Introduction
For the last 25 years the welfare of domestic animals has been an important aspect of applied ethology, i.e. that part of behavior biology that studies domestic animals. Consequently, scientists have searched for objective ways to evaluate welfare. As part of this evaluation, various ways to measure stress have been proposed, in that the demonstration of stress indicators could provide an objective information about how well farm animals kept under production conditions have adapted to their environment (e.g. Ladewig, 1994; 2000). Over the years many such potential indicators have been suggested, both behavioral and physiological stress indicators, as well as production, fertility, and disease related indicators. The main result of this research is that we have realized that we cannot answer these complex questions with the information provided by a single or a few stress indicators. In behavior science two separate disciplines focus on the behavior of domestic animals. Apart from applied ethology, i.e. the discipline that works primarily with the behavior of farm animals (cattle, pigs, and chickens), another discipline, usually referred to as behavior therapy of companion animals, works with the treatment of behavior problems in dogs and cats. Although the two disciplines use somewhat different approaches, they both attempt the same, to identify behavior problems in domestic animals, to search for the reason for them, and to correct them in various ways. In applied ethology the behavior problems are related to the way farm animals are housed and managed. Apart from fear related problems, the most common problems are aggression because of high stocking density, large group size, and competition over limited resources and stereotypies due to housing in a restricted and sterile environment. In companion animals behavior problems are also related to the way the animals are housed and managed, although cat and dog owners use different terms than housing and management. The most common problems are separation anxiety and aggression towards humans in dogs and urination problems in cats, all problems that, to a large extent, are related to fear. In applied ethology the approach to solve the problems is mostly experimental, i.e. scientific studies are first conducted in an experimental setting before they are transferred to practice. The approach in behavior therapy is primarily clinical, i.e. based on case stories and only to a limited extent on scientific experimentation.
To date the two disciplines have had surprisingly little contact with one another. This lack of collaboration is, however, unfortunate, because many of the problems that exist in the two areas are quite similar. A dog's fear of being left home alone is principally not much different from a fattening pig's fear of being attacked in its social group. A horse's anxiety of being ridden in traffic is in principle not much different from a dairy cow's hesitation to enter a milk robot. The problems we face when trying to diagnose these states are similar. Moreover, the way these states can be treated or prevented undoubtedly has much in common, too.
In fact, the problems that these two behavior disciplines face are not much different from the problems that veterinary medicine face in general. To identify a problem, whether a behavior problem or a health problem, is a process that in veterinary medicine is called to make a diagnosis. This diagnosis is based on a clinical examination and on various diagnostic tests. Correction of the problem once the diagnosis has been made is called therapy and procedures that reduce the probability that it occurs sometime in the future are called preventive medicine.
Comparison between animal health and animal welfare
The search for symptoms and the conduction of diagnostic tests to reach a diagnosis, plus therapy and prevention of the problem is the procedure used in medicine. It is also the way behavioral problems are handled in the behavior clinic. There are many reasons why we should use a similar procedure when we tackle welfare problems in farm animals.
Health is a condition characterized by the absence of disease. To be healthy, an individual must be free of bacterial and viral infections, of parasite infestations, it must live on a balanced diet, not be exposed to noxious agents plus many other conditions. Consequently, in order to guarantee the health of an animal (e.g. in order to issue a health certificate), a veterinarian looks for symptoms of a disease, such as changed behavior, fever, pale or reddened mucus membranes etc. plus conducts various diagnostic tests (such as serological tests, x-ray examination, etc.). If no symptoms of diseases are found in general or in specific, a limited warrantee can be given that the animal is healthy.
In the same way that health is characterized by the absence of disease, welfare is characterized by the absence of a number of aversive states such as pain, suffering, frustration, boredom etc. Consequently, to guarantee the welfare of an animal, we must search for symptoms (or stress indicators) such as changed behavior, altered secretion of stress hormones, altered heart rate, reduced fertility and many more. In addition, we may conduct various behavior tests (open field test, elevated plus maze test, human approach test, tonic immobility test and many more). If no symptoms of reduced welfare are found, we can give a limited warrantee concerning the welfare status of the animal.
In our search for symptoms of reduced welfare it is important to realize that just like "disease" cannot be diagnose by one or a few signs common for all types of sicknesses, so it is not possible to diagnose reduced welfare by a few symptoms either.
Instead of searching for common signs of stress that can indicate the presence of a wide variety of aversive conditions, we must reverse the process and use the approach that is common in medicine. When a health problem occurs, the search for the cause begins. In addition, specific signs or symptoms of the disease are defined and, if necessary, specific diagnostic tests are developed to help diagnosis of the disease. The next step in the process is to develop an effective therapy of the disease and finally the effect of preventive measures are tested out.
