HOW DOES TEMPERAMENT AND

INFLUENCE LEARNED AVERSION TRAINING

IN DOMESTIC ?

Robyn Louise Taylor

Murdoch University

School of Veterinary and Life Sciences

Bachelor of Science Honours in Conservation and Wildlife Biology

Western Australia

This dissertation is submitted for the degree of Honours of Conservation and Wildlife Biology

December 2017 How does Temperament and Breed Influence Learned aversion Training in Domestic Dogs? Robyn Louise Taylor October 2017 Murdoch University

2 Word Template by Friedman & Morgan 2014 Morgan & Friedman by Word Template

Thank you to my supervisors, Tracey Kreplins and Trish Fleming from the school of Veterinary and Life Sciences, Murdoch University.

Thank you to my family and friends for their constant support and encouragement while I pursued this part of my life and career.

To Damon van der Linde, thank you for volunteering to be my assistant during the course of my training, and for keeping me company on the long drives.

i Word Template by Friedman & Morgan 2014 Morgan & Friedman by Word Template

“It was the tea, the morning jogs, my family and my art that kept my mind focussed and calm on the task at hand.” Robyn Taylor.

ii Word Template by Friedman & Morgan 2014 Morgan & Friedman by Word Template

Declaration

This dissertation is the result of my own work and includes nothing, which is the outcome of

work done in collaboration except where specifically indicated in the text. It has not been

previously submitted, in part or whole, to any university of institution for any degree, diploma,

or other qualification.

In accordance with the School of Veterinary and Life Sciences, this thesis is does not exceed

25,000 words, and it contains less than 20 figures.

Signed:______

rd Date:______3 October 2017

Robyn Louise Taylor Bachelor of Science Honours Student

Murdoch University

iii Word Template by Friedman & Morgan 2014 Morgan & Friedman by Word Template Temperament and breed influences on learned aversion in domestic dogs

Summary / Abstract

Accidental poisoning of domestic dogs is an unfortunate occurrence when using poison baits to control introduced pests such as feral cats, red foxes, and wild dogs. This study aimed to investigate how (i.e., toy, , sporting, domestic working and working dogs) and temperament influence aversion methods. Using non-poisonous baits and a device which emits a small electrical correction, domestic dogs were trained to avoid commercially-available non- toxic FoxOff® baits. Fifty-six dogs were recruited through online media services (e.g.,

Facebook). Each dog underwent four sessions of ‘one-on-one’ learned aversion training, spread across 6 weeks where individual dogs were subjected to a small electrical correction after having touched a non-toxic bait that has been attached to the training device and earthing rod

(this created an electrical charged bait). Each dog’s behaviour, temperament and level of trainability were monitored during each session, based on temperament and behavioural test guidelines and procedures. Dog were categorised based on their level of trainability; easy

(i.e., only one to two training sessions and one to two repetitions of electrified baits during those sessions were required), moderate (i.e., only two to three training sessions and two to three repetitions of electrified baits during those training sessions were required), and difficult (i.e., all four training sessions and three repetitions of electrified baits during those four training sessions were required). The results indicated that 50 dogs were successfully trained to avoid the bait and that temperament and breed significantly influenced their level of trainability.

Moreover, specific temperaments of obedience, excitability, playfulness and boldness were significantly related to each of the dog breeds respectively. were classified as having a difficult level of trainability; sporting and domestic working dogs displayed a moderate level of trainability, and working and toy dogs presented with easy levels of trainability based on their behavioural responses during the four training sessions. Furthermore, the temperament traits boldness (negatively correlated with trainability), fearfulness and anxiousness (positively correlated with trainability) significantly influenced dog trainability. This study demonstrated that learned aversion training with domestic and working dogs of different breeds and

4 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs temperaments has future potential in relation to the development and implementation of practices involving learned aversion training methods for dogs.

