Literature Review

Temple Grandin
Dept. of Animal Science
Colorado State University
Fort Collins, Colorado 80523 USA

Dissertation, University of Illinois, 1989


A. Effects of Environmental Restriction and Complexity

Behavioral effects of environmental restriction

This will be a review of the scientific literature on the effect of either isolation or sensory restriction on animals after weaning. In all cases, prior to weaning, the animals were reared by their natural mothers. Therefore, any effect of sensory restriction was not confounded with the effects of maternal deprivation, which profoundly alters later behavior (Harlow and Zimmerman, 1959; Mason, 1960).

Isolated postweaning rearing conditions will increase the reactivity and excitability of rodents (Korn and Moyer, 1968; Valzelli, 1973; Riittenen et al., 1986). In the Korn and Moyer (1968) experiments on adult rats, four weeks of isolation in standard laboratory cages was less detrimental in these respects than fourteen weeks. Isolation heightened emotionality; isolated rats reacted violently when they were picked up.

Hyperexcitability has been documented in dogs kept in a sensory-restricted environment. Pairs of puppies residing in barren kennels become unusually aroused and excited when exposed to something new (Melzack, 1954), These puppies could hear and smell dogs in adjacent kennels, and the kennels were illuminated continuously so the puppies in a pair could see and interact with each other. Almost a year after being returned to a residential environment with a human family, the subjects were still hyperexcitable (Melzack, 1954) and electroencephalographic patterns from their brains indicated extreme arousal (Helzack and Burns, 1965).

Animals living in a barren or restricted environment will seek stimulation. During a four-hour experiment, rhesus monkeys in an opaque cage worked hard to discriminate between blue and yellow cards to operantly open a window for a brief glimpse into the rest of the laboratory (Butler, 1953). Rats residing in a barren environment will bar press more than those in an enriched environment (Ehrlich, 1959). The restricted environment consisted of five rats in a 60 x 30 x 23-cm wire cage. Enriched environment rats were raised in groups of ten in a two-tiered environment consisting of a 120 x 76 x 60-cm cage filled with playthings such as corks, tunnels, ramps, and platforms. During testing, each rat was placed individually in a Skinner box. Reinforcement for bar pressing consisted of either clicks or turning on a light.

Pigs residing in crates which were just wide enough to allow turning around at a flared end, turned around 11 to 12 times per day (McFarlane et al., 1988). Even animals which did not have to turn around to gain access to feed or water nevertheless did turn around about the same number times as did those which had to do so. This is presumptive evidence of the pig's need for a certain level of physical activity.

Pigs in intensive housing systems spend long periods sleeping, punctuated by brief periods of intense excited activity, as when a door slams or a person walks into the room. This behavior is similar to that of Melzack's puppies; in the restricted environment, they became very excited when minor changes were made in their cages (Melzack, 1954). Stallions kept in stalls without exercise were harder to handle than those given daily exercise (Dinger and Noiles, 1986).

Pigs reared in indoor pens with minimal contact with people were more excitable and difficult to load into a trailer than those reared outside with frequent contact with people (Warriss et al., 1983). Stolba and Wood-Gush (1980) found that pigs in barren pens with concrete floors reacted more strongly to and played longer with a tire than did those in straw-bedded pens.

Wood-Gush and Beilharz (1983) found that pigs in cages spent less time lying when their environment was enriched with a trough filled with dirt. Maybe pigs in more barren pens sleep longer to reduce the arousal level of an overly aroused nervous system. Zentall and Zentall (1983) suggested that autistic children withdraw from stimulation to prevent further arousal of an already overly aroused nervous system.

Central nervous system effects of environmental restriction

Environmental restriction has long-term effects on an animal's central nervous system (Helzack, 1969). Six months after being released from a sensory-restricted environment, young dogs still had electroencephalographic patterns which indicated an abnormally high level of arousal (Helzack and Burns, 1965).

