The Social Nature of European Rabbits (Oryctolagus cuniculus)

Abstract

Increasing concerns regarding the wellbeing of laboratory animals have caused biomedical research stakeholders to reconsider traditional housing of laboratory species and to provide social companionship for social species. European rabbits (Oryctolagus cuniculus) are commonly individually housed in research facilities despite the occurrence of social groups in the wild. Here we review the current literature to provide a comprehensive description of the social behaviors and preferences of rabbits in the wild and in captivity. The implications of these studies regarding social housing of laboratory rabbits are discussed.

Increasing concerns regarding the wellbeing of nonhuman animals and the publication of the 8th edition of the Guide for the Care and Use of Laboratory Animals³¹ have prompted the biomedical research community to reexamine long-held beliefs regarding animal housing in laboratories. The Guide’s new emphasis on providing social animals with companionship and the subsequent changes in regulator and accreditor focus have shifted the paradigm of many research institutions that historically housed social animals in single cages. Following widespread success with transitioning singly housed NHP into social pairs or groups in larger, more enriched cages and pens and the resulting dramatic improvement in wellbeing,¹⁶ the conversation has broadened to consider all species found in social groups in the wild but often housed singly in research facilities. However, not all species are social by nature, and forced group housing in captivity may have deleterious effects on the wellbeing of such animals. Therefore, stakeholders must consider which laboratory species might benefit from social housing in captivity before enacting widespread changes that might actually be counterproductive to improving laboratory animal wellbeing.

European rabbits (Oryctolagus cuniculus) are the third most widely used USDA-regulated laboratory animal species in the United States, with more than 20,000 animals reported in 2014.⁶⁹ As adults, rabbits are most commonly housed singly in research facilities, yet wild rabbits are often found in social groups.¹¹ This apparent contradiction has led many stakeholders to advocate shifting rabbits to social housing situations, whereas others believe that adult rabbits are too aggressive for social housing.⁷⁵ Interestingly, many of the same arguments against providing social interaction for adult rabbits, such as physical injuries as a result of aggressiveness, territoriality, and resource guarding, were also made by those against social housing for adult NHP, especially males.¹⁶ Those concerns did not materialize once managers accounted for the ethologic needs of the species.¹⁶ But is this experience likely to translate to rabbits? The gregarious nature of rabbits in the wild seemingly suggests that they, like NHP, would benefit from social companionship in the laboratory setting. However, anecdotal reports of aggression in rabbits, often without an obvious cause and resulting in serious injuries, during socialization attempts in the laboratory have prevented widespread social housing of the species.⁷⁵ The literature on the subject is relatively brief, considering the number of animals used in research facilities each year, and many of the published reports are subjective assessments rather than well-controlled studies that provide conclusive evidence regarding social housing benefits, risks, and methods for this species. In the current report, we review the available literature on the social preferences of rabbits in the wild and in captivity with the goal of characterizing the social needs of the species and the implications for housing in the laboratory. Importantly, the majority of the studies cited observed wild populations of O. cuniculus or captive populations of New Zealand white rabbits. It is possible—perhaps even likely—that the inbreeding of laboratory rabbits has resulted in significant behavioral changes, and stakeholders should consider differences in behavioral profiles between wild and captive-reared rabbits and between different breeds of rabbits used in research when reading this review.