Contents

Summary / Abstract ...... 4 Contents ...... 5 List of tables ...... 6 List of figures ...... 7 Glossary ...... 8 General Introduction ...... 9 Aims ...... 11 Review of existing literature ...... 12 Conclusion ...... 27 Authors ...... 28 Affiliations ...... 28 Abstract ...... 28 Keywords ...... 29 Introduction ...... 29 Materials and methods ...... 33 Results ...... 48 Discussion ...... 58 Conclusion ...... 66 Acknowledgements ...... 68 Appendix ...... 69 Appendix 1: Training device information and set up ...... 69 Appendix 2: Temperament Categories derived from C-BARQ and ...... 72 Appendix 3: Learned aversion Trainin Certificate ...... 80 Appendix 4: Table of breed group percentages ...... 82 References ...... 87

Robyn Taylor – October December 2017 5 Temperament and breed influences on learned aversion in domestic dogs

List of tables

Table 1. Ten of the main dog breeds within Australia classed as the top ten most popular Australian dog breeds and the top ten most intelligent Australian dogs, (adapted from; ("Most popular dog breeds," 2003; "Popular breeds of 2011," 2012) ...... 14 Table 2. Australian dog breeds and in relation to their temperament types and trainability levels, (adapted from; (Turcsán, Kubinyi, & Miklósi, 2011)...... 14 Table 1. Dogs participant classifications Gregory, 2011), (modified by Robyn Taylor, 2017). . 34 Table 2. Trainability levels based off bait inspectional behaviour and bait activity...... 37 Table 3. Dog behavioural cue measurements...... 38 Table 4. Methodology key for data collected and analysed for the C-BARQ and Temperament tests ...... 40 Table 5. Univariate Tests of Significance between trainability (bait activity) and temperament traits)...... 57 Table 6. Dog Stranger Aggression levels based on behavioural cues that are measured on a Likert Scale, adapted from; (Vas et al., 2005)...... 74 Table 7. Dog Fear and Anxiety levels based on behavioural cues that are measured on a Likert Scale, adapted from; (Beerda et al., 1998)...... 76 Table 8. Dog bold levels based on behavioural cues that are measured on a Likert Scale, adapted from;(Beerda et al., 1998)...... 78 Table 9. Dog playfulness levels based on behavioural cues that are measured on a Likert Scale, adapted from; (Beerda et al., 1998)...... 79 Table 10. Table comparing the first and last training session of breed group percentages ...... 82

6 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

List of figures

Figure 1. Body postures of the dog, adapted from Gregory (2011)...... 18 Figure 1. Annual Impact of Pest Species in order of cost, (McLeod et al., 2004) ...... 30 Figure 2. Flow chart illustrating the testing and training programs ...... 39 Figure 3. Non-metric MDS Temperament plots ...... 50 Figure 4. MDS Non-metric Multivariate graph showing domestic dog a). Trainability and b). Behaviour scores...... 54 Figure 5. Scatter plots showing the percentages of a). time (in minutes) and b). distance (in meters) that each dog breed group spent with the baits (bait inspectional behaviour) within their first and last learned aversion training session. Column graphs showing percentages of behaviour and learned aversion types...... 54 Figure 6. Generalised linear repeated measures model showing a). Bait activity (bait touches and contact with bait), and b). Bait inspectional behaviour (distance and time spent with bait)...... 56 Figure 7. Generalised linear repeated measures model showing how the five domestic dog breed group's changed in a). Body positioning and b). Behaviour across their four training sessions...... 56 Figure 8. Self- training device. (a) The unit itself, (b) the self-training device including the energizer, cords, earthing rod and non-toxic version of the FoxOff Bait and, (Kreplins et al., Draft)...... 70 Figure 9. The set-up of the training device, bait and wires, (images by Robyn Taylor, 2017). .. 71 Figure 10. Learned aversion training certificate for dog owners, (designed by Robyn Taylor, 2017)...... 81

Robyn Taylor – October December 2017 7 Temperament and breed influences on learned aversion in domestic dogs

Glossary

Aversion training: Aversion training is a form of conditioning where the subject is trained to respond or behave in a certain manner due to an unpleasant stimulus (i.e., a small correction) which leads to discomfort towards or avoidance of a particular object (Peckmezian & Taylor,

2015).

