 |
|
|
2.2.1. Variables
Cage cleaning is usually a rather disturbing husbandry routine for rodents. This procedure can be associated with a significant:
Gray and Hurst (1995) observed that aggression among the five members of group-housed CFLP mice escalated whenever the animals were briefly removed and subsequently placed back in their own dirty cages. The aggression eliciting effect of their own odor cues was replicated when the animals' home cage was not completely cleaned and deodorized but the soiled sawdust merely replaced with fresh sawdust. When the cage cleaning process involved the replacement of the soiled home cage with a new cage and fresh sawdust, inter-male aggression was substantially reduced, but not eliminated (Hurst, 1990).
Ambrose and Morton (2000) videotaped groups of five and six male BALB/c mice for one hour immediately after their cages were cleaned. The mice had their cages changed twice a week and were removed during this process from the soiled cage directly into a new cage.
A significant reduction in aggression was also achieved by furnishing each cage with a glass jar that was thoroughly cleaned and deodorized as part of the cage cleaning process.
These findings confirm the aggression-inducing effect of familiar, territorial scent marks deposited on objects (Jones and Nowell, 1973, 1975; Mugford, 1973; Gray and Hurst, 1995). They also demonstrate that the provision of species-appropriate enrichment such as objects for seeking shelter (box, tube, bottle) or objects for gnawing (wood block) can reduce the incidence of inter-male aggression and the associated risk of social distress and serious wounding, as long as they do not carry familiar odor cues. It should be noted that structural additions to the cage that do not provide escape routes/options can increase rather than decrease agonistic interactions among male mice right after cage cleaning (Van Loo et al., 2002).
Van Loo et al. (2000, 2004) videotaped trios of male BALB/c and CD-1 mice for one hour after their cages were cleaned. When the mice were transferred along with some of their soiled nesting material into clean cages, overt aggression was significantly reduced, and the animals' urine corticosterone concentrations were significantly lower compared with mice without nesting material. One may infer from this study and those of Gray and Hurst (1995) and Ambrose and Morton (2000) that specific scent marks deposited on the ground or on objects trigger territorial aggression while odor cues adhering to nesting material fail to induce, or perhaps even mitigate, aggression in male mice.
Armstrong et al. (1998) did not add enrichment objects but provisioned the weekly exchanged cage of eight male BALB/c mice with a 2.5-cm layer of fresh cornhusk. Observations carried out four days after each cage change revealed that the animals had significantly fewer wounds resulting from aggression than control animals, presumably because the mice could avoid conflicts relatively easily by breaking visual contact with each other in the husks. Van Loo et al. (2002) found in a subsequent study with trios of male BALB/c mice that agonistic interactions were significantly reduced immediately after cage cleaning when odor-free paper tissues presumably along with fresh sawdust bedding were added to the fresh cage.
Duke et al. (2001a) moved single male Sprague-Dawley rats to clean cages that contained new substrate plus a small quantity of the soiled bedding material from their previous cages. This failed to have a calming effect, and the subjects still showed significant cardiovascular (increased heart rate and blood pressure) and behavioral (arousal) responses. Repeated cage changing did not produce any lessening of these stress responses, suggesting that the rats could not adapt to this standard husbandry procedure.
2.3. Transfer to an Unfamiliar Location and Separation from Cagemates
2.3.1. Variables
Being removed from the home environment and transferred to an unfamiliar location is a very disturbing experience for captive animals, just as it is for humans. It has been documented in numerous reports that rats and mice experience significant changes in the resting values of their physiological parameters when they are moved in their home cage to a different location (Friedman and Ader, 1967; Brown and Martin, 1974; Euker et al., 1975; Pfister and King, 1976; Kvetnansky et al., 1978; Gärtner et al., 1980; York and Regan, 1982; Damon et al., 1986; Ursin and Murison, 1986; Cabib et al., 1990; Drozdowicz et al.,1990; Tuli et al., 1995a; Barrett and Stockham, 1996; Tabata et al., 1998; Sharp et al., 2003). Surprisingly, such reports are missing for other rodents and rabbits, who are also likely to be stressed when they are transferred to different living quarters.