Rabbits in the Wild and in Seminatural Conditions

In the wild, European rabbits often live in large groups in which a dominant buck might share a territory with several females and subordinate males around a multientranced burrow system, known as a warren.²⁵ This sociality is unique among rabbits and hares, and the European rabbit is the only leporid species known to form stable social groups.¹¹ Wild populations of this species divide themselves into social groups that might include multiple animals of the same sex.^61,65 For instance, one free-living rabbit population consisted of 11 to 14 breeding groups, with 89% of male and 96% of female rabbits living in a social group that contained at least one other adult of the same sex.¹⁰ A strict linear hierarchy develops in both sexes of these social groups.^49,73 In such groupings, the dominant buck routinely patrols a territory, and as often as daily, requires a submissive act from all other rabbits, both male and female, sharing the same space.³⁸ These submissive acts are manifested by the subordinate animals retreating or fleeing from the dominant buck's advances. Retreating is required to demonstrate submission, and any failure to retreat prompts an attack.³⁸ Whereas one author observed a minimal safety distance of 1 m between dominant and secondary bucks was required to demonstrate submission,³⁸ another observer reported males kept an average distance of 25 m between one another and never maintained close proximity during daily activities.⁷² In contrast, does spent 30% of their time at identical Cartesian coordinates as another doe during daytime activities in that same population.⁷²

Despite these observations, male rabbits apparently do have some preference for conspecific interaction. Although males have larger home ranges, there is extensive overlap, at least in space, with other rabbits of both sexes.¹⁴ One author concluded that a buck placed alone in a large pen grazed almost exclusively along the fence separating him from the rest of the colony, in order to graze as close as possible to conspecifics.³⁸ However, it is possible that the animal was actually patrolling the perimeter. Reportedly, male rabbits initiate many encounters with conspecifics, and these encounters are generally friendly to young and female rabbits but can be antagonistic toward other males.³⁶ Another recent study found that rabbits displayed no behavioral response to the scent of a neighbor but discriminated against the scent of a stranger, suggesting that they are not threatened by the presence of accustomed neighbors.⁴⁵

Some level of aggression apparently is normal in rabbit populations, given that aggression and dominance become evident in both males and females as they age. However, most observers have noted that, although aggressive chasing and submissive retreat remain common, overt fighting becomes rare after the order of dominance is established, as long as the group composition remains intact.^36,38,49 In one study population, the social order was much more affected by the removal of the top-ranking buck than the dominant doe, but fighting was otherwise rare in the stable group.⁴⁹ In contrast, intensive fighting resulting in serious injury persisted through the observation period in another population.⁷³ Importantly, although a rank order is apparent after 10 wk of age,³⁶ fights have not been observed before 12 wk,²⁷ and both sexes continue to engage frequently in affiliative behaviors (for example, cuddling up) until 17 wk of age.³⁶ Interestingly, aggression is not the cause of dispersion, given that aggressive interactions within natal groups do not increase at this time, even though dispersal tends to occur around 5 mo of age.³⁴ Even during the time of dispersal, male rabbits continue to have amicable interactions with one another.³⁴

Once dispersed, new animals to established groups are not generally welcomed by members of either sex, and dominant bucks attack new males.⁴⁹ If animals are not able to become established in a new group, they move to previously unoccupied territories or become nonterritorial satellites.³⁴ Rather than joining a group, dominated male rabbits may inhabit a buffer zone between the territories of dominant bucks, and they do not defend this forced home range.⁷¹ Although male rabbits have large home ranges, they extensively overlap with another,¹⁴ and a dominant male's home range decreases as populations increase, because additional animals require more attention to maintain dominance.³⁸ This pattern may suggest that even dominant bucks can tolerate the presence of other males. Females share a home range with one another and do not actively defend a territory,⁷¹ but numerous authors have observed that does fight as strongly as bucks (or even more so) for specific resources such as preferred nesting sites.^36,38,48,61 In one population, males were only aggressive toward one another when subordinate bucks approached too close to a dominant's mate.⁶¹ Similarly, in another study, male rabbits were not territorial unless groups of females were present.¹² In contrast, all adult animals demonstrated aggressive behavior toward one another, with no difference in frequency of aggressive behavior patterns between males and females in a large fenced-in area.³⁶ Interestingly, does may be aggressive toward males and drive them away if their advances are unwanted.⁶¹ However, male and female rabbits do establish pair-bonds with one another, and secondary bucks and does have been observed to remain in close association with a preferred partner for periods exceeding 1 y.³⁸

