Fear, avoidance and safety signals as rewards

48

Journal of Veterinary Behavior, Vol 5, No 1, January/February 2010

in Italy. Age is never the main selection criterion; however, ponies tend to start at a younger age (6.5 versus 13 years, Z5-3.7, p , 0.001; Mann-Whitney-U), and to be younger (13 years vs. 18; Z5-4.3, p , 0.001; Mann-Whitney-U) than horses. Retirement age varies depending on health. Thirty-six of the horses/ponies have worked in EAIs for more than 5 years. Although working conditions in EAIs are very diverse, none of the horses get specific training, apart from habituation sessions to EAI stimuli in 6 stable yards. Workload varies from 1 to 10 30-minute sessions/ horse/week. In 17 stables the same animals also work as riding school horses. All horses are housed in individual stalls. Horses have free access to a paddock at two stables. In 12 stables access to a paddock is restricted to part of the day. The public has no access to the stables/paddocks in 9 yards. Possible discomfort or stress is never systematically assessed. Increased latency in obeying aides, disobedience, stopping, mild bucking are the most frequently reported problems. Nine stables report aggressive reactions (threats through flattening ears, vocalizing, backing) towards some clients. These are sources of concern and warrant further investigation. Key words: horse; equine-assisted intervention, welfare

Acknowledgments We acknowledge the help of Dr. Guzzo and all respondents.

FEEDING ROUTINE RISK FACTORS ASSOCIATED WITH PRE-FEEDING BEHAVIOR PROBLEMS IN UK LEISURE HORSES E. Creighton1, J. Hockenhull2,* 1 School of Agriculture, Food and Rural Development, Newcastle University, Newcastle NE1 7RU, UK 2 Anthrozoology Unit, Chester Centre for Stress Research, University of Chester, Chester CH1 4BJ, UK *Corresponding author: [email protected] Domestic horses are typically provided with a diet and feeding regime that differs markedly from their evolutionary requirements. Modern feeding practices have been implicated as risk factors for a number of physiological and behavioral problems. Various anticipatory behaviors can be performed prior to feeding and are subsequently reinforced by the arrival of food. Such behaviors are often viewed negatively by owners and once established maybe performed in contexts outside feeding. This study used a large-scale Internet survey to quantify the performance of feeding-related problem behavior in UK leisure horses and to identify feeding practice risk factors for this behavior. Behavior data were collected for 890 horses; 70% of these performed some form of feeding-related behavior problem. Principal components analysis identified three groups of

problems: frustration behavior (49% of horses), aggressive behavior (44%) and stereotypic behavior (39%). Feeding routine risk factors associated with each of the behavior components were explored using logistic regression analyses. Forage availability was associated with two components; restricted access increased the risk of frustration behavior, while feeding ad lib reduced the risk of pre-feeding aggression. Pre-feeding stereotypies were affected by the frequency concentrate feeds were provided and the presence of other horses during feeding. The use of multiple dietary supplements, including nutritional calmers, was associated with increased risk of pre-feeding aggression and stereotypic behavior. The performance of behaviors attributed to feeding tit-bits by hand was associated with an increased risk of all components, suggesting these behaviors are indicative of generic feeding-related problems. Outside the feeding routine, participating in regular work reduced the occurrence of all three behavior problem components. Pre-feeding behavior problems are associated with feeding routines at odds with horses’ evolutionary requirements. The performance of tit-bit-related behavior is indicative of wider feeding-related problems and owners should be alerted to reconsider their overall feeding practices. Key words: horse; feeding practices; feeding behavior

FEAR, AVOIDANCE AND SAFETY SIGNALS AS REWARDS Robert Boakes* School of Psychology, University of Sydney, NSW 2006, Australia *Correspondence author: [email protected] To explain how animals acquire avoidance behavior, i.e. a response that prevents the occurrence of some aversive event, two-factor theory was developed on the basis of experiments with dogs and rats. The original version appealed to two learning processes: animals first learn to fear a stimulus – a warning signal – indicating that an aversive event such as a shock is likely to occur in the near future; and, second, when they make an appropriate response, this response is reinforced by reduction of fear, i.e. negative reinforcement. When combined with some minor additional principles, this description provides a good account of how an animal first acquires an avoidance response. However, it does not give a satisfactory account of how avoidance behavior is maintained. In particular, well-trained animals can continue to perform some response without displaying any fear. If no longer fear, then what motivates continued good performance? The answer to this question has been to add the concept of a conditioned inhibitor of fear or ‘safety signal’. This is a stimulus signalling that an otherwise expected aversive event will not occur. Such signals not only inhibit fear, but also serve as powerful positive rewards, a function that is highly resistant to extinction. The provision of clear safety signals, namely, immediate feedback that an animal

Abstracts has made the required response, can increase the speed with which an animal learns some new behavior and can reduce stress during training. Key words: avoidance learning; negative reinforcement; safety signal; 2-factor theory; conditioned inhibitor