Inequity aversion: Inequity aversion is a form of conditioning where the subject is tested to investigate whether they respond to fairness and equality in rewards or prefer incidental rewards with the potential to receive an over or under reward (Horowitz, 2012).

Personality: Personality is defined as the inherited, early developing inclinations that expand throughout the life of a subject and has often been considered as the foundation base for temperament development within that subject (Jones & Gosling, 2005).

Animal Temperament: Temperament is defined as a constructed model that is used to describe and measure the premise that animals have individual differences within a species. Individual animals tend to differ in regard to their thoughts, feelings and behaviour (Ley & Benett, 2007).

Operant conditioning techniques: Operant conditioning techniques are a form of learning where the behaviour of a subject is controlled by consequences (McKinley & Young, 2003).

The key concepts of this method involves positive and negative reinforcement as well as positive and negative punishment (McKinley & Young, 2003).

Standardised tests: A test which allows researchers and investigators to compare the differences and similarities within and between individual participants in relation to their performance in the trials (Diederich & Giffroy, 2005).

1080: Sodium fluoroacetate is an odourless, tasteless and colourless compound that is found within many native Australian flora (Government of Western Australia, 2017; Twigg et al.,

2000). 1080, a synthetic version of the poison found within those plant species, is used as a poison to aid in the management and control of pest vertebrate species (such as Felis catus, Sus scrofa, Canis familiars, Vulpes vulpes and Oryctologus cuniculus) in Australia (Government of

Western Australia, 2017; Twigg et al., 2000).

8 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

General Introduction

1080 poison baiting (sodium monofluoroacetate) occurs throughout Western Australia (WA) to reduce the number of invasive predators for the conservation of native species and livestock protection (Allen, 2015; Saunders, Coman, Kinnear, & Braysher, 1995). Beside Australia, a few other countries have been known to use 1080 within their vertebrate pest poisoning programs, including New Zealand, Mexico, Japan, Korea, Israel and a number non-restricted areas within the United States (Government of Western Australia, 2016 ). Other countries have not approved of the use of 1080 in their pest management programs due to concerns regarding the potential poisoning of non-targeted species (Government of Western Australia, 2016 ).

Native species in Western Australia have a high level of tolerance to 1080, however introduced vertebrate pest species, including feral cats, red foxes, wild and domestic dogs, are susceptible to the poison (Twigg et al., 2000). The success and effectiveness of programs involved with the deployment of 1080 baits to control and manage vertebrate pest species is influenced by the ingestion of the baits by non-targeted species such as domestic dogs (Glen, Gentle, & Dickman,

2007). Domestic dogs located near baited areas tend to be at risk of first and secondary degree poisoning, either from consuming the poisoned bait directly or ingesting the carcass of another animal that had previously consumed the bait (Meenken & Booth, 2010). It is prohibited for dogs to enter baited conservation areas , however, there are occasions when domestic dogs (i.e., conservation dogs) need to enter baited areas and are at risk of accidental poisoning (Kreplins et al., Draft).

Accidental consumption of poison baits within the Western Australian region is the most common cause of domestic dogs being poisoned (Goh, Hodgson, Fearnside, Heller, &

Malikides, 2005).The accidental consumptions of poisoned baits by domestic dogs is possibly as a result of the appetising way the 1080 poison baits are designed (i.e., often sausage style baits high in fat) to entice invasive pests to consume the baits (Glen et al., 2007). Household

Robyn Taylor – October December 2017 9 Temperament and breed influences on learned aversion in domestic dogs such as domestic dogs are likely to be attracted to the bait as well (Glen et al., 2007). Even when the poison baits might be inaccessible to dogs, other animals may move and drop the bait in areas where domestic dogs are located, thus providing dogs access to the bait (Glen et al.,

2007). As a result, unfortunate occurrences of accidental domestic dog poisoning occurs in a range of locations (Glen et al., 2007).

