Male management: coping with aggression problems in male laboratory mice

14
Male management: coping with aggression problems in male laboratory mice P. L. P. Van Loo 1 , L. F. M. Van Zutphen 1 & V. Baumans 1,2 1 Department of Laboratory Animal Science, Utrecht University, PO Box 80.166, 3508 TD Utrecht, The Netherlands and 2 Karolinska Institutet, 17177 Stockholm, Sweden Summary In a laboratory environment, aggressive interactions between male mice may exceed normal levels leading to negative effects both on the well-being of the animals and on the validity of experimental results. In this paper we review results from the literature and our own research with regard to coping with excessive aggressive behaviour in male laboratory mice. Based on this review practical recommendations concerning the housing and care of male laboratory mice are formulated. In short, it is recommended to avoid individual housing, to transfer odour cues from the nesting area during cage cleaning and to apply nesting material as environmental enrichment. Furthermore, group size should be optimized to three animals per cage. Further research, in particular into the effects of frequency, duration, type and severity of disturbances during an experiment on the degree of aggression, is recommended. Keywords Male mice; aggression; social housing; husbandry; group size; environmental enrichment Laboratory animals spend the vast majority of their life in their home cage. It is in the interest of both the animals and the researcher, that in this home cage an envir- onment is created that provides for the ani- mals’ physiological and ethological needs. The animals will then be better able to cope with stressful procedures, and experimental results will be more reliable. Legally, we are also obliged to adapt the environment as well as possible to the physiological and behavioural needs of laboratory animals (EC Directive 1986). For these reasons, research into housing conditions for laboratory animals that also take into account experi- mental, economic, and ergonomic precondi- tions is highly prioritized (Rodent Re nement Working Party 1998, Kornerup-Hansen 1999). The present standardized housing of laboratory animals often ful ls the needs of its inhabitants only marginally. As a result, the animals may behave abnormally and their well-being may be compromised to a great deal (Moberg & Mench 2000). One of the pro- blems often encountered in laboratory animal facilities is excessive aggression in group- housed male mice. Since mice comprise about 40% of all animals used in research (Ministry of Health, Welfare and Sports 2001), a solution to this problem would have great impact on a large number of animals. In this review we try to tackle the problem of excessive aggression in view of existing literature and our own experiments, and try to compose a set of recommendations for housing male mice in experimental settings balancing the behavioural and physiological needs of the animals with scienti c, eco- nomic and ergonomic demands. REVIEW ARTICLE Accepted 25 March 2003 # Laboratory Animals Ltd. Laboratory Animals (2003) 37, 300–313 Correspondence to: P. L. P. Van Loo E-mail: [email protected] by guest on May 29, 2015 lan.sagepub.com Downloaded from

Transcript of Male management: coping with aggression problems in male laboratory mice

Male management: coping with aggressionproblems in male laboratory mice

P. L. P. Van Loo1, L. F. M. Van Zutphen1 & V. Baumans1,2

1Department of Laboratory Animal Science, Utrecht University, PO Box 80.166, 3508 TD Utrecht,The Netherlands and 2Karolinska Institutet, 17177 Stockholm, Sweden

Summary

In a laboratory environment, aggressive interactions between male mice may exceed normallevels leading to negative effects both on the well-being of the animals and on the validityof experimental results. In this paper we review results from the literature and our ownresearch with regard to coping with excessive aggressive behaviour in male laboratory mice.Based on this review practical recommendations concerning the housing and care of malelaboratory mice are formulated. In short, it is recommended to avoid individual housing, totransfer odour cues from the nesting area during cage cleaning and to apply nesting materialas environmental enrichment. Furthermore, group size should be optimized to threeanimals per cage. Further research, in particular into the effects of frequency, duration, typeand severity of disturbances during an experiment on the degree of aggression, isrecommended.

Keywords Male mice; aggression; social housing; husbandry; group size; environmentalenrichment

Laboratory animals spend the vast majority oftheir life in their home cage. It is in theinterest of both the animals and theresearcher, that in this home cage an envir-onment is created that provides for the ani-mals’ physiological and ethological needs.The animals will then be better able to copewith stressful procedures, and experimentalresults will be more reliable. Legally, we arealso obliged to adapt the environment aswell as possible to the physiological andbehavioural needs of laboratory animals (ECDirective 1986). For these reasons, researchinto housing conditions for laboratoryanimals that also take into account experi-mental, economic, and ergonomic precondi-tions is highly prioritized (Rodent Re� nementWorking Party 1998, Kornerup-Hansen 1999).

The present standardized housing oflaboratory animals often ful� ls the needs of itsinhabitants only marginally. As a result, theanimals may behave abnormally and theirwell-being may be compromised to a greatdeal (Moberg & Mench 2000). One of the pro-blems often encountered in laboratory animalfacilities is excessive aggression in group-housed male mice. Since mice comprise about40% of all animals used in research (Ministryof Health, Welfare and Sports 2001), a solutionto this problem would have great impact on alarge number of animals.

In this review we try to tackle the problemof excessive aggression in view of existingliterature and our own experiments, and tryto compose a set of recommendations forhousing male mice in experimental settingsbalancing the behavioural and physiologicalneeds of the animals with scienti� c, eco-nomic and ergonomic demands.

REVIEW ARTICLE

Accepted 25 March 2003 # Laboratory Animals Ltd. Laboratory Animals (2003) 37, 300–313

Correspondence to: P. L. P. Van LooE-mail: [email protected]

by guest on May 29, 2015lan.sagepub.comDownloaded from

Aggression: bene� ciary ormaladaptive?

In their natural habitat, male mice (Musmusculus) live in despotic social groupsexisting of one dominant male, females withtheir progeny, and subordinate males. Thedominant male defends his territorycontaining resources such as females, foodand a nest site. Familiar subordinate malesare generally tolerated inside the territorialboundaries (Crowcroft 1966, Mackintosh1970, 1973, Hurst et al. 1993). Intermaleaggression within a socially stable groupmay be part of a broader behavioural strat-egy, in which all behaviour of an animal isaimed at a broader goal, namely to be � ttestfor a speci� c environment (Benus et al.1990a,b, 1991, Sluyter et al. 1995). In thewild, mice show rhythmic populationnumbers, going up to a certain optimumlevel, followed by a rapid decline and anincrease to the optimum again. It appearsthat when mice are living in small numbers,it is bene� cial to possess a ‘sturdy’ char-acter. Sturdy involves being aggressive andless � exible, showing routine-like beha-viour, which enhances the chance of survi-val. When population size increases,however, it is bene� cial to be more � exible.It appears that in the wild, mice of the more� exible type will migrate to unfamiliargrounds, and will be more successful in� nding and maintaining a new territory(Busser et al. 1974).

Although potentially damaging, aggres-sion can be regarded as having bene� ciaryeffects in the short run. In such situations,physiological and behavioural mechanismsat work within and between animals pre-vent aggression from escalating to levelsthat are actually damaging. When malemice are group-housed in the laboratory,a certain level of intermale aggression canthus be regarded as normal or natural(Bisazza 1981, Brain & Parmigiani 1990).However, aggression levels may be so highthat the welfare of the animals is jeo-pardized (Van Oortmerssen 1971, Bisazza1981, Brain & Parmigiani 1990; Fig 1).Several factors are important in contributingto this speci� c welfare problem.