Being separated from familiar social companions is distressing for any social animal, including rodents and rabbits. This experience is typically accompanied by behavioral fear responses such as freezing, reduced drinking and eating and associated loss of body weight, sustained increase in heart rate and blood pressure, and altered hypothalamic-pituitary-adrenal function and electroencephalic activity (Hadley, 1927; Fenske, 1990; Ehlers et al., 1993; Lawson and Churchill, 2000).
Dobrakovavá and Jurcovicova (1984) tried to habituate male Wistar rats caged in groups of four to being transferred in their home cage to another room, left there for a few minutes, and returned to the original location. This was repeated once every day for a period of 15 days. The animals were not able to adapt to this common procedure but showed significant increases in plasma corticosterone and prolactin which were not lower on day 15 than on day one.
Fenske (1992) was able to eliminate the typical freezing response to separation in group-housed male guinea pigs by confining the experimental subject in a small test cage which was placed in the familiar large home cage. This simple procedural adjustment allowed the subject to maintain uninterrupted auditory and olfactory contact with his cagemates.
2.4.1. Variables
Being involuntarily handled and forcefully restrained by the human "predator" is a powerful stressor for rodents and rabbits. It jeopardizes not only their well-being but also the scientific validity of data collected from them (Balcombe et al., 2004). Späni et al. (2003) and Kramer et al. (2004) showed that merely entering an animal room without touching a cage can be sufficiently alarming to trigger significant endocrine and cardiovascular stress responses that bear the risk of affecting subsequently collected stress-sensitive parameters, even before the actual experimental procedure is done with one of the research subjects of that room. Kramer et al. (2004) noticed that in individually caged C57BL/6 male mice do not adapt to this everyday event. Subjects showed significant heart rate elevations and significant body temperature increases in response to personnel entering their room at 9:30 a.m., even after repetitions of this disturbance on 12 consecutive days.
2.4.2.1. Habituating to Procedures
Dobrakovavá and Jurcovicova (1984) caught group-housed Wistar rats and handled each animal daily for one minute over a 15-day study period. Subjects showed elevated plasma corticosterone concentrations, but these were significantly lower on day 15 than on day one, indicating that the animals had adapted.
Sharp et al. (2005) recorded the heart rate and blood pressure of singly caged Sprague-Dawley rats who were given a subcutaneous injection once a day on four consecutive days. Injection resulted in a significant increase of both cardiovascular parameters and was accompanied by agitated movement. These responses did not change over the course of the four days, suggesting that the animals were not able to adapt to this common procedure within the given time frame.
Tuli et al. (1995b) demonstrated that singly caged BALB/c mice did not adapt to conventional restraint in perspex tubes. Even after 21 daily one-hour restraint sessions, elevated plasma corticosterone concentrations did not differ from those of subjects who were restrained only one time.
2.4.2.2. Training to Cooperate during Procedures
Gastric intubation for oral drug administration is probably the most distressing procedure that rodents and rabbits are subjected to routinely (Bonnichsen et al., 2005). The animal is usually exposed to two or three humans who apply forceful restraint/immobilization during an extremely uncomfortable, life-threatening, often injurious and sometimes even deadly procedure (Murphy et al., 2001; Murphy, 2001).
Huang-Brown and Guhad (2002) trained 57 Wistar rats living in trios to cooperate during daily oral administration of indomethacin and celecoxib, two anti-inflammatory drugs. An amount of medication equal to ten doses was mixed into approximately 500 mg of softened chocolate, and then divided into ten aliquots. The subjects were first allowed to develop a taste for pure chocolate by carefully placing a pellet into their mouth using a 14-gauge gavage needle. They were handled gently to avoid association of chocolate with aversive stimuli. After eight days of training only five percent (3/57) of the animals failed to cooperate, while 95 percent (54/57) of them displayed "eager anticipation" of the decoy whenever the cage door was opened. The rats' cooperative response did not change when the chocolate pellets contained the test drugs: they took and swallowed them without hesitation. This refined gavage method provided consistent, reliable, easy and accurate dosing. There was no need for keeping the animals individually. Housing them in small groups did not interfere with this new treatment technique.