It is not only 1080 poison baits causing poisoning in domestic dogs; many other forms of baits for rodents and invertebrates are toxic to dogs as well. An analysis of enquiries and research into the Australian Veterinary Poisons Information Services (AVPIS), indicated that between the years of 1985 and 2010, there had been one case where 772 dogs had been treated after being poisoned by metaldehyde slug bait. Additionally, Bromadiolone (an anticoagulant rodenticide) was involved with 288 cases of dogs suffering accidental poisoning (Bates, Sutton,

& Campbell, 2012; Services, 2015). Of these dogs, it was recorded that 25 dogs died and 28 had to be euthanised due to the accidental poisoning (Bates et al., 2012; Veterinary Poisons

Information Services, 2015).

This study could be beneficial to reduce the fatalities of domestic dogs within Western Australia

(WA) from bait toxicity and accidental poisoning which can occur due to a variety of poisons and toxic substances; such as (Ewing, 2006):

1. Poisoning from human medications: Dogs are attracted to the sweet, candy- like taste of

the top coated layer covering most tablets taken by humans, e.g., birth control and anti-

depressant medications.

2. Rodent and snail baits: Metaldehyde is a main ingredient found in rodenticide as well as

slug and snail bait. The poison directly affects a dog’s nervous system, often leading to

the dog becoming extremely ill or dying. There is an additional risk of secondary

poisoning to domestic dogs due to the consumption of a rodent or slug that has already

ingested the poison.

10 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

3. Chocolate: The most frequently treated poison ingested by domestic dogs; chocolate

tends to cause severe thirst, diarrhoea, shaking and seizures in dogs following the

consumption of the substance. Dark chocolate has been classified as the most toxic form

of chocolate due to the percentage of Theobromine, often causing extreme fatalities

within the domestic dog population.

4. Insecticides: Dogs can often be exposed to toxic chemicals such as insecticides used on

agricultural lands, community lawns as well as household gardens, often leading to

vomiting, fevers, anxiety, and depression and muscle tremors.

Examination of veterinary records and previous research has shown that domestic dogs are prone to accidental poisoning. Particularly, conservation working dogs who have a specific purpose of working within baited areas (i.e., feral pig detection). In theory, the introduction of learned aversion training to both domestic and populations would be useful in lowering the likelihood of dogs being accidentally poisoned.

Aims

The aim of this thesis is to investigate, how learned aversion methods are influenced by breed and temperament of dogs. This project involved:

1. Quantification of individual domestic dog’s temperament.

2. Identifying individual dog responses to the 1080 aversion training.

This project investigated the hypothesis that the learned aversion method is more effective for particular dog breeds and temperament types. It is likely that certain dog breeds possess certain temperaments which will influence the dog’s trainability and behaviour towards the learned aversion. If dogs are trained to avoid poison baits, land owners may be able to increase their bait deployment, thereby reducing invasive predators and enabling successful long-term conservation and livestock production outcomes.

Robyn Taylor – October December 2017 11 Temperament and breed influences on learned aversion in domestic dogs

Review of existing literature

Many different techniques are utilised to train domestic and working dogs. Learned aversion is an important training method to ensure that aversion to baits in the absence of the dog’s owner occurs. Previous studies relating to training method and animal temperament measures were reviewed to determine how different domestic dog breeds and temperaments influence the outcome of domestic .

