Refining dog husbandry and care

7/27/2019 Re fining dog husbandry and care

1/14

tion, and are considered by the WorkingGroup to be totally inadequate for housingof dogs, even for short periods.

There are also differences in the legalstatus of the various guidelines and in howthese are enforced. For example, within the

European legislation, both Appendix A tothe Convention ETS 123 and Annex II to theDirective 86/609/EEC are guidelines offeringadvice to Member States on how aspects ofthe legal documents could be interpreted.Annex II offers guidance on the interpreta-tion of Article 5 of the Directive which setsout the principles of accommodation in verybroad terms, namely that all experimentalanimals shall be provided with housing, an

environment, at least some freedom ofmovement, food, water and care which areappropriate to their health and well-being.In some Member States the guidance inAnnex II has been incorporated into nationallegislation, enabling the minimum recom-mendations to be enforced, whereas othercountries have not adopted these guidelines.In the UK, breaches of the codes of practicefor housing and care can lead to withdrawal

of the licence to keep animals.

Recommendations:

National legislation should set standardsfor the housing and care of dogs, andthese should be enforced. Such legisla-tion should provide detailed guidance on

the interpretation of the legislativerequirements, but not be so prescriptiveas to impede the development andadoption of innovative husbandry or carepractices of demonstrable welfare benetto the animals.

Efforts should be made to harmonizeminimum standards at a national andinternational level. This should be donein Europe through the Council of Europe

Convention ETS 123 and EuropeanCommunity Directive 86/609/EEC, andmore internationally through theOrganisation for Economic Co-operationand Development (OECD), theInternational Conference onHarmonisation of TechnicalRequirements for Registration ofPharmaceuticals for Human Use (ICH),and the World Health Organization

(WHO).

Rening dog husbandry and care S1:11

Laboratory Animals Ltd Laboratory Animals (2004) 38 (Suppl. 1)

Part 2

4 The natural history and behaviourof the dog in relation to itshusbandry and care

The process of domestication makes the doga tractable and therefore convenient animalfor use in the laboratory. Its status as acompanion animal and the particularrelationship the species has with humanshave resulted in it being given special

protection in some countries legislation onanimal experiments. For example, specialjustication is required for dog use in theUK (UK Government 1986). However, thefact that the dog is so familiar to humanscan lead to problems regarding its welfare in

the laboratory. Assumptions about its needsbased on so-called common sense and pastpractice may be awed. Therefore, as with

all other laboratory animals, there is acontinuing need to reassess dog husbandryand care in line with natural history andbehaviour and continuing advances in thescientic literature.

4.1 Ancestry and domestication

The domestic dog (Canis familiaris) isprobably derived from the Asiatic wolf

(Canis lupus pallipes) and hybrids withother races of wolf (Hemmer 1990), and it islikely that the process of domesticationbegan at least 14 000 years ago(Clutton-Brock 1995). It has been suggestedthat domestication leads to suppression of


2/14

the animals perceptual world so that fearand stress responses become reduced ortolerated and docility is increased (Hemmer1990), and that these changes are associatedwith reductions in the size of the body andhead (Clutton-Brock 1995). There is little

evidence, however, that domesticatedanimals are less intelligent than their wildcounterparts (Nicol 1996). Articialselection has produced dramatic changes tothe anatomy and behavioural repertoire ofvarious dog breeds (Coppinger & Coppinger1998). As a consequence, some breeds havepartially lost their ability to communicate(Goodwin et al. 1997), which could lead toincreased aggression in kennels (see Section

4.4 for ways of preventing and managingaggression).

Little is known about the social habits ofC. lupus pallipes, but studies of dog behav-iour in feral conditions indicate that, whilethey may appear solitary (Bernman & Dunbar1983), feral dogs form loose and changeablesocial associations strongly inuenced byfood supply (Scott & Causey 1973, Daniels& Beckoff 1989, Macdonald & Carr 1995).

However, prior to birth and during earlydevelopment of their pups, feral bitchesdistance themselves from their conspecics(Macdonald & Carr 1995) and parturitiontakes place within a den or a hidden area (seeSection 11.3). Unlike wolves, feral dogs showno evidence of cooperative care of young orhunting, but they apparently do cooperate indefence of territory and do interact as socialgroups (Boitani et al. 1995, Macdonald &

Carr 1995). It is well established that socialcontact is extremely important for domesticdogs, which need to be reared and kept in asocial environment in order to develop andbehave normally (see Section 13). That said,as a carnivore, the risks of aggression arehigh so particular attention is needed tomaintain dogs in harmonious social groups(see Sections 4.4 and 14).

As a feeding and survival adaptation,

dogs are inquisitive and actively seekinformation about their surroundings. Pensshould therefore be designed to allow goodvision out of them (see Section 6.1.2). Incommon with other carnivores, dogs spendmuch of their time resting (Hubrecht et al.

1992) but this should not lead to an under-estimation of their requirement for physicaland social stimulation during their activeperiod (see Sections 5, 6 and 8).

4.2 Senses and communicationIn order to understand dogs and providethem with a good husbandry system, it isimportant to understand how they perceivethe world and communicate betweenthemselves. The following two sectionscontain information on dog senses and theinterpretation of dog signals.