Despite the existence of these structured social groups, the benefits of sociality to rabbits are unclear, given that neither lifetime reproductive success in females nor increased survival in males is achieved by group-living.¹⁰ Sociality in rabbits may actually be more related to resource limitation than to an intrinsic social nature. Rabbits rely on access to cover as the primary means of predator escape, and range sizes are related to available cover.¹¹ In addition, the availability of suitable habitat for warren construction appears to control group sizes in at least some populations. For example, in light soils in which warrens can be constructed easily, tight rabbit groups were not observed.^47,54 In another study of several populations in different habitats, rabbits aggregated around burrows in habitats where cover was limited, but they dispersed in habitats where cover was readily available.⁴² Similarly, adult groups of rabbits did not form in the absence of multientranced warrens in another study of 2 rabbit populations.⁵⁶ In fact, does tended to space burrows evenly from one another rather than clump them, suggesting that they do not prefer to breed in close proximity to one another. Other authors observed that female rabbits do not share burrow systems when isolated burrows are available.¹² Similarly, bucks may share spatial ranges,⁶¹ but they avoid each other in time.⁵⁶ These observations suggest that although rabbits may accept the presence of conspecifics, they would evenly disperse if given the choice and optimal conditions.¹¹ However, dispersion did not occur in one study population in which new warrens were rarely constructed despite apparent overcrowding and the availability of suitable habitat nearby.⁵⁰ A study involving wild rabbits implanted with radiotelemeters found subordinate rabbits had a chronically elevated heart rate compared with dominant animals, thus suggesting that subordinates experience chronic stress.¹⁹ This observation is further supported by another study that found that rabbits of higher social rank have a much higher lifetime fitness than do subordinate animals.⁷³ In any case, free-range rabbits, both male and female, maintain an average interindividual distance of 15 to 20 m during daytime activities, suggesting that the animals, even when living in a social group, require large areas for recreation.⁷²

Arguments Against Single Housing of Rabbits

According to the Guide for the Care and Use of Laboratory Animals, animal housing should “provide sufficient space as well as supplementary structures and resources required to meet physical, physiologic, and behavioral needs (p 51).”³¹ Yet the Guide’s engineering standards for rabbit primary enclosures may be insufficient to meet this performance standard. For example, standard cages in which rabbits are typically singly housed do not allow species-typical behavior, such as hopping and rearing (Figure 1), and may contribute to poor health and wellbeing.^35,43,64 Rabbits singly housed in small cages can develop osteoporosis of the femur,³⁵ osseous hypoplasia,⁵⁸ and reduced bone strength,¹⁸ compared with those in large floor pens. Another study comparing floor types in individual and group settings found that rabbits housed individually have thinner tibial cortices than do those in housed in small groups, regardless of flooring type, and the authors hypothesized that increased activity accounted for the difference, although the basis for this hypothesis is unclear given that activity level was not measured.⁷ In this regard, another author found that doubling the housing space provided to an individually caged rabbit did not increase activity and actually resulted in increased abnormal behaviors, such as bar biting and paw scraping.³³ Furthermore, behavioral expression was limited in rabbits singly housed in 1.68-m² pens compared with those in 3.35-m² pens.¹⁷ Interestingly, the ‘small’ pen used in that study was 3 times larger than the minimal floor space currently required by the Guide. These observations suggest that large, tall pens are more appropriate for rabbits than are traditional laboratory cages, because pens allow rearing, stretching, hopping, and running behaviors that are not possible in single cages.^44,63

Figure 1.

(A) Caged and (B) penned rabbits engage in rearing behavior in response to a sound outside the housing room, but the caged rabbit is unable to become completely erect due to cage-height restrictions, even though the cage is compliant with all current regulations.

