Summary in English        Summary in German

The purpose of this site.  Scientists communicate research results by publishing articles in scientific journals dedicated to their particular field of interest.  When their findings affect the welfare of the public, they are also expected to make the facts more generally understood.  This is especially necessary when there are important public misunderstandings concerning these facts.  Conditioned taste aversion (CTA) is important to the general public since it could be applied to wildlife management problems, including the preservation of wolves and other endangered species.  Indeed, it might well be the only reasonable solution to some of these problems.  But public understanding of CTA has become grossly contaminated with misinformation.  Attempts by several scientists, including myself, to clarify the issues within traditional professional channels have not improved public understanding and have failed to move those who have been the principal source of the confusion.  Now this misinformation has been posted onto the World Wide Web for public consumption.  As a leading authority on the ecological applications of CTA, I feel that I must try to correct the public record.

I undertake this with reluctance. The web has none of the traditional restraints of editorial and peer review that usually maintain focus on the facts in scientific publications.  My efforts to clarify may be lost since without these constraints emotionalism and distortion can very quickly dominate the debate, rendering the issue incomprehensible to the reader.  I will therefore present the facts so that the reader may check the sources and form an independent opinion.  In the interest of brevity and clarity I have oversimplified content most likely to give the general readership trouble.  I hope I may be forgiven for this. I make it clear when I am merely offering an opinion while statements of fact about which there might be some question are referenced with numbers in parentheses corresponding to scientific research articles listed in the back.  To do all this properly requires a bit of explaining. You will be better off reading the whole text from fore to aft, but you can also browse by clicking onto the main headings below.
 
An alternate explanation of Conditioned Taste Aversion
For those of you interested in a description of Conditioned Taste Aversion from a more naturalistic, biological perspective or just interested in a deeper understanding of the process and its history, I have prepared an alternative explanation here.
 
For those of you interested in the actual documentation of the claims I am making, I have included a selection of the most relevant scientific literature which you can find here.
 

• Is it "real"?
• How is CTA important to the public?
• What exactly is CTA?
• What you must know of brain structure to understand how CTA does what it does.
• How is CTA learned?
• CTA is not repellency.
• CTA is not aversive conditioning.
• What evidence is there that CTA works in actual practice?
• Our working model.
• If CTA can control predatory behavior in so many different settings, why is it not being used right now?
• CTA versus the American wildlife management heirarchy.
• How might Gustavson’s work be replicated?
• Finally...an appeal for support...27 years and counting.
• Acknowledgements
• Referenced literature.

Is it “real”?  Among humans, there is a universal but vague familiarity with the process of CTA.  When reminded of the process, hardly any adult will fail to volunteer an occasion when a familiar and preferred food (or drink?) cannot now be consumed after having been once associated with illness.  Anyone who has searched the subject of conditioned taste aversion (or, CTA; also referred to as the “Garcia effect”, conditioned flavor aversion, taste aversion conditioning and sometimes “aversive conditioning”) will have noticed that there is much available on the subject.  A casual search reveals many web sites (for example www.sciencesoftware101.com/AP-Biology), and closer examination shows discussions of CTA in basic psychology and animal behavior texts (1, 2, 3).  A thorough search discloses a vast and very broad scientific literature reporting many thousands of research articles in which hundreds of scientists have described how it works and what effects it can have upon behavior of all animals (4).  CTA has been produced in invertebrates as different as sea anemones and gastropods and in a bewildering variety of other animals, including humans (5, 6).  Although at first many rejected the idea because it was so revolutionary, CTA is now probably supported by more solid empirical evidence than is any other behavioral process. With the sole exception of the wildlife management hierarchy in the United States, there is now no serious controversy within the scientific community concerning whether or not CTA exists as a unique and most powerful form of learning, or whether CTA can quickly produce long-term changes in predatory behavior.

BACK TO SUBJECT HEADINGS

 
How is CTA important to the public?   Wild animals kill livestock and consume agricultural crops.  CTA can be used to reduce losses in at least some livestock and agricultural crops that are now being damaged by animal “pests”.  CTA is a non-lethal form of predation control, so when these “pests” are themselves threatened or endangered species as is the case of wolves in the U.S. and many species of rare or endangered animals that consume agricultural products in the third world, CTA could be used to spare both the animal and the farmer.  CTA could also be used to reduce predation upon rare and endangered species that are prey of wild predators.  When large predators are induced to approach humans because of associated foods, this threat could also be reduced by CTA without the need to kill the predator. 