4.2.1 Olfaction and tasteDogs have a highly developed sense of smellcompared with humans, which is used inhunting and communication. Like manyother mammals, they inhabit a perceptualworld where things are as likely to beunderstood and remembered by their smellas by their sight. Excellent reviews of dogolfaction and taste can be found in Fox andBekoff (1975) and Thorne (1995); Alderton

(1994) also discusses the subject.All canids examine the faeces, urine, and

anal, genital, ear and mouth regions ofconspecics (Fox 1971). Dog urine, as wellas the secretions of the scent glandsbetween the digits and in the anal region,may have a unique identity and therefore beindividually recognizable (Doty & Dunbar1974, Simpson 1997). Bitches in oestrusrelease a pheromone, possibly methyl

p-hydroxybenzoate, which is attractive tomale dogs (Sommerville & Broom 1998).

Dogs and bitches scent-mark to transmitinformation about identity and social status.Social marking clearly takes place in kennels,even in groups as small as two. Hubrechtet al. (1992) showed that investigative snifngof the ground was more frequent in grouphousing compared with single housing, andhence it would seem that social housing in

addition to its other benets is a valuablemethod for increasing interest in the physicalenvironment (see Section 6.1.4). Visualdisplays such as scratching of the ground orlifting of the hind leg may accompany urineor faecal marking (Bekoff 1979), although

Laboratory Animals (2004) 38 (Suppl. 1) Laboratory Animals Ltd

S1:12 BVAAWF/FRAME/RSPCA/UFAW Joint Working Group on Renement


3/14

scratching may also add odour from theinterdigital scent glands (Simpson 1997).Scratching is not effective on kennel ooringand Hubrecht (unpublished data) did notobserve it in beagles housed on slatted ooror solid oor laboratory accommodation.

Canids often prefer to urinate and defaecatein a discrete area in captivity, away fromtheir sleeping area (Fox 1971); hence this isan argument for providing dogs with enclo-sures which allow adequate space for separatesleeping and activity areas (see Section 6.1.3).

The dog has a rhinarium, kept moist bysecretions from the lateral nasal glands,which serves to increase the humidity ofinhaled air with associated scents. Its

bilateral vomeronasal (Jacobsons) organs arewell developed. The elongated snout with itsinterior scrolls of bone (turbinates) providesa large area for the olfactory epithelium(75 cm2 in the beagle compared with 3 cm2

in humans) which contains around 20 to 40times as many primary receptor cells as thehuman nose (Syrotuck 1972). Also, theabsolute number of olfactory nerve cells perspecic receptor type is greater in dogs than

in humans (more than 100 fold) and theolfactory bulb in the dog is dramaticallylarger than it is in man. These adaptationsresult in the dog being sensitive to com-pounds at one thousand to one hundredmillion times lower concentrations than aredetectable by man. It is known that dogsrespond to odours such as oil, tobacco andother animals excreta (von Heimburger1959, reviewed by Fox & Bekoff 1975). It is

unfortunate that human insensitivity toodour has meant that the study and applica-tion of appropriate osmatic enrichment ofthe dogs captive environment has beenmuch neglected (see Section 8.2).

More impressive than their remarkablesensitivity to trace odours, is the ability ofdogs to pick out odours of interest from awelter of competing smells and to matchand distinguish them. Thorne (1995) reviews

evidence showing the astonishing ability ofdogs to track men by odour, and Steen andWilsson (1990) and Thesen et al. (1993)show that dogs can determine the directionof tracks, possibly by comparingconcentrations. Although dogs are clearly

able to identify odours from differentindividuals, it may be important in animalhusbandry to remember that odours fromdifferent parts of the body vary and dogs arenot able to generalize scents from differentparts of the body to identify an individual

(Brisbin et al. 1991). There is some evidencethat dogs sample odours from different partsof the human body for different purposes.Millot et al. (1987) presented familiar andunfamiliar clothes on a dummy to dogs, andsuggested that anogenital snifng is used toidentify strange humans, whilst snifng theface may be an attempt to determine thehumans emotive state or behaviouralintentions. Fox (1986) mentions that dogs

may respond sexually to certain perfumes,and although there are no research data onthis, the potential response of dogs to novelodours on animal handlers should beconsidered by animal care staff.

Olfaction is used in food selection and isan essential component of taste. Dogs havefour main groups of taste receptors thatrespond to sugars, acids, nucleotides andfuranols (Thorne 1995). The last two

probably respond to meaty and fruity-sweetcompounds in the diet respectively. Unlikecats, dogs are insensitive to salt. Dogsfood preferences and their implicationsfor laboratory diets and feeding in thelaboratory are discussed in Section 7. Dietsin laboratories may not be as palatable asthose of wild dogs, or even of those providedto companion dogs, and variety is muchmore likely to be restricted.

4.2.2 Hearing and vocalization

Dog hearing has a wider frequency rangeand is more sensitive than that of humans;dogs are most sensitive to sounds between120 kHz, compared with 15 kHz inhumans, and they can hear sounds up to50 kHz compared with an upper limit of1520 kHz in humans (Sander et al. 1977,

Heffner 1983, Fay 1988). Their hearing rangemay reect a predatory adaptation to ndingsmall rodents. Below 250 Hz dog hearingsensitivity is similar to that of man, butin the range of 500Hz to 16 kHz theirthreshold of sensitivity can be 24 dB lower




4/14

than that of humans (Fay 1988). Thismeans that they can hear sounds that areup to four times quieter than the humanear can detect. It is probably no coincidencethat the peak energy of a bark falls withinthis range (Sales et al. 1996).

