To address the species-typical behavior of horses in research institutions, both their social organization and their daily activity patterns must be taken into consideration. Free ranging (Berger, 1986) and wild (Boyd, 1998) horses live in bands composed of several mares, one or, occasionally, more stallions, foals and yearlings. The adult composition of these bands is quite stable. Members within established groups have preferred associates with whom they spend most of their time. This social organization should be respected as much as possible in equine housing, in order to safeguard both the welfare of the species and the validity of the data obtained. Horses communicate to one another and to us with their ears, tail and general posture, as well as vocally.

Feral (Salter and Hudson, 1979) and pastured domestic (Crowell-Davis et al., l985) horses spend 40- 60% of their time grazing both day and night (Keiper and Kennan,1980). Grazing combines feeding with exercise because the pattern is for the horse to take a few steps, prehend, bite off and chew several tufts of grass and then take a few more steps before repeating the process. Horses rarely trot or canter unless startled, so exercise need not be strenuous to preserve their welfare. They stand up most of the time but must lie down for deep sleep (Ruckebusch, 1972); because this takes place in the early morning hours of darkness, one may not appreciate that the animals lie down and need a suitable surface to do so. Horses drink in association with feeding. Since they eat frequently they also drink frequently (Laut et al., 1985).
Horses may allogroom one another. The occurrence of this behavior varies among pairs of horses and with the season (Clutton-Brock et al., 1976). Other care of the coat includes biting at the coat, rolling, rubbing on inanimate objects or rubbing the face on the forelegs. Provision must be made for these activities to maintain the horse's welfare.

Figure 1. Grazing [horse walks a few steps, prehends food with his head down, walks, etc.] is the major natural activity of horses.

• Pasture is typically the best environment for horses because it allows them to graze (Figure 1) and live in a group (Figure 2). There must be a source of shelter from both the sun and from rain or snow. The fences must be adequate to restrain the animals, but not dangerous. Barbed wire should not be used. Post and rail or board fences are traditional, but the latter require upkeep and may be eaten by the horse. Woven wire topped by wood is suitable. Gates must be properly designed to avoid injury (Figure 3).

Figure 2. Pasture with at least one other horse is the most species-appropriate environment.

Unfortunately, there are not many experiments that can be performed with pastured horses, but pasture should be used for a holding area between experiments. Paddocks are suitable providing they are covered with quality grass. They are also valuable for exercise when the horses are released in a group. When allowed to control their own environment, horses will spend the majority of their time outside in a paddock rather than in a stall (Houpt and Houpt, 1988).

Figure 3. Horses interacting across a gate. The nostril-to-nostril greeting is often followed by squealing and striking with the foreleg. The foreleg could become entrapped between the rails. This gate is safe because the rails are rounded.

• In contrast to the natural social group, domestic horses, in general, and research horses, in particular, are kept either singly or in groups that frequently change composition. Social isolation is a disturbing experience for horses, and isolated subjects show behavioral and physiological stress reactions (Mal et al., 1991a,b) that may influence research data collected from them. When they have the option, horses will spend about half of their time in contact with other horses (Houpt and Houpt, 1988). Some experiments, particularly those dealing with nutrition, ectoparasites or infectious diseases, may require temporary isolation, an unnatural diet or no bedding. In such cases, environmental enrichment must be provided. The enrichment should be something, such as foraging, that simulates normal equine behavior. A food dispensing apparatus (Figures 4a & b) was first described by Malpass and Weigler (1994) and has been shown to be effective in maintaining normal behavior in stabled horses (Winskill et al., 1996).

Figure 4a (above). A food dispensing cylinder [Pasture Pal®] that can be used as an enrichment device. The horse must push it to obtain pellets or grain from holes in the device (Photo by Txema Peralta). Figure 4b (right). Horse foraging from the Pasture Pal® (Photo by Txema Peralta).

When horses are kept individually, they may be housed in box stalls (Figures 5 & 6) that allow them to turn around but usually prevent any social contact, or they may be housed in tie stalls (Figure 7) that are probably more species-appropriate because they allow social contact (McDonnell et al., 1999).

Figure 5. A box stall. This stall has the advantage that the horse cannot easily injure himself. The disadvantage is that the animal has social contact only distantly through two sets of bars with the horse across the aisle. Arrow indicates automatic waterer supply line.	Figure 6. This type of box stall allows visual and tactile contact between horses. Note that the dividing wall between the stalls has a screened window to allow more visual contact.

Figure 7. A Straight or tie stall with straw bedding. Note that there is a screen between the horses to allow social contact, but not biting. The partitions should be solid, although ideally still removable in an emergency, as the horses will lean on them; wood is generally more suitable than metal even if it is harder to keep clean.

• Because horses who are confined for a long time develop edema, and because after confinement in a stall without the opportunity to exercise, horses show a compensatory increase in activity when released from their stalls (Houpt et al., 2001), horses should be released daily, preferably in a group on a pasture or paddock. When given a choice, horses spend twice as much time in a paddock when other horses are around as when they are released on it alone (Lee et al., 2001).
     