Stereopathies and other abnormal behaviors, both indicators of poor wellbeing, are associated with single housing of rabbits in cages.^24,25,37 For example, singly housed does spent more time engaged in abnormal behaviors, including sham digging, floor chewing, and bar biting, than did paired does.⁹ In contrast, rabbits in pens exhibited more species-typical behaviors , including rearing, stretching, hopping, and running.⁴⁴ Another study noted that the trichophagy and stereotypic behaviors reported by others^9,17,25,55 in singly housed does did not occur in group-housed does.²⁹ In addition, single rabbits in cages reportedly develop abnormal patterns of locomotion and resting.³⁹ For example, singly housed rabbits are generally less active but more restless than are their-group housed counterparts.⁵⁵ At least 2 studies have compared the incidence of stereotypies in single- and group-housed does and found that group-housed does displayed no such behaviors, whereas behaviors such as bar chewing or biting, trichophagy, and hair pulling were relatively common in the single-housed animals.^24,55 Importantly, stereopathies in rabbits typically occur during the quietest period of the day, so that unless additional observational methods such as videotaping are used, the prevalence of stereopathies in rabbits may be underreported.²⁵ A recent study found that, compared with paired and penned rabbits, rabbits housed individually in cages exhibited the highest fear levels and incomplete behavioral patterns, suggesting that they were in the most stressful of the 3 housing conditions tested.⁶⁷

Social Preferences of Rabbits in Captivity

The obvious detrimental effects of singly housing rabbits in cages coupled with the ability of wild animals to at least tolerate the presence of conspecifics suggest that rabbits might benefit from social housing in the laboratory. This hypothesis has been supported by several observational reports of captive rabbits. In a production setting, all rabbits under observation frequently ‘cuddled up’ to a conspecific until 120 d of age, and this behavior was independent of both age and sex.³⁶ When given the choice, does in a large fenced-in field spent as much as 90% of their daylight resting periods in physical contact,⁶² and does in laboratory pens spent 79% of the time in close proximity with others.²⁴ Another study found that paired rabbits in cages rested in body contact with one another for as much as 58% of their total resting time.⁶⁴ In addition, does worked almost as hard for limited social contact as they did for food, suggesting that a high value was placed on social interactions.⁶⁰ Similarly, bucks in cages preferred social interaction with a conspecific through perforated panels that allowed visual but only minimal physical contact.⁴¹ More recently, an anecdotal report suggested that unrelated 10- to 12-wk-old bucks can be successfully pair-housed for at least 3 to 4 wk when placed from a shipping carton immediately into a cage together and that these animals choose to remain in close proximity to one another.⁵⁹ Like does, unrelated male rabbits that are group housed as adults engage in affiliative behaviors, including resting in physical contact and allogrooming (Figures 2 and 3), although agonistic behavior also occurs.¹⁵ In our experience, paired adult bucks in cages similarly engage in affiliative behaviors including food sharing (Figure 4), but pairing adult animals directly in cages is typically unsuccessful.¹⁵

Figure 2.

Two adult male rabbits group housed in a pen rest in contact with one another, while a third animal explores the pen.

Figure 3.

Two adult male rabbits engage in allogrooming.

Figure 4.

A pair of caged adult male rabbits share a food hopper. This pair was initially group housed with 4 other rabbits as adults and then transferred to cages as a pair.

This evidence suggests that laboratory rabbits may prefer social companionship. However, does engage in agonistic behaviors to form a social hierarchy,⁹ and mixing adult does frequently results in high levels of injurious aggression.¹ Not surprisingly, subordinate does²¹ and bucks²⁰ had higher levels of cortisol than do more dominant animals, suggesting subordinates experience some degree of chronic stress. High-ranking female rabbits are more active than are low-ranking ones, suggesting that subordinate animals may be inhibited behaviorally by dominants.² In contrast, social rank has not been associated with changes in humoral immunocompetence in does, a common measure of chronic stress,⁶⁸ or differences in other physiologic and immunologic measurements.⁷⁴ Furthermore, pair housing of does reduces variance in growth and serum chemistry parameters, which could decrease animal use as a result of decreased interanimal variability.⁵³ Another study found that group housing of breeding does is associated with higher kit mortality and shorter doe lifespan, suggesting that continuous social housing of this group of animals is suboptimal.⁶⁶ Similarly, subordinate group-housed bucks in a large, outdoor fenced field had lower body masses, higher adrenocortical activities, and higher heart rates than did dominant animals,⁷³ mirroring findings in wild bucks.¹⁹ These endocrine and immune responses suggestive of chronic stress in socially housed rabbits conflict with behavioral evidence that animals sometimes choose social interactions.

