAGEMENT PROBLEMS

Successful captive and field studies. 

Although much was known of the physiological, laboratory, and clinical aspects of conditioned taste aversion before the mid-1970’s, nothing at all was known of its potential to alter predatory behavior in captivity or in the field until the first research was conducted by Guvstavson.  Since then, much has been learned in both captive and field studies.   Gustavson found that captive coyotes (Canis latrans) that ate sheep meat baits wrapped with lambs wool and with the interior meat laced with a hidden dose of Lithum Chloride (LiCl) subsequently avoided live sheep while others that formed aversion to rabbits in the same manner avoided live rabbits (67).  That is, illness lasting approximately one hour following consumption of an inanimate meat bait with the same taste and scent characteristics as of live prey suppressed not only feeding but also predatory attack upon live prey under captive conditions.  These conditions included moderate food deprivation, the expectancy to feed among the predators, and lengthy close contact with otherwise totally defenseless live prey.  These results were later extended to demonstrations of suppressed feeding and predatory attack upon live sheep among captive wolves (Canis lupus) (68).  Conditioned taste aversion of captive hawks (Buteo sp) suppressing attack upon live mice was also reported at about this time (76).  Gustavson (72) summarized the comparative literature on conditioned taste aversion up to that time in which it was becoming evident that conditioned taste aversion was a fundamental process wide spread in the animal kingdom.  Successful captive studies intended to examine aspects of potential field applications of conditioned taste aversion have continued (e.g. 3, 4, 53, 56, 77, 78, 79, 80).

After initial captive work with wolves and coyotes, Gustavson distributed LiCl-laced sheep meat baits into the field to control coyote predation upon sheep.  Predation rates as reported by cooperating woolgrowers and confirmed by visual inspections declined in regions receiving the conditioned aversion treatment in a manner predicted by the captive studies (81).  These results were independently confirmed in another field study of the effects of conditioned taste aversion upon coyote predation upon sheep (82).  A small-scale evaluation of conditioned taste aversion to cattle among freely ranging wolves in northern Minnesota appeared also to have been successful (83).  A larger scale and longer term field application of conditioned taste aversion also reported substantial reductions in coyote predation upon sheep in treated locations compared with those that did not receive treatment (84).

In this field work there was no opportunity to actually observe the establishment and then the expression of conditioned taste aversion among freely ranging coyotes and so the evidence of treatment effect, though reasonable, was indirect.  In 1979 Nicolaus (52, 85) reported the first field study in which individually identified freely ranging predators were observed interacting with prey before, during, and after the establishment of conditioned taste aversion.  In this work, Nicolaus observed free-ranging raccoons (Procyon lotor) killing and eating live chickens before conditioned taste aversion and then observed the same individuals interacting with live chickens after eating chicken carcasses carefully injected with moderate amounts of LiCl (6 g LiCl per 2 Kg chicken).  The same individuals that previously killed chickens avoided live chickens after treatment while newly arrived individuals that had not been treated with the LiCl chicken baits killed and ate them freely.  Treated raccoons began to abandon the treatment site where live chickens were offered and so on randomly-chosen nights dog food induced them to continue visiting.  When they arrived to consume the dog food, they engaged in behavior suggesting that they identified and rejected the chickens at a distance primarily on the basis of scent.  Even though they were treated with all female white leghorn chicken carcasses, they avoided these and male red bantam chickens as well.  Thus, the conclusion seemed inescapable that when raccoons consumed chicken carcasses containing a successfully hidden dose of LiCl, they acquired a conditioned taste aversion resulting in suppressed predatory attack upon untreated live chickens.  In addition, there was evidence that the aversion also much reduced the willingness of treated raccoons to approach the location where live chickens were placed.  Though treated raccoons were present when untreated raccoons repeatedly killed and ate chickens, they did not resume predatory attacks.  These results were consistent with captive research on wolves, coyotes, hawks and other species and with the field research on coyotes, and consistent also with the theory that conditioned taste aversion is a fundamental process common to all vertebrates.   In a brief subsequent study of these same predators it was evident that LiCl could not be successfully hidden in such small and bland prey as eggs.  Raccoons initially opened and consumed salty-tasting LiCl-treated eggs but on subsequent nights rejected both LiCl and Sodium Chloride (NaCl) treated eggs.  They did so by opening and tasting all eggs, rejecting the salty taste and consuming eggs untreated with either salt.  That is, they obviously formed a conditioned taste aversion but because they detected the aversion agent in the eggs, they discriminated between treated and untreated prey (unpublished data).  This was later confirmed when free-ranging mongooses (Herpestes auropunctatus) discriminated between treated and untreated eggs (86) and when a variety of egg predators also discriminated between LiCl treated eggs and untreated eggs (55).