Dogs, like most predators, possess a fairlyprecise ability to tell the direction a soundis coming from. Their sound localizationability corresponds to an ability todistinguish arrival times in the left and rightears of as little as 55 microseconds(Kalmykova 1981). The mobility of the earsallows them to scan the environment forsound and then to collect the sound waves.

Dogs bark to communicate (see Fox &

Bekoff 1975, Simpson 1997), but there aredifferences between breeds. Husbandry staffshould be adept at identifying the meaningof various vocalizations, from distress callssimilar to the whines or whimpers ofpuppies to the high-pitched bark of a dogseparated from its owner or conspecics(see Section 4.3).

It is possible that some dog vocalizationsmay have become partially divorced from

their communicatory function. It is acommon nding that aroused or exciteddogs bark, and this may be due to selectionby humans for a behaviour which wouldserve as an early warning of danger or allowhunting dogs to be followed. It has also beensuggested that barking may become stereo-typed when dogs are housed in suboptimalconditions (see Section 5). Social facilitationoften results in the barking spreading to

other dogs (Fox 1971, Adams & Johnson1994), and in some animals the act ofbarking may act as a stimulus to furtherbarking (Scott & Fuller 1965).

4.2.3 Vision

Dogs have good eyesight that is in someways inferior but in other ways superior tohuman vision; their visual system is

optimized to exploit a different environmen-tal niche than humans. Dogs are not as ableto discriminate form and pattern as well ashumans, although their performance isimproved when the patterns move ratherthan when they are static (Fox & Bekoff

1975). Nonetheless, herding dogs canrespond to hand signals at distances of onemile, so vision is evidently an importantsense.

A number of studies have been conducted,using a variety of different methodologies to

examine visual acuity in dogs. Generally,dogs cannot see clearly anything narrowerthan about a sixteenth of a degree of arc,which corresponds to 20/75 vision on theSnellen fraction; what a normal human cansee clearly at 75 feet (22.86 m), a dog wouldgenerally need to be 20 feet (6.10 m) away tosee in detail (Miller & Murphy 1995,Murphy et al. 1997, Tanaka et al. 2000).

Dogs have a much more limited power of

accommodation than humans, generally notmore than two or three diopters (Duke-Elder1958, Murphy et al. 1992). This means thatthey can focus and accurately image closeobjects only if they are no nearer thanaround 33 to 50 cm; anything closer will bea blur. This may explain why dogs generallytry to sniff or touch objects at close range;they simply cannot see them very well andso augment vision with other senses.

Dogs have a wider eld of view thanhumans do, increasing their ability to scanthe horizon, but the range of overlap of thetwo eyes is smaller, limiting three-dimensional vision to a narrower sector(Miller & Murphy 1995). Dogs are probablymore aware of activity occurring aroundthem than humans are because of this widereld of view. Also, simply by virtue ofvisual perspective, the world will look very

different to a beagle which stands at around33 to 41 cm at the withers.

The retina of the dogs eye is composedprimarily of rod photoreceptors and theyalso possess a tapetum giving them goodnight vision, although probably at theexpense of visual acuity (Miller & Murphy1995). The minimum light threshold of dogsis well beyond that of humans, allowingdogs to function in both bright and dim

light (Walls 1963, Kemp & Jacobsen 1992).Dogs, like most mammals, possess and usecolour vision. They are dichromatic andhave difculty in discriminating betweenmiddle and long wavelength lights (Jacobs1983, 1993, Neitz et al. 1989, Jacobs et al.




5/14

1993). The two photoreceptor cone cells inthe dog retina have their peak response atlight wavelengths that correspond to thecolours perceived by humans as violet (429to 435nm) and yellowgreen (about 555nm).Although it is not known whether dogs

perceive these two colours in the same wayhumans do, the canine visual spectrum maybe divided into two hues; one in the violetto blueviolet range (430 to 475 nm), whichis probably seen as blue by dogs, and asecond in the range seen by humans asgreenish-yellow, yellow and red (500 to620 nm), which is probably seen as yellowby dogs (Neitz et al. 1989). Dogs also appearto have a narrow band in the bluegreen

range (475 to 485 nm) that is without colourand likely seen simply as shades of whiteor grey.

A practical consequence of these ndingsis that many of the toys manufactured fordogs may be the wrong colour. A brightredorange dog toy stands out dramaticallyagainst a green lawn to trichromatichumans, but to a dog its colour is presum-ably far less readily distinguishable from the

green grass; a blueviolet object would be amuch better choice when the background isgreen. Similarly, it may be problematic toteach dogs to distinguish among red, orange,yellow and green objects solely on the basisof colour (Miller 2001).

Dogs use visual signals to communicate,and posture is an important component incommunicating emotional state. It isimportant that laboratory staff are able to

interpret visual signals as they provide anindication of the dogs emotional state andsocial status (see Section 4.3).