Factors in� uencing aggressivebehaviour

Genetic background

Laboratory mice differ from wild mice intheir genetic composition. To achieve geneticstandardization, intensive selective breedinghas been used to create inbred strains of micewith minimal genetic variation. Further-more, breeding schemes have been selectedfor several morphological, physiological orbehavioural traits to create mouse models forhuman diseases. During the process ofinbreeding, several mouse strains havebecome highly aggressive, either as a sideeffect of selective breeding, or becauseaggressive behaviour was the main selectioncriterion (Bisazza 1982, Mondragon et al.1987, Guillot & Chapouthier 1998,Parmigiani et al. 1999). Odour cues, whichare important for kin recognition and adver-tisement of social status and essential for themediation of aggressive behaviour, appear tobe similar between families within inbredstrains of mice, which may hamper indivi-dual recognition (Nevison et al. 2000). Thusdisturbed social behaviour, such as excessiveaggression between group members, mayhave a genetic background.

Environmentally induced disturbedsocial behaviour

Laboratory mice that live in a barren con-� ned space such as a laboratory cage may beunable to respond to each other in a proper

Fig 1 Male mouse subjected to severe aggression ofa cage mate

Aggression in male mice 301

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

social way. Subordinate mice are unable to� ee from the dominant’s sight, or migrate outof the territory. When the proper behaviouralresponse is frustrated, the animal’s attemptto cope can be deemed to fail, causing a stateof suffering in the subordinate mouse. Fur-thermore, the dominant male may respondwith more extreme aggression than naturally,in an attempt to achieve the desired effect(i.e. disappearance of the subordinate). In pigskept at high stocking density, Ewbank andBryant (1972) showed that the close proxi-mity of the subordinate pig continued toelicit aggression from the dominant.

Frustration and lack of control andpredictability

In a laboratory environment behaviouralresponses may be frustrated, i.e. they aretriggered to be performed, but critical factorsto complete it are missing. This may, forexample, be the case when humans areapproaching the cage, and escape out of sightis not possible. In these instances, an animalmay direct its energy into aggression againstnearby animals (Broom & Johnson 1993). Inan experiment in which rats received electricfoot shocks but were unable to escape, therats’ escape behaviour was redirected intoaggression towards each other (Ulrich &Azrin 1962). Furthermore, unpredictablesituations such as experimental proceduresand routine husbandry procedures to whichthe animal is not able to anticipate mayoccur regularly in the laboratory. Aggressionis a common response to such unpredictablesituations (Broom & Johnson 1993). Cagecleaning, for example, often induces highlevels of aggression in male mice (Gray &Hurst 1995).

Severe aggression in group-housed malemice may thus be an indicator of poor wel-fare of the aggressor, although this need notnecessarily be the case in all situations. Inany case, severe aggression can certainly beregarded as affecting the psychological andphysiological well-being of the mice beingthe target of the aggression, either by frus-trating an appropriate behavioural responseor through pain and distress resulting frominjuries (National Research Council 1992).

Additionally, severe aggression may hamperthe validity of experimental results. Differentkinds of measures are taken to avoid severeaggression between male mice in thelaboratory.

Use of females

From strains known to be aggressive, pri-marily female mice can be used for experi-mentation. This is common practice nowand, mainly due to this gender preference,many male mice are euthanized beforeweaning (on average 65%, which can cumu-late to 80% for highly aggressive strains suchas FVB mice, Laboratory Animal ScienceAssociation 1998). This waste is highlyundesirable since the killing of animalswith no apparent purpose creates an ethicalproblem.

Use of docile strains

The use of males from a docile strain may, insome instances, be a good solution. However,mouse models should be suitable in the � rstplace for solving the scienti� c problem beingstudied which could be in con� ict with thechoice of a docile strain. A wealth of litera-ture has revealed that behavioural char-acteristics of an animal, such as high or lowaggressiveness, do not act alone, but are partof a variety of characteristics that form theanimal’s coping style (see Koolhaas et al.1997 for a review). Selecting a model on onecharacteristic may therefore bias the experi-mental outcome, render the experimentinvalid, and the use of animals morallyunjusti� able.

Separation of males or permanentindividual housing

Males can be separated when aggressionincreases to unacceptable levels. Also,directly after weaning, male mice may behoused individually. Individual housing isrecommended for several highly aggressivestrains such as Swiss=CD-1 and FVB (Com-mittee on Infectious Diseases of Mice andRats 1991, Mouse Genome Database 2001).However, the lack of opportunity to interact

302 Van Loo, Van Zutphen & Baumans

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

with conspeci� cs greatly in� uences thebehaviour and physiology of both mice andrats and has frequently been referred to as‘isolation stress’ or ‘isolation syndrome’. Itshould be noted at this point that ‘individualhousing’ is not the same as ‘isolation’. In thecase of individual housing, the animals canstill obtain visual, auditory and olfactorycues from conspeci� cs in adjacent cages. Theterm isolation should only be used in thosecases where animals lack all sensory cues ofconspeci� cs. Both terms, however, have beenused in the past to describe effects ofindividual housing. Animals that sufferfrom ‘isolation syndrome’ are more dif� cultto handle and become more aggressive,sometimes show stereotyped behaviour pat-terns and may suffer from convulsions. Fur-thermore, several studies indicate thatindividual housing induces reduction ofstress resistance and immunocompetence,higher tumour incidence, hypersensitivity totoxic agents and increased pathology such as‘scaly tail’ (Chance & Mackintosh 1962, Ader& Friedman 1964, Hatch et al. 1965, Barrett& Stockman 1966, Gartner 1968a,b, Baer1971, Brain 1975, Haseman et al. 1994).

It is evident that individually-housedanimals differ from socially-housed animalsin behavioural, neuro-endocrinological andneuro-physiological parameters, and as suchmay not be appropriate as a research model.However, it has also been proposed thatindividual housing may not be as stressful formale mice as it may seem. Comparisons ofsocially-housed dominant male mice andindividually-housed male mice revealed thatseveral neuroendocrine parameters indicat-ing sympathetic activity of the adrenal glandwere similar between these males(Maengwyn-Davies et al. 1973, Haemisch &Gartner 1996). The wild counterpart of mostlaboratory mice (Mus musculus) is territorial,and housing male mice individually may beregarded as providing them with their ownterritory (Brain 1975).

Evidence exists that both rats and miceprefer social to individual housing. In rats ithas been shown that animals that werereared in group-housed conditions and werelater individually housed would choose tofeed and sleep in close proximity with others

with maximal body contact rather than alone(Gartner 1968a,b). In the same review,Gartner concludes that individual housing ofrats will have many undesirable effects andshould be avoided if possible. Sherwin (1996)conducted a preference test in which labora-tory mice were willing to spend increasingamounts of energy to obtain access to certaintraits, one being companionship with thecompanion housed in a nearby cage.