Marr et al. (1993) developed a simple training technique to gain the cooperation of ten NZW rabbits for voluntary oral administration of the test drug tosufloxacin. For five days, the animals were offered sucrose water daily from a syringe. The tip of the syringe was coated with sucrose granules. Most rabbits spontaneously approached the syringe when it was inserted through the bars of the cage, tasted the tip of it and/or drank the fluid. Rabbits who did not initially seek the syringe usually did so with minimum encouragement. These training sessions were repeated three times a day, for a total of 15 minutes per session, until all animals swallowed the sucrose solution content of the syringe. The antibiotic solution was then substituted for the sucrose solution, while the tip of the syringe remained coated with sucrose granules for each subsequent daily administration of the drug. Within two days, eight of the ten rabbits were seeking the syringe when research staff entered the room.
They "would stand with their paws on the front of the cages, protrude their faces from between the bars, and appear to beg for the syringe containing the antibiotic solution" (Marr et al., 1993, p 48).
Needless to say, these rabbits did not experience any fear, distress or undue discomfort during this refined oral administration procedure.
There are no reports of training attempts to gain the cooperation of rodents or rabbits during injection or blood collection in order to reduce the significant stress reactions of the subjects resulting from enforced handling and restraint (Krulich et al., 1974; Moynihan et al., 1990; Brockway et al. 1993; Tuli et al., 1995c; Tabata et al., 1998).
The restraint stress experienced by the handled subject can be buffered under certain circumstances by the presence of one or several compatible conspecifics (social buffer) and by appropriate environmental modifications.
Ader and Friedman (1964) recorded the behavioral responses of Sprague-Dawley rats to being gently picked up by a person. Animals kept alone were more fearful of the handling person and showed significantly more alarm vocalization and resistance to being picked up than animals who shared a cage with five other rats. It was considerably more difficult to handle rats who were caged alone than those caged with other rats. Giralt and Armario (1989) found in Sprague-Dawley rats that their stress response to acute immobilization as measured by the increase in plasma corticosterone concentration was significantly greater when they were housed alone than when they lived in groups of four.
Sharp et al. (2002, 2003) worked with Sprague-Dawley rats bearing telemetry transmitters. Each subject was tested in single-housing and group-housing (four same-sex animals) conditions, and his/her heart rate monitored both during and after a subcutaneous injection. For this procedure, the cage was placed on a workbench, the water bottle and cage lid were removed, and the target subject was gently picked up and placed on the bench surface. The investigator held the rat with one hand, lifted the loose skin at the nape of the neck and injected 0.2 cc of sterile saline into the skin pocket by using a 26-gauge needle. The rat was placed back into the cage, which was then returned to the rack. This sequence of events required 20-30 seconds. All rats showed a significant cardiovascular stress response to the involuntary handling and injection that did not return to baseline within 60 minutes, but rats housed alone had significantly greater increases in heart rate than did rats housed with other rats. This effect could not be confirmed during the tail vein injection procedure, but it was evident during vaginal lavage with females from groups vs. females living alone.
Moncek et al. (2004) compared the stress response of Wistar rats under various conditions to gentle one-minute handling sessions. One set of rats was kept in groups of three or four in barren cages, while the other subjects were taken from groups of ten animals living in cages five times larger and enriched with various toys, tunnels, swings and a running wheel. The large-group enriched rats showed significantly lower ACTH, corticosterone and adrenaline responses to handling than the small-group nonenriched rats. It is not clear whether the stress-buffering effect was due to the larger number of group members, the more space available to each subject, the enrichment or as is presumably the case a combination of these factors.
Belz et al. (2003) studied vein-cannulated, individually housed female Sprague-Dawley rats who lived in 1100-cm2 cages that were either barren or enriched with rubber toys for gnawing and squares of compressed cotton fiber for shredding. The living space was the same for all animals, but those in enriched cages were not only easier to handle but also showed a significantly lower adrenocorticotropin stress response to intraperitoneal injection. This effect could not be verified in male rats.