The experiments reviewed in this paper support the hypothesis put forth by Walsh and Cummins (1975) that animals living in an enriched environment usually are less excitable than animals in barren surroundings. Restricting sensory input makes the nervous system of both humans and animals more sensitive and more reactive to external stimulation. Schultz (1965) stated: "When stimulus variation is restricted, central regulation of threshold sensitivities will function to lower sensory thresholds. Thus, the organism becomes increasingly sensitized to stimulation in an attempt to restore balance." Walsh and Cummins (1975) concluded that animals in an enriched environment are subject to greater arousal during rearing and therefore are less likely to exhibit over arousal in a novel or highly stimulating environment.

Sensitization to external stimuli occurs within the central nervous system. Placing a small cup on a person's forearm to block tactile stimulation for one week increased tactile sensitivity on the opposite (unshielded) forearm (Aftanas and Zubek, 1964). This effect was quite persistent, as increased sensitivity was still present three days after the cup had been removed.

The sensitizing effect of restricted sensory input also can occur across sensory modalities. The brain needs sensory input to maintain normal reactivity levels. Zubek et al. (1964a) found that a person who is blindfolded or wears translucent goggles for one week has increased tactile, pain, and heat sensitivities, respectively. Even partial deprivation of visual input will sensitize the skin (Zubek et al., 1964b). Trimming the whiskers of baby rats causes the areas of the brain that receive sensory input from the whiskers to become more excitable (Simons and Land, 1987), and this effect persisted. Receptive fields were still enlarged three months after the whiskers had regrown.

Other indicators of increased central nervous system arousal are the effects of depressant and convulsive drugs, respectively, on animals residing in different environments. Rats living singly in small cages were compared to those in groups, and the isolated rats had higher thresholds for anesthetics. Juraska et al. (1983) administered both depressant and convulsive drugs to rats isolated in small laboratory cages (isolated condition, IC) or living in a group of 12 in an enriched environment (enriched condition, EC) which consisted of a large enclosure with many different toys. The IC rats injected with sodium pentobarbital took longer to lose the righting reflex. In a second experiment, when rats reared in IC and EC, respectively, were given the convulsant drug, pentylenetetrazol, the IC rats went into light flash-induced convulsions at lower doses.

Bombarding an animal with excessive stimulation can cause detrimental signs similar to those caused by restricted stimulation. In the IC/overstimulated condition, mice were subjected involuntarily to intense sound, light, and vibration, whereas those in the EC were allowed to initiate and control their interaction with toys. Both IC and IC/overstimulated mice were more irritable than were EC mice, and both forced under stimulation and forced overstimulation increased irritability.

Concept of optimal stimulation

Zentall and Zentall (1983) believe that organisms modulate incoming sensory stimulation to maintain the nervous system at an optimal level of arousal. In some cases, for example, stereotyped behavior may be an attempt by a disturbed biological system to restore homeostasis (Damasio and Maurer, 1978). Environments offering low levels of stimulation tend to cause behavior in animals which is stimulus-seeking. Conversely, animals living in an overstimulating environment sometimes attempt to reduce their central nervous system arousal by engaging in repetitive activity (Zentall and Zentall, 1983).

Dantzer and Mormede (1983) and Dantzer (1986) suggested that repetitive stereotypic behavior can serve a de-arousal function. For example, food-deprived pigs will engage in repeated chain-pulling in between food deliveries which occurred every four minutes, and pigs which had access to the chain had lower plasma corticosteroid levels. Stereotypes also occur in pigs living in barren environments which provide low levels of sensory stimulation (Ewbank, 1978; Markowitz, 1982). Therefore, stereotypes may occur in both under stimulating and overstimulating environments. In under stimulating environments, of course, they probably serve to increase sensory input (Ewbank, 1978; Harkowitz, 1982). Young children deprived of normal hugging will engage in excessive masturbation, which stopped when tactile contact with parents was increased (McGray, 1978).

Even though animals living in a barren environment may attempt to reduce central nervous system arousal level, the central nervous system nevertheless may remain in an abnormally aroused state. Perhaps some of the environments that have been imposed experimentally deprived the animals beyond their physiological ability to cope. For example, extremely barren environments would never be encountered in nature.