In order to obtain further insight into whatthe animals would prefer, we investigated thesocial preferences of male BALB=c mice inrelation to the level of aggression,dominance hierarchy, age, familiarity andkinship (Van Loo et al. 2001b, andunpublished data; Fig 2). These preferencetests were conducted to further explore thequestion whether individual housing formale mice is acceptable or not. Resultsindicated that, in general, all mice preferredthe vicinity of a cage mate during restingperiods, especially when mice were familiarto each other (i.e. had been housed togetherpreviously). Social preference during activeperiods increased with age. During theseperiods, the mice clearly showed increasingappetitive behaviour: they regularly sniffed atthe openings in and alongside the partitionseparating them from each other, and dugaway the sawdust near it, as if attempting toestablish social contact. One might assumethat the inability to actually establish con-tact could in� uence the mice’s choice ofbehaviour. The mice were, however, able tocommunicate by sight, hearing, smell andnose contact. The hierarchy between twomale mice that are not able to be in bodilycontact does not cease to exist when closeolfactory and visual contacts are possible(Parmigiani et al. 1989, Hurst et al. 1993),and isolation effects on behaviour and phy-siology do not develop (Kudryavtseva 1991,1994, 2000, Kudryavtseva et al. 2000). Fur-thermore, by preventing access to the cagemate, not all social interactions can takeplace, i.e. consummatory responses affectingthe rewarding properties of social contactcannot interfere with the anticipatory beha-viour which may indicate the need for socialcontact (Spruijt 2001). We therefore assumethat the conducted tests provided valuable

Aggression in male mice 303

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

information on the preferences of the micethemselves for social or individual housing.

The general conclusion drawn from theseexperiments was that male mice prefer socialhousing to individual housing, independentof their level of aggression and social status.Together with the existing evidence of nega-tive side effects of individual housing on thewell-being of mice these results support therecommendation that individual housingshould be avoided whenever possible.

Individual housing may be required by theexperimental set-up, e.g. in metabolic cages.However, is expected excessive aggressionbetween male mice enough reason to makeindividual housing a necessity? It may beargued that excessive aggression in group-housed male mice provides a serious welfareproblem for both dominant and subordinatemice, and that individual housing will be lesscostly from a welfare perspective. On theother hand, both individual housing andgroup housing in combination with excessiveaggression have obvious negative con-sequences for the mice. These kinds ofcost–bene� t analyses can also be found infarm animal welfare research. In laying hens,the welfare bene� t of preventing cannibalismis weighed against the welfare cost of beaktrimming. To objectively make suchcost–bene� t analyses for animal welfare is an

impossible task, due to the complexity andinterdependence of aspects in� uencinganimal well-being (Appleby 1997, cf. Broom1997, Fraser 1997, Newberry & Estevez 1997).Since both alternatives may induce animalsuffering, why should we opt for either ofthese? Fox (1983=84) points out that there is alimit to the extent that we can adapt an ani-mal to a captive environment. It may there-fore be better to seek for possibilities to adaptthe environment to its inhabitants ratherthan the other way around.

Changes in housing and husbandryprocedures

Controversies

Both scienti� c studies and anecdotal evi-dence from animal caretakers have beenreported concerning the modi� cation ofhousing and husbandry procedures todecrease aggression in group-housed malemice. Many of these studies and reports,however, contradict each other. Partial cagecleaning or transfer of dirty sawdust to aclean cage has been reported to increaseaggression (Gray & Hurst 1995) but also todecrease aggression (McGregor et al. 1991).Others have investigated the availability ofspace (Poole & Morgan 1976, Vestal &

Fig 2 Preference test system. IC ˆ inhabited cage with a mouse behind a partition, ECˆ empty cage,M ˆ middle cage, T ˆ tunnel, D ˆ infrared detector, P ˆ perspex partition with holes

304 Van Loo, Van Zutphen & Baumans

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

Schnell 1986, McGregor & Ayling 1990) andgroup size (Welch & Welch 1966, Greenberg1972, Poole & Morgan 1973, Butler 1980,Barnard et al. 1994). In these studies too, bothincreases and decreases in aggression havebeen reported, and effects of group size andcage size seem to be interdependent. Most ofthe studies in this area, however, havefocused on the effects of environmentalenrichment on aggression, but again withvarying success. Several authors have indi-cated that environmental enrichment leadsto an increase in aggression when malelaboratory mice are housed together, andconclude that the enrichment may actuallybe unfavourable to their well-being in thisrespect (McGregor & Ayling 1990, Haemisch& Gartner 1994, Haemisch et al. 1994,Haemisch & Gartner 1997). Others havefound that cage enrichment or environmentalcomplexity does not alter, or decreases theamount of aggression between male mice(Vestal & Schnell 1986, Chamove 1989a,Ward et al. 1991, Armstrong et al. 1998,Ambrose & Morton 2000).

The identi� cation of factors explaining theapparent controversies in the studies men-tioned above is complicated by the widevariety of parameters used to quantifyaggression and by the differences in experi-mental design. Aggression levels are based onnumber of wounds, attack latency time,number or duration of � ghts or agonisticencounters. Furthermore, the degree offamiliarity between the mice varies fromnewly acquainted to being familiar frombirth, and test environments vary from novelto familiar.

In further experiments, we tested a selec-tion of adaptations to the housing and hus-bandry of male laboratory mice for itsmodulating properties to aggression andsocial stress. The experiments have beendescribed in original papers (Van Loo et al.2000, 2001a,b, 2002, 2003, 2004). Adaptationswere introduced in a variety of husbandryaspects that potentially in� uence intermaleaggression. Evidence for this was primarilybased on personal experience of animalcaretakers and researchers, and on the sci-enti� c reports mentioned above. These hus-bandry aspects were routine husbandry

procedures, group composition, cage size, andcomplexity of the cage. Apart from theiralleged aggression modulating properties, theadaptations were chosen for three other mainreasons. First, adaptations to laboratory ani-mal facilities should preferably be econom-ically and ergonomically feasible. Theadaptations chosen can all be implementedin the existing housing design, with minimalcosts with regard to material and personnel.Second, the success of changes in laboratoryanimal husbandry is directly related to themotivation of the animal care staff, whereimportant factors are extra workload andobvious bene� ts to animal welfare (Benn1995, Markowitz & Gavazzi 1995, Van deWeerd & Baumans 1995). With the choice ofadaptations to be tested and recommended tomodulate aggression, this has been kept inmind. Third, the aim was to provide a set ofhusbandry adaptations each capable ofdecreasing aggression and=or social stress,with a possible synergistic effect whencombined. The variety of aspects in themice’s environment that was tested were notinterdependent, which made a combinationof any set of adaptations possible.