The standard practices of shooting, poisoning or otherwise killing animal “pests” are not very selective, are increasingly resisted by the public, and have often failed to produce long-term reduction in depredation.  For example, for more than a century coyotes have been extensively killed in the U.S. but both their range and their populations are actually expanding.  Fences and other physical barriers are expensive, difficult to maintain and usually completely out of reach in poorer nations where the majority of endangered species exist and where subsistence agriculture makes every loss to wild animals much more acutely-felt than in developed nations.  Repellents largely do not confer protection to prey unless they can be continually and perfectly applied to target prey.  Frightening devices often become quickly ignored.   Guard animals (dogs and llamas) seem, however, to deter small predators such as coyotes and the potential and the limits of their usefulness should be pursued. 

Although there are some circumstances in which CTA is inapplicable, there are many others involving agricultural losses and declining populations of endangered species in which CTA may be the only solution.  I believe efforts to reintroduce wolves and other carnivores into the wild in the face of opposition by agriculturalists are excellent examples of this.

BACK TO SUBJECT HEADINGS

 
What exactly is CTA?  CTA is the process by which animals associate the taste of a food with an illness that occurs after its consumption.  They show that they have acquired this learning or, conditioning, by avoiding the taste of the particular offending food wherever it is encountered long after they have fully recovered from the illness.  It is an extremely powerful form of learning and is quite different from what most people experience as “conscious” learning.  It has the following characteristics (7):

1. Illness is associated with food taste even with a substantial time interval between eating and illness onset.
2. Illness and food taste are linked by the way the nervous system is organized and so except as described in #5 below, animals do not learn to avoid non-food/non-taste things or places in their external world through illness as a punishment.
3. Only a single taste-illness pairing may be sufficient to produce the aversion.
4. The aversion is extremely resistant to forgetting, or extinction and does not require repeated illnesses for it to be remembered long after recovery from the initial illness.
5. As long as the taste of food is present during the meal, other non-taste food cues such as food scent or, in the case of birds, visual food cues can be conditioned by illness, too.  In this way, a taste-illness event results in avoidance of the food from a distance on the basis of its scent or appearance (8, 9, 10).
6. The process superficially seems irrational.  For example, if a cancer patient receives chemotherapy or radiotherapy and then eats a meal (of even familiar, preferred food) before the onset of illness induced by the cancer treatment, the patient will likely form a very powerful aversion to the food (11).  The patient can consciously “know” that the illness was due to cancer treatment rather than the food and can discuss this relationship intelligently, but will still be unable to consume this same food and is usually nauseated at a distance by its smell.  It has been observed that wolves and coyotes that have consumed a mutton bait containing an undetectable dose of an illness-causing substance will, long after having recovered from the illness, avoid live, moving, bleating sheep as long as they taste and smell like the baits.  This is so even though the bait was not living and moving when it was eaten (12, 13).
7. Very different animals that have acquired CTA behave in remarkably similar ways.  Mammals presented with the offending scent or taste close by may shake their heads from side-to-side, wretch, urinate and retreat (12, 13, 14, 15, 16).  Since birds may avoid food from a distance on the basis of its appearance, the response may be much less dramatic; a simple act of ignoring an offending object.  If they taste the food, however, they engage in just the same kind of head shaking and other behaviors seen among mammals (17).
8. Location and specific contexts are largely irrelevant to the formation and expression of CTA.  That is, it doesn’t much matter where or under what particular circumstances a meal that induces illness is eaten.  The taste (and scent or appearance) of the food associated with the illness will later evoke rejection wherever and under whatever circumstances it is encountered.

BACK TO SUBJECT HEADINGS

 
What you must know of brain structure to understand how CTA does what it does.  To understand how CTA works, you must consider a few facts about the anatomy and functioning of the nervous systems of animals.  The gross brain structures of vertebrates (fish, amphibians, reptiles, birds, and mammals…including humans) all have a few things in common.  One commonality is that the brain is divided into two interconnected but anatomically distinct and functionally different parts (3).  These parts are structurally referred to as the Brain Stem and the Fore Brain (each of these two portions is sub-divided into several interconnected regions and tracts which we will ignore here).

     The Fore Brain.  This portion of the brain is not principally involved in the formation of Conditioned Taste Aversion.  Since most people badly confuse the learning of CTA with the learning that occurs in the fore brain, we must dispose of it first to avoid this difficulty.  The fore brain is composed of the outer cerebral cortex, where most of our conscious “being” is experienced, together with sub-cortical structures that contribute to the functioning of the cortex.  Taken together, the fore brain is built to receive input from the external world and make decisions that guide the animal through its external environment.  It uses routines enabling it to improve its performance through learning, and so we and other animals can change the way we respond to and “view” the world on this basis.  In general, this learning is a process by which the brain sorts out irrelevant stimuli from relevant signals that are precise in space and time.  The learning is rather like trial and error/success, since the external world has so much irrelevant noisiness in it.  The rules it follows to “make sense” out of all the confusion through a process known as “classical conditioning” are as follows (18):