Sharp et al. (2005) assessed cardiovascular stress responses to common handling procedures in SH rats, who were individually housed in 930-cm2 cages that were barren or furnished with a simulated burrow, a feeding enrichment gadget and a shredding-and-nesting item. While blood pressure was not affected by enrichment, heart rate responses to subcutaneous and tail vein injection were significantly lower in enriched vs. non-enriched rats, indicating that enrichment had a stress-buffering effect. This could not be replicated in Sprague-Dawley rats.
Van de Weerd et al. (2002) scored the behavioral responses of RIVM mice kept in groups of eight during routine handling procedures. Animals living in 840-cm2 cages enriched with a nest box, wood-wool, climbing structures and gnawing blocks showed significantly fewer signs of stress and resistance than animals living in 375-cm2 unfurnished cages. The stress-buffering effect may have been due to the enrichment, the larger cage size or a combination of these two factors.
2.5.1. Variables
Multi-tier caging is the prevailing housing system for rodents and rabbits. It bears intrinsic variables that have the potential to influence research data (Figure 4):
Of all the commonly considered environmental factors, light intensity within cages (Clough, 1982) and level of the living environment on the multi-tier cage rack are probably the most variable. The different degrees of illumination resulting from the housing of rodents and rabbits at different levels of the animal room may be one of the explanations for variation in experimental results (Lockard, 1962). If cages are placed at different levels of the room at different distances from the light source, the investigator may be measuring behavioral rather than experimental results (Mulder, 1971, 1976).
In the wild, rodents and rabbits lead a nocturnal or crepuscular life style and change their general activity level as a function of light intensity (Aschoff, 1960; McClearn, 1960; Ross et al., 1966). They show less emotionality, as reflected in decreased defecation and urination in an open field, under low illumination than under high illumination (Ross et al., 1966). There is ample scientific evidence that differences in illumination affect not only behavior and general activity but almost all physiological systems as well, thereby influencing the results of behavioral and physiological experiments and toxicological tests (Chance, 1947; Ross et al., 1966; Porter, 1967; Wurtman, 1967; Weihe et al., 1969; Ott, 1974; Weihe, 1976; Saltarelli and Coppola, 1979).
Rodents and rabbits are terrestrial animals who avoid predators such as humans by retreating under covered structures close to the ground or into burrows under the ground. In laboratories, being confined high above the ground in an upper-row cage without a refuge area is likely to be more disturbing for them than being confined close to the ground in a lower-row cage, which also does not offer a refuge, but is at least relatively dark, hence more secluded.
|
|
|
|
Many authors do not mention in scientific publications at which level of the room the research subjects were housed (Davis et al., 1973; Lang and Vesell, 1976; Gamble, 1979; Clough, 1982), suggesting that they ignore the possibility that this variable could influence data in their scientific studies. Likewise, there are only a few studies assessing the influence of cage position in a multi-tier rack on behavioral and physiological parameters.
Galef and Sorge (2000) provisioned the cages of individually housed Sprague-Dawley rats with PVC tubing. The animals were never seen inside the tubes during the night, but they used them during the day about 60 percent of the time when their cages were located on the top shelf versus only eight percent of the time when their cages were located on the bottom shelf. It stands to reason that the rats used the tubes as protection against overhead light exposure, which was significantly more intense on the top shelf than on the bottom shelf.
Kaliste-Korhonen et al. (1995) observed singly caged Wistar rats from five different shelf levels in an open arena. Rats from the top shelf showed a significantly longer latency in rearing than rats from lower shelves. This was interpreted as a sign that the animals on the top shelf were better habituated to the high intensity of light to which they were exposed in the open-field test arena.
Izidio et al. (2004) mention that cage position influenced the behavior of singly caged Lewis rats and SH rats in the open-field test, with animals housed in top cages appearing less anxious than those housed in bottom cages.