Cage cleaning

Cage cleaning, although essential for hygiene,disrupts odour cues and stimulates activity,leading to an increase in aggression (Gray &Hurst 1995, Rodent Re� nement WorkingParty 1998). Aggression between male mice ismediated by social odour cues, and bothsource and distribution of odour cues appearto be important in this respect (Bishop &Chevins 1987, Hurst et al. 1993). Since micekeep their nests clean of urine and faeces(Blom et al. 1993), the nesting area will pri-marily contain pheromones originating fromplantar and other body glands. In contrast tomost urinary pheromones, these have anaggression-inhibiting effect (Mugford 1972,Jones & Nowell 1975, Stoddart 1980, Brown1985). The � rst experiment (Van Loo et al.2000) revealed that simple changes in routinecleaning procedures signi� cantly affected thelevel of aggression. The transfer of usednesting material from the old cage to the newcage clearly reduced aggression, while the

Aggression in male mice 305

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

transfer of sawdust soiled with urine and=orfaeces seemed to intensify aggression. Therecommendation resulting from this experi-ment is that material derived from the nestingarea should be transferred to the clean cage.

Group composition

Two aspects of group composition—kinshipand group size (Van Loo et al. 2000, 2001a)—were tested for aggression modulatingproperties. Evidence exists that socialinteractions between male mice can beaffected both by kinship (Kareem & Barnard1982, Hayashi & Kimura 1983, Kareem 1983,Kareem & Barnard 1986) and by group size(Poole & Morgan 1973, Butler 1980, Barnardet al. 1994). In our experiments, however,kinship did not affect the level of aggression.A possible explanation is that familiarizationbetween the non-littermates being housedtogether while they were still young, mayhave overruled effects of kin recognition(Kareem & Barnard 1982), or because kinrecognition in inbred strains such as theBALB=c strain may be disturbed byinbreeding (Nevison et al. 2000). Group sizeon the other hand, affected aggression to amarked degree. Groups composed of threemice showed the least aggression ascompared to groups of � ve or eight mice. Inthe latter, not only the dominant mouse, butalso the subordinate mice showed moreaggression than in groups of three mice. Thisseems to indicate that the dominancehierarchy is more stable in smaller groups.The housing of male mice in small groups istherefore recommended.

Cage size

For the effects of cage size on the level ofaggression, the literature reports contra-dictory results (Welch & Welch 1966, Poole& Morgan 1976, Vestal & Schnell 1986,McGregor & Ayling 1990, Jones 1991). In theexperiments described in Van Loo et al.(2001a), mice were housed in cages of twodifferent sizes: one providing the minimalamount of space per mouse that is legallyrequired, the other providing about 50% extraspace. The smaller cages appeared to invokeless aggression than the larger cages. These

results should be interpreted with care sincethe relationship between cage size andaggression may actually be curvilinear withan optimum (Polley et al. 1974). In differentspecies evidence exists that crowding mayinduce a reduction in aggression (Welch &Welch 1966, Ewbank & Bryant 1972, Hughes& Wood-Gush 1977). Crowding effectsreported in mice and rats are generallystress-related; these include increasedcorticosterone levels, impaired skinhomeostasis and diminished immuneresponse (Brown & Grundberg 1995,Csermely et al. 1995, Denda et al. 1998).

Environmental enrichment

Studies investigating the effect of cagecomplexity by the addition of environmentalenrichment on the well-being of laboratorymice are numerous (cf. Van de Weerd &Baumans 1995). In the majority of cases, theincrease of complexity or the addition ofenvironmental enrichment enhances thewell-being of the mice being studied. For malemice, however, several studies report anincrease in aggression when cages are morecomplex (McGregor & Ayling 1990,Haemisch & Gartner 1994), while othersreport no effect (Vestal & Schnell 1986) or adecrease in aggression (Chamove 1989a, Wardet al. 1991, Armstrong et al. 1998, Ambrose &Morton 2000). The effects of cage complexityon aggression were tested in an experiment(Van Loo et al. 2002) by adding two types ofenvironmental enrichment to the cage (Fig 3).One type of enrichment (shelter) was rigid,and structured the cage into speci� c areaswith properties meeting the mice’s needs, i.e.a dark and secluded sleeping area, an escaperoute, wire mesh to eat from and to use asvoiding area (Blom et al. 1993, Schlingmannet al. 1994, Van de Weerd & Baumans 1995,Baumans 2000). The other enrichment (nest-ing material) could be manipulated, and gavethe mice the opportunity to build a nest forsleeping and thermoregulation (Van de Weerdet al. 1997, 1998). Results revealed thataggression and stress-related parametersincreased when mice were housed with ashelter, while a signi� cant decrease inaggression was apparent when mice were

306 Van Loo, Van Zutphen & Baumans

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

housed with nesting material. Both enrich-ments provided for the already mentionedimportant features in the mice’s cage.Additionally, with the nesting material micecould actively structure their environmentat their own will. It has been argued thatactive behaviour may have a reward value initself, i.e. that an animal’s needs may lie asmuch in the ‘doing’ as in the ‘achieving’(Wemelsfelder 1997a). Items providing for thisneed, such as nesting material, would thus bemore suitable enrichment items than rigidstructures (Poole 1998). The recommendationresulting from this experiment is that malemice should be provided with nestingmaterial to actively structure their cage.

Long-term consequences

When combining the results from the latterexperiments, a reduction in aggression wasexpected to be most signi� cant in groups ofthree mice, provided with nesting materialthat was transferred after cage cleaning. Wehave also tested this combination ofhusbandry adaptations with provenaggression-reducing properties (Van Loo et al.2003, 2004). This experiment differed in twoimportant aspects from the previousexperiments, namely experimental designand choice of strain.

Experimental design

In the previous experiments, groups ofmice were alternately subjected to different

housing or husbandry procedures, with theexception of the experiment studying groupsize. The advantage of this experimental set-up was that groups of mice served as theirown control and even slight changes in theusually quite variable levels of aggressioncould be detected. The disadvantage was,however, that no long-term effects of changesin housing and husbandry procedures onaggression and stress could be examined.Long-term effects may differ from short-termeffects, since animals may habituate to theadapted situation or loose interest in intro-duced enrichment items after some time(Chamove 1989b, Bloomsmith et al. 1991,Van de Weerd & Baumans 1995). Therefore,husbandry procedures of experimental groupsand control groups of mice were not changedduring this experiment.

Choice of strain

For the previous experiments, BALB=c hadbeen the strain of choice. Males of this strainare known to be moderately aggressive andanimals may � ght � ercely. The main goal ofthis study was to � nd ways of preventingexcessive aggression arising in group-housedmale mice. In previous experiments, how-ever, levels of aggression hardly ever reachedlevels beyond what could be considered nor-mal. It may thus be argued that the animalmodel used was not the best model to studyenvironmental in� uences on excessiveaggression since such levels of aggressionnever arose. However, mechanisms

Fig 3 The two enrichment items tested: Kleenex tissues (left) and the Utrecht Shelter (right)

Aggression in male mice 307

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

underlying the development of excessiveaggression may not differ distinctly frommechanisms regulating normal levels ofaggression. Excessive aggression merely maybe a state in which the mechanisms toprevent normal levels of aggression toescalate, dysfunction. Environmental stimuliin� uencing normal levels of aggression maytherefore also in� uence the development ofexcessive aggression. Furthermore, foraggression modulating properties of husban-dry changes to be investigated, we needed tohouse male mice in groups for a relativelylong period. When using a highly aggressivestrain there would have been a distinctpossibility that males would have to beseparated early during the experiment due toexcessive aggression. Still, it was consideredof main importance to examine whether theaggression reducing adaptations to the hus-bandry procedures, as found in previousexperiments, would also decrease aggressionin strains known to be highly aggressive, tothe extent that group housing is optional.Therefore, the combination of presumedaggression reducing husbandry adaptationswere tested not only in the BALB=c strain,but also in a strain known to be highlyaggressive (Swiss-derived CD-1; Parmigianiet al. 1999).

