Comfortable Quarters for Laboratory Dogs

Introduction
The dog is most familiar to us as a pet or companion animal, and is, in fact, one of the oldest domesticated animals. For many people their pet dog is one of the family, sometimes talked to as a human, and usually provided for, to the best of the owner's ability. As a consequence, in some countries the dog when used as an experimental animal receives special protection. For example, in the UK, specific justification has to be provided before dogs can be assigned to a study involving pain, suffering or distress. Such dogs have to originate from designated supplying establishments which are inspected by the Home Office (HMSO 1989). In the US, the Animal Welfare Act requires that experimental dogs are provided with exercise (USDA 1991).

The day-to-day careperson should have experience of pet dogs outside the laboratory. This can provide valuable insights into the dogs' needs, variation in character, and a better understanding and rapport with the dogs in her or his care. A good careperson will empathize with dogs from a basis of a sound understanding of their biology.

Planning new dog accommodation
All animal accommodation is the result of a compromise between the needs of the investigator, the needs of the day-to-day careperson and the needs of the animal. However, a casual inspection of most existing dog housing would suggest that all too often little or no attention is paid to the dog's needs.

Enclosures are usually barren simple affairs designed primarily for easy husbandry and to maintain the animal's physical health. These are, of course, essential elements of good design, but the dog is the prime user of the enclosure and spends very much longer periods in that environment than the staff who services it. It is vital, therefore, that designers consider the dog's normal behavior and are aware of the extent to which the enclosure might restrict such behavior.

Unfortunately, there has been rather little research carried out to address the challenge of appropriate kennel design that satisfies the dog's needs.

Flexibility of pen design is extremely important. Groups of pens within a room should be arranged in such a way that it is possible to move an animal to another pen temporarily during wet cleaning, and so avoid exposing the subject to buckets of water, high pressure hoses or other aversive stimuli. It should be possible to open doors or pop holes between pens so that larger "super-pens" can be created as desired, either to allow the animals more space and complexity or to allow groups of dogs to run together (Figure 1).

It is also important to remember that dogs have different personalities which are a product of their genetic makeup and personal experiences. Hence, there will be an inevitable variation in the dogs' responses to kennelling conditions. Ideally, we should be aiming to design an environment in which the most nervous dog in the colony can live without being unduly stressed.

Space considerations
Dogs kept in laboratories are generally confined in much more restricted spaces than their domestic counterparts. It is worth remembering that in wild or feral conditions dogs may roam over areas of up to 2850 ha (7040 acres). It is impossible to emulate such freedom in the laboratory, and in any case, dogs ranging over these large areas would probably only use them in order to obtain sufficient food. Nevertheless, provision of adequate space is essential for laboratory dogs as it affects not only their behavior, but also determines whether the animals can be housed in social groups and whether there is sufficient room for enrichment devices.

Confinement in pens or cages intrinsically restricts the dog's ability to perform species- typical behaviors, and to adjust social contact with other dogs (Bebak and Beck 1993). Small or shallow-depthed pens may also not allow the dog to retreat from events that he or she considers alarming at the front of the pen. If the dog experiences difficulties in coping with the kennelling condition he or she may show various behavioral abnormalities such as locomotory stereotypies or excessive barking (Luescher et al. 1991). If dogs develop stereotypies there is good reason to reexamine the particular type of housing and attempt to improve it.

Stereotypies are not always easy to recognize as they will often cease when a person enters the room. It can be useful to employ closed-circuit television (CCTV) or video techniques to monitor the dogs from another room. Running a videotape in fast forward will make it easier to detect abnormally repetitive behavior patterns.

Socialization during development
Dogs are social animals. It is therefore not surprising that socially restricted rearing conditions lead to the development of behavioral abnormalities and crippled behavioral repertoires (Thompson et al. 1956; Fuller 1967). Similarly, if dogs are not provided with an adequate early experience of humans they will later be fearful of people and as a result can be difficult to handle (Freedman et al. 1961; Scott and Fuller 1965; Wolfle 1987). A reasonably complex physical environment during the puppies' development is also necessary to ensure that the animals do not become excessively nervous, institutionalized dogs (Fox and Stelzner 1966; Fox and Spencer 1969; Wright 1983). Puppies go through a stage between the 3rd and 12th week of life—the so-called primary socialization period—when it is particularly easy for them to develop relationships with other individuals (Scott and Fuller 1965; Wright 1980). At the same time, the puppy becomes attached to the area that it knows. While there is debate as to whether this is really a "critical period," it does seem to be a time of special importance in the puppy's development (Markwell and Thorne 1987).