Raccoons are perhaps the only species of North American mammalian predators willing to aggregate in large numbers and habituate to observers in a manner making this kind of study possible.  Field research on conditioned taste aversion to characterize the process closely must nevertheless rely upon actually observing large numbers of freely ranging predators before, during, and after treatment.  Breeding birds provide this opportunity because they are diurnal, occupy breeding territories with prominent displays and the same individuals can be found in the same places for long periods of time.  Because most of them defend their territories aggressively, there are few incursions by birds from one territory to any other and so experimental events that occur in one territory are assuredly independent of other such territories.  In the spring it may be possible to locate many such bird territories and birds of the same species strongly tend to occupy territories of very similar size and habitat.  Thus, it is possible to randomly assign large numbers of territories to experimental treatments and controls in which each territory is independent of all the others, contemporaneous and ecologically equivalent.  That is, it is possible to conduct simple but elegantly designed experiments by using breeding territories as experimental replicates.   This was first achieved in studies of American Crows (Corvus brachyrhynchos) in North Dakota and Minnesota (58) in which a new and less detectable aversion agent, Landrin, was used to treat eggs as baits.  Since the Crows remained largely on breeding territories and were conspicuous, preliminary observations made it possible to match the number and distribution of treated baits to the numbers of consumers and to document how many baits were required to produce complete and permanent avoidance of prey at a distance.  This was achieved when a mean of 3 Crows per territory (a breeding pair and one or more “nannies”) consumed a mean of 4.8 treated green-colored eggs (1.6 baits per individual) before they completely and permanently avoided these baits at a distance.  Studies of this kind also reveal details of the responses of predators with conditioned taste aversion useful in interpreting and predicting future field research on the subject.  For example, once these avian predators had acquired the aversion, their foraging shifted on the basis of the visual characteristics of prey so that they readily found and consumed untreated white eggs randomly distributed among the green eggs but simply flew over and ignored the green eggs.  Since these birds avoided green eggs at a distance as a class of prey it no longer mattered what the eggs later contained.  This was confirmed when these same birds were offered untreated green eggs located in new patterns.  In fact, without alternative safe white egg prey, the Crows actually abandoned the new locations with green eggs to feed on insects in wheat fields at substantial distances elsewhere.  Throughout this time Crows in “control” territories where neither kind of egg induced illness, ate eggs of both types freely and they quickly found and fully consumed eggs placed in new locations.

This procedure was applied to Common Ravens (Corvus corax) on a wildlife refuge.  In their breeding territories Ravens were offered mimic eggs of Greater Sandhill Cranes (Grus Canadensis) treated with Landrin (57).  Preliminary observations revealed locations of active Raven breeding territories, the dimensions of the territories, the distances separating adjacent territories and the numbers of Ravens occupying each territory (a mean of 2.2 Ravens per territory).  Thus, as in the Crow study it was possible to place an appropriate number of baits into appropriate places in order to conduct the study.  Within designated treatment territories, Ravens were offered pairs of model “crane” eggs designed to induce time-delayed illness after consumed and in other places, alternate chicken eggs that never induced illness.  Ravens quickly found and fully consumed an average of 2.5 treated “crane” eggs on the first few days and then for a few additional days injured or moved and so therefore “killed” an additional mean of 4.0 eggs per territory.  Predation upon “crane” eggs attributable to Ravens then abruptly ceased but in two territories treated Ravens continued to consume alternative prey while in two others, avoided these as well.  Ravens in control territories where neither of the two kinds of eggs was treated consumed both freely.  So, as in the previous Crow study, aversions were established among Ravens that consumed only about 1 treated egg each.  Ravens remained in their territories completely ignoring the “crane” eggs offered to them each day in conspicuous locations on conspicuous “crane” nests.  In fact, these breeding Ravens not only ceased predated the “crane” eggs themselves, but they also very aggressively defended their breeding territories from incursions by other breeding and non-breeding Ravens.  In this way, the survival of “crane” eggs was doubly enhanced compared with eggs placed outside of the breeding territories of Ravens with conditioned taste aversions.  This territorial “baby sitting” effect was predicted earlier by Gustavson (personal communication) and was systematically confirmed for the first time in this study.  The results of this work demonstrated that if properly prepared convincing treated mimics of the eggs of endangered birds are placed in appropriate numbers and locations within the breeding territories of avian predators some time before the endangered birds lay their eggs, conditioned taste aversion could quickly, completely, and permanently suppress predation upon the eggs of the endangered species.