From the point of view of husbandry, itis a common observation that dogs willmake considerable efforts to gain visualinformation about their surroundings.This is an adaptive behaviour for manycarnivores to ensure that vision ofapproaching threats or prey is not overly

obstructed. There is some evidence thatwolves use high vantage points (Murie1985), and many zoos now provide moundsor platforms in carnivore enclosures. Raisedplatforms have been found to be benecialfor laboratory dogs (see Section 6.1.2).

4.2.4 TouchTouch is the primal sense in dogs andextremely important for well-being. Touchcan have dramatic effects on the emotionalstate and physiological responses of a dog,and tactile experiences are important in the

context of socialization and habituation,especially in the very early days and weeksof life when other senses are not yet func-tional (see Section 13). The tactile experi-ences of the dam during pregnancy arereported to affect the development of thehypothalamicpituitaryadrenal (HPA) axisof the unborn puppies and their subsequentability to deal with stress as adult dogs. Itis therefore important to consider tactile

interaction with pregnant bitches andensure that such interaction is positive andrewarding from their perspective.

Dogs rest in contact and this is probably ameans of maintaining social cohesion incanine groups. This behaviour may beassociated with pleasant sensations experi-enced during puppyhood. Hubrecht et al.(1992) found that beagles aged between 68months averaged up to 10% of their day

resting in contact. Even adult dogs in rela-tively unestablished social groups spent anaverage of more than 2% of their day restingin contact. It is likely, however that thisbehaviour can be inuenced by social status.

Stroking or petting dogs is good for bothdog and handler. Odendaal and Lehmann(2000) found that phenylethylamine plasmaconcentrations changed signicantly duringpositive interaction in both humans and

dogs. They concluded that it is possiblethat, during stroking, dogs experience thesame feeling of elation from theamphetamine-like neurotransmitter as dohumans stroking dogs. Hennessy et al.(1998) found that stroking can reduce dogcortisol levels during sequential blood sam-pling. They therefore propose that sympa-thetic handling/stroking can be a valuabletechnique to reduce the dogs stress during

aversive procedures (see Section 16). Lynchand Gant (1968) showed that the dogs heartrate also fell during stroking and that theresponse was easily conditioned to a tone.However, human social contact can also beaversive to individual animals that have not




6/14

been adequately socialized during theirprimary socialization period and thereforeadequate attention must be paid to the levelof social interaction with puppies between 3and 14 weeks of age (see Section 13).

Recommendation: The way in which dogs use their sensory

and perceptual abilities to interpret andassess physical environments and socialinteractions should always be consideredwhen designing all aspects of dog hous-ing and husbandry within a laboratorycontext.

4.3 Interpreting dog signals

It is important that laboratory staff are ableto recognize and accurately interpret dogsignals, not only in order to assess thewelfare of the dogs in their care but also toensure that they respond appropriately wheninteracting with the animals. See Simpson(1997) for an in-depth appraisal of caninecommunication.

Olfaction, one of the most importantmeans of communication in the dog is,

unfortunately, the sense that humans havemost difculty comprehending. Visualsignals (i.e. body postures and facialexpressions) and vocalizations will thereforebe the most important signals in terms ofhumandog interactions. These methods ofcommunication provide an indication of adogs emotional state and can provideinformation about the relative social statusof dogs within the same social group. They

can also provide information about the dogsintended action in response to a situation orinteraction, and accurate interpretation isinstrumental in minimizing pain and dis-tress. For example, a relaxed dog will showan alert facial expression with ears carriedhigh and forward (Fig 2), whereas a dog thatadopts a low posture with head and earsheld low is indicating a lack of self-condence (Fig 3). Such an individual may

require extra habituation to a procedure ormore specic training in order to minimizeits fear responses. It may be instinctive forstaff to offer reassurance to these individuals,but care must be taken not to inadvertentlyreinforce fear in this way.

Since dogs are only capable of usingcanine communication systems, the signalsthat they use in interactions with other dogsare also used to communicate with humans(but the potential role of conditioning in themodication of these signals must be appre-ciated). The success of interspecies commu-nication is dependent on the way in whichhumans interpret and respond to dog signals,

as well as the way in which dogs interpretand respond to signals, intentional orotherwise, given to them by humans viatheir use of body posture and eye contact.

During social interactions withconspecics and humans, dogs willfrequently use communication signals totest the response of the receiver and will



Fig 2 An alert facial expression with ears carried

high and forward is indicative of a relaxed emotionalstate

Fig 3 A crouched body posture and lowered headis indicative of a fearful emotional state. The dog is

also showing anticipation of getting off the table


7/14

then react accordingly. The expectedresponse will be one which is based on theoriginal motivation for the signal, for exam-ple a fearful body posture would be expectedto result in a response of retreat on the partof the person or other dog. In dogdog

communication, there is usually a commonunderstanding of the signals used anddeference on the part of one dog will lead toan assumption of higher social status on thepart of the other. Unfortunately, doghumaninteractions often result in responses thatthe dog considers to be unexpected andunjustied; this can lead to problems ofmiscommunication and a breakdown in therelationship between the two parties. An

example would be the continuation ofa scientic procedure, which the doginterprets as potentially harmful, whenthere has been clear canine signalling ofpain and distress which should, from thedogs perspective, have resulted intermination of the interaction.