Domestic dog training

The domestic dog (Canis lupus) is considered to be the oldest form of a domesticated animal

(Maejima et al., 2006). This domestication occurred more than 14,000 years ago, since then dogs have served mankind for the purposes of hunting, military and guard use, companionship

(i.e., pets) and even as source of food (Maejima et al., 2006). Today, domestic dogs play significant roles within modern society, such as guide dogs, drug detection dogs and rescue dogs (Maejima et al., 2006).

Dog training occurs in many forms; negative reinforcement and learned aversion training are only two of the many training techniques employed. Domestic dogs are typically trained through operant conditioning techniques where they are taught to respond with a desired behaviour, reinforced via a reward or punishment system, (i.e., positive reinforcement and negative reinforcement) (Deldalle & Gaunet, 2014; Maejima et al., 2006; Ziv, 2017). McKinley and Young, (2003) considered the use of negative reinforcement (i.e. electrical aversive measures) to be an effective form of dog training, as the technique enabled dog breeds to perform better in obedience related training programs. Conversely, Ziv, (2017) argued that the use of electrical aversive measures were associated with increased levels of aggressive behaviours within dog breeds subjected to the technique. Ziv, (2017) continued to contend that there was little evidence to suggest the use of electrical aversive methods were more effective

12 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs than positive reinforced training techniques, reporting that dogs subjected to negatively reinforced techniques using electrical aversive methods associated the shocks with the owner’s presence rather than the stimulus which they were being training to avoid (Ziv, 2017). There were major concerns expressed regarding the welfare of the dogs being trained using this method, especially in relation causing chronic stress onto those dogs involved (Ziv, 2017). It was recommended that positively reinforced and reward based training methods should be used in place of negative and/or electrical aversive training techniques to facilitate the dogs’ welfare

(Ziv, 2017). Nonetheless, there are cases where positively reinforced and reward based training techniques would not provide the preferred response from dogs, for example, pig detection dogs would not be able to learn to avoid poisoned baits if they were trained through a reward system

(Peckmezian & Taylor, 2015). It is be recommended that pig detection dogs should be trained utilising negatively reinforced or learned aversion methods, given their work requires them to work independent of their owners when they detect feral pigs in bushland (Deldalle & Gaunet,

2014; Peckmezian & Taylor, 2015; Range, Horn, Virany, & Huber, 2008). If pig detection dogs were trained via a reward system method, they would become susceptible to poisoned baits once they entered a baited area, as they would be reliant on the presence of their owner as well as a command to leave the baits, which once obeyed they would expect a reward or a treat for compliance (Peckmezian & Taylor, 2015). It is impractical for a pig detection dogs’ owner to follow their dogs into a baited area, negatively reinforced and learned aversion methods are considered to be more appropriate (Peckmezian & Taylor, 2015). The overall purpose of dog training programs, whether it is conditioned training, reinforced training, or learned aversion training, is to receive a desired response from the dog which is dependent on the owner’s preferences.

It is likely that temperament and breed will affect the trainability of a domestic dog. A variety of studies involving dog temperament traits and its influences on trainability, has suggested that a dog’s level of submission and willingness to obey their owner’s commands impacts a dog’s level of trainability (i.e., dogs who were found to be submissive and obedient to their owners

Robyn Taylor – October December 2017 13 Temperament and breed influences on learned aversion in domestic dogs were considered to be highly trainable) (Serpell & Hsu, 2005). These temperament traits were identified as crucial to the successful training of dog breeds, as well as maintaining their health status and relationship with their owners (Serpell & Hsu, 2005). Studies have shown that canine breeds have characteristic temperaments, making certain breeds more trainable than others

(Vandeloo, 2009).