Ader et al. (1991) assessed emotionality in single- and group-housed NOD mice. Mice caged in the top shelf were significantly more emotional as evidenced by vocalizing, struggling, urinating and defecating than mice caged in the middle shelf, who in turn were more emotional than mice caged in the bottom shelf. Garner et al. (2004) noticed in mice of various strains, who were kept in different-sized groups in clear plastic cages, that barbering was significantly more severe (higher percentage of body denuded) in upper-row caged groups than in lower-row caged groups. Unfortunately, no ethological data were collected to support the implicit inference that mice caged in upper shelves spend more time barbering than mice caged in lower shelves.
Greenman et al. (1983) found in a large sample of BALB/c mice that animals kept on the top shelf consistently consumed more food, yet had consistently lower body weight gains than animals kept on lower shelves. Greenman et al. (1984) also assessed the distribution of spontaneous and chemically induced tumors in BALB/c mice in relation to the shelf level of their home cages. The animals were assigned to groups of four per cage. Shelf level significantly influenced five of six major spontaneous neoplasms. Lagakos and Mosteller (1981) noticed in CD-1 mice that the incidence of reticulo-endothelial tumors increased conspicuously from the bottom (17 percent) to the top shelf (32 percent), and warned that failure to take shelf level into account in the design of carcinogenicity studies can easily lead to erroneous conclusions.
It should be emphasized here that US animal welfare regulations stipulate that indoor housing facilities provide lighting that is "uniformly distributed" (United States Department of Agriculture, 2002, p 72 and 81).
Even though it is very costly, cages in a multi-tier arrangement can be illuminated uniformly by mounting light fixtures at the level of each tier on the wall rather than on the ceiling. Behavioral and physiological differences due to cage position, however, would not be affected by this refinement. As long as the primary enclosures of caged animals are stacked on top of each other in racks, differences in the animals' distances from the ground are unavoidable. This variable is intrinsic to the multi-tier caging system.
2.6.1. Variables
The noise environment of rodents and rabbits is a daily variable that is usually uncontrolled and overlooked in the methodology section of scientific articles, even though it is likely to have important implications not only for the animals' well-being but also for the reliability of research data collected from such animals (Gamble, 1979, 1982; Pfaff, 1974; Milligan et al., 1993). The pattern of physiological changes elicited by noise in rodents and rabbits is the same as it is in humans. Noise associated with cage cleaning, general maintenance, and especially with construction and remodeling work can be uncomfortable for attending animal care personnel, but it must be overwhelming for the caged animals. The noise can be so intense that certain areas of animal-holding facilities can be officially designated "hazardous noise environments" by Health and Safety Departments. Confined animals may experience the noise not only as annoying to the ears but also as alarming for their sense of security. It is well established in rats, mice, guinea pigs and rabbits that exposure to loud noise is associated with profound alterations in the neural, endocrine and cardiovascular systems (Anthony et al., 1959; Anthony and Harclerode, 1959; Anthony, 1963; Zondek and Tamari, 1967; Lockett, 1970; Armario et al., 1985), which can manifest in mice and rats as potentially fatal seizures (Bevan et al., 1951; Iturrian and Fink, 1968; Wada and Asakura, 1970).
Barrett and Stockham (1996) observed that Wistar rats had elevated plasma corticosterone levels during an experiment on days when the animal attendant had cleaned out the cage racks. A controlled study was then conducted in which singly caged rats were deliberately exposed to ten minutes of loud whistling and talking accompanied by banging of food hoppers, cage doors and fecal dropping trays. This resulted in a significant increase in the rats' corticosterone concentrations.
There are many strategies to systematically reduce noise in animal facilities that have yet to be implemented. Consultation with a qualified acoustical engineer can lead to specific solutions for all but the most recalcitrant noise abatement problems (Peterson, 1980; Johnson et al., 2005).
Carlton and Richards (2002) took steps to control at least some noise. Using readily available industrial and architectural sound absorption panels, and fitting acoustical covers on electrical motors of cagewashers reduced noticeably the noise level throughout the whole facility by 3-5 decibels (dB), in hallway areas by 5-7 dB, and in cagewash areas by 8-10 dB.
It is also possible to mask noise peaks in animal rooms with relatively loud background noise/music, but the effect on the caged animals is not known (Pfaff and Stecker, 1976).