• Horses prefer bedding to a bare surface (Hunter and Houpt, 1989; Figure 7). The preference for bedding is most marked when the horses are lying down, which is not surprising because that is the time when their soft tissues are in contact with the substrate. Horses prefer straw to wood shavings in some studies (Mills et al., 2000), but not others (Hunter and Houpt, 1989). Wood shavings are more absorbent and more insulating than straw (Airaksinen et al., 2001) but don't provide a source of oral enrichment.
       
• Horses should have water available ad libitum. Automatic waterers are often provided, but, although they are labor saving, horses do not drink as much from automatic waterers as from buckets, probably because a deep receptacle is more natural for a suction drinker such as the horse (Nyman and Dahlborn, 2001). Automatic water lines may be broken by horses, and the waterers may malfunction, either depriving the horse of water or flooding the stall (see Figure 5, arrow). If automatic waterers are used, care must be taken that a newly introduced horse can operate them. It is always a good idea also to provide a bucket until the new animal has been observed to drink from the automatic waterer.
    
• There are several oral stereotypies in captive horses; wood chewing and cribbing are the most common. Diet can modify them. Wood chewing decreases when hay rather than pellets is fed (Willard et al., 1977). Cribbing occurs most frequently subsequent to eating grain - the sweeter the feed the more the horse cribs. Increasing the amount of hay or decreasing or eliminating sweet feed will reduce cribbing (Kusunose, 1992; Gillham et al., 1994). Providing ad libitum access to hay is the best way to avoid the development of cribbing and wood chewing in stalled horses (Figure 8).

Figure 8. A mare and foal fed ad lib hay from a rack. Note that the ground around the rack is devoid of grass. Racks prevent wastage of hay but are not a natural way for a horse to eat (cf., Figure 1). Horses may inhale particulate matter including pathogens and allergens when the hay is above their chest.

• Visual contact with conspecifics can have positive or negative effects on the horse's comfort. Weaving - a locomotor stereotypic behavior - decreases in proportion to the numbers of sides of a stall that have windows (Cooper et al., 2000). Presumably, it is the sight of other horses that increases the horse's comfort because mirrors are also effective in decreasing weaving (McAfee et al., 2002). Negative effects are perception of threats from another horse and anxiety, when a horse who is usually present is removed. As a general rule, horses should have a view of other compatible horses.
      
• Horses will work for light in a dark stable (Houpt and Houpt, 1988). Their living quarters should, therefore, be well illuminated. Contrast in light between environments can be a problem in that horses often refuse to enter a barn that is markedly darker than the outdoor environment.
      
• Grooming lowers the heart rate of a horse, suggesting that it has a relaxing, calming effect (Feh and de Mazìeres, 1993). A stalled horse, however, is denied the opportunity to allogroom (Figure 9) and may, in addition, even be denied the opportunity to autogroom. Some horses are reluctant to roll in their stalls, and stalls often provide no or ineffective rubbing surfaces. Grooming by humans is probably a suitable substitute.

Figure 9. Mutually grooming [allogrooming].

Horses should be groomed several times a week. This is particularly important if they are dirty from rolling in the mud and when they are shedding in the spring. Grooming is also an opportunity to examine the horse's body; abnormalities may be detected that otherwise would not be noticed until they have become more serious. Hooves should be cleaned daily.

• Perhaps the most important contribution to a research horse's comfort is the caretaker. The caretaker should be willing and able to handle horses gently, but effectively. He or she should be alert to the horse's environment [e.g., sharp edges where a horse can be injured or unlatched stalls], to changes in water and food intake and to social relations among the horses.
  
  Since we usually do not keep horses in stable groups or pairs in research facilities, aggression frequently occurs even between horses in separate stalls or paddocks. If the aggressor is highly motivated or the pair evenly matched, even very sturdy fences can be kicked down, resulting in damage to the animals and/or the facility. For these reasons, caretakers have to be aware of the relationship between the horses in their charge and make sure that only compatible partners are placed next to each other. The caretaker can determine whether aggression is imminent if two horses, after touching nostrils (Figure 3), squeal and strike with a foreleg or if one or both partners approach with flattened ears, or if their tails are lashing. It is important to watch the horses at feeding time because that is when aggression is most apt to occur. Another critical situation is when horses are being taken from a paddock or corral. Often, the animals will fight to be the first to leave. The best rule here is to release the most dominant horse [the one who can displace all the other horses with impunity] into the paddock last, and return him back into the stable first so that he won't prevent others from entering or leaving.

The essentials of adequate housing for horses in research institutions are: Access to hay and water ad libitum, a soft substrate for lying, visual contact with other horses and opportunity to exercise at will with other horses. Handling should involve firm and gentle restraint, but not fear or pain. Grooming is necessary for comfort as well as cleanliness.