Given the often contradicting findings, the apparent social preference of rabbits in some of the studies discussed may actually reflect a preference for the larger, more enriched cages that are usually used for group housing over the smaller, more barren cages typically used when animals are singly housed. For example, does showed a significant preference for a solitary pen when given the choice between 2 pens in which the only difference was the presence or absence of other does.²⁸ In a second experiment comparing an enriched, group pen with a smaller, more barren cage (resembling the current standard laboratory housing), all does, including the lower ranked ones, showed a strongly significant preference for the group pen.²⁸ These preferences may reflect the animals’ desire to perform different behaviors in the different environments, but it alternatively might indicate that the does, both high- and low-ranking, actually preferred solitary pens when appropriately enriched. In contrast, a modified traditional caging system consisting of 2 side-by-side cages joined by a special access port was used to pair house unrelated adult female rabbits.³⁰ No aggression was noted in the 8 pairs formed for that study, and the rabbits spent 88% of the observation intervals occupying the same cage.³⁰ A more recent study comparing individual and paired does in conventional cages similarly found that does pair-housed in cages developed fewer abnormal behaviors than did individually caged does suggesting that wellbeing is higher in socially housed animals.⁹ Additional studies are warranted in this area to more definitively assess rabbit preference for cage size.

Implications for Social Housing of Laboratory Rabbits

Despite some successes, many attempts at social housing rabbits have been unsuccessful leading the majority of stakeholders to relegate rabbits to single housing in research facilities. Unfamiliar rabbits commonly demonstrate aggressive behavior on introduction,⁴⁶ but these fights for rank order are a normal social response,¹ given that agonistic interactions are the basis for social hierarchy formation in rabbits.⁴⁹ To be acceptable, these interactions must result in a stable social group in which the wellbeing of all the animals is improved by the grouping. In other words, if rabbits are likely to continue fighting or subordinates are likely to experience chronic stress, single housing may be preferable for this species. Maintaining group stability is more complicated in rabbits, because any disruption of the established social order, as might occur when conspecifics are removed for experimental manipulations, is a potent trigger of severe agonistic behavior.⁴⁹ For example, even when the group composition was maintained, group housed does who had been separated engaged in agonistic interactions when returned to group housing, although returning animals to the original group resulted in fewer lesions than did forming new groups.⁴ However, adult bucks recovering together from pentobarbital anesthesia, but not halothane or methoxyflurane, did not attack one another as long as they remained pair-housed, suggesting that anesthesia might not incite aggression as long as all animals in the group recover simultaneously.⁸ The problem, however, is that even in apparently stable groups aggression often persists or unexpectedly escalates without any apparent provoking factor.^9,22,46 Because rabbits are crepuscular, a significant portion of social interactions occur at night,⁵² such that close daytime monitoring may fail to indicate potential problems before injury occurs. Even among does, as many as 20% of established compatible pairs need to be separated eventually, due to high levels of aggression.²² Numerous authors have suggested a variety of factors in successful group housing of rabbits, including cage size,⁷⁰ prior social experience,⁴⁶ age of the animals,⁴⁶ relatedness,¹³ and the presence of visual barriers,²² but only recently have authors begun to examine these potential contributors systematically.