Traditional housing, husbandry and handling practices for rodents and rabbits jeopardize not only the welfare of the animals but also the scientific validity of research data collected from them. Most of these risk factors confinement, cage cleaning, transfer, restraint and noise are sources of stress and distress that cannot be avoided in the research laboratory setting, but practices can be refined so that the animals and science are less affected by them. One risk factor the multi-tier caging system cannot be refined but could be avoided altogether without adverse effects on the animals and on science.
A good management program not only provides the environment, housing and care that foster the animals' well-being, but also minimizes variables that can influence research data. This will have the added advantage of decreasing the number of animals required to achieve statistical significance in the scientific results (Russell and Burch. 1959; Home Office, 1989; National Research Council, 1996; Öbrink and Rehbinder, 1999). Confinement in barren living quarters is probably the most prevalent extraneous variable.
The distress and fear associated with confinement can be buffered by the presence of one or several companions. This has been demonstrated in both sexes in rats and mice, and there is good reason to believe that the same holds true for other rodents and female rabbits. Rather than forcing these animals to permanently live under conditions of social deprivation, thereby jeopardizing their well-being and the validity of research data collected from them, compatible pair- or group-housing should be the norm.
The stress- and distress-mitigating, comforting influence of companionship has been confirmed in several other species including goats (Lyons et al., 1988), sheep (Baldock and Sibly, 1990), chickens (Jones and Merry, 1988), nonhuman primates (Coe et al., 1982; Shively et al., 1989; Coelho et al., 1991), and human primates (Bovard, 1959; Epley, 1974, Kawachi and Berkman, 2001). It is reasonable to assume that being able to engage in positive social behaviors contributes to the well-being of any social animal (Institute for Laboratory Animal Research, 1992), including rodents and rabbits.
Distress and fear due to confinement can also be ameliorated by increasing the complexity of the living space. Objects and structures that can be used as hiding places are particularly effective in enhancing the confined subjects' ability to cope with distress by taking refuge in a secluded place. This has been shown in rats and mice, both in single- and in group-housed animals. It remains to be investigated whether confinement distress can also be buffered by specific environmental modifications in the case of hamsters, gerbils, guinea pigs and rabbits.
Maladaptive behaviors are generally regarded as objective signs of inadequate housing conditions. Rodents and rabbits develop such behaviors even in groups kept in large, well-furnished enclosures. This underscores the fact that enforced permanent confinement is an intrinsic stressor for them.
Behavioral problems become particularly evident during the animals' activity phase, which in rodents and rabbits is the night. They may be overlooked completely during the day when the animals sleep. To assess the behavioral health of nocturnal animals, it is essential to conduct observations during the night without being a source of distraction or disturbance. Such observations are lacking in the literature, so it is very possible that behavioral pathologies are much more common in rodents and rabbits than has been usually assumed.
In rodents, the incidence and the frequency of maladaptive behaviors can be reduced but not eliminated by the provision of cover-providing structures and substrates. These environmental modifications are likely to increase the animals' sense of security and well-being, thereby decreasing their need to engage in bizarre behavior patterns that may help them cope with inner tension arising from inadequate living conditions.
In rabbits, maladaptive behaviors are best treated with companionship and/or provision of hay or straw. These additions to the environment offer species-appropriate distraction, thereby decreasing the time that the animals could spend engaging in maladaptive behaviors.
Preliminary evidence suggests that some maladaptive behaviors can be "imprinted" in the subject's neurophysiological system (Garner, 2005). It has been demonstrated in voles (Ödberg, 1987), pigs (Arellano, 1992), horses (McAfee et al., 2002), chickens (Jones et al., 2004), nonhuman primates (Chamove et al., 1984; Kessel and Brent, 1998), human primates (Christenson and Mansueto, 1999) and other species (Moon-Fanelli et al., 1999) that once established, these pathologies are extremely resistant to treatment. Species-appropriate environmental modifications may alleviate but not eradicate them, while pharmacological intervention may stop them temporarily.
