- Email: [email protected]
Genetic evaluation of behaviour in dogs
Abstracts / Journal of Veterinary Behavior 10 (2015) 438e448
3 Genetic evaluation of behaviour in dogs PER ARVELIUS* Swedish University of Agricultural Sciences, Department of Animal Breeding and Genetics, P.O. Box 7023, SE-750 07 Uppsala, Sweden *Corresponding author: [email protected] A dog’s behavioural characteristics are important for the dog, for the dog owner and for society as a whole. Because behavioural traits are heritable and can be changed by breeding, they should be included as an important part of the breeding goal. To be effective when selecting breeding animals, good methods for measuring behaviour are essential; the measurements must show genetic variation and also be genetically correlated to the breeding goal. Behavioural data can be collected in many different ways. For example, measurements can be more or less objective, and the ratings can refer to behaviours displayed in a specific situation or to an overall interpretation indicating the degree of expression of predefined traits. Choice of method can be expected to affect the usefulness of the measurements from a breeding perspective. The potential genetic progress also depends on how the measurements are utilized. In livestock breeding, effective techniques for evaluating animals genetically have been developed and extensively used with great success. Instead of selecting breeding animals based on their phenotypic performance (which still is the most common practice in dog breeding), these techniques allow for estimating an animal’s breeding value by adjusting the phenotype for environmental factors and by taking information on relatives into account. Thereby, breeding animals can be more accurately selected for their genetic qualities. Advantages and disadvantages with different methods for measuring dog behaviour for breeding purposes will be discussed, as well as the prospects for improving dog behaviour by applying modern techniques for genetic evaluation. Keywords: behaviour; breeding; dog; genetic evaluation; recording
4 Measuring working dog performance N.J. ROONEY* Animal Welfare and Behaviour Group, School of Veterinary Science, University of Bristol, Langford, UK BS40 5DU *Corresponding author: [email protected] If we are to select and breed for optimal working ability, it is vital that we have meaningful and reliable measures of performance. For any working dog role, there are multiple dimensions to performance and the literature abounds with different, often arbitrary, ways of measuring or rating these. Using arms and explosives search dogs as an example, we have taken a systematic and evidence-based approach to derive methods to quantify working ability. First, derivation of a meaningful vocabulary can be achieved through systematic workshops, interviews and using psychosocial techniques to prioritise the most important attributes of performance. Standardised testing sessions can be useful tools to measure multiple aspects of performance, providing they represent the range of roles the dog is required to perform. When well designed test outcomes correlate well with trainer’s’ assessments they are less prone to rater error (e.g. “halo” effects), and hence discriminant validity is improved. However, to truly measure working performance, one needs to monitor dogs in their daily work and training. This needs to be done by handlers and trainers, although there is variability in rating ability with the former sometimes showing leniency towards their own dogs. Interactive training in the use of scales and benchmarking of scales can improve this ability, and is therefore a vital
439
component of any monitoring scheme. It is only through the derivation of honest, well-validated performance measures that we can ensure maintenance and improvement of working dogs’ performance. Keywords: working dog; performance; reliable measures
5 Performance assessments in dogsedetermining ’good’ behavioral measures and phenotypes BJÖRN FORKMAN* Dept of Large Animal Science, University of Copenhagen, Grønnegårdsvej 8, 1870 Frederiksberg C, Denmark *Corresponding author: [email protected] When trying to predict the behavior of a dog in a specific situation we are assessing the underlying motivational tendency of the dog. Motivations cannot be observed, only inferred. There is no test that can test the motivation directly, and to be able to assess it correctly it is therefore important to use more than one measure/one test even when trying to predict the behavior in a specific situation. The more well defined the situation to be predicted is the easier it is to predict. In many cases the situation cannot be well defined however and it is known that even minor changes, e.g. meeting an unknown man as opposed to meeting an unknown woman is sufficient to change the behavior of the dog. Any behavioral test has a certain sensitivity and a certain specificity. A high sensitivity means that the test is likely to find all cases of a given behavior, a high specificity that it will only measure that behavior and not others. Because there is a tradeoff between sensitivity and specificity it is important to determine what is most important when constructing the test and interpreting the results. With few exceptions these concepts are not used in the behavioral literature but instead agreement (measured as correlation) is used as a measure of how well the test performs, e.g. between aggression in the test situation and previous history of bites. Depending on the aim of the assessment this may lead to erroneous conclusions. Keywords: canine performance; good measures; behavioral phenotypes; dogs
6 Middle latency response testing for auditory cognition in canines PETER M. SCHEIFELE 1, *, KRISTINE E. SONSTROM 1, KAREN L. OVERALL 2, ARTHUR E. DUNHAM 2 1 University of Cincinnati FETCHLAB, 3202 Eden Ave, Cincinnati, USA 45267-0379 2 Biology Department, University of Pennsylvania, Philadelphia, PA, USA *Corresponding author: [email protected] Event Related Potentials (ERPs) using Mismatch Negativity (MMN) and the Auditory Middle Latency Response (AMLR) measure changes in cognitive brain activity in direct response to auditory stimuli. The MMN and AMLR generate responses from the auditory cortex, thalamus and frontal cortex, areas with connections to the hypothalamus, hippocampus and amygdala. Specifically, MMN was identified as a non-attentive response to an oddball “deviant” tone presented within a series of tones. The subject does not need to attend to the stimulus in any of these tests because the response is derived from auditory discrimination processing in auditory primary and association areas. Hence, MMN and AMLR testing assays central auditory perception, auditory memory, and attentional processes associated with auditory sensory input to conscious perception and higher forms of memory. When combined with Brainstem Auditory Evoked Response (BAER) testing to measure