Although the general time of this socialization period is agreed upon, there is surprisingly little known about how much human-contact time is needed to adequately socialize a dog with humans. Some studies suggest that socialization with humans can be achieved through relatively small amounts of time: 40 minutes or less per week spent with a litter (Scott and Fuller 1965), or 5 minutes per week spent with each puppy (Wolfle 1990). Hubrecht (1995) worked with beagle puppies who had been considered by a pharmaceutical firm to have already received adequate socialization. Even so, an extra two minutes spent in the pen each weekday combined with 30 seconds of petting each puppy (i.e., 2.5 minutes intensive contact with the human handler per puppy per week) resulted in behavioral changes in male puppies 6-11 months later. These changes could be interpreted as intensified seeking of human contact.

Group/single-housing
A good kennel design should be large enough and flexible enough to house dogs socially in harmonious groups and to allow them to be separated where necessary. Single-housing for prolonged periods is likely to be deleterious to the dog and is associated with an increased incidence of behavioral abnormalities (Hetts et al. 1992). Single-housing may sometimes be necessary for reasons of health or aggression, but the duration should always be kept to a minimum.

There are no clear data indicating what an optimal group size might be. Pair-housing seems to be a reasonable compromise, as dogs in pairs spend a similar proportion of their time interacting with each other as dogs kept in groups of 5-11 animals (Hubrecht 1993). In the group-housing situation aggression can be a serious problem, which may lead to the death of animals. Therefore, it is important to ensure that there is an adequate husbandry routine to monitor the dogs and forestall potential problems. A video and sound monitoring system can be very helpful.

Exercise
One obvious effect of confinement is to restrict locomotory behavior. Small enclosures not only discourage exercise because there is no possibility of travelling to another location, but they also restrict the type of locomotion that is possible. Increasing pen dimensions beyond the minimum standards acceptable in the US (USDA 1991) does not seem to make much difference to the dog in terms of physical fitness (Clark et al. 1991), aggression or play (Beback and Beck 1993). Nonetheless, a study of the behavior of mixed breed dogs housed in pens with spacious, complex outdoor runs of 744 m2 has shown that both the activity of the dogs and the range of species-typical locomotory behaviors was greater than that shown by dogs in small standard pens of less than 7 m2 (Hubrecht et al. 1992).

Dogs in too small enclosures simply do not have enough space in which to run in a dog-specific manner, but we do not know as yet how important this might be to them. However, it is known that small enclosures are associated with a higher prevalence of circling and other stereotypies than relatively large enclosures (Hubrecht et al. 1992). This indicates that too small living areas affect the dogs' behavioral health and hence their general well-being. It is reasonable to recommend that as an absolute minimum the enclosure must provide adequate space for a dog to locomote for more than a few paces in a straight line. This can be provided through a dimension of the pen that is at least four times the dog's length.

Minimum space allowances
Minimum dimensions of pen design are recommended as stipulated by the Home Office (HMSO 1989). The allocated floor areas provide a reasonable compromise between economic costs and allowing sufficient room for locomotion, socialization and enrichment purposes.

It is arguable whether pen size should be based on body weight, however, body size is an equally poor predictor of a dog's space requirements. For example, an active puppy of, say, eight weeks might well require more space than an old, relatively inactive dog.

The space recommendations listed in Table 1 are based on professional experience. They have the benefit of encouraging pair-housing because the minimum sizes for single cages provide sufficient room for two adult dogs. For example, the minimum floor area for adult beagles of the weight category 10-25 kg is 4.5 m2, both for a singly housed dog and for two pair-housed dogs. The addition of a third dog, however, necessitates that the floor area be increased by 2.25 m2, i.e., to 6.75 m2.

The height of the enclosure should at least allow the dog(s) to stand on hind legs without touching the roof. However, with pens of these sizes, the vertical dimension should also allow the careperson to stand in the pen without having to stoop. In this respect a height of 150 cm may be too small (Table 1).