When a similar approach is used to solve behavior problems, we search for the cause of the problem. Usually behavior scientists agree that this step is an important one. But when it comes to defining specific signs of the problem or to develop specific behavior tests that could help identification of the problem, we are often less specific and less goal oriented.
Epidemiology, health, and welfare
One thing that behavior scientists can agree upon is that the cause of most problems is multifactorial. Unfortunately, it is an extremely rare situation where we are able to pinpoint one or a few reasons for a problem. On the contrary, in most cases we have to consider a multitude of factors as well as the interaction between them, when we analyze the reasons for a problem. Needless to say that more than once have we lost our way in the chaos towards the solution of the problem. But also in this aspect it can be helpful to turn towards the health profession to look for ways to analyze problems of a complex nature.
Many health problems nowadays are not caused by a single agent, the presence of which means that the animal is sick, and the absence of which means that the animal is healthy. Instead, more and more diseases are caused by two or more states that must be "just right" in order for the disease to manifest itself. The analysis of the interaction between involved factors is one of the subject areas of epidemiology.
Since most behavior problems are of a multifactorial nature, the use of multivariate techniques, such as factor analysis, is an important addition to behavioral research. As a consequence, more and more behavior studies include such methods, at least as far as the causation of a behavior problem is concerned (e.g. Alban et al., 2000; Moinard et al., 2000). There is little doubt that the inclusion of such methods is an absolute necessity in our future work.
Diagnostic tests or behavior tests
In clinical work it is often the case that symptoms revealed by a general examination (e.g. body temperature, pulse, blood pressure etc) are not specific enough to justify an exact diagnosis. Principally, this situation is similar when it comes to a behavior problem. Most behavior problems in cats and dogs are not seen in the behavior clinic. Similarly, many behavior problems in the barn are only performed when the stock person is not present, or the symptoms of the problem are so subtle that the animal must be put in a special situation to reveal their existence.
Just like the health profession has developed specific diagnostic tests (serological tests, x-ray examination etc), so have behavior scientists over the last decades developed various behavior tests, the aim of which it is to support a diagnosis. Thus, many fear tests have been developed (e.g. the human approach test in different species, tonic immobility in chickens, the glove test in mink), or tests for general activity (e.g. the open field test). Only a few of these tests, however, have been sufficiently evaluated according to the methods used for evaluation of diagnostic tests. Obviously, before a diagnostic test can be used in connection with the eradication of a disease, it is necessary to know to what extent the test can be trusted. In other words, the sensitivity, the specificity, and the prevalence of the test must be calculated.
As far as behavior tests are concerned, the same evaluation must be done before we can trust our tests. Obviously, before we use a test to say something about the level of fear in a herd, we must know how good our test is to indicate both fearful and calm individuals. Again, the tools used in epidemiology can be applied for such an evaluation.
Therapy: Solutions here and now
One of the goals of our study of the behavior of domestic animals is to improve welfare. An important aspect of this effort is to reduce the frequency of behavior problems. In principle this reduction can be achieved either by treating animals in some way after the problem has developed or by somehow changing aspects of the environment so that the problem does not develop. The first approach is similar to therapy in the clinic, the second is similar to prevention.
In applied ethology we have seen many instances where the therapy of a behavior problem has been done by others that the behavior scientists. Thus, producers are often quite effective in developing methods by which a problem is solved. Tail biting in fattening pigs is reduced by cutting part of the tail off. Feather pecking in laying hens is avoided by amputating part of the beak in chickens. Crushing of the piglets is avoided by putting rails around the lactating sows, to name a few examples. Crib biting in horses is suppressed by placing a tight collar around the throat of the horse. Aggression between group members of many different species is avoided by keeping animals in individual housing.
Probably all behavior scientists can agree that these "treatments" are symptomatic, i.e. they are aimed at reducing the animal's possibility to perform the unwanted behavior, but do nothing to alter the animal's motivation to perform the behavior. However, as long as the result of an untreated behavior problem (e.g. infection, infliction of severe pain) is worse than the symptomatic treatment, it seems meaningless not to allow such treatment, provided that sufficient effort is made to find and institute more appropriate methods. We have to be more aware that, just like it is not acceptable simply to restrain an animal that is motivated to perform some unwanted behavior, so it is not acceptable either to "restrain" the producers in their activity, before acceptable alternatives are available.