Nervous systems are endowed with a certain amount of plasticity so the animals can deal with changing environmental demands. The ability to adjust the relative sensitivity of sensory systems to different environments would be useful to an animal for survival in the wild. Simons and Land (1987) concluded that sensory input affects the development of the brain's somatosensory cortex.

Structural changes in the brain

Stimulation and opportunities for activity provided by an enriched environment cause changes in dendritic branching in the brain. Groups of weanling rats residing for four weeks in a large enclosure with toys and objects to climb (enriched condition, EC) had greater dendritic branching in the visual cortex than did those living singly In standard laboratory cages (isolated condition, IC) (Greenough and Juraska, 1979; Volkmar and Greenough 1972).

More recent research results indicate that new synapses may be formed by neural activity induced by the enriched environment (Greenough et al., 1985). Researchers also have compared IC and EC with (social condition, SC), the latter consisting of two animals living together in a 22 x 25 x 30-cm laboratory cage. Floeter and Greenough (1979) found no differences in dendritic branching in the cerebellums of SC and IC monkeys (Macaca fascicularis). The EC monkeys had the most dendritic branching.

In rats, SC animals had intermediate levels of dendritic branching in the visual cortex (Volkmar and Greenough, 1972). The IC rats had the least branching, the EC rats the most. Weanling rats in EC also had heavier brains than did those in IC (Bennett et al., 1964; Diamond et al., 1964). Rats in EC also had a thicker cortex and additional glial cells (Bennett et al., 1964; Diamond et al., 1964; Diamond, et al., 1966). In these experiments, the control rats were housed in trios in 32 x 32 x 20-cm laboratory cages, and all rats were in their respective environments for at least 29 days.

Very short (four-day) periods of differential housing affected cortical depth in the dorsal-medial part of the brain's occipital cortex (Diamond et al., 1976). Cortical depth differences in weanling rats were induced mainly by environmental impoverishment, whereas in adults they were induced mainly by enrichment.

Neural hypertrophy does not necessarily mean that an animal is in an environment that is higher quality overall. Research indicates that neural hyptertrophy can sometimes be detrimental. Rats exposed to continuous lighting had greater spine density in the visual cortex (Parnavelas et al., 1973), but their retinas were damaged (Bennett et al., 1972, 0'Steen, 1970). Stimulation of the hippocampus with an implanted electrode induced axonal growth and reorganization of synapses, but these new connections increased excitability and seizures developed (Sutala et al., 1988)

Farm animals

Enriching pigs' environments may help prevent changes in the central nervous system which lead to hyperexcitability and stereotyped behavior. Casual observations of pigs residing in indoor pens indicates that they often exhibit an excessive startle reaction to door slamming or other disturbances. Farmers have learned that playing a radio with a variety of music and talk reduces pigs' reactivity to extraneous sounds. Observations at slaughter plants indicate that some pigs are more excitable and difficult to handle than others (Grandin, 1986).

Pigs living in indoor pens with minimal contact with people were more excitable and difficult to load than those residing outside with frequent contact with people (Warris et al., 1983). Pigs from 1.2 x 1.2-m relatively barren pens also were slower to approach a novel object or human strangers compared to pigs that had had access to toys and frequent contact with people (Grandin et al., 1983).

Pigs which are either balky or excitable are more difficult to transport and handle at the slaughter plant. Those that at reluctant or refuse to move are more likely to be subjected to excessive electric-prodding. Electric prod-induced excitement is detrimental to pork quality (Barton-Gade, 1985; Grandin 1986). Repeated electric-prodding also will increase bloodsplashing (hemorrhage) in the pork carcass (Calkins et al., 1980). Further, it is detrimental to the pig's welfare; heart rate increases progressively with successive prods, and if it is continued, the pig's heart rate will reach dangerously high levels and death can result (van Putten and Elshof, 1978; Mayes and Jesse, 1980). A pig will stop and lie down when its heart rate is near the lethal point (Hayes and Jesse, 1980).