Main domestic dog breed types

In Australia, in excess of 3.4 million dogs are kept as companion animals; these include a range of breeds with varying temperament types and trainability levels. Refer to Table 1 ("Most popular dog breeds," 2003; "Popular breeds of 2011," 2012) and Table 2 (Turcsán, Kubinyi, &

Miklósi, 2011

Table 1. Ten of the main dog breeds within Australia classed as the top ten most popular Australian dog breeds and the top ten most intelligent Australian dogs, (adapted from; ("Most popular dog breeds," 2003; "Popular breeds of 2011," 2012) . Top 10 most popular Australian dog breeds Top 10 most intelligent Australian dog breeds in (main breed types in descending order). descending order. 1). Staffordshire bull Terrier 1). Border collie 2). Labrador 2). Poodle 3). 3). German shepherd 4). Cavalier King Charles 4). Golden retriever 5). Golden retriever 5). Doberman pincher 6). Border collie 6). Shetland sheepdog 7). American Staffordshire Terrier 7). Labrador retriever 8). Poodle 8). Papillion 9). Pug 9). 10). Rottweiler 10). Australian cattle dog

Table 2. Australian dog breeds and in relation to their temperament types and trainability levels, (adapted from; (Turcsán, Kubinyi, & Miklósi, 2011). Breed Type Temperament Trainability Type Level Cocker Spaniel, Terrier, Beagle, Collie, Bulldog, Golden Retriever, Calm, sociable Highly Mountain dog, Pointer, Pug, Labrador Retriever, Sheepdog, Shih and bold. Trainable. Tzu, Husky, Bull Terrier, Staffordshire Terrier, Australian Shepherd, Mountain , Cocker Spaniel, Poodle, Rottweiler, Border Collie, Boxer, Great Dane, Mountain Dog, Doberman, Dalmatian, Dachshund and Jack Russell Terrier. Miniature Dachshund, Miniature Poodle, Ridgeback, Shepherd, and Aggressive, not Highly Hound. very social and Trainable anxious. Bull Terrier, Maltese, German , Chihuahua and Yorkshire Aggressive, Low level of Terrier. antisocial and trainability. timid.

14 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Malamute, Shepherd dog, English and Irish Wolfhound. Calm, not very Low level of social and trainability. anxious.

For decades, dogs have been bred for purposes such as hunting or herding and have been influenced by human preferences for particular physical and mental characteristics (Ley &

Benett, 2007). This resulted in over 400 dog breeds that are recognised in the world today (Ley

& Benett, 2007). Through the breeding of domestic dogs, several species have undergone a series of changes in relation to their behaviour as a result of human preferences (Hart, 1995).

For example, certain domestic dog breeds were initially selected and bred for the purpose of hunting and trained to elicit desired behavioural responses conducive to proficiency in hunting, e.g., obedience, trap target species but not injure it (Hart, 1995). Another group of domestic working dog breeds, were selected and bred on the basis of their ability to perform complex tasks such as herding sheep or cattle (e.g., Kelpie), whilst other domestic dog breeds have been selectively bred based on their behaviour towards protecting livestock from predation (e.g., the

Mareema sheep dog); (Hart, 1995). These examples of breeds with a specific function have led to the characterisation of the domestic dog breeds within current our society that originated from the enhancement of their “native” behaviour rather than the development of a new behavioural characteristic from within breed types (Hart, 1995). Conversely, there have been dog breeds considered to function as working dogs which have previously had their hunting drives and instincts suppressed due to the selection of preferred behavioural features through historical human domestication (Trut, Plyusnina, & Oskina, 2004). Trut et al., (2004) have demonstrated that when selective pressures are directed onto preferred behavioural traits in dogs, domestication occurs. Therefore, selection pressures placed on dog breeds act as the key to the transformation and domestication of preferred behavioural traits and temperaments within breed groups.

The main areas of research regarding dog breeds involve the temperament traits of each breed as well as their trainability levels (Hart, 1995). Environmental factors, genetics and owner

Robyn Taylor – October December 2017 15 Temperament and breed influences on learned aversion in domestic dogs behaviour also influence trainability levels and temperament traits (Hart, 1995). For example,

Hart (1995) found that of six domestic dog breeds tested, Cocker followed by Beagles were the easiest breed to be trained, due their obedience and submission to their owners.

