Layout

The use of real-time heart rate monitors to assess arousal levels during canine behavioural testbatteries DIAnA RAyMenT1, DR BeRT De GRoeF1, DR RICHARD PeTeRS2 AnD DR LInDA MARSTon1,31 Department of Agricultural Sciences, La Trobe University, Bundoora Victoria2 Department of Zoology, La Trobe University, Bundoora Victoria3 GOTAFE Department of Animal Sciences An accurate, objective and reliable test to identify testers having worked and trained together for behavioural traits in pet dogs is thought to be an several years. This exemplifies the issues plaguing a important and useful tool in companion dog behaviour-only approach to developing test protocols.
management and welfare. However, disagreement The accurate interpretation of subtle canine between interest groups as to how such tests should behaviours is difficult, especially if the dog is be constructed, administered and interpreted, has apparently doing little. Understanding of canine lead to confusion, distrust of results and worse, in behaviour and communication requires attention to some situations inaccurate interpretations resulting subtle body language as well as gross body in an animal’s death. This is partly due to a variation movements, but the sheer volume of information in focus between purebred and sporting enthusiasts, incorporating these completely impractical for general professionals (Taylor & Mills, 2006), and partly due to use (Figure 1). Furthermore, even with extensive a general lack of trust for test results due to the training and experience, a tester's previous haphazard way tests have been developed to date experiences with dogs or breeds, and their own (Bräm, Doherr, Lehmann, Mills, & Steiger, 2008; understanding of canine nature, will inevitably Diederich & Giffroy, 2006; Mornement, Coleman, influence results where any subjective interpretation of behaviour is required or allowed. Tests must bebased on easily observable, objective measures if the Despite several decades of research into canine results of tests carried out by one person, are going behaviour, and much discussion about how tests to aid interpretation of tests carried out by others. It should be developed and assessed, we are still a long seems unlikely that either of these two issues will be way from having a practical, objective test that is addressed fully while the focus of test developers capable of providing reliable information about dogs.
Taylor and Mills (2006) stress the need forbehavioural tests to be objective, reliable and valid inorder for tests to be of useful, regardless of whetherthe test is being used for identifying sporting orworking dogs, or selecting a suitable pet from ashelter or pound. They also highlighted the differencebetween behavioural and temperament tests, theresults of the former being highly influenced bylearning and situational variables, while the latter(confirmed by assessing test-retest reliability overtime) being more reflective of a dog's inherent'personality'. Finally, the authors suggested thatsimilarities in the tests currently used by differentcanine interest groups, may indicate that a suitableseries of subtests could yield information useful inthe selection of dogs for pet, sporting and workingroles.
Figure 1 In many current behavioural tests, the behaviour ofthe dogs in this photo would score similarly due to the fact Studies following that of Taylor and Mills (2006) have that they are both lying down with similar body positions, yielded reliability and validity results for test despite their facial expressions and postures indicating protocols, but even the most comprehensive tests quite different responses to the situation. In order to and evaluations have encountered issues with accurately describe each dog's response, detailed reliability - inter-rater reliability and concurrent validity information about the size and shape of their eyes, the in particular. For example, the test protocol direction and focus of their gaze, the position of their presented by Valsecchi et al. (2011) showed mouth and tongue and their ear-set would all need to be significant agreement between testers for many behavioural traits, but not in two important areas,sociability with both people and dogs and the Taylor and Mills (2006) proposed that the use of interpretation of passive responses, despite the relevant physiological measures might help reduce issues with inter-rater reliability in behavioural tests.