Results of this experiment indicated thatlong-term enrichment with nesting materialand the repeated transfer of nesting materialwhen cleaning the cages did not affect inter-male aggression levels (Van Loo et al. 2003).Since aggression levels between groups werehighly variable, and groups of mice could notserve as their own control, a possible effecton aggression may have been too low todetect. Another explanation may be that thenovelty effect, rather than the differences inhusbandry procedures per se, were respon-sible for the aggression decreasing effectsfound earlier. The perception of ‘change’ or‘novelty’ may be regarded as importantfactors providing for variety in a captiveanimal’s environment (Wemelsfelder 1997b,Poole 1998). Novel objects stimulateinvestigative or manipulative behaviour thatdecreases over time as the animal habituatesto the presence of the object (Shepherdsonet al. 1998), and frequent change of objects to

avoid boredom has been recommended (Benn1995, Markowitz & Gavazzi 1995). Thealternate application of different housing andhusbandry procedures earlier may have sus-tained the effect of novelty. From the resultsof this experiment, it is not possible to � ndout whether detection levels were too low, orwhether aggression-inhibiting effectsdeclined after the mice became familiarizedwith their enriched environment. Stress-related parameters such as corticosteronelevels, thymus weight and food and waterconsumption, however, indicated reducedlevels of stress in enriched-housed conditions(Van Loo et al. 2004). This implies that thehusbandry changes as recommended in theprevious experiments may also havebene� ciary effects on the well-being of group-housed male mice in the long term.

Concluding considerations

An interesting phenomenon is that overallaggression levels were low in all experi-ments, both in BALB=c mice and in the Swissderived CD-1 mice. In the experiment testingpreference of male mice for social housing inwhich mice were housed in pairs, the highestlevels of aggression were observed, and sev-eral pairs of mice needed to be separated oreuthanized. In the experiment investigatinggroup size, in several groups of � ve and eightmice, animals were badly wounded as well.In the � nal experiment, with a group size ofthree mice, only four out of 20 CD-1 groupswere moderately wounded. In three of thelatter, aggression returned to acceptablelevels after several weeks. In the other groupsof Swiss:CD-1 mice, all animals lived inharmonious groups. This is remarkable,since the general presumption is that adultSwiss males will kill each other when grouphoused (Bisazza 1981). The experiments inwhich aggression remained withinacceptable levels differed in two aspects fromthe experiments in which aggression levelsreached excessive levels and studies of othersreporting high levels of aggression (cf.Parmigiani et al. 1999), namely group sizeand nature of disturbances.

308 Van Loo, Van Zutphen & Baumans

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

Group size

In general, mice were housed in groups ofthree. As mentioned before, housing malemice in groups of three clearly invokes lessaggression than in groups of � ve or eightmice, since the hierarchy in the latter groupsappears to be less stable. The combinedresults of the social preference experimentand the experiment on group size imply thathousing in groups of three would also invokeless aggression than pair housing. Evidencefor this has also been provided by Bisazza(1981), who observed severe � ghting in pairsof Swiss mice, while these mice had peace-fully shared nests previously when they werehoused in groups of three. When male miceare housed in pairs, the subordinate male isthe only animal at which aggression can bedirected. In groups of three mice, aggressioncan be directed at two subordinates. Thesesubordinates may � nd social support witheach other that may reduce the level of socialstress they experience. Social support afterencountering a (socially) stressful situationappears to be essential for successful copingin animals as well as in humans (Medalie1985, Cowie 2000). Ruis et al. (1999) showedthat male rats that had been subjected tosocial defeat without post-defeat social sup-port developed depression-like behaviouraland physiological symptoms, while rats thatwere group-housed after social defeat, werecoping well. Keeney and Hogg (1999) usedsocially-defeated mice that lived individuallywhile the unfamiliar mouse that defeatedthem lived opposite behind bars, as an animalmodel of depression. Mice that are sociallyhoused immediately after surgery appear torecover quicker than mice that arehoused individually (Baumans, personalcommunication).

Nature of disturbances

In most experiments described in this review,animals were subjected to routine handlingsuch as cage cleaning, weighing and a visualdaily check, and in some instances, a rela-tively short disturbance e.g. to obtain urinesamples. These disturbances may be regardedas relatively low-stress situations. In theexperiment on social preferences, however,

pairs of mice were removed from their homecage and physically separated for a period of48 h on three occasions. Evidence exists thatintermale � ghting in mice increases withprogressive isolation, and that effects onsocial behaviour and aggression in both miceand rats already become apparent after aperiod of 24 h of isolation (Goldsmith et al.1978, Cairns et al. 1985, Varlinskaya et al.1999). Although the mice used in thisexperiment were not isolated completely,they were unable to perform normal socialbehaviour during the periods of separation. Ingeneral, more than 40% of the experimentsconducted with mice in The Netherlandsinvolve stressful procedures, such as indu-cing traumatic physical, chemical or psy-chological stimuli, in� ammations orinfections, subjection to radiation, etc. Thedegree of discomfort for these mice is cate-gorized as moderate to extremely severe(Ministry of Health, Welfare and Sport 2000).Mild disturbances such as cage cleaning,already induce an increase in aggressionbetween male mice (Gray & Hurst 1995, VanLoo et al. 2000, 2001a, 2002), and exposure toa repeated or even a single extreme stressormay lead to excessive aggression (Pant &Nath 1993). Once excessive aggression hasarisen, it may not return to normal levels soeasily. Therefore, re� nement, in particularminimizing the number and duration ofstressful procedures not only directly affectsthe well-being of the animals, but alsoindirectly, by reducing the chance of trigger-ing excessive aggression.

Recommendations for the housing andcare of male mice in laboratories

Based on the information presented in thisreview, the following recommendations canbe put forward:

(1) Individual housing of male mice shouldbe avoided whenever possible.

(2) When cleaning the mice’s cages, materialfrom the nesting area should be trans-ferred to the clean cage. These materialsseem to contain aggression-reducingodour cues. Transferring bedding or othermaterials contaminated with urine

Aggression in male mice 309

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

and=or faeces should be avoided, sincethese may contain aggression-elicitingodour cues.

(3) Male mice can best be housed in groups ofthree animals. Larger group sizes maydecrease the chance of a stable hierarchydeveloping, while pair housing mayincrease aggression, and lack of socialcomfort may induce depression-likesymptoms in the subordinate.

(4) When applying environmental enrich-ment, nesting material would be theenrichment of choice. It provides theopportunity to indulge in species-speci� cbehaviour. When � rst introduced, itdecreases intermale aggression, and, inthe long term, it may help the animals tobetter cope with stressful situations.