Structures within the dog enclosure
Larger enclosures allow greater possibilities for the provision of structures and enrichment devices. They also allow the provision of separate sleeping and exercise areas so that the more complex and interesting and gives the animal(s) the option of exercising some choice. Most importantly, a large enclosure permits social housing.

Chews/toys are sometimes objected to because they can trigger aggression, cause hygiene problems or simply because the dogs lose interest in them. Appropriate presentation of such enrichment items can overcome these objections. For example, the toys can be suspended from the ceiling by sprung chains (Figure 1) which prevents soiling and makes it impossible for one dog to monopolize them (Hubrecht 1993). Dogs are very motivated by food. Toys or chews that have an appetizing aroma or taste are therefore likely to receive a lot of attention. Such items can reduce the time during which the animals are inactive and decrease destructive behavior aimed at cage fixtures and furnishings. In a study with beagles, toys/chews (rawhide, plastic pipe, tugtoy) were used by puppies during 64% of the time, and by subadult dogs (7-13 months old) during 24% of the time (Hubrecht 1993; Hubrecht 1995). No habituation effect was noticed.

Any enrichment device has to be practical, provide a measurable benefit for the animals and must not interfere with the aims of the research protocol.

Dogs have a natural interest in their surroundings. Accordingly, kennels should not overly restrict the animals' ability to obtain information about the surroundings. High walls or partitions result in the dogs being unable to see to the end of their rooms. This can prompt them to spend a relatively high proportion of their time on hind legs or in apparently repetitive, possibly stereotypical jumping behavior. Obvious ways around this problem include reducing the height of solid partitions between pens for at least a portion of their length, or providing platforms (Figure 1) within pens. Hubrecht (1993) has shown that platforms are extensively (55% of observation time) used by laboratory dogs to play and rest on, and that they do not pose any risk, even for dogs with gastric fistulas.

Platforms increase the complexity of pens, thus allowing the dogs more choices within their environment. If properly installed so that they do not block the existing floor area, platforms also provide additional horizontal space. In effect, they make use of the otherwise inaccessible third dimension of the pen (Figure 1).

Dog housing should always be designed so that the occupants can retreat to an area that provides them with a sense of security. This need not cause a problem of visibility for the day-to-day careperson, as it can simply be an area with a few barriers shielding an animal from view on two or three sides. It is particularly important to offer such structures when dogs are housed in large social groups to give the animals better control over their social interactions.

Social contact of adult dogs with animal care staff
It is reasonable to assume that dogs who have been socialized to humans while puppies will wish to socialize with them as adults. It is generally agreed by professionals that human socialization with adult dogs improves handleability and provides a form of enrichment (Fox 1986). However, in kennels housing large numbers of dogs, the pressures on staff are often such that contact time with the dogs becomes very limited (Hubrecht et al. 1992). Managers should be aware that this can be a serious problem and should establish and implement a routine socialization program.

Minimization of stress during interactions between people and dogs
The easiest way to minimize stress in the dog, is to ensure that he or she reacts well to handling. All dogs should, therefore experience adequate socialization with other dogs during the "primary socialization period" at the age of 3-12 weeks, and receive regular, gentle and sympathetic handling thereafter. The handling of an animal should always be carried out by the same member of the staff or, if this is not possible at times, by another who applies the same handling techniques.

Training dogs so that they become used to experimental and clinical procedures is very important in order to avoid stress responses (Figures 2 -4). If, for example, a procedure involves temporary restraint—which is known to be potentially stressful—the dog should first be gently introduced to this situation by the familiar handler. The handler should always remain with the animal during procedures. If this is not possible, the personnel carrying out the procedure should also pick up the dog from the pen and return the dog afterwards back to pen in the same skilled manner as the regular handler would normally do.

Staff training is crucial. Staff should be proficient in basic handling techniques such as those shown in Figures 5-10 (MacArthur 1987). Care personnel should know how to approach the dog steadily and quietly make confident contact, and should frequently reward the dog by petting and talking quietly to him/her. The staff's demeanor while in the animal rooms should always be calm, confident and quiet. The aim is to establish a bond of trust with the dogs so that handling is a positive rather than a negative experience for all parties involved (cf., Figures 2-10).