Considering the social behaviors of the rabbit in the wild, especially their cues for dominance and submission, cage size may be a significant contributor to sustained aggression. Although domesticated rabbit species are used for biomedical research, they retain many of the behaviors of their wild counterparts⁶⁵. Observational studies of wild populations indicate that rabbits demonstrate submission by fleeing or retreating,³⁸ and the cages and pens currently used for social housing likely do not provide sufficient space for sufficient retreat and that the dominant animal attacks in response to this perceived challenge.³ For example, increasing the number of animals forced to live together, thus reducing the possible interindividual distance, increases fighting.⁵¹ In contrast, when a 5 × 4 m enclosure was used to group house 4 unrelated adult males (5 m² per rabbit), fighting was rare after the hierarchy was formed, an effect that the author attributed to the ability of the animals to actively escape due to the large size of the enclosure.²⁰ This hypothesis was supported by a recent study that found that a larger enclosure reduced aggression between unfamiliar does.⁷⁰ Interestingly, that study⁷⁰ also highlighted the importance of prior social experience, because pairs first tested in the small enclosure engaged in more aggressive behaviors during subsequent meetings in the larger enclosure than did pairs initially tested in the larger enclosure. Similarly, bucks socially housed first in pens (1.3 m² per rabbit) were less aggressive toward a partner from the group when a pair was transferred to a smaller cage (0.5 m² per rabbit) than when bucks were paired initially in cages.¹⁵ According to data from wild populations, male rabbits appear to require a distance of at least 1 m to demonstrate submission,³⁸ so it seems reasonable cages used for social housing should allow for at least that minimal interindividual distance. Floor space itself may not be as important as is the configuration of the space available, especially if escape distances are narrowed.⁵¹ For example, one author reported no aggression after group formation in square pens that maximized the animals’ ability to escape,²⁰ whereas others noted sustained aggression for 2 wk after group formation in more narrow, rectangular pens at which time the animals were separated.¹⁵ Still, floor space likely contributed to the observed sustained aggression, given that the pens in that study¹⁵ only provided 1.3 m² per rabbit compared with the 5 m² area in the study²⁰ that noted no aggression. More research is needed to determine the optimal floor space, density, and cage configuration for socially housed rabbits of both sexes.

Aspects other than cage size related to the primary enclosure may also contribute to the success or failure of social grouping. Grouping does in a novel pen without odor markings to prevent territorial aggression has been recommended,^43,46 but another study found no significant reduction in the number or duration of aggressive behavior in does grouped in a novel pen compared with adding animals to the home pen of an existing group;²³ in light of their observations, the authors hypothesized that dominance aggression is likely more important than is territorial aggression in does. However, male rabbits typically disperse to avoid conflict⁴⁹ and tend to disperse farther and have larger home ranges than do females,³⁴ suggesting that bucks may be more territorial than are does and that housing in a neutral space may be more important in bucks than does. Supporting that hypothesis, unrelated adult bucks reportedly begin to defend a territory when individually housed in an area for as little as a few hours,⁴⁰ and pair housing of bucks by removing a divider between 2 cages in which animals had been housed for only one hour resulted in almost immediate and persistent aggression that required separation to avoid injury.¹⁵ In another study, both does and bucks attacked an intruder animal placed in their home cages, and the home animal was almost always the winner, suggesting that neutral sites may be preferable to a known territory when initiating group housing of rabbits.⁵¹ In addition, visual and physical barriers with the pen may be useful to diffuse aggression²¹ (Figure 5). For example, hiding places helped to reduce, but did not eliminate, severe skin injuries in one study of does,⁵⁷ but the effect of such interventions has not been systematically evaluated in bucks.

Figure 5.

Two adult male rabbits use visual barriers to avoid aggression immediately after introduction. The dominant rabbit chased the subordinate animal around the pen until they rested on either side of the barrier, out of visual contact.