(5) Disturbances during an experimentshould be limited as far as practicallypossible since they may create situationsin which excessive aggression can arise.Further research is necessary to investi-gate if, and how frequency, duration, typeand severity of disturbances in� uence thedevelopment of excessive aggression.

References

Ader R, Friedman SB (1964) Social factors affectingemotionality and resistance to disease in animals:IV. Differential housing, emotionality, and Walker256 carcinosarcoma in the rat. PsychologicalReports 15, 535–41

Ambrose N, Morton DB (2000) The use of cageenrichment to reduce male mouse aggression.Journal of Applied Animal Welfare Science 3,117–25

Appleby MC (1997) Life in a variable world: beha-viour, welfare and environmental design. AppliedAnimal Behaviour Science 54, 1–19

Armstrong KR, Clark TR, Peterson MR (1998) Use ofcorn-husk nesting material to reduce aggression incaged mice. Contemporary Topics in LaboratoryAnimal Sciences 37, 64–6

Baer H (1971) Long-term isolation stress and itseffects on drug response in rodents. LaboratoryAnimal Science 21, 341–9

Barnard CJ, Behnke JM, Sewell J (1994) Socialbehaviour and susceptibility to infection in housemice (Mus musculus): effects of group size,aggressive behaviour and status-related hormonalresponses prior to infection on resistance toBabesia microti. Parasitology 108, 487–96

Barrett AM, Stockman MA (1966) One or manyanimals in a cage? Nutrition Reviews 24, 116–19

Baumans V (2000) Environmental enrichment: a rightfor rodents! In: Progress in the Reduction, Re�ne-ment and Replacement of Animal Experimenta-tion (Balls M, Van Zeller EM, Halder ME, eds).Amsterdam: Elsevier, pp 1251–5

Benn DM (1995) Innovations in research animal care.Journal of the American Veterinary MedicalAssociation 206, 465–8

Benus RF, Bohus B, Koolhaas JM, Van Oortmerssen GA(1990a) Behavioural strategies of aggressive andnon-aggressivemale mice in response to inescapableshock. Behavioural Processes 21, 127–41

Benus RF, Den Daas S, Koolhaas JM, VanOortmerssen GA (1990b) Routine formation and� exibility in social and non-social behaviour ofaggressive and non-aggressive male mice. Beha-viour 112, 176–93

Benus RF, Bohus B, Koolhaas JM, Van Oortmerssen GA(1991) Heritable variation for aggression as a re� ec-tion of individual coping strategies. Experientia 47,1008–19

Bisazza A (1981) Social organization and territorialbehaviour in three strains of mice. BollettinoZoologica 48, 157–67

Bisazza A (1982) Hereditary differences in socialbehaviour of male mice (Mus musculus L.).Bollettino Zoologica 49, 207–11

Bishop MJ, Chevins PFD (1987) Urine odours andmarking patterns in territorial laboratory mice(Mus musculus). Behavioural Processes 15, 233–48

Blom HJM, Witkamp ACP, Schlingmann F,Hoogervorst MJC, Van de Weerd HA, Baumans V,Beynen AC (1993) Demonstration of preference forclean versus soiled cages as expressed by laboratorymice. In: Evaluation of Housing Conditions forLaboratory Mice and Rats (Blom H, PhD thesis).Febodruk, Enschede, pp 81–99

Bloomsmith MA, Brent LY, Schapiro SJ (1991)Guidelines for developing and managing anenvironmental enrichment program for nonhumanprimates. Laboratory Animal Science 41, 372–7

Brain PF (1975) What does individual housing mean toa mouse? Life Sciences 16, 187–200

Brain PF, Parmigiani S (1990) Variation in aggressive-ness in house mouse populations. BiologicalJournal of the Linnean Society 41, 257–69

Broom DM, Johnson KG (1993) Stress and AnimalWelfare. London: Chapman & Hall

Broom DM (1997) Welfare evaluation. AppliedAnimal Behaviour Science 54, 21–3

Brown RE (1985) The rodents II: suborderMyomorpha. In: Social Odours in Mammals,Vol. 1 (Brown RE, Macdonald DW, eds). Oxford:Clarendon Press, pp 345–428

Brown KJ, Grundberg NE (1995) Effects of housing onmale and female rats: crowding stresses male butcalm females. Physiology & Behavior 58, 1085–9

310 Van Loo, Van Zutphen & Baumans

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

Busser J, Zweep A, Van Oortmerssen GA (1974)Variability in the aggressive behaviour of Musmusculus domesticus, its possible role in popula-tion structure. In: The Genetics of Behaviour(Van Abeelen JNF, ed). Amsterdam: North-Holland,pp 185–99

Butler RG (1980) Population size, social behaviour,and dispersal in house mice: a quantitativeinvestigation. Animal Behaviour 28, 78–85

Cairns RB, Hood KE, Midlam J (1985) On � ghting inmice: is there a sensitive period for isolationeffects? Animal Behaviour 33, 166–80

Chamove AS (1989a) Cage design reduces emotion-ality in mice. Laboratory Animals 23, 215–19

Chamove AS (1989b) Environmental enrichment: areview. Animal Technology 40, 155–78

Chance MRA, Mackintosh JH (1962) The effects ofcaging. In: Collected Papers Vol. 11: The Environ-ment of Laboratory Animals (Laboratory AnimalsCentre, MRC Laboratories). Surrey: Carshalton,pp 59–64

Committee on Infectious Diseases of Mice and Rats(1991) Noninfectious skin conditions important indifferential diagnosis. In: Infectious Diseases ofMice and Rats. Washington DC: NationalAcademic Press, pp 195–7

Cowie H (2000) Bystanding or standing by: genderissues in coping with bullying in English schools.Aggressive Behavior 26, 85–97

Crowcroft P (1966) Mice all Over. London: FoulisCsermely P, Penzes I, Toth S (1995) Chronic

overcrowding decreases cytoplasmic free calciumlevels in T lymphocytes of aged CBA=CA mice.Experientia 51, 976–9

Denda M, Tsuchiya T, Hosoi J, Koyama J (1998)Immobilization-induced and crowdedenvironment-induced stress delay barrier recoveryin murine skin. British Journal of Dermatology138, 780–5

European Council Directive (1986) Directive on theapproximation of laws, regulations and adminis-trative provisions of the Member States regardingthe Protection of Vertebrate Animals used forExperimental and other Scienti�c Purposes(86=609=EEC)

Ewbank R, Bryant MJ (1972) Aggressive behaviouramongst groups of domesticated pigs kept atvarious stocking rates. Animal Behaviour 20, 21–8

Fox MW (1983=84) Animal freedom and well-being:want or need? Applied Animal Ethology 11, 205–9

Fraser D (1997) Science in a value-laden world:keeping our thinking straight. Applied AnimalBehaviour Science 54, 29–32

Gartner K (1968a) Sammelreferat A: Zur Soziologieder Laboratoriumsratten, physiologischePsychologie der Gruppen- und Einzelhaltung.Deutsche Tierarztliche Wochenschrift 2, 45–8

Gartner K (1968b) Sammelreferat B: Zur Soziologieder Laboratoriumsratten, physiologische