Additionally, this study showed noticeable differences between dog breeds and their temperament traits towards a series of behaviour and trainability tests (Hart, 1995). Additional research indicated that working dog breeds such as Border Collies have an easy trainability factor which was attributed to their high obedience levels towards their owners, as well as characteristics e.g.,bite and grip, prey drive, tracking and retrieving instincts as a result of selective breeding programs that enhanced the breed’s suitable for pest management programs within WA (i.e., pig detection dogs); (Vandeloo, 2009). In relation to trainability and temperament, obedience and submission were considered to be favourable temperament traits required of a dog in order to be considered highly trainable (Hart, 1995; Vandeloo, 2009) .

Temperament in animals

Animal temperament is defined as the consistent individual differences in behaviour, a definition that has been well established in the study of dogs (Fratkin, Sinn, Patall, & Gosling,

2013). Temperament not only differs between dog breeds, but it also differs between individual dogs (Ley & Benett, 2007). King, Marston and Bennett (2012) verified that common temperament traits are found in dogs within the same breed group, and that the expressions of those temperament traits differ between individuals. Temperament is acknowledged as inherited inclinations that develops throughout the life of an individual dog, and with the addition of experiences and selective breeding programs, distinct characteristics within breed groups have been produced (Jones & Gosling, 2005; King, Marston, & Bennett, 2012). This is significant as it indicates that dog temperament can be derived and observed from their behavioural responses towards certain stimuli and situations (King et al., 2012). In turn, this significant outcome can aid in the identification of a dog’s primary temperament trait as well as their possible reaction to certain situations and stimuli. This would enable the creation of temperament and behavioural

16 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs criteria from which dog breeds can be selected for particular purposes, such as working and guide dogs (King et al., 2012).

Results from studies on canine temperament are reliant on the individual dogs studied. A comprehensive review of Ley and Benett’s (2007) study of canine temperament identified a gap in the literature due to the lack of diversity in the types of dog breeds that were initially utilised and studied in this research. Serpell & Hsu (2005) identified a distinct ‘trainability’ factor, which was characterised by the willingness of the dog to respond to its owner’s commands

(referred to as obedience). The authors demonstrated that subsequent selection of desirable temperament traits such as obedience for more specialised working skills, could have possibly emphasised the trainability factor in many breeds used today (as working and social dogs) when compared to other breeds (Serpell & Hsu, 2005). Marshall-Pescini et al., (2016) examined the effects of differences in trainability levels and breed groups on problem solving behaviours in dogs, and suggested that an important factor affecting a domestic dog’s temperament and social cognitive abilities was their individual life and training experiences. A major limitation of studies involving canine temperament appeared to be the lack of use of standardised testing measures, which resulted in inconsistencies within the dog breeds studied, (Marshall-Pescini,

Frazzi, & Valecchi, 2016). The use of standardised measures would allow for direct comparison across similar studies in terms of the effect of training and breed group on problem solving behaviours in dogs (Marshall-Pescini et al., 2016).

Measurement of temperament in dogs

Temperament in domestic dogs has been measured through tests designed to primarily assess predictable and useful tendencies characteristic of companion and working dogs (Taylor &

Mills, 2006). These tests can be used to select dogs suitable for particular roles, such as rescue dogs (Svartberg & Forkman, 2002; Taylor & Mills, 2006). Behavioural tests are used in order to identify traits that influence canine temperament (Taylor & Mills, 2006). For example, a dog’s reaction to strangers by either ‘fleeing’ from prey-like objects or from a variety of fear and

Robyn Taylor – October December 2017 17 Temperament and breed influences on learned aversion in domestic dogs aggression inducing stimuli (Svartberg & Forkman, 2002). Svartberg & Forkman (2002) used a standardised test on 164 dog breeds in which dogs were exposed to several scenarios, including being introduced to strangers; play tests and a variety of aggression and fear inducing tests.