1. The time between the “signal” (a bell, say) and the reward/punishment (getting dinner/an electric shock) must be very short in order for the animal to learn that the signal really does predict a relevant outcome.  Otherwise, irrelevant confusing “noises” (other sounds, the wind, etc.) intervene and confuse.
2. Even with a short interval between signal and outcome, most animals need several trials that pair the two before they “get it”, or learn the connection.  After all, while the bell was ringing, a lot of other things were going on in the outside world too and some of these might just as likely be a signal to dinner or the shock as was the bell!
3. Once the connection between signal and outcome is learned, the animal shows that the signal now has some special relevance by responding to the signal in some special way (approach eagerly, run away).  But if the signal is not then followed by the now “expected” outcome (reinforced with repeated reward/punishment), then the animal eventually stops responding and the learning is said to have extinguished ( been forgotten).
4. Some things are not learned easily by this part of the brain.  The examples of “conscious” learning above include signals and outcomes that are external to the animal. This part of the brain does not learn much when signals and outcomes are internal.  These specifically, are food taste (inside the mouth) and illness (inside the body)! (7)

        The Brain Stem.  The unique and powerful features of CTA are understandable only if you understand the general functioning of the brain stem and how it regulates the internal melieu of the body. Among other things, the brain stem receives input concerning the internal working and condition of the body and makes executive “decisions” regulating that internal working.  It controls such “vegetative” functions as movement of the gut to aid digestion, respiration, and constriction of blood vessels to fine-tune blood flow (3).  Most people are surprised to hear that the brain stem, together with its associated structures, can also “learn” some things much better than the conscious forebrain.  It is the brain stem, together with some structures that project above it, that “learns” what foods are good to eat (that is, improve the internal environment) and what foods are dangerous (damage the internal environment: illness). This is where decisions concerning what foods you prefer to eat and what foods you prefer not to eat are made.

It is structured as follows (19): The nerves in the tongue that bring taste sensation to the brain do not go to the conscious forebrain.  Instead, they first go to a part of the brain stem in the medulla called the nucleus of the solitary tract (NST).  Converging at the same place is a branch of one of these nerves (the vagus nerve) that also brings sensory impulses from the upper gut.  Taste information from the medulla is brought forward to an area beneath the fore brain, and a little of this is allowed to reach the cortex where we can consciously report that we “taste” something.  This information is very limited, however (we consciously taste only “sweet”, “sour”, “salt”, and “bitter”, often confusing sour with bitter).  Most food flavor we experience consciously is really scent, not taste.

[We consciously know little about our unconscious selves, but we can make contact with our powerful unconscious “other person”.  For example, consciously decide to hold your breath for as long as you possibly can and a very different inner person will soon oppose you.  Or, you might consciously decide to jam your finger down your throat, or chew on an aspirin full of bitter alkaloid and you will soon see what your brain stem thinks of the idea.  By doing the last two of these, you are also getting into contact with what it is like for an animal that is confronted with the taste of a food for which it has acquired a conditioned taste aversion!] 

BACK TO SUBJECT HEADINGS  

 
How is CTA learned? For predators, eating follows an inevitable sequence.  It starts with food seeking (appetitive) foraging behavior controlled largely by the fore brain.  It is flexible behavior because it has to be.   Foragers almost never know where food actually is or what it will do once it’s located.  By the time food actually gets into the mouth the brain stem largely takes over (salivation and gastric secretion begin) and this includes food taste sensation going to the brain stem.  If the food has not previously been linked with illness, then the brain stem will allow it to remain in the mouth and to be swallowed.  Next, it begins to be digested further.

If the outcome of digestion and absorption is “good” (nutrient repletion), then all is well and this taste will be “remembered” by the brain stem as acceptable and so again be welcomed in the future.  Since the food is still desirable, external cues associated with this food will continue to be sought by the animal as it forages in the external environment.  That is, the unconscious brain stem decides what food the conscious fore brain “wants” to pursue.   As they say, “You might be able to control what you do, but you can’t so easily control what you want to do!”   

If the food is poisoned, then something very different happens.  Many poisons produce gastric distress.  These signals are brought by the vagus nerve from the gut to the nucleus of the solitary tract (NST) of the brain stem, which is also known as the “emetic center”.  Sufficient stimulation here results in vomiting which purges the poison.  Poisons that don’t irritate the gut may remain long enough to be absorbed into the blood.  If so, then vomiting occurs when another small area in the brain stem near the NST called the area postrema (AP), is triggered by blood-borne poisons.  If the poison is removed by vomiting or by the lengthier and usually more punishing process of detoxification, then the animal may fully recover.  The only lasting effect will be that a “memory” for the taste of food eaten before the onset of the illness will be retained by the brain stem.  As described earlier, other senses like food scent or appearance can also be conditioned and so the offending food may be avoided at a distance.    