3 Genetic evaluation of behaviour in dogs PER ARVELIUS* Swedish University of Agricultural Sciences, Department of Animal Breeding and Genetics, P.O. Box 7023, SE-750 07 Uppsala, Sweden *Corresponding author: [email protected] A dog’s behavioural characteristics are important for the dog, for the dog owner and for society as a whole. Because behavioural traits are heritable and can be changed by breeding, they should be included as an important part of the breeding goal. To be effective when selecting breeding animals, good methods for measuring behaviour are essential; the measurements must show genetic variation and also be genetically correlated to the breeding goal. Behavioural data can be collected in many different ways. For example, measurements can be more or less objective, and the ratings can refer to behaviours displayed in a specific situation or to an overall interpretation indicating the degree of expression of predefined traits. Choice of method can be expected to affect the usefulness of the measurements from a breeding perspective. The potential genetic progress also depends on how the measurements are utilized. In livestock breeding, effective techniques for evaluating animals genetically have been developed and extensively used with great success. Instead of selecting breeding animals based on their phenotypic performance (which still is the most common practice in dog breeding), these techniques allow for estimating an animal’s breeding value by adjusting the phenotype for environmental factors and by taking information on relatives into account. Thereby, breeding animals can be more accurately selected for their genetic qualities. Advantages and disadvantages with different methods for measuring dog behaviour for breeding purposes will be discussed, as well as the prospects for improving dog behaviour by applying modern techniques for genetic evaluation. Keywords: behaviour; breeding; dog; genetic evaluation; recording
4 Measuring working dog performance N.J. ROONEY* Animal Welfare and Behaviour Group, School of Veterinary Science, University of Bristol, Langford, UK BS40 5DU *Corresponding author: [email protected] If we are to select and breed for optimal working ability, it is vital that we have meaningful and reliable measures of performance. For any working dog role, there are multiple dimensions to performance and the literature abounds with different, often arbitrary, ways of measuring or rating these. Using arms and explosives search dogs as an example, we have taken a systematic and evidence-based approach to derive methods to quantify working ability. First, derivation of a meaningful vocabulary can be achieved through systematic workshops, interviews and using psychosocial techniques to prioritise the most important attributes of performance. Standardised testing sessions can be useful tools to measure multiple aspects of performance, providing they represent the range of roles the dog is required to perform. When well designed test outcomes correlate well with trainer’s’ assessments they are less prone to rater error (e.g. “halo” effects), and hence discriminant validity is improved. However, to truly measure working performance, one needs to monitor dogs in their daily work and training. This needs to be done by handlers and trainers, although there is variability in rating ability with the former sometimes showing leniency towards their own dogs. Interactive training in the use of scales and benchmarking of scales can improve this ability, and is therefore a vital
439
component of any monitoring scheme. It is only through the derivation of honest, well-validated performance measures that we can ensure maintenance and improvement of working dogs’ performance. Keywords: working dog; performance; reliable measures
5 Performance assessments in dogsedetermining ’good’ behavioral measures and phenotypes BJÖRN FORKMAN* Dept of Large Animal Science, University of Copenhagen, Grønnegårdsvej 8, 1870 Frederiksberg C, Denmark *Corresponding author: [email protected] When trying to predict the behavior of a dog in a specific situation we are assessing the underlying motivational tendency of the dog. Motivations cannot be observed, only inferred. There is no test that can test the motivation directly, and to be able to assess it correctly it is therefore important to use more than one measure/one test even when trying to predict the behavior in a specific situation. The more well defined the situation to be predicted is the easier it is to predict. In many cases the situation cannot be well defined however and it is known that even minor changes, e.g. meeting an unknown man as opposed to meeting an unknown woman is sufficient to change the behavior of the dog. Any behavioral test has a certain sensitivity and a certain specificity. A high sensitivity means that the test is likely to find all cases of a given behavior, a high specificity that it will only measure that behavior and not others. Because there is a tradeoff between sensitivity and specificity it is important to determine what is most important when constructing the test and interpreting the results. With few exceptions these concepts are not used in the behavioral literature but instead agreement (measured as correlation) is used as a measure of how well the test performs, e.g. between aggression in the test situation and previous history of bites. Depending on the aim of the assessment this may lead to erroneous conclusions. Keywords: canine performance; good measures; behavioral phenotypes; dogs
6 Middle latency response testing for auditory cognition in canines PETER M. SCHEIFELE 1, *, KRISTINE E. SONSTROM 1, KAREN L. OVERALL 2, ARTHUR E. DUNHAM 2 1 University of Cincinnati FETCHLAB, 3202 Eden Ave, Cincinnati, USA 45267-0379 2 Biology Department, University of Pennsylvania, Philadelphia, PA, USA *Corresponding author: [email protected] Event Related Potentials (ERPs) using Mismatch Negativity (MMN) and the Auditory Middle Latency Response (AMLR) measure changes in cognitive brain activity in direct response to auditory stimuli. The MMN and AMLR generate responses from the auditory cortex, thalamus and frontal cortex, areas with connections to the hypothalamus, hippocampus and amygdala. Specifically, MMN was identified as a non-attentive response to an oddball “deviant” tone presented within a series of tones. The subject does not need to attend to the stimulus in any of these tests because the response is derived from auditory discrimination processing in auditory primary and association areas. Hence, MMN and AMLR testing assays central auditory perception, auditory memory, and attentional processes associated with auditory sensory input to conscious perception and higher forms of memory. When combined with Brainstem Auditory Evoked Response (BAER) testing to measure