Animal Boredom - A Model of Chronic Suffering in Captive Animals and Its Consequences For Environmental Enrichment

FRANÇOISE WEMELSFELDER

The author describes how the term "boredom" is frequently used to interpret the abnormal behaviour of animals who are permanently housed in small, barren cages. Many scientists, however, regard this term as an unfound projection of human values upon animal behaviour. The author proposes some starting points for a scientific understanding of animal boredom. She defines boredom as the impaired ability to actively focus attention upon, and interact with, the environment. Suggestions for the experimental investigation of boredom based upon this definition are provided. The practical consequences of the proposed model for the alleviation of animal boredom through environmental enrichment are discussed. Housing of laboratory rabbits is used as an example to illustrate how simple enrichment measures can be highly effective if adequately applied. She argues if such measures are combined to enhance the active, flexible character of species-specific behaviour, both the well-being of captive animals and the quality of scientific research benefits significantly.

INTRODUCTION

Millions of laboratory animals are presently housed in small, extremely barren cages, in which opportunities for species-specific interaction with the environment are largely absent. In such an environment, animals develop various forms of abnormal behaviour, which many people intuitively take as a sign the animals are intensely bored. The question is whether this interpretation can be supported by a scientific understanding of abnormal behaviour, and, if so, what does such an understanding tell us about possible ways to prevent boredom in captive animals? This paper aims to provide some starting points for answering these questions.

First, I will provide a short overview of the abnormal behaviour patterns shown by captive animals. I will then point out some problems which arise in the interpretation of these patterns, and propose a model of animal boredom. The second half of the paper is concerned with the practical applications of the proposed model of animal boredom. Using the housing conditions of laboratory rabbits as an example, I will argue that environmental enrichment may not only enhance the well-being of animals, but may also improve the quality of animal management and research.

A THEORY OF ANIMAL BOREDOM

Abnormal Behaviour In Captive Animals

Generally speaking, animals housed in a barren environment show an overall decrease in interaction with the environment. This comes to expression in a variety of symptoms (Wemelsfelder, 1990). The animals lie down and sleep more, and spend significantly more time sitting. On the other hand, they over- react to novel and/or unexpected events with fearful and aggressive responses. Furthermore, the animals may develop stereotyped patterns of behaviour. Such patterns consist of high repetitive and uniform sequences of behaviour which seem to be of no direct functional value to the animal. Examples are bar-biting in tethered sows (Figure 1), stereotyped pacing shown by zoo animals such as polar bears and wolves, and various locomotory stereotypes in laboratory primates. Sometimes such behaviour can be damaging to other animals; licking and nibbling tails and ears of offspring may for example induce cannibalism in rats and mice.

As time of confinement proceeds, such patterns tend to become increasingly directed towards the animal's own body or products thereof. Primates may spend long periods of time masturbating, rocking their own body, or eating and regurgitating their own feces. Rats may chase their own tail, tethered sows may show long bouts of chewing air, with no other apparent effect than producing large amounts of saliva. Such tendencies may eventually develop into various forms of compulsive self-mutilation. Laboratory monkeys gnaw at their own limbs or genitals, while parrots will pull out their feathers until completely naked. In summary, the overall decrease in interaction shown by captive animals comes to expression in decrease in behavioural variability and an increase in self directed behaviours (Dantzer, 1986).

The Interpretation of Abnormal Behaviour

Behavioural scientists mostly refer to the behaviour patterns described in the previous paragraph as "abnormal," because these patterns, by and large, do not occur in the wild. Moreover, they bear a strong resemblance to behaviour pathologies in human beings.  It is generally accepted that in human beings, abnormal behaviour may be a sign of depression, or other forms of subjective suffering. In animals, however, it is by no means regarded as self-evident that abnormal behaviour involves suffering.