What we now understand about the structure and functioning of the brain stem can, in theory, explain why CTA is so powerful and different from classical conditioning.  Each of the unique characteristics of CTA listed above under   What exactly is CTA?   now makes more sense:

1. A delay in time between a meal and illness still produces CTA while such a delay would ordinarily prevent classical conditioning. Since this part of the brain only receives information concerning taste and internal condition of the body, there is little opportunity for confusing and competing information to interfere with making the connection between the remembered taste signal and illness that follows some time later.

2. Since taste and illness information converge at the important brain stem regions, there is a special link between taste events and illness events.  This is why electric shock does not alter preference for food taste and bells and whistles don’t get avoided when they precede an illness (7, 20, 21).

3. Since there is much less confusing “noise” in the taste-illness system, only a single pairing of taste and illness may be sufficient for the learning to occur.

4.  Again, since the taste-illness axis is insulated from confusion, there is less interference that might result in the kind of “forgetting” associated with classical conditioning.  The brain stem will grimly remember the taste of a food that was associated with illness for a very long time without any need to be “reminded” by reinforcement.

5. Animals that have experienced a taste-illness pairing can avoid the offending food/prey on the basis of non-taste stimuli (scent, visual cues) because in some way the taste component of the meal allows other, non-taste cues that were simultaneously present with the taste to have access to the taste-illness system (8, 9, 10).  Non-taste food cues don’t have access unless there is at least some taste cue present during feeding.    Precisely how this process works is a matter of continuing research.

6. CTA only seems irrational if one only considers familiar, conscious learning and thinking processes.  It makes much better sense once you understand that unconscious processes that “listen” only to the internal condition of the body decide food preference.

7. Since location and the particular circumstances surrounding a meal that induces illness do not have access to the brain stem, memory of these details is not retained by it.  The food taste (and scent/appearance) associated with illness will be avoided wherever encountered.

8. The remarkable similarity of behavior among such a wide variety of animals now makes sense, also.  Of all of the parts of the central nervous system, the brain stem has evolved the least and so has remained similar among vertebrates.  Structural similarity in the nervous system strongly suggests functional similarity.

BACK TO SUBJECT HEADINGS

 
CTA is not repellency.  From the foregoing, it should be clear that once a predator has a conditioned aversion to some prey item it is the taste and other cues associated with the prey itself that the predator will avoid.  A repellent like hot sauce, on the other hand, is a noxious substance that presents a barrier between the predator and the prey the predator still wishes to consume.  Thus, any behaviorally plastic predator will exploit or even create a break in the barrier.  Even the mosquito with its limited behavioral repertoire is ready to resume feeding whenever and wherever the repellent is rubbed off.  In general, tests using noxious repellents to protect prey have been unsatisfactory (21, 22) for just these reasons.

BACK TO SUBJECT HEADINGS

 
CTA is not aversive conditioning.  “Aversive conditioning” refers to any occasions in which animals associate some cue with some punishment (18).  CTA can be classified within this category and it was very early on referred to in this way, but there are many important differences between CTA and what is ordinarily understood as aversive conditioning.  Most aversive conditioning involves some form of external punishment such as an electric shock or rubber bullets that are used to punish an animal just as it is engaging in some kind of forbidden behavior. Since both the signals and the punishment are external to the animal, the learning follows the rules listed above for classical conditioning rather than the learning rules for CTA.  External punishments such as electric shock can greatly affect how animals respond to external cues such as location, physical objects, lights and sounds. But, the procedure absolutely requires that the punishment be applied many times, and always right while the animal is engaging in the offending “act”.  This learning may be highly situation-specific and so may not reliably generalize to new situations in other settings and new places.  Since this learning is very different from CTA, it also fails to alter preference for food taste (7, 18, 20, 21).  Furthermore predators must be maintained in captivity to be conditioned before they may be released, but CTA can be produced merely by distributing treated baits into the field, as you will see.  Although it has not “approved” the use of CTA, the U.S. Fish and Wildlife Service has “approved” the use of shock collars and rubber bullets to condition wolves to avoid livestock in spite of these difficulties.  For more complete information on these procedures see the following: (http://www.forwolves.org/ralph/no16dead.htm and http://www.enn.com/news/enn-stories/2000/07/07062000/rubberwolf_14471.asp). 