Based on the results of the Swedish Dog Mentality Physiological measures such as heart rate are both objective and can be easily measured. However, proposed a shy-bold 'super trait' in dogs, showing that these measures will only be reflective of behavioural 'bold' dogs performed better across a range of traits that are strongly affected by arousal. Almost a working dog trials including tracking, search and decade prior to the suggestion by Taylor and Mills protection work, and that boldness scores correlated (2006), Vincent and Leahy (1997) noticed that there with scores for playfulness, curiosity-fearfulness and was a relationship between heart rate variability of sociability. Several other authors have studied the guide dogs while on a training walk and how the dogs shy-bold continuum in dogs, showing somewhat normally responded to novelty and stress. Dogs comparable results and indicating that these traits described by trainers as 'calm and non-stress prone' are generally stable over short periods of time. Traits showed low mean baseline heart rates and smaller that vary independently of the shy-bold super trait variation in overall heart rate during training walks, include chase-proneness, intelligence and overall than those described by trainers as 'excitable and aggression. While no studies correlating the shy-bold stress-prone'. Similarly, Wright, Mills and Pollux axis with physiological variables have yet been (2012) showed that owner reports of poor impulse published, Svartberg (2002) draws parallels between control and a reduced tolerance of delayed rewards in shy-bold axis and the emotionality-coping style model pet dogs during a learning task, correlated with low proposed by Koolhaas et al. (1999), which linked urinary levels of serotonin and dopamine. Low activity of the sympathetic and parasympatheticnervous systems to stress reactivity (termed circulating levels of serotonin have been linked with 'emotionality') and characteristic behavioural patterns anxiety disorders, characterised by intense or for dealing with stressors (termed 'coping style') prolonged stress during typically non-stressful (Figure 2). While Starling, Branson, Thomson, and situations, in people and other animals. Both of McGreevy (2013) also suggest that aspects of the these results indicate that a simple measure of shy-bold continuum could be analogous to the arousal, such as heart rate, could be applied within a emotionality-coping styles model, the relationship behavioural test protocol to aid understanding of the between these two models remains unclear, limiting the application of findings from studies of the shy- Within the field of canine behavioural research, study bold continuum to those based on the emotionality- into the biological basis of temperament and behavioural traits stems primarily from two areas.
Firstly, behavioural biochemistry has been studiedfrequently as the basis for diagnostic and treatmentoptions for maladaptive stress-related and aggressivebehaviours in dogs, treatment of which now oftenincludes psychotropic drugs like Valium and Prozac.
Studies in this area have tended to focus on thefunction (or lack of function) of key neurotransmitters,such as serotonin and dopamine, occurring alongsidespecific behavioural syndromes (Riva, Bondiolotti,Michelazzi, Verga, & Carenzi, 2008; Rosado, García-Belenguer, León, Chacón, Villegas, & Palacio, 2010).
Secondly, selection of working dogs has lead to asteadily increasing focus on the heritability ofmeasurable temperament traits (Kubinyi, Sasvari-Szekely, & Miklosi, 2011; Meyer, Schawalder, Gaillard,& Dolf, 2012; Takeuchi, Hashizume, Arata, Inoue- FIGURe 2 The dual axes model of 'emotionality-coping style' Murayama, Maki, Hart, & Mori, 2009; Wilsson & proposed by Koolhaas, De Boer, Buwalda, and Van Reenen Sundgren, 1997). Due to the different focuses of (2007). In this model, the typical behavioural response of a each area of research into the biochemical and stressed animal falls along a horizontal 'coping style' axis, genetic basis of temperament, no clear picture of the while the vertical 'emotionality' axis indicates how easily physiology controls temperament has emerged.
one limitation of the current literature on the ‘shy- However, the broad traits of fearfulness/fearlessness, bold’ axis in dogs is that test methodologies have reactivity/impulsivity and sociability have consistently focused upon 'proactive' coping styles, i.e. those appeared to correlate with each other, which could be characterised by active avoidance or approach of the indicative of a 'higher-order' trait observed in a stressor (Koolhaas et al., 2007) which fall on the far number of mammalian species termed the ‘shy-bold right of the graph in Figure 2. Behavioural tests axis’, suggested to have some physiological basis commonly presented in the shy-bold literature are scored based on intensity of gross behaviour repertoires, placing animals that actively avoid a development of questionnaires based on personality stressor at one end of the continuum with those descriptions by people well known to the dog (Ley, actively approaching a stressor at the other. However, Bennett, & Coleman, 2009; Ley, McGreevy, & these tests fail to include subtle communications of Bennett, 2009), as these have proved to be quite passively responding dogs during tests and instead reliable when used for human studies. However, group all passively responding dogs together variation in definitions of traits like 'friendly and regardless of arousal level (Horváth, Igyártó, Magyar, sociable' between studies limits the use of these & Miklósi, 2007; Svartberg, 2002). This presents a questionnaires for behavioural test development, as significant limitation in the broader application of we cannot compare directly between personality these findings, as differentiation between dogs traits described owners and behaviours observed in experiencing a negligible level of stress (i.e. those in tests. These limitations highlight the importance of the 'docile' quadrant of Figure 2), and those that are objective test measures in obtaining accurate stressed but respond in a passive manner (i.e. those information about dogs, although it appears that, at in the 'shy' quadrant of Figure 2), is not taking place.