I contend that current models of animal behaviour in fact have no place for concepts such as boredom and depression. In such models, animal behaviour is generally conceived as an adaptation to external environmental circumstances. That is, the organization of behaviour is primarily regulated by external stimuli; this occurs directly, or through inheritance of pre-programmed rules encoded in the nervous system. These rules may be very complex, in which case one may speak of cognition. The principle remains the same, however, namely that behaviour is caused by external stimuli. Thus, behaviour is conceived as an essentially passive phenomenon (McFarland, 1989). In such a framework, abnormal behaviour is seen as an adaptation to barren environments and their lack of environmental stimulation. Most scientists do accept that the adaptive process may initially induce frustration; yet the possibility of chronic suffering is either left undiscussed, or is explicitly denied (e.g. Wiepkema, 1987; Dantzer, 1986). This is hardly surprising. By regarding adaptation to the environment as the ultimate goal of behaviour, the possibility of chronic suffering is priori excluded. There are no criteria by which to qualify behaviour as abnormal in such a theory.

I submit, however, the passive, pre-programmed aspects of animal behaviour, though important, constitute only half the story. Animal behaviour is, in fact, predominantly active and flexible. Ample evidence exists that rather than being determined by external stimuli, animals actively determine the impact of such stimuli upon the behaviour. They direct their attention to the environment through a continuous stream of subtle orienting movements. Furthermore, they seek novel stimulation through exploration and play. This indicates behaviour is not solely guided by pre-programmed rules. Animals actively imply these rules in the here-and-now of new situations. As a consequence, behaviour acquires a principally innovative and flexible nature.

Most behavioural scientists accept this is true for higher vertebrates such as chimpanzees, however, they strongly object to regarding the behaviour of lower vertebrates and invertebrates as similarly innovative. Indeed, numerous examples of innovative behaviour have been described in chimpanzees. Jane Goodall, for example, describes how a male chimpanzee became the leader of a group not by fighting, but by using four-gallon paraffin cans to impress competing males (Kummer & Goodall, 1985). However, innovative patterns of behaviour which are no less "clever" than those seen in chimpanzees have been observed in other species as well. Many ingenious examples of tool use in birds are reported, both in observations in the wild and from laboratory studies (Beck, 1980). Various bird species use sticks for different purposes, notably food gathering. Furthermore, Beck (1980) reports of an owl who acquired the habit of wiping the bloody beaks of her young with a leaf after a meal. Indications of innovative learning are also found "lower down" on the evolution scale. Honeybees, for example, learn to avoid unpleasant sensations induced by the anatomy of the alfalfa flowers by inventing new ways to enter those flowers (Gould & Gould, 1982). Insects also use tools; ants use bits of leafs, mud, and sand to transport soft foods such as honey to and from the colony (Beck, 1980).

Generally, scientists take for granted that in "lower" animals, such behaviour must be genetically pre-programmed, and cannot be regarded as innovative in the true sense of the word. However, it is well established that social insects and birds acquire a substantial part of their behaviour repertoire through learning, either by imitating their parents, or through trial and error. On the other hand, primates such as chimpanzees do not invent novel behaviours out of the blue, they must try out and practice various skills through play and social interaction. There is therefore, no valid reason to assume the behaviour of some species is merely a clever mechanical imitation of the "real" intelligence of other species. All species appear to be capable, within their species-specific range of behavioural possibilities, of actively dealing with their environment to adapt that environment to their needs and preferences. They learn this at various levels of complexity, but in all cases, such learning appears to be an active and flexible process.

I propose that in long term captivity, animal behaviour gradually loses its active and flexible character. In the development of abnormal behaviour such as stereotypies, we see behaviour becomes increasingly rigid, and loses its flexibility. The self-directedness of the behaviour shows animals close themselves off from their environment, rather than interact with it. One may object and hold that such a process can still be regarded as an adaptation to a barren environment. The non-adaptive character of abnormal behaviour comes, however, best to expression in the response to novel stimuli or a novel environment.

Various studies indicate that animals who have developed stereotyped behaviour patterns no longer actively interact with novel stimuli (Wood-Gush et al., 1983). Most poignantly, this appears in animals who after years of close confinement are transferred to a more enriched environment. These animals often fail to pay any form of active attention to the new environment. Instead, they persist in the performance of previous established abnormal patterns (Meyer-Holzapfel, 1968). Generally, animals raised in more impoverished environments show less tendency to manipulate novel objects than those raised under enriched conditions; isolated rats, for example, do not move or climb objects, but investigates them in a less active, more generally explorative fashion (such as sniffing) (Einon & Morgan, 1967; Mason & Green, 1962; Renner & Rosenweig, 1986). Such observations suggest the attention abilities of animals in prolonged captivity may gradually become impaired, and it is meaningless to interpret abnormal behaviour as a form of adaptation. To investigate this hypothesis, novel objects should be presented to captive animals over a prolonged period of time and their response toward these objects should be recorded. Such research is presently under way at the Scottish Agricultural College in Edinburg , Scotland .