BACK TO SUBJECT HEADINGS

 
What evidence is there that CTA works in actual practice?   Nearly all of the information concerning CTA described above was known by 1974, when Carl Gustavson first proposed that CTA might be used to alter the feeding behavior of free-ranging predators (12, 13).  Captive studies of black bears, domestic cats, ferrets, cougars, hawks and other predators had all demonstrated that these animals could easily form CTA (6), but the most relevant were Gustavson’s captive studies of wolves and coyotes.  He demonstrated both would avoid the taste and scent of live prey after recovering from the effects of consuming a meat bait laced with an undetectable dose of an illness-causing substance (Lithium Chloride). In neither of Gustavson’s early studies were there any complex or special inducements to produce the effect.  Healthy and unimpaired predators were simply allowed to kill and consume live prey in large enclosures to demonstrate that they were able and willing to engage in normal predatory behavior.  Some days later they were given a meat bait with the taste and scent of the live prey that was laced with an undetectable amount of an aversion agent: lamb wool wrapped around a large amount of ground mutton with a moderate amount of Lithium Chloride well mixed with the meat.  They were observed to become ill, were allowed to recover for several days on their usual dog food diet, deprived of food until they were very hungry, and then again placed into the enclosure with live prey.  Predatory attack was suppressed, resulting in the prey surviving unharmed over post-test durations vastly exceeding predator-prey confrontations in nature.   Later, CTA was tested in field trials to reduce coyote predation upon sheep and to induce coyotes to abandon campgrounds where they had been pestering tourists for food (23, 24).  In these trials, treated sheep baits were distributed wherever free-ranging coyotes were most likely to find and consume them and changes in reported predation upon sheep were recorded.  Although in these field trials other interpretations of the results were possible because it was not possible to make direct observations of predator-prey interactions, the simplest explanation was that CTA was produced and this resulted in the treatment effects that were reported.  CTA to control coyote predation upon sheep was again successfully tested in a 3-year field project in Saskatchewan (25).

In 1977, my students and I found that free-ranging raccoons in a natural setting could be easily induced to cease attacking and killing live chickens by consuming chicken carcasses laced with small amounts of Lithium Chloride very well distributed in the baits (14).  The same individuals were observed before, during, and after eating baits and since the only raccoons that attacked chickens after treatment were those not present during the treatment, the results seemed clear-cut and lasted as long as we continued the study (over a year).

Since then, research involving free-ranging predators has proceeded, but its rate has been slowed by lack of funding; a subject I will discuss briefly at the end.  In all of the studies involving free-ranging predators the predators were entirely outside of captivity, entirely untouched and entirely free to move wherever they wished.  In these field studies, baits were simply placed in the most conspicuous locations possible to optimize the chances that these free-ranging predators would find, consume them, and thus be treated.  In much of this field research designed to produce and evaluate CTA among free-ranging predators, eggs were used as prey. This is because eggs are familiar and preferred prey of a very wide variety of predators and aversion-producing substances can be added to the interior. Eggs are also totally without their own defense against predation. Therefore, whether or not CTA can suppress predatory attack among free-ranging predators is a much more clear-cut event than would be the case if eggs were capable of escaping predation without the protection of CTA. 

By randomly assigning breeding territories of predatory birds into treatment and control sites in 1981, we demonstrated that free-ranging American Crows very quickly acquired CTA by eating green-colored eggs treated with a small amount of well-mixed illness-inducing agent (26).  This not only suppressed attack upon eggs of a particular color in places where their consumption produced illness, but also untainted eggs of the same color in other locations as well.  All the while, Crows occupying other breeding territories consumed untreated eggs freely.  I repeated the procedure in 1982 with Common Ravens that developed CTA to illness-causing eggs designed to look like those of Greater Sandhill Cranes in a national wildlife refuge (27).  At the refuge, breeding pairs of Ravens defended their territory against intrusions by the many non-breeding Ravens.  In effect, treated territorial Ravens that refused to consume the eggs themselves, “baby sat” the surrogate eggs by protecting them from predation by others that had not been treated.  Again, Ravens in breeding territories assigned to the control group ate “Crane” eggs freely.  CTA suppressed attack upon eggs among free-ranging American Crows in two further studies (28, 29) and among a wide variety of free-ranging mammalian egg predators as well (30, 31).  In a modified replication of the old raccoon study, individually ear-tagged free-ranging raccoons were observed in the full sequence of consuming untreated egg prey, consuming treated eggs, and then returning to be observed to ultimately avoid untreated eggs at a distance during a post-test spanning two years (15).  More recently, CTA has been established among free-ranging nuisance black bears (32) and among free-ranging black birds that consumed insecticide-tainted insect prey (33).  On only one occasion in any of my studies have predators failed to acquire a long-term aversion that resulted in avoidance of untreated target prey.  Mongooses in the Virgin Islands learned to discriminate between treated and untreated eggs because in the tropics treated eggs began to decompose within 2-3 hours and so both scent and taste of these eggs differed from that of untreated eggs (34). 