this point in time, using multiple questionnaires to Koolhaas et al. (2007) also points out that despite allow comparisons to be made between multiple animals having a characteristic behavioural style for dealing with stressors, or 'default strategy', which is our research project seeks to evaluate a number of indicative of their temperament, the behaviour of an behavioural test battery subtests currently used in individual is highly situation-dependent. This is behavioural tests or proposed in the literature, using because animals will choose the strategy perceived to measures of both behaviour and real-time heart rate work most effectively in reducing stress, a choice that recordings. A total of 16 interactive and 25 sound- is influenced by both previous learning and the level based subtests that have shown some promise for of threat perceived by the animal during the eliciting a variety of behaviours in a test situation, encounter. Few studies in the shy-bold continuum were selected from the literature. one hundred and literature acknowledge these limitations, failing to eighty dogs of various breeds and cross-breeds were address the requirements for multiple subtests to brought into the testing facility by their owners and determine the 'default' strategy of the animal, while were subjected to the same 60-minute protocol by ensuring that all animals tested experience equal the same tester. All dogs wore a Polar RS800CX levels of stress during each subtest. Careful test telemetric heart rate monitor throughout the tests, design and an objective assessment of arousal levels and were also video recorded to allow accurate to indicate perceived stress, such as heart rate, could scoring of behaviours at a later date. Blood samples potentially address both of these issues in were collected on arrival at the facility and directly following the tests, to allow heart rate results to be To further complicate test development, behavioural compared to changes in plasma cortisol and prolactin tests of dogs are often compared against owner levels over the test, as additional indicators of stress reports of their dogs' typical behaviour in order to intensity. While the tests were conducted, owners assess how accurately the test reflects the dogs' true were asked to fill out a questionnaire about nature. This information is often gathered via a themselves and their dogs, which included questionnaire, asking the owners to indicate how demographic questions, the Canine Behavioural and often their dog behaves a certain way, or how they Research Questionnaire (CBARQ), the Dog Impulsivity think their dog 'feels' about a particular situation. In Assessment Survey (DIAS) and the Monash Dog- an effort to overcome issues of owner bias and poor owner Relationship Survey (M-DoRS). Results quality descriptions of pet dog behaviour as reported obtained during the tests will be compared to results by Mariti et al. (2012) and Tami & Gallagher (2009), obtained from the owner-questionnaires, in order to design of owner report questionnaires has trended ascertain whether the behaviour seen during the test away from subjective descriptions of behaviour, like was typical of owner perceptions of the dogs' 'my dog appears happy when visitors arrive', in favour of physical descriptions of behaviour such as 'my dog While no formal evaluations of test results have been freely approaches visitors with his tail wagging and carried out as yet, a wide variety of behaviours and mouth open', with the frequency or intensity of the patterns were observed and several interesting behaviour scored against numerical scales (Hsu & preliminary results were noted during testing. Dogs Serpell, 2003). While this approach limits subjectivity with a reported history of separation-related anxiety of owner reports to a degree, the focus on frequency maintained high baseline heart rates throughout the of gross dog behaviours such as barking, growling, test, with a noticeable drop in heart rate at the onset baring of teeth and lunging (Hsu & Serpell, 2003) and of each subtest that produced a response. This relative lack of subtle behavioural indicators of pattern was opposite to that shown by all other dogs, stress, such as lip licking or averting of gaze, means who maintained a lower baseline heart rate with an this method also suffers from an inability to properly increase at the start of each subtest that produced a identify stressed, but passively responding dogs.
stress response. However, these dogs showed less Some research focus has been directed towards the variation overall than dogs who appeared to experience a similar level of overall stress throughout Koolhaas, J., De Boer, S., Buwalda, B., & Van Reenen, K. (2007).