The initial stage of attention impairment, I submit, may be characterized as boredom. As the animal is deprived means for behavioural interaction, his or her attention becomes increasingly dispersed towards inappropriate stimuli, such as another animal's tail, or his or her own limbs. One could say in that stage the animal "does not know what to do." The animal suffers from a general lack of meaningful behavioural goals, and becomes increasingly listless and withdrawn. In the final stage of impairment, attention disintegrates. The animal may respond chaotically to his or her environment, or become apathetic. In both cases, he or she becomes virtually helpless to cope with and\or change the situation. This stage may be regarded as evidence of depression and\or anxiety.

Many scientists have difficulty applying such terms, derived from human psychiatry, directly to animals. One may argue animals do not possess conscious awareness, and are therefore incapable of experiencing emotions such as boredom or depression. Unfortunately, this is not the time and place to discuss the complex and difficult problem of animal consciousness. Elsewhere, I have considered the problem in its own right, and its consequences for models of animal welfare (Wemelsfelder, 1993a,b). I do want to submit at this point, however, that because of the active nature of their behaviour, animals cannot be regarded as automatic adapting devices. First and foremost, they should be seen as subjective beings, to whom a meaningful relationship with the environment matters fundamentally. The break down of that relationship may then be regarded as direct evidence of suffering. When rigorously defined (in terms of decreasing attention), concepts such as boredom do not necessarily imply an unscientific projection of human feelings and values unto non-human animals. Instead, they provide the basis for a scientifically valid and experimental approach to animal suffering.

PRACTICAL APPLICATIONS: 

ENVIRONMENTAL ENRICHMENT IN LABORATORY ANIMALS

How can we prevent the development of states such as boredom and depression in captive animals? The notion of animal boredom outlined in this paper suggests successful enrichment should enhance the active character of behaviour, resulting in an increasingly frequency of interactive behaviours such as exploration, manipulation, play, and social interaction. At the same time, abnormal patterns of behaviour should largely disappear. The question arises whether great effort and expense must be made to achieve these aims, or whether small and relatively simple beginnings can be made. I suggest the latter is the case; if they are applied in a way which is meaningful to the animal, simple and inexpensive materials can have far reaching effects on the well-being of animals. Moreover, application of such materials does not necessarily lead to an undesirable increase in the workload of animal caretakers and technicians. I will illustrate this with an example.

An Example: Environmental Enrichment in Laboratory Rabbits

Rabbits are a relatively neglected laboratory species. They are, at least in the Netherlands , the only species still largely individually housed in cages which do not provide sufficient room for the animal to take a single hop. As a result, rabbits develop spinal deformations, a wide range of stereotyped behaviour, cannibalism, breeding problems, and overly aggressive and anxious reactions to human caretakers. It therefore seems no luxury to pay more attention to the behavioural needs of laboratory rabbits, both from the rabbit's and the scientist's point of view.

Wild rabbits, like most other rodents, spend a large proportion of their time gnawing and nibbling at potential food sources. Furthermore, they are social animals and live in elaborate systems of underground burrows. The provision of substrate for appetitive and feeding behaviours, on one hand, and social companions, on the other hand, may fail to be effective if applied separately. If applied in combination, however, they may succeed in permanently activating the rabbit's behavioural repertoire.