BACK TO SUBJECT HEADINGS

Our working model. The working model we have always followed in predicting that a taste-illness pairing that produces CTA should suppress the killing of prey, wherever they are encountered, is based upon observations of encounters between prey and predators with CTA.  Prey survive these encounters because predators with CTA respond to them with “disgust” rather than killing and eating them. Long after having recovered from eating a treated sheep bait, captive wolves and coyotes gag and retreat as though punished again by the mere scent of live sheep (12, 13).  Free-ranging predators in field studies have been observed to respond in a similar manner on occasions when they have been observed to come into close contact with offending prey (14, 15, 32, 33).  Predatory attack in field studies has generally been absolutely suppressed, presumably because both predators with CTA and their prey are free to avoid each other at greater distances than in captivity.   

Sometimes only the taste component is conditioned and so we find a hungry coyote whose nose and eyes say “YES!” and so it attacks, but then whose tongue says “NO!”.  What is seen then is the coyote with CTA to the taste of lamb wretch and run away upon biting lamb (12).  When again confronted with live lamb, predator avoids at a distance.  

BACK TO SUBJECT HEADINGS

If CTA can control predatory behavior in so many different settings, why is it not being used right now?  I have been asked this question many times.  My response has, until quite recently, been a shrugging “ask the governmental agencies with the responsibility for wildlife management”.  In the process I know it sounds like I am hiding some flaw in CTA.  There is no hidden flaw.  But, the U.S. Fish and Wildlife Service, and now the U.S. Department of Agriculture have widely pronounced that we have been in error and that CTA fails to produce desired results. Before continuing, we must consider:

A word about scientific ethics.  Whenever a scientist believes that some other scientist has made false or erroneous reports, it is his ethical duty to replicate the questionable work in order to see whether or not the original reports are justified before publicly announcing that the reports are false.  In fact, it is the scientific tradition that no new discovery ever becomes widely accepted until it has been independently replicated many times by other scientists and so undergone just this kind of careful scrutiny.  It is precisely for this reason that every rigorous scientific journal demands that all reports it publishes include clear descriptions of the methods used to produce results.  Only in this way can other scientists know how to replicate the results. 

Claims made by the U.S. government that CTA failed to “work” were founded on tests of CTA that failed to replicate work reported by Gustavson and others.  When the tests reporting “failures” of CTA were first published, it was our opinion that the reported results were perfectly reasonable…given the procedures that were used.  The problem was these tests failed to replicate Gustavson’s studies.  That is, in each case where it was possible to ascertain what the procedure was, the procedures differed from those Gustavson used and recommended. These differences usually took the form of including some feature of the experimental procedure that enabled predators to discriminate between the food cues associated with bait and those of the live prey.  Lithium Chloride tastes salty but live prey do not and so Gustavson recommended care in hiding a sufficient quantity of Lithium in the baits to induce illness in such a way that the taste and scent of baits did not differ from that of live prey.  Captive or free-ranging predators that consume an illness inducing bait with food taste that differs greatly from that of live prey will discriminate between the two by avoiding the former and continuing attack and consumption of the latter.  That is, an aversion to food item “A” does not dupe a predator into avoiding food item “B” if “A” and “B” differ in ways relevant to CTA.  Thus, while we recommended that meat baits be carefully constructed in order to hide 3.0-8.8 g of salty-tasting Lithium Chloride per Kg of meat bait  (average: 4.0 g/Kg bait) (12, 13, 14, 35), studies reporting failure of CTA used from 13.5-45.0 g Lithium Chloride per Kg bait (average: 22.6 g/Kg bait, or a factor of 5.5 times more salty Lithium) (36, 37).  Predictably, predators that consumed salty-tasting meat baits, refused to eat salty meat baits again after recovering from illness (showing that CTA had indeed been produced), but continued to attack and consume non-salty baits or live prey.  In other “failed” studies, where substances other than Lithium Chloride were used, differences between bait and target prey continued to be included (38, 39), or the research was arranged in such a way as to make it literally impossible to produce a measurable CTA (reviewed in 35).

The flaws in the studies reporting “failures” were fundamental, very easy to spot and were thoroughly critiqued in published scientific review articles (35, 41).  But, to date no state or federal agency in the U.S. has published the results of a legitimate replication of procedures used by Gustavson and others to successfully produce CTA from as early as 1974.  Instead, the flawed studies reporting “failed CTA” continue to be used to support the claim that there is controversy concerning whether or not CTA can “work”. 

The only hopeful sign was in 1995 during a conference on repellents (42) where a paper was given supporting the use of baits that successfully mimicked live prey.  Thus, our enjoiners to dupe predators by maintaining the same taste and scent cues between baits and live prey were finally adopted, but for some reason were repackaged under the unwieldy reference to “Conditioned food aversions based on deception or, CFABD.   The author correctly made much of the need to be sure baits retained the same food cues as those of the live prey we wished to spare from predation.  Curiously, however, none of the work in which this principle was so flagrantly violated was cited as obvious illustrations of why such “deception” was so necessary.  The idea was near to the right track but unfortunately as seen below, this move, like all of our enjoiners, has not much influenced the leadership in the wildlife management hierarchy.