When interpreting dog signals it isimportant to remember that they arecontext-specic and that their meaning will

be inuenced by the response of the receiveras well as the intent of the sender. As aconsequence of this it is impossible toattach one rigid meaning to each signal andinappropriate to adopt inexible responsesto signals that are given. We should bear inmind also that breed differences can have asignicant effect on forms of communication,and since beagles are relatively small dogs itcan be easy to miss subtle changes in their

body posture.

4.3.1 Visual signals

Staff responsible for the day-to-day care ofdogs need to be able to recognize caninevisual signals and understand the functionof such signals in demonstrating rank andtesting the hierarchy within a group of dogs.In addition, the role of canine visual signals

in conveying information about the dogsemotional states must also be considered.

Demonstrating rank and testing thehierarchy: The relative rank structure ofcanine society, which is often referred to as

a hierarchy, is designed to remove the needfor confrontation and conict; aggression istherefore less likely to occur within a stablesocial group. Staff need to understand theirown role in maintaining canine socialstability and controlling stress within and

between cages or pens, through theprovision of calm and condent signalling(see Section 4.4). It is important to appreci-ate that rank is a dynamic concept and thata state of dominance is given to an animalby the subordinate actions of anotherindividual, rather than taken through theiruse of so-called dominant signalling.Consequently, bullying tactics areunnecessary and unsuccessful in terms

of establishing social hierarchies, andconfrontational communication has noplace in stabilizing social groups.

When dogs are kept in conned spacesthey are under increased social pressure andthe range of defence strategies at theirdisposal may be limited. For example, thedefence strategy of ight in reaction to asituation of potential conict may not bephysically possible, and dogs, of any rank,

which would normally opt to walk awayfrom potential confrontation may be unableto do so. It is important also to realize thatsubtle visual signals, which are used by dogsto diffuse conict at a distance, may be oflimited use within a conned space andtherefore more exaggerated (or frequent), andpotentially challenging signals are likely tobe used. As a consequence of losing ight asa potential response, dogs rely on the

alternative defence strategies of freezing,appeasing or ghting to resolve conict, andthis may be signicant when consideringthe issue of aggression in group housing (seeSection 4.4).

Emotional condence or lack of behav-ioural inhibition in dogs is shown by anincrease of physical presence indicated by ahigh body posture, raised tail, fully stretchedfront legs, upright head and neck, and erect

and forward-pointing ears (Fig 4B). When acondent individual is challenged, or threat-ened, intense eye contact and exposure ofthe canine teeth may be seen as additionalsignals (Fig 4F). When an inhibited or lesscondent individual is challenged he/she




8/14



Fig 4 Expressive social responses in the dog (Reproduced from Shepherd (2002) in BSAVA Manual ofCanine and Feline Behavioural Medicine, with permission from the British Small Animal VeterinaryAssociation)


9/14

will respond by adopting a lowered bodyposture, or cowering, the head and neck canbe tilted, eyes are averted, ears folded back,and the tail is tucked in under the belly(Fig 4H).

Two behaviours that are often interpreted

as the ultimate gesture of passive surrenderto a superior are rolling over and lyingbelly-up (Fig 4J). Whilst these signals areextreme forms of appeasement behaviour,which are designed to diffuse conict, theydo not indicate that the individual has givenup completely. If the threat or challengecontinues, individuals in these positionswill still respond with aggressive defence,and can snap and bite if pushed too far. The

normal canine response to an individualwhich has rolled over to expose its bellywould be to sniff the inguinal region andthen turn away, thereby conrming therelative rank of the individuals and avoidingescalation into confrontation.

When two dogs approach one anotherobliquely rather than head on, the signalappears to be one of calm social interactionbut there is no evidence of any value of this

signal in terms of social status or relativerank.

Both uninhibited and inhibited individu-als can use vocalization and piloerection aspart of their signalling repertoire and thiscan make interpretation of these signalsdifcult. In addition, facial expressions canappear very similar between the two groups.Differentiation can rely on subtle differences,including the degree to which the mouth is

opened, the angle and shape of the corner ofthe mouth, how far back the lips are pulled,how the nose and face are creased and howthe teeth are presented (Feddersen-Petersen1986). Dogs lacking in condence mayretract the corners of their lips, pulling theirmouths into a shape that often looks to uslike a smile (Fig 5). It can even result in theteeth showing, which can be confusing; butthe dog smile is invariably accompanied by

other appeasement gestures, such as low-ered, attened ears or a bowed neck. In con-trast the fang-bearing threat of a condentindividual involves raising the forelip toexpose the canine teeth, which does nothappen in the appeasing smile.

Fear: Fearful dogs may exhibit behaviourswhich are designed to avert potential threat,be that real or perceived, and these mayinclude not only visual signals of apparentappeasement but also signs of aggression.Problems arise when these signals, which

are a response to the perception of threat,are misinterpreted as signs of relative status.As part of a canine fear response individualswill carry their ears down and back andtheir heads down with the eyes wide (Figs4H and 6). Fearful individuals have aheightened startle response and will usuallyattempt to ee. If this is not possible theymay freeze in place, with stiffened forelimbsand their weight distributed to the back of

the body (Figs 4I, 6 and 7). Sometimes theyface what frightens them (Fig 8); whenreached for or approached by a human, suchdogs are likely to bite, particularly whenunable to escape. In extremes of fear, somedogs may urinate, defaecate or empty theiranal glands. See Simpson (1997) fortechniques to reduce fear in dogs and otheruseful responses to dogs in a care-giving orveterinary setting.