B. Environmental Enrichment Methods

Choice of environmental enrichment

Stolba (1981) observed pigs in a semi-natural environment help determine suitable environmental enrichment devices. The studies were an important first step in determining choices pigs make from among a variety of natural objects. It is important that the play objects provided pigs be things that they appreciate, and the objects must support sustained interest. Research can be used to determine the type of objects pigs prefer.

Choice tests have been used to determine animals' preferences for type of flooring, social groupmates, and mate (Hughes and Black, 1973; Hughes, 1976; Dawkins; 1982; Farmer and Christison, 1982; McGlone and Morrow, 1987); to test sheep's preferences for type of handling facilities (Hitchcock and Hutson, 1979; Hutson, 1981) and restraint methods (Grandin et al., 1986; Rushen, 1986); and to determine pigs' preferred angle and flooring surface for ramps (Phillips et al., 1988).

When choice tests are used, it is essential that they be controlled adequately to prevent previous experiences from influencing the choices. For example, previous experience with different pastures can alter pasture preference in sheep (Arnold and Maller, 1917), and the environment in which they were reared can affect caged hens' subsequent flooring and social preferences (Hughes, 1976).

Another potential problem with preference tests is that something that is preferred on a short-term basis may have detrimental long-term effects on the animal's well-being (Hughes and Black, 1973; Duncan, 1978; van Rooijen, 1982) For example, animals may prefer one type of flooring during a two-hour preference test, but the preferred flooring may injure the animals' feet after several months.

Piglets usually prefer plastic-coated expanded-metal floor over concrete (Farmer and Christison, 1982). In practice, plastic-coated expanded-metal flooring is excellent for farrowing crates and nursery pens but causes problems when used for a finishing floor. Casual observations indicated that pigs finished on either flattened or plastic-coated expanded metal had excessive hoof growth, and they balked and were difficult to drive into a high-speed slaughter line (Grandin, 1988).

Preference tests are one useful method for determining practical objects that producers could use to enrich a pig's environment. Unlike flooring, feed, or housing, object choices are unlikely to have adverse long-term effects on the pigs' health or performance. There is a need for information on the pig's short- and long-term preferences for means of environmental enrichment so that recommendations can be made to producers.

Play in Farm Animals

Farmers have offered pigs various objects in attempts to enrich their environment and reduce vices. van Putten (1969) recommended placing chains or rubber hoses in pens to occupy the pigs' attention and reduce tail-biting. Weanling pigs will play with many kinds of objects, and they vigorously chewed, tugged, and shook suspended strips of cotton cloth (Grandin et al., 1983). They also rooted and rolled balls, wooden boards, plastic boxes and other objects when these items were first placed in their pen. When these objects became soiled with excreta, however, the pigs lost interest in them. Suspended strips of cloth seemed to hold the pigs' interest longer than did suspended chains, clean objects more than did soiled.

Fagen (1984) described play as follows: "Play means behavioral performance that emphasizes skills for interacting with the physical and social environments and that occur under circumstances under which the function of the exercised skills can not possibly be achieved." Play occurs in many farm animals, including cattle, sheep, pigs and horses (Brownlee, 1954, 1984; Tyler, 1972; Schoen et al., 1978; Sachs and Harris, 1978; van Putten, 1978; Dobao et al., 1984-85; Crowell-Davis et al., 1987), as well as a relative of the domestic pig, the collared peccary (Byers and Beckoff, 1981; Byers, 1984). It is interesting that in the collared peccary both juveniles and adults play, because ordinarily in ungulates only juveniles and subadult males play (Byers, 1984).

Object play, including "tug-of-war", has been observed in dogs, cats, horses, and jackals (van Lawick and Goodall, 1971; West, 1977; Fagen, 1981; Martin, 1984). Captive bushdogs and foxes play with sticks (Biben, 1982). A captive rhinoceros will root a ball and other objects (Inhelder, 1978; Hediger, 1968). Piglets also engage in object play (Gundlach, 1988; Fraedrich, 1974). Weanling pigs tugged on cloth strips in a manner similar to that of dogs playing "tug-of-war" (Grandin et al., 1983). Foals will pick up sticks and carry them around and toss them in the air (Crowell-Davis et al., 1987).