BACK TO SUBJECT HEADINGS

CTA versus the American wildlife management hierarchy.  In spite of hearing many rumors during these years, I have not attributed to malice what could be explained by incapacity.  After all, authors of earlier work reporting “failures” seemed to have misunderstood CTA.  For example, from the first CTA was confused with repellency and “aversive conditioning” (35, 41).  Even now in spite of our warnings, repellency and CTA are often treated as synonymous in wildlife research reports.

Two events have changed my mind concerning the nature of the problem, however.  The first of these was that the Central Science Laboratory of the British Ministry of Agriculture, Fisheries, and Food has recently undertaken to do just the sort of studies that have not been undertaken by the U.S. Government (16, 43).  They undertook them principally because they recognized the potential of CTA, but were troubled by the “controversy” surrounding the process. Athough they are not in a position to work with large predators and so cannot replicate Gustavson’s wolf research, their tests of CTA have revealed that the kinds of results first reported by Gustavson and later by me can be easily and reliably obtained.  They report that predators as different as rats, cormorants, and red foxes quickly formed CTA and avoided referent prey at a distance over many trials; findings very much at  odds with their American counterparts.  Their work confirmed again that our successes in producing CTA were not merely due to some special talent that only we possessed. The other event was my discovery of a web site maintained by the federal agency with which those reporting “failed” CTA are most consistently associated.  This is the National Wildlife Research Center of the U.S. Department of Agriculture.  I invite all to read this page.   It is: www.aphis.usda.gov/ws/nwrc/flavor.htm.  It is anonymous and cites no published evidence for its many claims.  Among these are the claims that CTA somehow does not suppress predatory attack and so cannot produce beneficial effects in a cost-effective manner and that CTA is too sensitive to “procedural variations”.  These amazing assertions contradict a very large body of scientific evidence (for a few examples: 5, 6, 12, 13, 14, 15, 16, 26, 27, 28, 29, 30, 31, 33, 35, 43).  Beyond its falsehoods, the web site seems to me to serve no purpose other than to try to extinguish, in the strongest possible language, any further discussion on the use of CTA.  Thus, the rumors I have ignored for many years now seem to have been publicly confirmed for the first time by the tone and content of this web site.  That is, it appears that resistance to CTA within the American wildlife management hierarchy indeed is, like CTA itself, more a matter of the gut than of reason.   

Why should this particular group regard a perfectly neutral and factually supported behavioral process such as CTA with such antipathy when no other group has difficulties with it?  Although it is not externally evident, this group is not perfectly homogeneous.  I personally know some members of this very wide spread consortium of wildlife management researchers who do not share the attitudes of those who influence the allocation of vertebrate animal damage research dollars and so the direction of research.  A few wildlife biologists, always peripheral to the policy makers, have even publicly criticized the low qualify of research work sometimes produced by wildlife biologists.  An example of this is a paper published under the pseudonym John MacNab (44).  So far, however, the public ranks of those working on animal damage control have remained remarkably closed.

Any attempt to explain the arbitrary, capricious, and in fact unethical behavior directed toward CTA must therefore focus upon the motives of a relatively few policy makers.  I am not a psychoanalyst and so cannot provide a precise opinion that can be confirmed with evidence, but an informed reading of the web site posted by the U.S. Department of Agriculture www.aphis.usda.gov/ws/nwrc/flavor.htm reveals more about the attitudes of those who produced it than I can say.   What ought to raise a few eyebrows is the public spectacle of a federal bureaucracy caught in the act of  “trying to strangle the baby in the cradle”.

Vertebrate animal damage control research is now, as always, dominated by the U.S. Fish and Wildlife Service and the U.S. Department of Agriculture and is mostly, in my opinion, a blizzard of muddled, expensive and predictably inconclusive studies of poisons and other lethal devices, repellents, aversive conditioning, scaring devices, and physical barriers.  The experiments reporting “failures” of CTA are, in my opinion, scientifically useless and so a waste of money and, more importantly, opportunity. They reveal little that is new and useful. The studies ignore the simple and effective procedures known to produce CTA and that suppress predatory attack upon prey. Instead, they stop short after demonstrating how one procedure or another might fail to confer protection from predatory attack. No attempt is made to use the data to develop means of making CTA succeed in any particular application.  Since there are situations involving real and immediate risk to endangered species and to agricultural losses that might be mitigated only by the use of CTA, agencies that feel free to ignore or discourage CTA while pursuing less promising measures appear to me to be “fiddling while Rome burns”.