Anxiety: Anxious dogs may show acombination of greeting signs, tail waggingand approach, while at the same timeshowing signs of trying to avoid theindividual or situation. This is sometimesreferred to as a state of approachavoidanceconict and it demonstrates a very unstableemotional state. These individuals may showsigns consistent with fear, such as a low

head carriage and raised hackles. They mightroll over and urinate, which is an extremesign of appeasement, or show more subtlesigns such as lip smacking and yawning.

Dogs that are traditionally referred to asexcitable often lack condence and signssuch as exuberance, barking, and jumpingup may be learned attention-seekingbehaviours which are being used asappeasement gestures in the context of

greeting human beings.

Tail wagging: Tail wagging is an indicatorof arousal and of intention to interact.In a domestic context the tail wag is oftenconditioned as a greeting behaviour and it is




10/14

a behaviour which is generally viewed as a

sign of happiness and well-being in thedog. However, tail wagging alone does notgive any indication of motivation and it isdangerous to always interpret a wagging tailas a friendly gesture. Instead it should beinterpreted alongside other communicationsignals, and factors such as the height of tailas it is wagged should also be taken intoconsideration. A medium tail carriagein conjunction with steady wagging and

condent approach is interpreted as ageneral expression of friendly greeting(Fig 4D), while a tail held high and wavingsteadily along its length indicates condentinteraction and may even be used as part ofa threat display. A tail held in a low position



Fig 5 This head posture with lowered ears and

concentrated facial expression is often accompaniedby lip smacking and is indicative of an anxious

emotional state

Fig 6 Fearful dogs may attempt to avoid eyecontact. This facial expression is accompanied by a

stiffened body posture, which can make it difcult

to place the dog in a sitting position on the table

Fig 7 A fearful dog with a lowered head posture

and stiffened forelimbs avoids eye contact with thehandler

Fig 8 Fearful dogs may retreat from peopleentering the home pen and attempt to get into the

corner of the pen or under the platform

and wagging rapidly and erratically,

particularly towards the tip, indicates a lackof condence and is often associated withother clear signs of fear and appeasement(Fig 4E).

Play bow: Play is a lifelong activity in dogs.Dogs that want to engage another in playwill bow down with the front legsoutstretched and hindquarters elevated in

the play-bow position (Fig 4C). This isunderstood to signal that the interactionthat follows is not to be taken seriously, andduring the play that follows, dogs oftenengage in aggressive signalling andambiguous social gestures.


11/14

Raised paw: The raised paw is ofteninterpreted as a positive signal and seen asan attention seeking behaviour or anindication of intent to play. However, unlessthere are other clear signs of play behaviourit is more likely that the elevated paw is

being used as sign of appeasement (Fig 4E).When it is seen in combination with othersigns of behavioural inhibition and lowlevels of condence its purpose is to avert athreat and indicate a desire for the presentinteraction to be terminated (Fig 4H). Incontrast, a condent individual who istolerant of approaches and handling will bemore likely to remain sitting or standingwith all four feet on the oor.

4.3.2 Vocalizations

Dog vocalizations can involve whines,whimpers, growls, barks and howls, or somecomplex combinations (Cohen & Fox 1976).These are all fairly non-specic signalsunless they are evaluated within the socialcontext in which they occur. Their meaningis best understood by considering what theyaccomplish (e.g. growling causes the

perceived threat to retreat and is thereforehighly effective as a defence strategy).

Husbandry staff should be experiencedenough to be able to identify the meaning ofvarious dog vocalizations and to understandthe role of vocalizations in their recognitionand monitoring of pain and distress (seeSection 17.6). For example, differentiationbetween the constant vocalization of anindividual demanding social interaction and

the intermittent howl of a dog that has beenseparated from his or her social companionsis important if staff are to take appropriateaction. It is important also to understandthe role of vocalization in dogdogcommunication, both in the context of playand within confrontational encounters.Misinterpretation of barking as a form ofthreat or conict can lead to prematurehuman intervention in dogdog interactions

and may even lead to an increase in theincidence of aggression within pens (seeSection 4.4).

Whines and growls: A whine is ahigh-pitched pure-toned sound, rising slightly

then falling quickly. Whines are used whenthe aim is to signal a non-threatening statusand to seek the non-violent approach of asocial partner. They are used in friendly orcare-soliciting contexts. Dogs can alsowhine when frustrated (such as when

desired movement is prevented) or whenexperiencing pain (see Section 17.6).

A growl is a low-pitched, acousticallyrough, harsh tone. Growls are usually usedin hostile contexts or distance-provokingcontexts where the aim is to repel animpending threat. Growls are often used inconjunction with warning facial and bodyexpressions (see Section 4.3.1) and should beinterpreted as early-stage threats, which

could be followed by more intense signs ofaggression if they go unheeded.

Whines and growls vary in intensity andpitch, and there is a correlation between theurgency of underlying motivation and thepitch of the signal. Whines shade intowhimpers, and growls into snarls, asurgency grows. Dogs use whines and growlsand all their variants in many differentcircumstances, and accurate interpretation

will rely on observation of all concurrentsignalling including body posture and facialexpression.