Prevention: Lasting solutions
Whereas treatment of a behavior problem in many cases means reducing the animals possibility to perform the behavior and, thus, is symptomatic, a far better and ethically more accepted way to proceed is to reduce an animal's motivation to perform the unwanted behavior. Needless to say that this realization is not new in applied ethology. A growing body of evidence indicates, however, that many behavior problems can be truly prevented, if a certain exposure of the animal occurs at a specific period of its life. In the behavior clinic it has been known for years that in order for cats and dogs to become good pets, they must be socialized towards conspecifics and humans during a so-called sensitive period while they are puppies or kittens (e.g. Scott & Fuller, 1965. But more and more studies on farm animals indicate that a similar phenomenon is at work in farm animals.
In a recent study on laying hens it was shown that hens that had been exposed to perches during the first four weeks of life used elevated nest boxes significantly more for egg laying than hens that had not been exposed to perches early in life. The result of the more frequent use of nest boxes was that fewer eggs were laid on the floor and that cloacal cannibalism occurred less frequently (Gunnarsson et al., 1999). Similarly, in this host institute it was shown many years ago that exposure of calves to slatted floor during a time when they had less of a problem lying down on the hard surface resulted in faster adaptation to this floor type later when the animals were adult (Pougin, 1982).
Preventive medicine means, among other things, inoculating an animal to alter its immune system with the specific aim to prevent the outbreak of a specific disease later on. In principle, this approach is no different from exposing an animal to certain stimuli (e.g. perches) to alter the central nervous system (e.g. the right synaptic connections in the brain to develop spatial orientation) with the specific aim to prevent floor laying and cloacal cannibalism later on. Whether we shall call this preventive ethology, as I am suggesting elsewhere (Ladewig, 2000), or something different is less important. What is important is that in our attempt to search for the true cause of a behavior problem, we need to be much more goal oriented.
Conclusion
In present day welfare research more and more studies are done which use multivariate techniques to identify important factors and correlations between them. Similarly, many new behavior tests are being developed that give much more specific information about the various states of the animals. There is no doubt that these attempts are important and necessary additions to behavior research. But if we think a little more clinically and maybe call this way of thinking clinical ethology, and if we think a little more preventively and maybe call this way of thinking preventive ethology, chances are that we will be more efficient in reducing the frequency of behavior problems, that producers will be a little bit more interested in and positive towards our results, and that the welfare of the domestic animals may end up being a little bit better.
References
Alban L, Ersb�ll AK, Bennedsgaard TW, Johnsen PF (2000) Validation of methods for welfare assessment at herd level: An example. Acta Agric. Scand., Sect. A, Animal Science, Suppl., In press.
Gunnarsson S, Keeling LJ, Svedberg J (1999) Effect of rearing factors on the prevalence of floor eggs, cloacal cannibalism and feather pecking in commercial flocks of loose housed laying hens. Brit. Poult. Sci., 40, 12-18.
Ladewig J (1994) Stress Kap. 15 in: Döcke, F. (ed.): Veterinärmedizinische Endokrinologie, 3. Ed., Verlag Gustav Fischer, Jena, pp. 379-398.
Ladewig, J (2000) Long term confinement as a stressor. In The Biology of Animal Stress (eds. G.P. Moberg and J.A. Mench) CAB International.
Ladewig, J (2000) Preventive ethology: A new way to solve old problems. Submitted.
Lebelt D, Benda C, Schönreiter S, Zanella A, Unshelm J (1999) Nociception, heart rate, ß-endorphin, and cortisol levels in cribbing and weaving horses. In: Current Research in Applied Ethology 1998.
KTBL-Schrift, 362:90-98.
Moinard C, Mendl M, Nicol CJ, Green LE (2000) Investigations into risk factors for tail biting in pigs on commercial farms in England, UK. Proc. 9th Symposium of the International Society for Veterinary Epidemiology and Economics, Colorado.
Pougin M (1982) Zur Beurteilung der Anpassung von Jungrindern an die Spaltenbodenhaltung mit Hilfe von Verhaltensmerkmalen sowie Veränderungen an der Klaue und Blutserumenzymen Ph.D. Thesis, Rheinischen Friedrich-Wilhelms-Universität zu Bonn.
Scott JP, Fuller JL (1965) Genetics and the Social Behaviour of the Dog. Chicago, University of Chicago Press.
Table 1. Example of behaviour problems in domestic animals
Table 2. Example of the behaviour problem agression in pregnant sows
Table 3. Example of the behaviour problem crushing of piglets in lactating sows
Table 4. Example of the behaviour problem tail biting in fattening pigs
Table 5. Example of the behaviour problem cloacal canibalism and floor laying in laying hens
Table 6. Example of the behaviour problem crib biting in horses
|