Psychologie der Gruppen- und Einzelhaltung.Deutsche Tierarztliche Wochenschrift 4, 97–100

Goldsmith JF, Brain PF, Benton D (1978) Effects of theduration of individual or group housing onbehavioural and adrenocortical reactivity in malemice. Physiology & Behavior 21, 757–60

Gray S, Hurst JL (1995) The effects of cage cleaning onaggression within groups of male laboratory mice.Animal Behaviour 49, 821–6

Greenberg G (1972) The effects of ambienttemperature and population density on aggressionin two inbred strains of mice, Mus musculus.Behaviour 42, 119–30

Guillot PV, Chapouthier G (1998) Intermale aggres-sion, GAD activity in the olfactory bulbs and Ychromosome effect in seven inbred mouse strains.Behavioural Brain Research 90, 203–6

Haemisch A, Gartner K (1994) The cage design affectsintermale aggression in small groups of malelaboratory mice: strain speci�c consequences onsocial organization, and endocrine activations intwo inbred strains (DBA=2J and CBA=J). Journal ofExperimental Animal Science 36, 101–16

Haemisch A, Voss T, Gartner K (1994) Effects ofenvironmental enrichment on aggressive behavior,dominance hierarchies, and endocrine states inmale DBA=2J mice. Physiology & Behavior 56,1041–8

Haemisch A, Gartner K (1996) Dissociation betweenadrenal tyrosinehydroxylase and phenylethanola-mine-N-methyltransferase activities followingrepeated experience of defeats in individuallyhoused male DBA=2J mice. Physiology & Behavior59, 1117–22

Haemisch A, Gartner K (1997) Effects of cage enrich-ment on territorial aggression and stress physiologyin male laboratory mice. Acta PhysiologicaScandinavica 161(suppl. 640), 73–6

Haseman JK, Bourbina J, Eustis SL (1994) Effect ofindividual housing and other experimental designfactors on tumor incidence in B6C3F1 mice.Fundamental and Applied Toxicology 23, 44–52

Hatch AM, Wiberg GS, Zawidzka Z, Cann M,Airth JM, Grice HC (1965) Isolation syndrome inthe rat. Toxicology and Applied Pharmacology 7,737–45

Hayashi S, Kimura T (1983) Degree of kinship as afactor regulating preferences among conspeci�cs inmice. Animal Behaviour 31, 81–5

Hughes BO, Wood-Gush DGM (1977) Agonisticbehaviour in domestic hens: the in� uence ofhousing method and group size. Animal Behaviour25, 1056–62

Hurst JL, Fang J, Barnard CJ (1993) The role ofsubstrate odours in maintaining social tolerancebetween male house mice, Mus musculusdomesticus. Animal Behaviour 45, 997–1006

Jones RB (1991) Reply to McGregor and Ayling 1991.Applied Animal Behaviour Science 27, 295–9

Aggression in male mice 311

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

Jones RB, Nowell NW (1975) Effects of clean andsoiled sawdust substrates and of different urinetypes upon aggressive behavior in male mice.Aggressive Behavior 1, 111–21

Kareem AM, Barnard CJ (1982) The importance ofkinship and familiarity in social interactionsbetween mice. Animal Behaviour 30, 594–601

Kareem AM (1983) Effect of increasing periods offamiliarity on social interactions between malesibling mice. Animal Behaviour 31, 919–26

Kareem AM, Barnard CJ (1986) Kin recognition inmice: age, sex and parental effects. AnimalBehaviour 34, 1814–24

Keeney AJ, Hogg S (1999) Behavioural consequencesof repeated social defeat in the mouse: preli-minary evaluation of a potential animal modelof depression. Behavioural Pharmacology 10,753–64

Koolhaas JM, De Boer SF, Bohus B (1997) Motivationalsystems or motivational states: behavioural andphysiological evidence. Applied Animal BehaviourScience 53, 131–43

Kornerup-Hansen A (1999) Report of the COEworking group on rodents and rabbits. In: Revistade Ciencia 23–24, Abstracts of Scienti�c Papers ofthe ICLAS-FELASA conference, May 1999, Palmade Mallorca, p 88

Kudryavtseva NN (1991) A sensory contact model forthe study of aggressive and submissive behavior inmale mice. Aggressive Behavior 17, 285–91

Kudryavtseva NN (1994) Experience of defeatdecreases the behavioural reactivity to conspeci�csin the partition test. Behavioural Processes 32,297–304

Kudryavtseva NN (2000) An experimental approachto the study of learned aggression. AggressiveBehavior 26, 241–56

Kudryavtseva NN, Bondar NP, Alekseyenko OV(2000) Behavioral correlates of learned aggression inmale mice. Aggressive Behavior 26, 386–400

Laboratory Animal Science Association (1998) TheProduction and Disposition of Laboratory RodentsSurplus to the Requirements for Scienti�cProcedures. Report of a LASA Task Force Meeting,12 June 1998

Mackintosh JH (1970) Territory formation by labora-tory mice. Animal Behaviour 18, 177–83

Mackintosh JH (1973) Factors affecting the recogni-tion of territory boundaries by mice (Musmusculus). Animal Behaviour 21, 464–70

Maengwyn-Davies GD, Johnson DG, Thoa NB,Weise VK, Kopin IJ (1973) In� uence of isolation and� ghting on adrenal tyrosine hydroxylase andphenylethanolamine-N-methyltransferase activi-ties in three strains of mice. Psychopharmacology28, 339–50

Markowitz H, Gavazzi A (1995) Eleven principles forimproving the quality of captive animal life. LabAnimal 24, 30–3

McGregor PK, Ayling SJ (1990) Varied cages result inmore aggression in male CFLP mice. AppliedAnimal Behaviour Science 26, 277–81

McGregor PK, Barnard C, Hurst JL (1991) Reply toJones RB 1991, Varied cages and aggression.Applied Animal Behavioural Science 27, 297–9

Medalie JH (1985) Stress, social support, copingand adjustment. Journal of Family Practice 20,533–5

Ministry of Health, Welfare and Sport (2000)Zodoende 1999, Yearbook on Animal Experimen-tation in The Netherlands, Tables 7–10

Ministry of Health, Welfare and Sport (2001) AnimalExperimentation in The Netherlands. Documen-tation 15, The Netherlands, March 2001

Moberg GP, Mench JA (2000) The Biology of AnimalStress. Basic Principles and Implications forAnimal Welfare. Wallingford: CABI Publishing

Mondragon R, Mayagoitia L, Lopez-Lujan A, DWaz J(1987) Social structure features in three inbredstrains of mice, C57Bl=6J, Balb=cj, and NIH:a comparative study. Behavioral and NeuralBiology 47, 384–91

Mouse Genome Database (MGD) (2001) MouseGenome Informatics Web Site. The JacksonLaboratory, Bar Harbor, Maine. (URL: http://www.informatics.jax.org/), January 2001

Mugford RA (1972) Intermale � ghting affected byhome-cage odors of male and female mice. Journalof Comparative and Physiological Psychology 84,289–95

National Research Council (1992) Recognition andAlleviation of Pain and Distress in LaboratoryAnimals. Washington DC: National AcademyPress