Activity patterns in livestock

There are few data available on the daily activity patterns of play in pigs. Domestic livestock do have daily patterns of activity for grazing and other behavior (Hughes and Reid, 1951; Squires, 1974; Arnold and Dudzindki, 1978). Pigs residing indoors are more active during the day (Morrison et al., 1968). In rich environments there are two peaks of activity -- around 0900 and 1700 hrs in pigs (Dantzer and Mailha, 1972; Dantzer, 1973; Bure, 1981). When pigs had an opportunity to control the temperature in their pen, a cooler environment was preferred at night (Curtis and Morris, 1982).

Aggression and environmental enrichment

Pigs are a relatively feisty species. When strange pigs are mixed they will fight one another vigorously, at least in part to establish a new social order (McBride et al. , 1964). Fighting among pigs causes injuries (Symoens and van der Brand, 1969; Jones, 1969) and will reduce weight gains if the pigs are overcrowded or have limited access to food (Graves et al., 1978; Sherritt at al., 1974).

Fighting among newly weaned piglets generally has little adverse effect on the pigs' long-term performance, but it can be detrimental to welfare. McGlone and Curtis (1985) found that fighting and injuries could be reduced by providing newly weaned pigs small hides where they could place their head and shoulders. Anecdotal reports suggest that providing pigs toys during times of social mixing also will reduce fighting.

Straw bedding did not reduce fighting in growing pigs fed ad libitum, but it had a tendency to reduce fighting in fasted pigs (Kelley et al., 1980) Another factor which will influence aggression is type of housing. Pigs in indoor pens were more aggressive than those residing outdoors (Meese and Ewbank, 1973). Crowding tends to increase pig aggression (Bryant and Ewbank, 1972; Kelley et al., 1980; Randolph et al., 1981), but it does not increase all kinds of aggression in a uniform manner. Reducing floor space reduces aggression which occurs away from the feeder, but it has no clear effect on aggression at the feeder (Ewbank and Bryant, 1972). Restricting lying space increased aggressive encounters when animals were standing but had little effect on aggression in other areas (Ewbank and Bryant, 1972). More information is needed about the effects of environmental richness on aggression in pigs.

C. Thesis Research Outline

Dendritic growth in the pig's cerebral cortex

In the first experiment to be reported in this thesis, the effect of environmental enrichment on dendritic growth in the pig's brain was studied. Heretofore, many such studies have been conducted with rodents (see review by Greenough and Chang, 1985), and three with primates (Struble and Riesen, 1978; Floeter and Greenough, 1979; Stell and Riesen, 1987). The study to be reported here was the first conducted with a domestic farm animal. Valuable knowledge about environmental effects on brain development may be able to be gained by studying animals that have a larger, more complex brain than does the rat.

Excitability and handling of the pig

The second series of experiments were designed to determine effects of environmental enrichment on the excitability and handling of pigs living in intensive systems. The aim was to determine effects of various kinds and extents of environmental enrichment on the excitability of pigs in space-restrictive, relatively barren pens.

0bject play by the pig

In the third series of experiments, choice tests were conducted to determine the types of play objects that are preferred by pigs, both short- and long-term. Daily activity patterns of toy play also were determined. In the last experiment, effects of access to play objects prior to or during social mixing of newly weaned piglets were determined to learn whether toys would reduce agonistic behavior.

The environmental enrichment methods used in all of these experiments were simple, practical procedures that swine producers could use. Results of these experiments may help answer questions as to whether relatively simple environmental enrichment procedures will improve the productivity and welfare of pigs residing in intensive production systems.


Click here to continue viewing the Dissertation on neural development.

Click here to return to the Homepage for more information on animal behavior, welfare, and care