I believe that things would very quickly take the road to clarity with legitimate replications of Gustavson’s captive studies of wolves and/or coyotes that are video taped, publicly conducted and whose results are publicly disclosed. Of course, these would have to be conducted with healthy, unimpaired animals and by a clearly neutral and reputable behavioral biologist familiar with the scientific literature on CTA and willing and able to follow Gustavson’s procedure with complete openness. 

BACK TO SUBJECT HEADINGS

How might Gustavson’s work be replicated? Since 1974 public support for predator-prey research in which predators kill and consume live prey has changed and so an absolutely pure replication of Gustavson’s research is no longer possible.  A modified procedure, however, could still legitimately replicate Gustavson’s procedure at least in so far as we currently understand the process of CTA, complete with suppressed attack upon live prey and in a way that meets any reasonable objection on humanitarian grounds. The essence of the work is to show that large predators that otherwise would attack and consume target prey cease doing so if they have acquired CTA.  To show that hungry predators are normal subjects for testing, it is necessary to document their attempts to attack target prey before CTA is established and to show that such attack ceases after CTA is established.  In these tests before CTA is established, physical barriers can protect sedated prey as the behavior of hungry predators attempting to get to them is documented.  Then, the first posttests after predators recover from having consumed appropriate treated baits may be made with prey carcasses, or with sedated prey protected by the same barriers, or both. Only after suppression of predatory attack is demonstrated may live prey be introduced with no barrier. Furthermore, lithium chloride was used in 1974 because it was the only aversion agent then available, not because it is an ideal substance to produce CTA.   Its salty taste requires that care must be used in constructing baits to prevent predators from discriminating between baits and live prey (36, 37) and it is also marginal since the illness it induces among predators is severe.  Fortunately, other substances such as thiabendizole and ethinyl estradiol that are very effective in producing CTA have been identified since 1974.   They are much less detectable and they also do not produce severe illness symptoms among the predators (15, 16, 30, 31, 32, 43).  With these new substances proven and available, I would now not dream of using Lithium in most captive and field applications. 

Although a legitimate replication of Gustavson's research will require some funding, this sort of research need not be prohibitively expensive.  CTA is a relatively low technology procedure that doesn’t require expensive bells and whistles.  I know this because to date, I have had to pay for most of my research out of my own pocket and I am not a rich man (see the acknowledgements sections of each of the research papers I have authored).

I presume that once it becomes generally known what a proper application of CTA can do, the wildlife hierarchy will not be able to continue to resist it with the impunity they have enjoyed for so long.  I believe the public deserves to receive what it pays for when it supports organizations responsible for wildlife research, though they will not receive it unless they are informed enough to demand it.  This should be nothing less than honest and prompt reporting of results of good-quality, honest, aggressive research into any process with prospects of success, regardless of its bureaucratic popularity.  Where research expertise outside of the organization exceeds that inside, it should be willing to support rather than resist these research efforts.  Mistakes of the past should be admitted even though such admission might embarrass a colleague. 

BACK TO SUBJECT HEADINGS     

Finally.  You are probably wondering why I am authoring this web page.  I have to author it.  Carl Gustavson died a few years ago, no one else has stepped in to correct an increasingly contaminated public record and there have been precious few competent scientists working on the problem of field applications of CTA.  I am now semi-retired and so have nothing to gain other than peace of mind.  From the first, CTA struck me as very interesting, and an honorable thing to spend my life researching.  This I have tried to honorably do and I will no longer remain silent while it is so grossly misrepresented.

I expect that every manner of objection will be raised by what is said here and fear that the main point may become obscured.  Here then, is the main point: ordinary scientific ethics require a legitimate and absolutely open public replication of research that has been called into doubt, especially if the outcome is as important as the survival of endangered species.  I urge that interested individuals and non-government organizations seriously consider financially supporting legitimate replication of Gustavson's work.  If ever a little money could make a difference, this is it.

YEARS WITHOUT A LEGITIMATE REPLICATION OF GUSTAVSON’S WOLF RESEARCH: 

27…AND COUNTING

BACK TO SUBJECT HEADINGS

Acknowledgements

I expect to be crucified for speaking the truth but I fear being crucified for saying something stupid.  I have therefore asked a variety of persons to read and comment upon these pages.  These referees range from very wise laymen to loving well-informed family and eminent scientists.  I would like to thank them all and would, under ordinary circumstances, do so by name.  But, the cyber-ground we now tread is not firm enough to assure fair treatment to every contributor.  Indeed, some have taken professional risks by participating.  To the anonymous, I would like to express my sincere thanks.  My relatives are already implicated and so I can name them with thanks and love: Jan my wife, John and Michael my sons, and Beth and Michelle my daughters.  And finally my thanks also to Lorraine and Brian Sinclair, for the German translation of the summary.  

BACK TO SUBJECT HEADINGS

Referenced Literature

BACK TO SUBJECT HEADINGS