Nevison CM, Barnard CJ, Beynon RL, Hurst JL (2000)The consequences of inbreeding for recognizingcompetitors. Proceedings of the Royal SocietyLondon, Series B 267, 687–94

Newberry RC, Estevez I (1997) A dynamic approach tothe study of environmental enrichment and animalwelfare. Applied Animal Behaviour Science 54,53–7

Pant KK, Nath C (1993) Dopaminergic involvement inthe effects of piracetam on foot shock inducedaggression in mice. Indian Journal of MedicalResearch 98, 155–9

Parmigiani S, Mainardi M, Brain PF, Haug M,Brunoni V (1989) Variation in aggressive behaviorand anatomo-physiological correlates generated bycrowding without physical contact in the housemouse. Aggressive Behavior 15, 191–200

Parmigiani S, Palanza P, Rodgers J, Ferrari PF (1999)Selection, evolution of behavior and animal modelsin behavioral neuroscience. Neuroscience andBiobehavioral Reviews 23, 957–70

Polley CR, Craig JV, Bhagwat AL (1974) Crowding andagonistic behavior: a curvilinear relationship?Poultry Science 53, 1621–3

312 Van Loo, Van Zutphen & Baumans

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from

Poole TB, Morgan HDR (1973) Differences inaggressive behaviour between male mice (MusMusculus L.) in colonies of different sizes.Animal Behaviour 21, 788–95

Poole TB, Morgan HDR (1976) Social and territorialbehaviour of laboratory mice (Mus Musculus L.) insmall complex areas. Animal Behaviour 24, 476–80

Poole TB (1998) A systematic approach to environ-mental enrichment using the ‘scan’ system.Animal Technology 49, 7–17

Rodent Re� nement Working Party (1998) Re� ningrodent husbandry: the mouse. Laboratory Animals32, 233–59

Ruis MA, ten Brake JH, Buwalda B, De Boer SF,Meerlo P, Korte SM, Blokhuis HJ, Koolhaas JM(1999) Housing familiar male wildtype rats togetherreduces the long term adverse behavioural andphysiological effects of social defeat. Psycho-neuroendocrinology 24, 285–300

Schlingmann F, Van de Weerd HA, Blom HJM,Baumans V, Van Zutphen LFM (1994) Behaviouraldifferentiation of mice housed on different cage� oors. In: Welfare and Science, Proceedings of theFifth FELASA Symposium, Brighton 1993, pp 355–7

Shepherdson DJ, Mellen JD, Hutchins M (eds) (1998)Second Nature, Environmental Enrichment forCaptive Animals. London: Smithsonian InstitutionPress

Sherwin C (1996) Laboratory mice persist in gainingaccess to resources: a method of assessing theimportance of environmental features. AppliedAnimal Behaviour Science 48, 203–14

Sluyter F, Bult A, Lynch CB, Van Oortmerssen GA,Koolhaas JM (1995) A comparison between housemouse lines selected for attack latency or nest-building: evidence for a genetic basis of alternativebehavioral strategies. Behavior Genetics 25, 247–52

Spruijt BM (2001) How the hierarchical organizationof the brain and increasing cognitive abilities mayresult in consciousness. Animal Welfare 10,S77–S87

Stoddart DM (1980) Dispersion and social integration.In: The Ecology of Vertebrate Olfaction. London:Chapman and Hall, pp 141–60

Ulrich RE, Azrin NH (1962) Re� exive � ghting inresponse to aversive stimulation. Journal ofExperimental Analysis of Behaviour 5, 511–20

Van de Weerd HA, Baumans V (1995) Environmentalenrichment in rodents. In: Environmental Enrich-ment Information Resources for LaboratoryAnimals. AWIC Resource Series, Vol. 2, pp 145–9

Van de Weerd HA, Van Loo PLP, Van Zutphen LFM,Koolhaas JM, Baumans V (1997) Preferences fornesting material as environmental enrichment forlaboratory mice. Laboratory Animals 31, 133–43

Van de Weerd HA, Van Loo PLP, Van Zutphen LFM,Koolhaas JM, Baumans V (1998) Strength ofpreference for nesting material as environmental

enrichment for laboratory mice. Applied AnimalBehaviour Science 55, 369–82

Van Loo PLP, Kruitwagen CLJJ, Van Zutphen LFM,Koolhaas JM, Baumans V (2000) Modulation ofaggression in male mice: in� uence of cage cleaningregime and scent marks. Animal Welfare 9, 281–95

Van Loo PLP, De Groot AC, Van Zutphen LFM,Baumans V (2001a) Do male mice prefer or avoideach other’s company? In� uence of hierarchy,kinship and familiarity. Journal of Applied AnimalWelfare Science 4, 91–103

Van Loo PLP, Mol JA, Koolhaas JM, Van Zutphen LFM,Baumans V (2001b) Modulation of aggression inmale mice: in� uence of group size and cage size.Physiology & Behavior 72, 675–83

Van Loo PLP, Kruitwagen CLJJ, Koolhaas JM,Van de Weerd HA, Van Zutphen LFM, Baumans V(2002) In� uence of cage enrichment on aggressivebehaviour and physiological parameters in malemice. Applied Animal Behaviour Science 76, 65–81

Van Loo PLP, Van der Meer E, Kruitwagen CLJJ,Koolhaas JM, Van Zutphen LFM, Baumans V (2003)Strain-speci� c aggressive behaviour of male micesubmitted to different husbandry procedures.Aggressive Behavior 29, 69–80

Van Loo PLP, Van der Meer E, Kruitwagen CLJJ,Koolhaas JM, Van Zutphen LFM, Baumans V (2004)Long-term effects of husbandry procedures onstress-related parameters in male mice of twostrains. Laboratory Animals (in press)

Van Oortmerssen GA (1971) Biological signi� cance,genetics and evolutionary origin of variability inbehaviour within and between inbred strains ofmice (Mus musculus)—a behaviour genetic study.Behaviour 38, 1–92

Vestal BM, Schnell GD (1986) In� uence of environ-mental complexity and space on social interactionsof mice (Mus musculus and Peromyscus leucopus).Journal of Comparative Psychology 100, 143–54

Varlinskaya EI, Spear LP, Spear NE (1999)Social behavior and social motivation inadolescent rats: role of housing conditions andpartner’s activity. Physiology & Behavior 67,475–82

Ward GE, Fiat RA, DeMille D (1991) Environmentalenrichment for laboratory mice (Mus musculus).Animal Technology 42, 149–56

Welch BL, Welch AS (1966) Graded effect of socialstimulation on d-amphetamine toxicity, aggres-siveness, and heart and adrenal weight. Journal ofPharmacology and Experimental Therapy 151,331–8

Wemelsfelder F (1997a) The scienti� c validity ofsubjective concepts in models of animal welfare.Applied Animal Behaviour Science 53, 75–88

Wemelsfelder F (1997b) Life in capacity: its lack ofopportunities for variable behaviour. AppliedAnimal Behaviour Science 54, 67–70

Aggression in male mice 313

Laboratory Animals (2003) 37 by guest on May 29, 2015lan.sagepub.comDownloaded from