Airaksinen S, Heinonen-Tanski H, Heiskanen M-L 2001. Quality of different bedding materials and their influence on the compostability of horse manure. Journal of Equine Veterinary Science 2, 125-130

Berger J 1986. Wild Horses of the Great Basin The University of Chicago Press, Chicago, IL

Boyd L 1998. The 24-h time budget of a Takhi harem stallion (Equus ferus przewalskii) pre- and post- reintroduction. Applied Animal Behaviour Science 60, 291-299

Clutton-Brock TH, Greenwood PJ, Powell RP 1976. Ranks and relationships in Highland ponies and Highland cows. Zeitschrift für Tierpsychologie 41, 202-216

Cooper JJ, McDonald L, Mills DS 2000. The effect of increasing visual horizons on stereotypic weaving: Implications for the social housing of stabled horses. Applied Animal Behaviour Science 69, 67-83

Crowell-Davis SL, Houpt KA, Carnevale J 1985. Feeding and drinking behavior of mares and foals with free access to pasture and water. Journal of Animal Science 60, 883-889

Feh C, de Mazières J 1993. Grooming at a preferred site reduces heart rate in horses. Animal Behaviour 46, 1191-1194

Gillham SB, Dodman NH, Shuster L, Kream R, Rand W 1994. The effect of diet on cribbing behaviour and plasma endorphin in horses. Applied Animal Behaviour Science 41, 147-153

Houpt KA, Houpt TR 1988. Social and illumination preferences of mares. Journal of Animal Science 66, 2159-2164

Houpt KA, Houpt TR, Johnson JL, Erb HN, Yeon SC 2001. The effect of exercise deprivation on the behaviour and physiology of straight stall confined pregnant mares. Animal Welfare 10, 257-267

Hunter L, Houpt KA 1989. Bedding material preferences of ponies. Journal of Animal Science 67, 1986-1991

Keiper RR, Keenan MA 1980. Nocturnal activity patterns of feral horses. Journal of Mammalogy 61, 116-118

Kusunose R 1992. Diurnal pattern of cribbing in stabled horses. Japanese Journal of Equine Science 3,173-176

Laut JE, Houpt KA, Hintz HF, Houpt TR 1985. The effects of caloric dilution on meal patterns and food intake of ponies. Physiology and Behavior 35, 549-554

Lee J, Floyd T, Houpt K 2001. Operant and two-choice preference applied to equine welfare. Proceedings of the 35th International Society of Applied Ethology International Congress, 110

Mal ME, Friend TH, Lay DC, Vogelsang SG, Jenkins OC 1991. Behavioral responses of mares to short-term confinement and social isolation. Applied Animal Behaviour Science 31, 13-24

Mal ME, Friend TH, Lay DC, Vogelsang SG, Jenkins OC 1991. Physiological responses of mares to short term confinement and social isolation. Equine Veterinary Science 11, 96-102

Malpass JP, Weigler BJ 1994. A simple and effective environmental enrichment device for ponies in long-term indoor confinement. Contemporary Topics in Laboratory Animal Science 33, 74-76

McAfee LM, Mills DS, Cooper JJ 2002. The use of mirrors for the control of stereotypic weaving behaviour in the stabled horse. Applied Animal Behaviour Science (in press)

McDonnell SM, Freeman DA, Cymbaluk NF, Schott HC, Hinchcliff KW, Kyle B 1999. Behavior of stabled horses provided continuous or intermittent access to drinking water. American Journal of Veterinary Research 60, 1451-1456

Mills DS, Eckley S, Cooper JJ 2000. Thoroughbred bedding preferences, associated behaviour differences and their implications for equine welfare. Journal of Animal Science 70, 95-106

Nyman S, Dahlborn K 2001. Effect of water supply method and flow rate on drinking behavior and fluid balance in horses. Physiology and Behavior 73, 1-8

Ruckebusch Y 1972. The relevance of drowsiness in the circadian cycle of farm animals. Animal Behaviour 20, 637-643

Salter RE, Hudson RJ 1979. Feeding ecology of feral horses in western Alberta. Journal of Range Management 32, 221-225

Willard JG, Willard JC, Wolfram SA, Baker JP 1977. Effect of diet on cecal pH and feeding behavior of horses. Journal of Animal Science 45, 87-93

Winskill LC, Waran NK, Young RJ 1996. The effect of a foraging device (a modified `Edinburgh Foodball') on the behaviour of the stabled horse. Applied Animal Behaviour Science 48, 25-35

Dr. K.A. Houpt graduated from the University of Pennsylvania with a VMD and PhD. She is a diplomate of The American College of Veterinary Behaviorists and director of the Animal Behavior Clinic at Cornell University College of Veterinary Medicine. Her main research focus is equine welfare.

Dr. T.S. Ogilvie-Graham’s credentials include DVM&S in Applied Equine Ethology from Edinburgh University. He is Lieutenant Colonel RAVC, in charge of the Royal Household Cavalry.

Back to Table of Contents