Substrates such as straw, hay and/or wood chips can have a multifunctional effect upon the animal's behaviour, if provided in sufficient quantity. First, food can be spread out in the substrate several times a day. This allows the animal to search for its food instead of gobbling it down. The animal will also nibble continuously at the substrate itself (Figure 2). Furthermore, hay and straw form a soft bedding to lay down upon, and can be used for digging, hiding, building nests and as a protection against too much light (Figure 3). In short, a few handfuls of substrate does not just fill the cage physically, but also psychologically, increasing the complexity of the space available and enhancing the functional structure of the animal's behaviour. However, a serious problem to providing substrate such as hay to singly-housed animals may be that it soils rapidly and forms a hygienic hazard. This problem is largely due to the lack of sufficient space for the animal to create a defecation area. Housing several rabbits together in a larger area may prevent such undesirable effects. Let us therefore consider the possibility of housing rabbits in social groups.

The presence of conspecifics in a social group provides an endless variety of meaningful stimulation to the animal. However, a well known problem with housing rodents in groups is the frequent outbursts of aggressive encounters, in which the animals may severely damage each other. We must recognize the fact that a bunch of animals together in one cage is not necessarily a well functioning social group. If no other substrate for behaviour is provided, the animals have no other choice than to direct attention toward their cagemates. They may then become stressed by too much stimulation and respond aggressively towards each other. Hay and straw may therefore play an important role in the regulation of social behaviour, both by facilitating appetitive behaviour, and providing a place to hide.

This assumption has been affirmed practically in several laboratories in the Netherlands . Rabbits have been successfully housed in cages of several square meters big, in which a nesting box and ample straw and/or hay is provided (Figure 4). Because the animals create a defecation area in one corner, the cage stays clean and the straw unsoiled for weeks. Another successful way of achieving group housing is to join several smaller cages into a chain of attached compartments, into which the animals can withdraw from other animals (as show in Figure 5, this system is equally suitable for house rats and mice). Straw is provided in several, but not all compartments, since the animals defecate in compartments without straw, no hygienic problems arise. An interesting observation was made one day, when straw was accidentally not provided, and aggressive behaviours increased dramatically. As soon as straw was given, aggression declined and peace was restored. This chance observation clearly demonstrates the interdependence of enrichment measures. If applied separately, a certain form of enrichment may cause considerable problems. A meaningful combination of measures, however, may prevent such problems and generate a situation which is generally easy to manage.

This principle is further demonstrated by the fact that rabbits housed in enriched group-housing appear to be easier to catch and handle than individually housed rabbits. Laboratory staff said they expected the rabbits to become "wilder" in group-housing condition, and to have to chase after them to catch them. The opposite, however, turned out to be true. The animals became much tamer, and would often voluntarily approach a staff member after he or she had entered the animal room. This effect could be enhanced by rewarding the animal with some food or attention (Figure 6). The staff attributed this change to the fact that the rabbits seemed much more relaxed and happy in their new environment and lost their fear of human caretakers as a result.

Overall, the following effect of enrichment on the well-being and management of laboratory rabbits was reported by laboratory staff from the various institutions:

Clearly then, adequate enrichment not only benefits the animal, but also benefits the well-being of personnel and contributes to efficient management. The conclusion must be that adequate environmental enrichment consists of a combination of relatively simple measures, which enhance the interaction with the environment. Only then does a situation arise which is stable and easy to manage, and in which boredom and depression do not occur.

Consequences For Scientific Experimentation

Finally, I would like to make a few comments regarding the consequences of such an approach to environmental enrichment for scientific experimentation. Many scientists are reluctant to approve of efforts to enrich laboratory cages. They dread an increase in behavioural variability may threaten the consistency of experimental results. Obviously, this is a legitimate concern, but I doubt it is actually the case. We must realize as the ability to control the environment decreases for the human experimenter, it actually increases for the animal. The animal learns to cope with novel and unexpected changes in the environment, and is less likely to be stressed by the laboratory procedures imposed, such as handling, restraint, and injections. A decrease in abnormal behaviours furthermore indicates a decrease in physiological and psychological stress.

The animal thus becomes a more "normal" model for scientific investigation. This may well imply his or her response to experimental situations is more, rather than less, stable and consistent. In one Dutch scientific laboratory, environmental enrichment led to a considerable decrease in the number of animals needed for experimentation. This effect was attributed to enhanced reliability of experimental results. This clearly shows caring for the well-being of experimental animals and seeking to understand their needs and requirements can very well go hand in hand with enhancing the quality of scientific research.

REFERENCES:

Françoise Wemelsfelder