the test, that was not related to being separated from Individual Variation in Coping with Stress: A Multidimensional their owners (i.e. dogs with pronounced fear Approach of Ultimate and Proximate Mechanisms. Brain, Behavior responses to sounds or novel objects). Dogs that were described by their owners as typically adapting Koolhaas, J.M., Korte, S.M., De Boer, S.F., Van Der Vegt, B.J., Van well to new situations and being relatively 'non-stress Reenen, C.G., Hopster, H., Blokhuis, H.J. (1999). Coping styles in prone' exhibited a steady decrease in baseline heart animals: current status in behavior and stress-physiology.
rate throughout the test, as well as a relatively low Neuroscience & Biobehavioral Reviews, 23(7), 925-935. doi:10.1016/s0149-7634(99)00026-3 level of variation in heart rate, particularly during theauditory section of the test in which the dogs were Kubinyi, e., Sasvari-Szekely, M., & Miklosi, A. (2011). Genetics and freely roaming the room (some even fell asleep!).
the social behaviour of the dog revisited: searching for genes Interestingly, dogs who appeared easily stressed relating to personality in dogs. In M. Inoue-Murayama (ed.), Fromgenes to animal behaviour (pp. 255-274). Budapest, Hungary: during the test, but were described by their owners as playful or somewhat chase-prone (i.e. they typicallyshow a pronounced predatory response to small Ley, J.M., Bennett, P.C., & Coleman, G.J. (2009). A refinement and animals or toys), did not consistently display those validation of the Monash Canine Personality Questionnaire (MCPQ).
Applied Animal Behaviour Science, 116(2–4), 220-227. doi: behaviours during the test, and showed a reduced behavioural repertoire overall. Dogs that weresomewhat relaxed throughout the protocol were more Ley, J.M., McGreevy, P., & Bennett, P.C. (2009). Inter-rater and test–retest reliability of the Monash Canine Personality Questionnaire- likely to show a fuller behavioural repertoire and were Revised (MCPQ-R). Applied Animal Behaviour Science, 119(1–2), more likely to exhibit play or predatory behaviours 85-90. doi: 10.1016/j.applanim.2009.02.027 during a number of subtests, even if they were notdescribed as highly playful or chase-prone by their Mariti, C., Gazzano, A., Moore, J.L., Baragli, P., Chelli, L., & Sighieri,C. (2012). Perception of dogs’ stress by their owners. Journal of owners. This indicates that playfulness is perhaps Veterinary Behavior: Clinical Applications and Research, 7(4), 213- not directly related to a dog's typical 'coping style' as suggested by previous literature, but rather that indogs falling on the higher end of the 'emotionality' Meyer, F., Schawalder, P., Gaillard, C., & Dolf, G. (2012). estimationof genetic parameters for behavior based on results of German axis, play and predatory behaviours are suppressed Shepherd Dogs in Switzerland. Applied Animal Behaviour Science, by their stress response in non-familiar situations or 140(1–2), 53-61. doi: 10.1016/j.applanim.2012.05.007 Mornement, K.M., Coleman, G.J., Toukhsati, S., & Bennett, P.C.
If the results of this project show promise in the use (2010). A Review of Behavioral Assessment Protocols Used by of the telemetric monitors during a test battery, we Australian Animal Shelters to Determine the Adoption Suitability of would like to proceed by refining the protocol and Dogs. Journal of Applied Animal Welfare Science, 13(4), 314-329.
producing a 'gold standard' test. Ideally, the aim is to produce a test that can be used to aid good dog- Riva, J., Bondiolotti, G., Michelazzi, M., Verga, M., & Carenzi, C.
owner matching in re-homing programs, and that (2008). Anxiety related behavioural disorders and could also be used to aid animal management neurotransmitters in dogs. Applied Animal Behaviour Science, professionals to identify problems with owned dogs 114(1–2), 168-181. doi: 10.1016/j.applanim.2008.01.020 and suggest appropriate, problem-specific Rosado, B., García-Belenguer, S., León, M., Chacón, G., Villegas, A., management and treatment options. Failing that, we & Palacio, J. (2010). Blood concentrations of serotonin, cortisol hope the results stimulate a broader focus during test and dehydroepiandrosterone in aggressive dogs. Applied Animal development in the future, accounting for both Behaviour Science, 123(3–4), 124-130. doi:10.1016/j.applanim.2010.01.009 behavioural and physiological responses in order toachieve a better understanding of dogs overall.