It should be possible to create and test housing arrangements and husbandry practices for rodents and rabbits that are so species-appropriate that the animals have no reason to develop maladaptive behaviors while they are young, and to keep them under conditions that do not prompt maladaptive behaviors once they are mature. Only then will it be possible to investigate specific factors to which the animals have difficulties adapting or to which they cannot adjust at all.
Aggression among social partners and defensive aggression against handling personnel is a common problem that has received surprisingly little attention in the literature. The provision of hiding options from threatening partners decreases the risk of injurious fight wounds not only in mice and hamsters, but probably also in gerbils, guinea pigs, and perhaps also in rabbits. The aggression-mitigating effect of structures that allow individuals to break visual contact with opponents has also been documented in pigs and nonhuman primates (Erwin, 1977; Waran and Broom, 1993; Maninger et al., 1998; Westergaard et al., 1999).
That group-housed hamsters are less aggressive than singly caged hamsters when being handled by personnel has important practical implications that also deserve exploration in other rodents and rabbits.
The scientific documentation that social companionship and species-appropriate furniture enhance normal brain function and stimulate brain tissue integrity in rodents highlights the folly of not housing the animals in compatible social arrangements in complex primary enclosures. A social animal who is permanently kept alone in a barren environment is literally crippled both neurophysiologically and ethologically. It is questionable whether species-representative data can even be obtained from such a subject.
The cleaning of their cages is a variable that regularly affects rodents and rabbits to a varying extent. Groups of male mice tend to react to this disturbance not only with stress but with conspicuous outbursts of serious aggression. This is probably the reason that refinement attempts have been focused on mice, while rats, hamsters, gerbils, guinea pigs and rabbits have been largely ignored.
Transferring male mice into completely cleaned cages with mouse-odor-free structures for escape and some old nesting material is probably the best option for minimizing inter-male aggression and mitigating the stress associated with this husbandry procedure. It is important to find out if similar techniques can be applied to other rodents and in rabbits.
The stress resulting from transfer to an unfamiliar location and separation from cagemates is significant, but too little research has been conducted to find out whether this stress can be mitigated or even avoided. Transferring an animal to an unfamiliar location or separating an animal from cagemates is often not necessary. Many non-invasive procedures can be carried out in the subject's home cage, and if the subject lives in a group setting he/she can be separated in an extra cage that still allows continual communication with the familiar group. There is also the possibility of moving an animal along with one or several companions, who will serve as a stress buffer in the unfamiliar location where the experiment takes place. These options need to be explored more systematically. They are likely to refine experimental methodology by reducing or eliminating the anxiety and fear of the subject.
The few studies that address the possibility of habituation suggest that the animals may adapt to being picked up and gently handled, but not to being forcibly restrained.
There are three options available to buffer the stress associated with restraint:
An animal caged on the top shelf lives in an environment that is much higher and receives different illumination than one caged on the bottom shelf. It is difficult to understand why this variable is generally overlooked. The few observations published strongly indicate that shelf level does have an impact, probably not only in rats and mice, but also in multi-tier caged animals in general. Ignoring this variable, because more animals can be kept in a room when the cages are stacked on top of each than when they are arranged at the same level of the room, contravenes scientifically sound research methodology.
It is a legal requirement and professional recommendation that laboratories should provide uniformly distributed illumination of sufficient intensity to permit routine inspection and adequate housekeeping practices, including the bottom-most cages in a rack (National Research Council, 1996; United States Department of Agriculture, 2002). No advice is given as to how this important stipulation can actually be met. Rotating cage position relative to the light source (National Research Council, 1996) will rotate the two variables distance from light source and distance from floor between the subjects, but it does not address the real problem of minimizing or eliminating them.
Noise is another variable that is commonly overlooked, even though there is scientific evidence that noise associated with husbandry procedures can be a source of stress, and that exposure to loud noise can affect the neuroendocrine system. There is the possibility of controlling some noise with sound absorption panels, but this option has not yet been systematically implemented, nor has its effect on the animals' physiological systems been studied.
Rodents and rabbits can hear sound frequencies that are inaudible to humans but which may affect the animals nonetheless. This circumstance has to be taken into consideration in efforts to control the effect of noise, within and outside the human range of sound perception, on research data.
Back to Table of Contents