RELIABILITY OF SALIVARY CORTISOL MEASURES IN DOGS IN TRAINING CONTEXT S. Ligout, H. Wright, K. van Driel, F. Gladwell, D.S. Mills, J.J. Cooper* Department of Biological Sciences, University of Lincoln, Riseholme Park, Lincoln, LN2 2PY, UK *Corresponding author: [email protected] Salivary cortisol has been widely used as a means of assessing stress in dogs (Beerda et al., 1988; Schalke et al., 2007), though several authors have questioned whether handling and the use of food/chews can affect measures (Kobelt et al., 2003; Dreschel and Granger, 2009). We investigated effects of food on salivary cortisol measures in a handling/training context. For the first study 10 dogs were used to investigate if food introduced unpredictable variation in salivary cortisol. For each dog tested, a control sample was taken, followed within 30 seconds by a sample with cheese (n 5 10), sausage (n 5 5) or chicken (n 5 5). Cortisol assayed with food tended to be lower (2.14 6 0.12ng/ml) than controls (2.59 6 0.23; t 5 2.08, p , 0.05). This may be due to dilution of samples by greater production of fluid, or because food contamination reduces the efficacy of ELISA. There was a high positive correlation between the measures from cheese samples and their controls (Pearson’s r 5 0.922, p , 0.001), but no correlation with the sausage and chicken. In the second study, we used cheese to encourage salivation and 10 dogs were used to investigate changes in salivary cortisol following training. Dogs responded to three common vocal commands; ‘‘come’’, ‘‘stay’’ and ‘‘leave’’, together with a fourth command of the owner’s suggestion, each given by the owner, and by a person unfamiliar to the dog. Saliva was sampled before training, then at 10, 15, 20 and 25 minutes following start of training. There was no difference in salivary cortisol measures between samples, but a high concordance across samples (Kendall’s W 5 0.858, p , 0.001). Our data suggests cheese does not introduce unpredictable variation is salivary cortisol measures, and consequently may be a more appropriate aid to encourage salivation than sausage, chicken or meat flavoured rope (Dreschel and Granger, 2009). Key words: dog training; salivary cortisol

References Beerda, B., Schilder, M.B.H., van Hooff, J.A.R.A.M., de Vries, H.W., Mol, J.A., 1998. Behavioural, saliva cortisol and heart rate responses

49 to different types of stimuli in dogs. Appl. Anim. Behav. Sci. 58, 365–381. Dreschel, N.A., Granger, D.A., 2009. Methods of collection for salivary cortisol measurements in dogs. Horm. Behav. 55, 163–168. Kobelt, A., Hemsworth, P.H., Barnett, J.L., Butler, K.L., 2003. Sources of sampling variation in saliva cortisol in dogs. Res. Vet. Sci. 75, 157–161. Schalke, E., Stichnoth, J., Ott, S., Jones-Baade, R., 2007. Clinical signs caused by the use of electric training collars on dogs in everyday life situations. Appl. Anim. Behav. Sci. 105, 369–380.

A CONCEPTUAL STUDY ABOUT ‘ANIMAL-ASSISTED INTERVENTIONS’ IN ANTWERP (BELGIUM) Tina Pira1,4, Cinzia Stefanini2, Lieve Meers1,4,*, Simona Normando3,4, William Ellery Samuels5, Frank ¨ dberg1 O 1 Ghent University, Department of Nutrition, Genetics, and Ethology, Merelbeke, Belgium 2 Associazione Vita da Cani, Arese (MI), Italy 3 Padua University, Dipartimento di Scienze Sperimentali Veterinarie, Legnaro (PD), Italy 4 University College Ghent, Faculty of Biosciences and Landscape Architecture, Melle, Belgium 5 City University New York, College of Staten Island NY, USA *Corresponding author: [email protected] For the past 20 years, a growing number of health care facilities and educational settings have started ‘animalassisted intervention’ programs (AAIs). These initiatives include ‘animal-assisted therapy’ in which client oriented therapeutic goals are set and progress records are kept, as well as ‘animal-assisted activities’ without specific goals for each client and ‘animal-assisted learning’ in which teachers have educational goals for each student. The aim of this study was to map the prevalence and the most common logistical issues for AAIs in Antwerp, a Belgian province. We collected data from 284 institutions (146 elderly homes, 18 centres for palliative care, 30 clinics, 65 institutions for people with challenges and 72 centres for youth at risk) using a semi-structured telephone interview. Eighty-five percent of the facilities offer AAIs, mainly involving residential or visiting animals that provide for various activities. Animals involved are dogs (36%), horses/ponies (20%), farm animals (15%), rodents (11%), birds (10%), or cats (4%). Other species (,1%) are dolphins, snakes, fish, spiders, falcons, snails, foxes and a ferret. Fifty-eight percent of the AAIs had goals such as improving physical activity (16%), cognitive skills (15%), psychological abilities (13%), rendering activities more interesting (11%), reaching individual goals and improving social contact (8%). Forty-two percent of the programs kept records on either the animals themselves or the work they perform. Twenty-five percent of the respondents reported occasional negative effects on the animals involved. Problems differ strongly between programs and include various forms of animal neglect, animal abuse, and aggression towards the animals, stress signals, hiding and running away