Starling, M.J., Branson, n., Thomson, P.C., & McGreevy, P.D. (2013).
“Boldness” in the domestic dog differs among breeds and breed groups. Behavioural Processes, 97(0), 53-62. doi:http://dx.doi.org/10.1016/j.beproc.2013.04.008 Bräm, M., Doherr, M.G., Lehmann, D., Mills, D., & Steiger, A.
Svartberg, K. (2002). Shyness–boldness predicts performance in (2008). evaluating aggressive behavior in dogs: a comparison of 3 working dogs. Applied Animal Behaviour Science, 79(2), 157-174.
tests. Journal of Veterinary Behavior: Clinical Applications andResearch, 3(4), 152-160. doi: 10.1016/j.jveb.2008.04.001 Diederich, C., & Giffroy, J.M. (2006). Behavioural testing in dogs: A Svartberg, K. (2005). A comparison of behaviour in test and in review of methodology in search for standardisation. Applied everyday life: evidence of three consistent boldness-related Animal Behaviour Science, 97(1), 51-72. doi: personality traits in dogs. Applied Animal Behaviour Science, 91(1– 2), 103-128. doi: 10.1016/j.applanim.2004.08.030 Horváth, Z., Igyártó, B.Z., Magyar, A., & Miklósi, Á. (2007). Three Takeuchi, y., Hashizume, C., Arata, S., Inoue-Murayama, M., Maki, different coping styles in police dogs exposed to a short-term T., Hart, B.L., & Mori, y. (2009). An approach to canine behavioural challenge. Hormones and Behavior, 52(5), 621-630. doi: genetics employing guide dogs for the blind. Animal Genetics, 40(2), 217-224. doi: 10.1111/j.1365-2052.2008.01823.x Hsu, y., & Serpell, J.A. (2003). Development and validation of a Tami, G., & Gallagher, A. (2009). Description of the behaviour of questionnaire for measuring behaviour and temperament traits in domestic dog (Canis familiaris) by experienced and inexperienced pet dogs. Journal of American Veterinary Medical Association, people. Applied Animal Behaviour Science, 120(3–4), 159-169.
Taylor, K.D., & Mills, D.S. (2006). The development andassessment of temperament tests for adult companion dogs.
Journal of Veterinary Behavior: Clinical Applications and Research,1(3), 94-108. doi: 10.1016/j.jveb.2006.09.002 Valsecchi, P., Barnard, S., Stefanini, C., & normando, S. (2011).
Temperament test for re-homed dogs validated through directbehavioral observation in shelter and home environment. Journalof Veterinary Behavior: Clinical Applications and Research, 6(3),161-177. doi: 10.1016/j.jveb.2011.01.002 Vincent, I.C., & Leahy, R.A. (1997). Real-time non-invasivemeasurement of heart rate in working dogs: a technique withpotential applications in the objective assessment of welfareproblems. The Veterinary Journal, 153(2), 179-183. doi:10.1016/s1090-0233(97)80038-x Wilsson, e., & Sundgren, P.e. (1997). The use of a behaviour testfor selection of dogs for service and breeding. II. Heritability fortested parameters and effect of selection based on service dogcharacteristics. Applied Animal Behaviour Science, 54(2–3), 235-241. doi: 10.1016/s0168-1591(96)01175-6 Wright, H.F., Mills, D.S., & Pollux, P.M.J. (2012). Behavioural andphysiological correlates of impulsivity in the domestic dog (Canisfamiliaris). Physiology & Behavior, 105(3), 676-682. doi:10.1016/j.physbeh.2011.09.019 About the authorDiana Rayment is a PhD candidate at La TrobeUniversity's Melbourne Campus, under thesupervision of Dr Bert De Groef, Dr Linda Marstonand Dr Richard Peters. She has several years'experience in veterinary nursing and as aprofessional dog trainer, where she developed herinterest and passion for companion dog behaviourand the human-dog relationship. She currentlylectures La Trobe's undergraduate students aboutcompanion animal issues in modern society.
Diana Raymentemail: D.Rayment@latrobe.edu.au

Source: http://www.aiam.com.au/resources/files/proceedings/AIAM%202013/AIAM%202013%20Diana%20Rayment.pdf