These effects were generally replicated when a new aversion agent, Carbachol, was tested on freely ranging American Crows (56).  In yet another much larger field study Crows in separate breeding territories were confronted with differing ratios of treated and untreated egg prey, including territories assigned to a control group in which all prey were untreated.  Where all of the green eggs were treated, Crows discriminated between green eggs, which they avoided, and untreated white eggs, which they continued to consume freely just as they did in 1983.  Where half of the green eggs were treated and half untreated, they avoided both the green eggs and the white eggs.  In this case, the situation was more complex.  White eggs were safe as were some of the green eggs while others caused illness.  Crows consumed both green and white eggs before time-delayed illness, could not taste the aversion agent in the green eggs and so they could not solve the task of attributing the illness to either kind of egg.  Evidently the value of the eggs was less than the threat of illness and so they avoided both kinds of eggs completely.  These same birds continued to avoid both kinds of eggs the next year when we again offered untreated eggs in these same breeding territories (54).  

In a series of experiments, Nicolaus and Lee (60) reported how organophosphate insecticide could produce conditioned taste aversion to insects among freely ranging insectivorous birds.  All Red-winged Blackbirds (Agelaius phoeniceus) that consumed a single meal of contaminated insects and then offered uncontaminated insects of the same kind thereafter immediately, completely, and permanently avoided all insects of this kind.  Conditioned taste aversion was produced even though the actual amount of insecticide birds consumed was too small to be detected by standard methods of assessing exposure to organophosphate insecticide. 

Conditioned taste aversion was successfully established in assemblages of freely ranging mammalian predators, including raccoons, badgers, opossums, coyotes and foxes when the aversion agent oral estrogen was applied in the field for the first time (55).  In places where predators were offered LiCl-treated eggs for comparative purposes, they discriminated between these salty-tasting, illness inducing eggs and untreated eggs just as the raccoons did in 1979 and as did the mongooses several years later (86).  But other predators in other locations failed to detect estrogen in eggs and so after a meal of estrogen treated eggs, avoided all eggs, treated or not. 

The details of estrogen-based aversion produced in freely ranging raccoons were reported by Semel and Nicolaus (59).   In this study, male raccoons that consumed a single meal of estrogen treated eggs avoided both treated and untreated eggs thereafter.  After female raccoons consumed a single meal of estrogen treated eggs, however, they continued to open but reject both treated and untreated eggs for a time before they too avoided eggs completely at a distance.  Treated raccoons ceased their nightly visits to the treatment site, apparently because of the now disgusting scent of eggs at the treatment site.  We had to be induced them to a new location without this scent in order for us to observe them avoiding untreated eggs during the lengthy post-test.  When examined a year later, raccoons known to have been previously treated with estrogen continued to avoid untreated eggs.
 