Barks: Barks are probably the most noisyand disruptive of the canine vocalizationsand are acoustically almost preciselyhalfway between growls and whines, in bothpitch and structure. Dogs produce distin-

guishable barks in a number of differentcontexts, including territorial defence,aggression, altering of others, anxiety,individual identity, social facilitation andplay solicitation (Simpson 1997, Yin 2002).Studies of companion animals indicate thathounds, such as beagles, or sporting dogsbred for high activity are particularly proneto barking when their owner is absent(Niego et al. 1990). Similarly, Hetts et al.

(1992) found that socially isolated laboratorybeagles (i.e. dogs having only auditorycontact with other dogs and contact withpeople only during routine husbandryprocedures) vocalize more than dogs thathave social contact.




12/14

Howls: Howls are long, melodious soundsthat sweep through many different pitches.Wolves howl under many differentcircumstances, and the full range ofmeanings of this vocalization are still notunderstood. However, it is certainly a form

of long-distance communication and hasbeen associated with such diversesituations as the location of other packmembers when the social group becomesdispersed and the deterring of members of arival social group who are encroaching onestablished territory. Howling may indeed bea natural response of dogs separated fromtheir social group or owners (Lund &Jrgensen 1999).

Moans and yelps: Some dogs moan inpleasurable contexts. Moans seem to belearned vocalizations, used only incommunication with humans and neverwith other dogs. Yelps are indicative ofdistress or actual pain (see Section 17.6).

Recommendations: Care-giving staff should be trained to

recognize and interpret canine visualsignals and vocalizations and to respondappropriately. They should be aware thataccurate interpretation of these isinstrumental in the minimization ofpain and distress.

Fearful dogs should be identied earlyand behavioural modication techniquesshould be used to reduce their fear.

Accurate interpretation of visualsignalling should be used to determinethe stability of a social group.

Visual signalling should never beinterpreted in isolation and should bemonitored for changes over time.

4.4 Aggression

Aggressive behaviour in dogs can be consid-

ered normal species-specic behaviour andis a function of a number of componentsincluding the health of the animal.Veterinarians and canine behaviour special-ists have classied many types of aggressionaccording to the context in which they occur.

Such labels may be helpful in some situa-tions, but the most important factor toconsider when faced with a problem of canineaggression is the underlying motivation forthe behaviour and the emotional state of theindividuals involved (see Section 4.3.1).

Defensive aggression is far more commonthan offensive aggression, and aggressiveresponses are far easier to provoke in insecureand fearful individuals (Overall 1997).Emphasis should therefore be placed onavoiding fearful animals by careful selectionof breeding stock (see Section 11.2) andsuitable socialization and habituation (seeSection 13).

Preventing and managing aggression:Since aggression is exhibited under manydifferent circumstances, effectivemanagement can vary depending on themotivation for the behaviour and thestimuli that trigger aggressive responses inthe particular dog. Attention should there-fore be paid to the timing of aggressiveincidents through the day and to anypotential link with levels of arousal in the

dogs. For example, activities within thepens, such as feeding and cleaning, may leadto increased levels of arousal and decreasedthresholds of tolerance toward other dogs. Ifdogs are allowed to run freely in the centralcorridors during these procedures, it is alsopossible that dogs may come into visual andeven tactile contact with individuals fromdifferent social groups; and the resultingarousal from these encounters could increase

tension within the pens. Care-giving staffshould adapt their routines where possibleto avoid any circumstances, such as movingand mixing dogs, that trigger or exacerbatesocial conict. In addition, it would bebenecial to take steps to decrease levels ofarousal at feeding and cleaning times, whichcould contribute to the onset of aggressiveincidents. Increased visibility has beenfound to reduce the level of arousal within

units when there is activity in the centralcorridor (see Section 6.1.2); this may helpreduce aggression at these times.

Whatever its proximate or ultimate cause,aggression can have serious complicationswhen dogs are housed in a group (e.g. injury




13/14

or shock). It is therefore critical to ensurethat there is an adequate husbandry routineto monitor the animals and forestall potentialproblems. A video and sound remotemonitoring system can be very helpful inthis regard, and alarms can be used to alert

personnel to particular instances of ghting.However, these should not be used as asubstitute for an adequate husbandry rou-tine and appropriate staff-to-dog ratio. Videois also helpful for investigating stereotypiesin the absence of an observer (see Section 5).

Territory can be dened as a space thatcontains important resources, such as food,shelter, resting areas and social interaction,and in the laboratory context the home pen

fulls this function. The potential forconict between individuals in relation togaining and maintaining access to thoseresources is something that needs to beconsidered when designing pens (see Section6.1) and when selecting individuals to behoused together (see Section 14). Large,complex pens with plenty of enrichmentmay help reduce the frequency of aggressionwithin the unit in the rst instance through

increasing the availability of high valueresources, such as resting places, in order todecrease their perceived value. It is also pos-sible that increasing the size of pens couldhelp foster and permit more effective caninesignalling and therefore decrease aggression(Beback & Beck 1993). Canine communica-tion signals are likely to be exaggerated inpen-housed dogs due to the close proximityresulting from the conned space that the

dogs are in (see Section 4.3.1). Defensive sig-nals are aimed at maintaining or increasingthe distance from a perceived opponent orthreat, and effective communication willresult in a decrease in aggressive incidents.If retreat to a safe distance is not possiblewithin the connes of the pen, this maylead to an increased incidence of aggression.