Factors related to the individual rabbits, such as age, relatedness, and prior social experience are also likely contributors to success with social housing. Maintaining littermates as social groups is a near universal recommendation for social housing.⁴⁶ If the need for animals can be forecast, it is possible to order pre-paired rabbits from a major vendor to increase likelihood of successful social housing.⁴⁰ Many authors have reported that they will not pair unrelated adult rabbits, given that high levels of aggression can occur.^1,40 In comparison, social housing rabbits prior to or as close to weaning as possible reportedly is more successful than is grouping adults, because fighting among rabbits has not been observed before 12 wk of age.²⁷ Because aggression precipitates as animals sexually mature, even in does,⁹ castration can be used as a tool to reduce aggression among males and strengthen the preference for affiliative social behaviors.^5,6,32 Relatedness has anecdotally been reported as important,⁴⁵ and one group reported that littermates that are never separated can live compatibly in pairs as adults, including intact males.³² In contrast, another researcher indicated that the rabbits did not have to be littermates as long as they were grouped before puberty.⁴² Group sizes of 4 to 8 animals have been recommended,^43,46 but the basis for this recommendation is unclear because wild rabbits tend to socialize in groups of 3 or fewer.⁷⁴

As far as procedures for actually forming groups, objective studies to determine best practices are unavailable, and most recommendations are based on anecdote and personal experience rather than true evidence. During group housing, neutral locations with multiple hide and escape opportunities are generally recommended.^6,26,40 Runs originally designed for large animals, or whole rooms when possible, may be ideal (Figure 6). The pen should have multiple sources of food and water as well to avoid conflict over resources.^26,71 Some authors have used a protected contact introduction period as an attempt to predict compatibility in full contact housing,^26,40 and rabbits may need to be introduced and separated daily over a period of up to 2 wk before a group is stable.^26,75 However, consistent with the observations of intense fighting after social instability in wild rabbits, daily separation has not been helpful in our experience, because the animals have to reestablish the hierarchy each day. Once groups are stable, manipulations and disruptions of the group should be minimized. Because removal of the dominant buck precipitates severe aggression, research strategies in which the subordinate animals are used first may contribute to ongoing success. As with all socially housed animals, socialized rabbits require close, continued monitoring to ensure all animals in the group remain compatible. Just as in other species, some individual rabbits may not be good social partners with one another, and alternative options, including single housing with protected physical or visual-only contact, should be sought when a group member is not thriving.⁴⁶

Figure 6.

Pen used for group housing, modified from a run used for large animals. Multiple food and water sources are present, as are enrichment items including wood shavings, visual barriers, and manipulable items.

Conclusions

Our understanding of the social preferences of rabbits in the wild remains incomplete, but resource limitation seems to be a major stimulus for group formation in this species. With adequate resources, being social does not appear to be a necessity for and may actually be avoided by wild rabbits. In contrast, single housing of rabbits in cages that do not allow normal exploratory and locomotor patterns such as rearing and hopping clearly is deleterious to rabbits’ wellbeing. However, social housing of this species is fraught with issues regarding wellbeing, especially for subordinate animals, which might experience chronic stress. Considering the behavioral patterns of rabbits, including cues for submission and dominance, rabbits appear to require large areas to obtain sufficient space to escape and retreat to avoid aggressive interactions. Much more research is needed to examine social preferences and behaviors and space needs in different housing configurations as well as optimal methods for pair or group formation in rabbits. These studies should be controlled, prospective studies. In addition, because the vast majority of the laboratory-based studies we referenced here describe New Zealand white rabbits only, future experiments should involve the various strains found in research institutions, given that different breeds likely will exhibit different behavioral patterns. In the meantime, the space restriction inherent in current rabbit caging may prohibit the expression of species-typical dominance–submission displays and thus prevent the establishment and maintenance of stable social groups. The wellbeing of laboratory rabbits might be improved through social housing of young, perhaps related and castrated, animals in very large, enriched pens; social housing of adult animals in typical cages is likely to result in injury and have a deleterious effect on wellbeing. When social housing of rabbits is pursued, frequent supervision of the animals and communication among all stakeholders is required to ensure that this practice does in fact improve the wellbeing of the animals.