Noise in kennels
The noise in kennels resulting from barking can be a nuisance and is potentially damaging to human hearing (van der Heiden 1992). Dogs often bark at each other through adjacent pens or at people passing by. Barking is also associated with feeding times and is probably the result of the generally high levels of excitement. Very often barking spreads to other dogs, and in some animals the act of barking may itself function as a self stimulus to further barking (Scott and Fuller 1965).

Dogs can detect sound ranging in frequency from 0.04 kHz (cycles per second) up to around 50 kHZ, which is well beyond the upper frequency limit of human hearing. They can hear sounds that are up to four times quieter than the human ear can detect.

Unfortunately, the possibility that noise might have similar effects on dogs as on humans has, until recently, not been considered. In dog kennelling, sound levels within the human hearing range can regularly reach values of 85 to 122 decibels (dB; Peterson 1980). The noise problem for humans can be addressed by protective wear such as hearing defenders, but these are of no value to the dog.

In a recent study, Sales et al. (in preparation) have surveyed noise levels, throughout the hearing range of dogs, in different kennel environments. The data showed that the dogs were regularly exposed to high sound levels throughout the working day and often well into the evening. The peak sound pressure level over the measuring period often exceeded 100 dB, and the sound level of a continuous steady tone (which would result in the same amount of sound energy as a fluctuating sound measured over the same period) often exceeded 95 dB.

The study also confirmed that the noise of barks from individual dogs commonly surpasses 100 dB. From this and the observations made during the investigation it would seem that most of the high level noise recorded in kennels is produced by the dogs themselves; although other events such as cleaning, high pressure hoses, doors banging and pagers may also contribute substantially to the acoustic environment (Sales et al. 1988).

The sound in dog kennelling is certainly at a level that is known to cause damage and stress not only in humans but also in other animals who have less acute auditory sensitivity than dogs (Gamble 1982; Milligan et al. 1993). If these levels are sustained throughout the day it would seem probable that dogs are also at risk from damage to their auditory system. Even if dogs were shown not to find high noise levels aversive, any physiological damage would certainly reduce the dogs' welfare and hence would be unacceptable.

Sound control should be a priority when designing dog kennels. Noise can be limited by the use of sound-absorbent materials. These must, however, be capable of being cleaned and should either be out of the dogs' reach or should be resistant to destruction. Efforts should also be made to reduce transmission of sound by the use of acoustic doors and cavity walls. Corridors can channel the sound from one area to another, but as sound travels less easily around corners, the use of straight corridors should be avoided. Major noise-producing equipment should be sited as far away from the animals as possible.

Conclusions
A good kennel design will:

• keep the animal(s) in good physical and mental health,
• permit easy handling of the dog(s) by personnel,
• be large enough to allow group-housing of compatible dogs,
• be flexible enough to allow pens to be joined together to create larger runs,
• permit choice of location and provide interest within the enclosure,
• provide some refuge from kennel mates through the use of visual barriers,
• allow the dog good visibility of the room or of the area outside the pen,
• reduce to a minimum sound egress and ingress, as well as reverberation through the room.

Dog supply and surplus animals
In the UK it is a legal requirement that dogs provided for scientific use are bred in and obtained from a designated breeding establishment (Editor's note: these correspond to class A dealers in the US ). There is much to be said for this provision as:

1. The dogs should be of a good and standard quality and free of disease.
2. They will not have experienced anything other than an institutional life and therefore will not suffer due to separation from a family life or because of loss of freedom;
3. The standards of a designated breeding establishment can be maintained through a process of licensing and inspection.
4. The public can be assured that their pets will not end up in research establishments. Former pets are not likely to do well in a laboratory environment. They will often be accustomed to frequent and positive human contact and to being taken out of the home for walks to defecate and urinate, and meet other dogs. It is also likely that they will have undergone a series of disruptive and frightening events before reaching the laboratory.

There are also problems in allowing dogs to be retired to homes after use in laboratories. If the research procedures cause lasting suffering or a risk of suffering, it may be more humane to destroy the animal. As with any rehoming program, it is worth remembering the importance of matching the dog closely with the new owner. Dogs who have spent a substantial portion of their life in institutions are likely to show a number of behavioral disorders when being rehomed.

These may include difficulties with house-training, separation anxiety problems, and aggression triggered by fear-inducing unfamiliar circumstances. In order to rehome research dogs safely and humanely, the institution must develop and maintain a comprehensive socialization and training program.