Increasing the level of human contactis sometimes suggested as a means of

controlling aggression. However, this canalso lead to increased value being given tohuman interaction and therefore increasedconict when human contact is available oranticipated (e.g. at the start of the day or atthe rst sight of human companions). It

must be recognized that dogs within alaboratory environment have certainrestrictions on the availability of socialinteractions with people, and therefore careshould be taken not to raise the expectationsof human contact to levels that are

unrealistic. Doing so may result in anxietyrelated to separation from humans and canbe detrimental to the welfare of individualanimals (see Section 13).

The supply of dogs in pre-existing groupsbefore they enter the research establishmentand the avoiding of regrouping on arrivalmay help to create stable groups and therebyreduce the risks of aggression (see Section14). The potential for conict may also be

reduced by the housing of individuals ofdifferent ages and/or body weights together.Dogs of like age, size and sex are more likelyto ght because of unclear demarcationbetween their respective resource holdingpotential (Wickens et al. 2001). Housingmales and females in separate rooms andpaying particular attention to reproductiveseasonality may also help prevent aggression.

Dealing with aggressive incidents:Aggressive incidents resulting in physicalinjury are relatively rare within well-managed laboratory environments. It is vitalto have accurate records of exactly what isoccurring and when, in order to implementtechniques to deal with aggressive incidentswhen they do occur. Records should includedetails of when the incidents occurred, whowas present in the unit at the time, what

activity was taking place (e.g. cleaning,feeding, worming, weighing), the level ofinjury to the individuals involved, and thebehaviour of those individuals and othermembers of the social group before, duringand after the incident.

Where aggression arises in an establishedgroup of animals it may be possible to sub-divide the group such that the sub-groupscan continue to live in harmony. Where pos-

sible, it is advisable to put physical spacebetween the pens of such sub-groups and tomove the groups into separate units, ratherthan across the aisle of the same unit, inorder to avoid continuing displays of socialconict (Canadian Council on Animal Care




14/14

1993). Alternatively, individuals that havebeen removed from a group because ofaggression can be introduced to dogs fromother social groups; such regrouping shouldinitially be done in large, novel pens that areinteresting to the dogs and away from the

home pen. Interactions between dogs shouldbe carefully supervised by suitably trainedstaff. When there is a history of aggressionbetween particular dogs, steps should betaken to avoid exercising them together.

While a stable hierarchy within a groupprevents actual ghts (and thus injuries)resulting from ritualistic displays, thedevelopment of such a hierarchy is oftenundermined by the lack of understanding in

humans. In the case of dogdog aggression,human interference often inadvertentlyencourages aggressive behaviour, andinappropriate support for a naturally lowranking individual against a perceivedbully may destabilize the hierarchy withina social group and inadvertently increase therisk of aggressive behaviour. It is importantfor staff to be trained in the recognition ofcanine signals so as to avoid premature

interference, which can lead to an escalationof aggressive signalling (see Section 4.3).

Animal care staff should be condent intheir ability to interpret canine communica-tion and to deal with aggressive dogs in anappropriate and safe manner. It is importantto maximize the stability of the group byreinforcing the natural social order withinthe group. Giving attention to the more con-dent individuals rst when greeting the

dogs, and allowing these dogs to be the rstof the group to go through doorways andreceive treats or food, may be benecial.Resisting the temptation to intervene pre-maturely in agonistic interactions betweenindividuals within a social group may alsohelp to stabilize the situation.

Recommendations:

Animal care staff should be trained torecognize aggressive signals and interpretthem accurately.

Managers of dog units should establishhusbandry routines which allow theadequate monitoring of aggression

within social groups and enable staff toforestall potential problems. A video andsound remote monitoring system can bevery helpful in this regard.

Pens should be of an adequate physicalsize and provide enough complexity,

both physical and social, to reduce thefrequency of aggression.

Care-giving staff should look carefullyat their husbandry routines, identifypossible causes of conict, and thenadapt their routines where possible toavoid any circumstances that may leadto aggression.

Staff should keep accurate records ofaggressive incidents and ensure that dogs

which are known to be aggressive toeach other are not exercised together.

Attention should be paid to reproductiveseasonality in order to reduce the risk ofaggression. It may be necessary tohouse male and female dogs in separaterooms.

Dogs should be supplied to user premisesin pre-existing stable groups to helpavoid aggression. The potential for

conict may be reduced by housingtogether individuals of different agesand/or body weights.

Where aggression arises in an establishedgroup of animals, consideration shouldbe given to subdividing the group suchthat the sub-groups can continue to livein harmony. Where possible, the pens ofsuch sub-groups should be distant fromone another.

When a decision is made to introducean individual to an alternative socialgroup, such regrouping should initiallybe done in large, novel pens that areinteresting to the dogs and away fromthe home pen.

Direct contacts between dogs should becarefully supervised by suitably trainedstaff who are condent in their ability tointerpret canine communication and to

deal with aggressive dogs in anappropriate and safe manner. The veterinarian, animal care staff and

canine behaviour specialist should period-ically review the mechanisms in place formonitoring and dealing with aggression.


Documents

Refining dog husbandry and care