THE CONDITIONED TASTE AVERSION “CONTROVERSY”

A thorough discussion tracing the history of objections to the suggestion that conditioned taste aversion might be of practical use in wildlife management problems is beyond the scope of this proposal.  All of this has been reviewed and explained at length elsewhere (See: 3, 87, 88, 89, 90, 91).  Those interested in this question should consult these sources.  There is, however, one objection to the use of conditioned taste aversion in wildlife management that has some bearing on this proposal.  This is the claim that conditioned taste aversion cannot suppress predatory attack.  This assertion first became known to us when it was placed for public consumption on the World-Wide Web by the National Wildlife Research Center as a matter of decree rather than of substantiated science (91).  We found this assertion difficult to rebut since it seems to exist more in the realm of rumor than in the realm of empirical science where data and methods are openly shared, conclusions strictly based upon these data, reports are peer reviewed by experts in the subject, and openly debated. We presumed that it originated from studies in which predators that detected the aversion agent in poorly constructed meat baits, formed aversion to the baits, but continued attacking and consuming untreated live prey.  Later, we learned from personal communications with some of those making the claim that scientific support comes from observations of laboratory rats that had acquired a conditioned taste aversion to mice, stopped eating the mice but continued to kill them.  This sort of thing has been observed before among laboratory rats and ferrets, but no one has attached such importance to these observations (92).  This is because these data could easily be attributed to the artificialities of captivity, including very close proximity to the mice.  It is also possible that since these animals obviously have a conditioned taste aversion to the mice, continuing attacks upon mice might have nothing to do with conditioned taste aversion.  That is, rats and ferrets might kill mice found nearby because rodents can be active predators upon their young.  The claim is poorly substantiated and seems to us contrived since those making the claim have access to wolves and coyotes, more than sufficient funding, and more than enough time since 1974 to conduct legitimate tests of their hypothesis with real predators under convincing conditions.  Yet, they have so far failed to do so. This claim entirely contradicts direct observations reported in previous captive and field studies. And, its proponents have made no attempt to explain why such a thing must be so, even though it would render meaningless the whole idea that conditioned taste aversion evolved as a defense against intoxication by prey with chemical defenses.  After all, why should a predator in nature incur the costs and risks of continuing to pursue and kill prey that it cannot in the end eat?

Nevertheless, the scientific question of this proposal is perfectly framed to clarify the issue.  That is, if a TBZ-based conditioned taste aversion protocol does suppress predatory attack among wolves, this should be evident in the fieldwork.  If our conditioned taste aversion protocol does not suppress predatory attack, then that should be evident also.
 

SUMMARY

When the requirements for successfully establishing conditioned taste aversion among captive animals are met and the expectations for the outcome reflect an understanding of the process of conditioned taste aversion, many studies have shown wide similarity in the responses among both avian and mammalian predators.  That is, predators form an aversion to the taste and scent of target prey after a single meal of meat bait containing a hidden dose of an aversion agent.  This procedure has been applied in the field under conditions allowing direct observation of predators before, during and after aversion treatment, just the same outcomes are reported.  Success in the field absolutely requires not only knowledge of the process of conditioned taste aversion, but also knowledge of the behavior, abundance and distribution of predators.  If anything, the results of field application of conditioned taste aversion can be even more robust and lasting than those obtained under captive conditions.  This may be because predators and prey may escape each other in the field.   Mammalian predators are free to avoid prey at great distances on the basis of what has become, for them, a noxious scent while avian predators are free to simply fly over and ignore prey they identify as no longer consumable at great distances on the basis of visual cues.
 

LITERATURE CITED

1)      The Merck Index. 1968. P.G. Stetcher (Ed). P. 1035.

2)      Thiabendazole. 1991. In  Veterinary Drug Handbook, Pocket Edition. D.C. Plumb PharmaVet Publishing. Pp392-395.

3)      Gustavson, C.R. and L.K. Nicolaus. 1987. Taste aversion conditioning in wolves, coyotes, and other canids: Retrospect and prospect In H. Frank (Ed), Man and Wolf:Advances, issues, and problems in captive wolf research. W. Junk, Boston, pp169-203.

4)      Gill, E.L., Whiterow, A., and D.P. Cowan. 1999. A comparative assessment of potential conditioned taste aversion agents for vertebrate management.  Appl. Anim. Behav. Sci. 67:229-240.

5)      Polson, J.E. 1983. Application of aversion techniques for the reduction of losses to beehives by black bears in northeastern Saskatchewan. Report to Department of Supply and Services, Ottawa, Canada SRC Publication C-805-13-E-83.