References
Bebak J, Beck AM 1993. The effect of cage size on play and aggression between dogs in purpose-bred beagles. Laboratory Animal Science 43: 457-459.

Clark JD, Calp JP, Armstrong RB 1991. Influence of type of enclosure on exercise fitness of dogs. American Journal of Veterinary Research 52: 1024-1028.

Fox MW 1986. Laboratory Animal Husbandry Ethology, Welfare and Experimental Variable. State University of New York Press: Albany.

Fox MW, Spencer JW 1969. Exploratory behaviour in the dog: Experiential or age dependent. Developmental Psychobiology 2: 68-74.

Fox MW, Stelzner D 1966. Behaviour effects of differential early experience in the dog. Animal Behaviour 14: 273-281.

Freedman DG, King JA, Elliot O 1961. Critical period in the social development of dogs. Science 133: 1016-1017.

Fuller JL 1967. Experiential deprivation and later behaviour. Science 158: 1645. Gamble MR 1982. Sound and its significance for laboratory animals. Biological Reviews 57: 395-421.

Hetts S, Clark JD, Calpin JP, Arnold CE, Mateo JM 1992. Influence of housing conditions on beagle behaviour. Applied Animal Behaviour Science 34: 137-155.

HMSO 1989. Code of Practice for the Housing and Care of Animals Used in Scientific Procedures. Her Majesty's Stationery Office: London.

Hubrecht RC 1993. A comparison of social and environmental enrichment methods for laboratory housed dogs. Applied Animal Behaviour Science 37: 345-361.

Hubrecht RC 1995. Enrichment in puppyhood and its effect on later behavior of dogs. aboratory Animal Science 45: 70-75.

Hubrecht RC, Serpell JA, Poole TB 1992. Correlates of pen size and housing conditions on the behaviour of kennelled dogs. Applied Animal Behaviour Science 34: 365-383.

Luescher UA, McKeown DB, Halip J 1991. Stereotypic or obsessive-compulsive disorders in dogs and cats. Veterinary Clinics of North America: Small Animal Practice 21: 401-413

MacArthur JA 1987. The dog. In The UFAW Handbook on the Care and Management of Laboratory Animals 6th Edition, Poole T (ed), pp 456-475. Harlow Longman Group: London.

Markwell PJ, Thorne CJ 1987. Early behavioural development of dogs. Journal of Small Animal Practice 28: 984-991.

Milligan SR, Sales GD, Khirnykh K 1993. Sound levels in rooms housing laboratory animals: an uncontrolled daily variable. Physiology and Behaviour 53: 1067-1076.

Petersen EA 1980. Noise and laboratory animals. Laboratory Animal Care 13: 340-350. Sales GD, Wilson KJ, Spencer KEV, Milligan SR 1988. Environmental ultrasound in laboratories and animal houses. A possible cause for concern in the welfare and use of laboratory animals. Laboratory Animals 22: 369-375.

Sales G, Hubrecht R, Peyvndi A, Milligan S, Shield B. Noise in dog kennelling: Is barking a welfare problem for dogs? (in preparation)

Scott JP, Fuller JL 1965. Genetics and the Social Behavior of the Dog. University of Chicago Press: Chicago.

Thompson WR, Melzack R, Scott TH 1956. "Whirling behaviour" in dogs as related to early exposure. Science 123: 393.

USDA 1991. Title 9, CFR (Code of Federal Register), Part 3. Animal Welfare; Standards; Final Rule. Federal Register 56: 6426-6505.

van der Heiden CV 1992. The problem of noise within kennels: what are its implications and how can it be reduced? The Veterinary Nursing Journal 7(1): 13-16.

Wolfle TL 1987. Control of stress using non-drug approaches. Journal of the American Veterinary Medical Association 191: 1219-1221.

Wolfle TL 1990. Policy, program and people: The three P's to well-being. In Canine Research Environment, Mench JA, Krulisch L (eds), pp 41-47. Scientists Center for Animal Welfare: Bethesda.

Wright JC 1980. The development of social structure during the primary socialization period in German Shepherds. Developmental Psychobiology 13: 17-24.

Wright JC 1983. The effects of differential rearing on exploratory behaviour in puppies. Applied Animal Ethology 10: 27-34.