6)      Ternent, M.A. and D.L. Garshelis. 1999. Taste-aversion conditioning to reduce nuisance activity by black bears in a Minnesota military reservation. Wilfdlife Society Bulletin, 27(3)720-728.

7)      Alcock, J. 2005. Animal Behavior: An evolutionary approach. 8^th edition. Sinauer Assoc. Inc. Sunderland, MA.

8)      Barnes, J.H. 1966.  The Indo-Pacific sea wasp, Chironex fleckeri, a dangerous syphozoan of the order Cubomedusa.  Pp. 131-139. In W.J. Rees (Ed.), The Cnidaria and their Evolution.  Academic Press, NY.

9)       Kohn, L. 1972. Buccal structures of Conus striatus. Science, 176: 49051.

10)   Whitley, G. 1943. Poisonous and harmful fishes. Bull. Coun. Sci. Industr. Res. No. 159. Melbourne.

11)   Brodie, E.D., and N.J. Smatresk. 1990. The antipredator arsenal of fire salamanders: Spraying of secretions from highly pressurized dorsal skin glands.  Herpetologica, 46:1-7.

12)   Goin, C.J. and O.B. Goin. 1971. Introduction to Herpetology. 2ned ed. San Francisco; W.H. Freeman and Co.

13)   Eisner, T. 1970. Chemical defenses against predation in arthropods, pp. 157-217. In E.Sondeimer, J.B. Simeone (Eds), Chemical Ecology, Academic Press, NY.

14)   Eisner, T., M. Eisner and M. Siegler. 2005. Secret Weapons: Defenses of insects, spiders, scorpions, and other many-legged creatures. Harvard University Press, Cambridge, MA.

15)   McIndoo, N.E. 1945. U.S. Dept. Agr. Bur. Entom. Plant Quarantine, ET661, 1-286.

16)  Kingsbury, J.W. 1964. Poisonous plants of U.S. and Canada. Prentice Hall, Englewood Cliffs, NJ.

17)   FDA poisonous plant data base. 2006.  Center for food safety and applied nutrition. www.cfsan.fda.gov/~djw/plantox .

18)   Harborne, J.B. 1982. Introduction to ecological biochemistry. 2ned ed. Academic Press, NY.

19)   Rothschild, M. 1985. British aposematic lepidoptera. Pp. 9-62 In J.H. Heath and A.M. Emmet (Eds), The moths and butterflies of Great Britain and Ireland.  B.H. and A. Harley, Essex, England.

20)   Bowers, W.S. 1980. Chemistry of plant/insect interactions. Pp. 613-633 In M. Locke and D.S. Smith (Eds), Insect biology in the future. Academic Press, NY.

21)   Smith, D.A.S. 1978. The effect of cardiac glycoside storage on growth rate and adult size in the butterfly Danus chrysippus. Experintia, 34:844-845.

22)   Bentley, E.W. 1972. A review of anticoagulant rodenticides in current use. Bull. WHO, 47:275-280.

23)   Jones, D.A. 1974. Co-evolution and cyanogenesis. pp. 213-242. In V.H. Heywood (Ed) Taxonomy and ecology. Academic Press, London.

24)   Keeler, R.F. 1975. Toxins and teratogens of higher plants.  Lloydia, 38:56-86.

25)   Bell, E.A. 1980. Alkaloid effects of anti-herbivory toxins. Prog. In Phytochemistry, 7:171-196.

26)   Garcia, J., McGowan, B.K., and K.F. Green. 1972.  Biological constraints on conditioning pp 21-42, In  M.E.P. Seligman and J.L. Hager (Eds), Biological boundaries of learning. Appleton-Century-Crofts, NY.

27)   Garcia, J and W. G. Hankins. 1977. On the origin of food aversion paradigms. pp 3-23, In L.M. Barker, M.R. Best, and M. Domjan (Eds). Learning mechanisms in food selection. Baylor University Press.

28)   Bell, R.H.V. 1971. A grazing ecosystem in the Serengeti. Sci. Amer. 224:86-93.

29)   Zahorik, D.M. and K.A. Houpt. 1977. The concept of nutritional wisdom: Applicability of laboratory learning models to large herbivores. Pp 45-71 In L.M. Barker, M.R. Best, and M. Domjan (Eds). Learning mechanisms in food selection. Baylor University Press.

30)   Raikow, R.J. 1976. The origin and evolution of the Hawaiian Honeycreepers. Living Bird, 15:95-117.

31)   Mitchell, D. Wells, C. Hoch, N. Lind, K., Woods, S.C. and L.K. Mitchell. 1976 Poison induced pica in rats. Physiol. Behav. 17:681-697.

32)   Hardman, J.G., Limbird, L.E., and A.G. Gillman. 2001. Goodman and Gillman’s Pharmacological basis of therapeutics. 10^th Ed. McGraw Hill, NY.

33)   Smith, S.M. 1975. Innate recognition of coral snake pattern by a possible avian predator. Science, 187:759-760.

34)   Brower, L.P. 1969. Ecological chemistry. Sci. Amer. 220:22-29.

35)   Garcia, J. and W.G. Hankins. 1975. The evolution of bitter and the acquisition of toxiphobia. Pp 39-47 In D.A. Denten and J.P. Coghlan (Eds), Olfaction and taste (Vol 5), Academic Press, NY.

36)   Hogan, J.A. 1977. The ontogeny of food preferences in chicks and other animals. Pp 71-98 In L.M. Barker, M.R. Best, and M. Domjan (Eds). Learning mechanisms in food selection. Baylor University Press.

37)   Gelef, B.G. 1977. Mechanisms for the social transmission of acquired food preferences from adult to weanling rats. Pp 123-150 In L.M. Barker, M.R. Best, and M. Domjan (Eds). Learning mechanisms in food selection. Baylor University Press.

38)   Capretta, P.J. 1977. Establishment of food preferences by exposure to ingestive stimuli early in life. Pp 99-122 In L.M. Barker, M.R. Best, and M. Domjan (Eds). Learning mechanisms in food selection. Baylor University Press.

39)   Alberts, J.R. 1982. Ontogeny of olfaction: Issues and perspectives from contemporary research. Pp 161-172 In W.Breipohl (Ed). Olfaction and endocrine regulation. IRL Press, London.

40)   Apfelbach, R. 1978. A sensitive phase for the development of olfactory preference in ferrets, Mustela putoris. Zeitschrift Saugeterkd, 43:289-295.

41)     Rozin, P. 1977. The significance of learning mechanisms in food selection: Some

 biology, psychology, and sociology of science. Pp 557-592 In L.M. Barker, M.R. Best, and M. Domjan (Eds). Learning mechanisms in food selection. Baylor University Press.

42)   Garcia, J. Kimeldorf, D.J. and R.A. Koelling. 1955. Conditioned aversion to saccharin resulting from exposure to gamma radiation. Science. 122:157-158.

43)   Wallace, A.R. 1867. Proceedings of the Entomological Society of London, 1867:80-81.

44)  Poulton, E.B. 1887. The experimental proof of the protective value of color and marking in insects in reference to their vertebrate enemies. Proceedings of the Zoological Society of London, 1887:191-274.

45)  Bates, H.W. 1862. Contributions to an insect fauna of the Amazon Valley, Lepidoptera: Heliconidae. Transactions of the Linnaean Society, London, 1862:495-566.

46)   Milgram, N.W. Krames, L. and T.M. Alloway. 1977. Food aversion learning. Plenum Publishing, NY.

47)   Riley, A.L. and L.L. Baril. 1976. Conditioned taste aversions: A bibliography. Animal Learning Behav. 4:1S-13S

48)   Riley, A.L. and D.L. Tuck. 1985. Conditioned food aversions: a bibliography. Pp 381-437 in N.S. Braveman and P. Bronstein (Eds). Experimental assessments and clinical applications of conditioned food aversions.  Vol. 443. Annals of the New York Academy of Sciences. NY.

49)   Riley, A.L. and K.B. Freeman. www.ctalearning.com

50)   Roll, D.L. and J.C. Smith. 1972. Conditioned taste aversion in anesthetized rats. Pp 98-102 In M.E.P. Seligman and J.L. Hager (Eds), Biological boundaries of learning. Appleton-Century-Crofts, NY.

51)   Bures, J., Bermudez-Rattoni, F. and T. Yamamoto. 1998. Conditioned taste aversion: Memory of a special kind. Oxford Science Publications, Psychology Series 31, Oxford University Press, NY.

52)   Nicolaus, L.K., Hoffman, T.E. and C.R. Gustavson. 1982. Taste-aversion conditioning in free-ranging raccoons, Procyon lotor. Northwest Science, 56:165-169.

53)   Massei, G., and D.P. Cowan. 2002. Strength and persistence of conditioned taste aversion in rats: Evaluation of 11 potential compounds. Appl. Anim. Behav. Sci. 75:249-260.

54)  Dimmick, C.R. and L.K. Nicolaus. 1990. Efficiency of conditioned aversion in reducing depredation by crows.  Jour. Appl. Ecol. 27:200-209.

55)   Nicolaus, L.K., Herrera, J., Nicolaus, J.C. and C.R. Gustavson. 1989. Ethinyl estradiol and generalized taste aversions to eggs among free-ranging predators.  Appl. Anim. Behav. Sci. 24:313-324.

56)   Nicolaus, L.K., Herrera, J., Nicolaus, J.C., and C.R. Dimmick. 1989. Carbachol as a conditioned taste aversion agent to control avian depredation. Agri., Ecosys, Environ. 24:13-24.

57)   Nicolaus, L.K. 1987. Conditioned aversions in a guild of egg predators: Implications for aposematism and prey defense mimicry. Amer. Midl. Natur. 117:405-419.

58)   Nicolaus, L.K., Cassel, J.F., Carlson, R.B., and C.R. Gustavson. Taste-aversion conditioning of crows to control predation on eggs. Science. 220:212-214.

59)   Semel, B., and L.K. Nicolaus. 1992. Estrogen-based aversion to eggs among free-ranging raccoons. Ecol. Appli. 2:439-449.

60)   Nicolaus, L.K. and H. Lee. 1999. Low acute exposure to organophsphate produces long-term changes in bird feeding behavior. Ecol. Appli. 9(3):1039-1049.

61)   Garcia, J., Hankins, W.G., and K. Rusiniak. 1974. Behavioral regulation of the milieu interne in man and rat. Science. 185:824-831.

62)   Garcia, J., and R.A. Koelling. 1966. Relation of cue to consequences in avoidance learning. Psychonom. Sci. 4:123-124.

63)   Garcia, J., Ervin, F.R, and R.A. Koelling. 1966. Learning with prolonged delay of reinforcement. Psychonom. Sci. 5:121-122.

64)   Rusiniak, K. Hankins, W.G., Garcia, J., and L.P. Brett. 1979. Flavor-illness aversions: Potentiation of odor by tase with toxin but not shock in rats. Jour. Compar. Physiol. Psychol. 96:527-539.

65)   Westbrook, R.F., Clarke, J.C. and S. Provost. 1980. Long-delay learning in the pigeon: Flavor, color and flavor-mediated color aversions. Behav. Neur. Biol. 28:398-407.

66)   Clarke, J.C., Westbrook, R.F., and J. Irwin. 1979. Potentiation instead of overshadowing in the pigeon. Behav. Neuro. Biol. 25:18-29.

67)   Gustavson, C.R., Garcia, J., Hankins, W.G., and K.W. Rusiniak. 1974. Coyote control by aversive conditioning. Science. 184:581-583.

68)   Gustavson, C.R., Kelly, D.J., Sweeney, M., and S. Thomas. 1976. Prey lithium aversions I: Coyotes and wolves. Behav. Biol. 17:61-72.

69)   Bernstein, I.L. and Webster. M.M. 1980. Learned taste aversions in humans. Physiol. Behav. 25:363-366.

70)  Bernstein, I.L. 1985. Learned food aversions in the progression of cancer and its treatment. Ann. New York Acad. Sci. 443:365-380.

71)   Bernstein, I.L. 1978. Learned taste aversions in children receiving chemotherapy. Science. 200:1302-1303.

72)   Gustavson, C.R. 1977. Comparative and field aspects of learned food aversions. Pp. 23-43 In L.M. Barker, M.R. Best, and M. Domjan. Learning mechanisms in food selection. Baylor University Press, Waco, TX.

73)  Garcia, J. and D.J. Kimeldorf. 1957. The temporal relationship within the conditioning of a saccharin aversion through radiation exposure. Jour. Compar. Physiol. Psychol. 50, 180-183.

74)   Burnes, R.J. and G.E. Connolly. 1980. Lithium chloride bait aversion did not influence prey killing in wolves. Proc. Ninth Vert. Pest Conf. Fresno, CA.

75)   Grill. J/K/ 1985. Physiological mechanisms in conditioned taste aversion. Pp67-88. In N.S. Braveman and P. Bronstein (Eds). Experimental assessments and clinical applications of conditioned food aversions. NY Acad. Sci. NY.

76)   Brett, L.P., Hankins, W.G. and J. Garcia. 1976. Prey-lithium aversions III: Buteo hawks. Behav. Biol. 17,87-98.

77)   Nicolaus, L.K., Farmer, P., Gustavson, C.R. and J.C. Gustavson. 1989. The potential of estrogen-based aversion in conrolling depredation: A step toward the “magic bullet”. Appl. Anim. Behav. Sci. 17:1-14.

78)   Cowan, D.P., Reynolds, J.C. and E.L. Gill. 2000. Reducing predation through conditioned taste aversion. Pp. 281-299. In L.M. Gosling and W.T. Sotherland (Eds). Behaviour and Conservation. Cambridge University Press, Cambridge.

79)   Massei, G., Lyon, A. and D.P. Cowan. 2003. Potential compounds for inducing conditioned taste aversion in ferrets. New Zealand Journal of Zoology. 30:95-100.

80)   Massei, G. Lyon, A.J., and D.P. Cowan. 2003. Conditioned taste aversion can reduce egg predation by rats. Jour. Widl. Manage. 66(4):1134-1140.

81)   Gustavson, C.R., Kelly, D.J., Sweeney, M. and S. Thomas. 1976. Final report to the Wahington Game Department, Olympia, WA.

82)   Ellins, S.R., Catalono, S.M., and S.A. Schechinger. 1977. Conditioned taste aversion: A field application to coyote predation on sheep. Behav. Biol. 20:91-95.

83)   Gustavson, C.R. 1982. An evaluation of taste aversion control of wolf (Canis lupus) predation in northern Minnesota. Appl. Anim. Ehol. 9:63-71.

84)   Gustavson, C.R., Jowsey, J.R. and D.N. Milligan. 1982. A three-year evaluation of taste aversion coyote conrol in Saskatchewan. Jour. Range Manag. 35:57-59.

85)   Nicolaus, L.K. Hoffman, T.E., and C.R. Gustavson. 1979. Taste aversion conditioning in raccoons (Procyon lotor). Paper presented to the Portland International Wolf Symposium, August, 1979.

86)   Nicolaus, L.K. and D. Nellis. 1987. The first evaluation of the use of conditioned taste aversion to control predation by mongooses upon eggs. Appl. Anim. Behav. Sci. 17:329-334.

87)   Bekoff, M. Gustavson, C.R., Kelly, D.J. and J. Garcia. 1975. Technical comment: Predation and aversive conditioning in coyotes. Science, 187:1096.

88)   Gustavson, C.R., Brett, L.P. Garcia, J., and D.J. Kelly. 1975. A working model and experimental solutions to the control of predatory behavior. In H. Markowitz and V. Stevens (Eds), Studies of captive wild animals. Nelson Hall, Chicago.

89)   Ellins, S.R., Gustavson, C.R. and J. Garcia. 1978. Conditioned taste aversion in predators: Response to Sterner and Shumake. Behav. Biol. 24:554-556.

90)   Quick, D.L.F., Gustavson, C.R. and K.W. Rusiniak. 1985. Coyote control and taste aversion. Appetite, 6:253-264.

91)   Nicolaus, L.K. www.conditionedtasteaversion.net

92)   Rusiniak, K.W., Gustavson, Hankins, W.G. and J. Garcia. 1976. Prey-lithium aversions: II Rats and ferrets. Behav. Biol. 17,73-85.