Factor analyses revealed five broad and narrow temperament traits found within dogs towards a variety of fear and aggression inducing stimuli, these traits included playfulness, curiosity, fearlessness, chase-proness, sociability and aggression (Svartberg & Forkman, 2002).

Figure 1. Body postures of the dog. (A, B) Neutral to alert attentive positions; (C) play bow; (D, E) active and passive submissive greeting-note tail wag and shift in ear position and in distribution of weight on fore and hind limbs; (F,G,H) gradual shift from aggressive display to ambivalent fear-defensive aggressive posture; (I) passive submission; and (J) rolling over and presentation of inguinal-genital region, adapted from Gregory (2011).

Generally, dogs tend to communicate their behaviour and temperament through a set of visual and audio cues as well as olfactory signals (Gregory, 2011). The positioning of the tail, ears and overall body stance of the dog tends to be an indicator of the dog’s behaviour, attention and temperament (Gregory, 2011). Consequently, thorough observation of the behaviour and body

18 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs positioning of individual dogs it is possible to determine a temperament for that individual

(Gregory, 2011).

Standardised methods were relied on to examine temperament traits and level of trainability in relation to specific breeds, especially those that have a working purpose (i.e. guide dogs; Hart,

1995). Hart (1995) suggested that specific temperament traits (i.e., high obedience and submissive behaviour towards the dogs’ owners) were related to working dog breeds and their high trainability levels. However, the data collected from this study was limited by the use of only one standardised testing method (Hart, 1995). Conversely, a study involving guide dogs completed a set of assessments on juvenile dogs that focussed on their behavioural suitability for future roles as working and guide dogs (Craigon et al., 2017). The findings highlighted that high trainability levels within dogs who have a working purpose was attributed to their obedience, low aggression, fearfulness, low stress levels and low energy levels..

Given that standardised methods and temperament surveys have been used to measure domestic dog temperaments, it was considered more appropriate to approach temperament measures with additional quantitative methods in order to improve the accuracy of the data collected and analysed by Hart (1995). As a result, Hart (1995) contested whether the scientific field would lose some, if not all, of its productivity if researchers were restricted to using only one form of standardised methodology when identifying temperament and trainability differences amongst domestic dog breeds. In turn, this limitation could potentially reduce the knowledge gained through studies regarding dog training and temperament behaviour (Diederich & Giffroy, 2005).

The advantages of using combined standardised and quantitative methodology tests in the investigation of domestic dog behaviour, temperament and trainability within breeds include time efficiency, results that are easy to interpret and discuss, and highlighted parameters that could affect the data outcome (such as neuter and sex status) (Diederich & Giffroy, 2005).

Robyn Taylor – October December 2017 19 Temperament and breed influences on learned aversion in domestic dogs

In relation to the effects of dog breed on trainability levels, Marshall-Pescini et al., (2016) demonstrated that there were significant differences in problem solving skills and behaviours amongst 128 dogs belonging to four breeds. The outcome of the study showed that dogs that possessed high levels of trainability were able to overcome the problem solving tasks at a faster rate than dogs that had a low trainability level (Marshall-Pescini et al., 2016). These results showed significant differences in trainability amongst breed groups and suggested that high trainability levels were strongly influenced by the individuals previous training experiences and temperament (Marshall-Pescini et al., 2016).

Studies on dog training and temperament traits established that domestic dog temperaments and trainability can be measured and tested through a series of dog owner questionnaires, reports and personal accounts (Ley & Benett, 2007; Serpell, 2017). Fratkin et al., (2013) demonstrated that, through the use of those dog owner questionnaires, such as the Canine Behavioural

Assessment and Research Questionnaire (C-BARQ), a predictability and consistency in the behaviour and temperament of dog breeds could be identified. This implied that the owner’s perspective of certain dog breeds played a crucial role in human-canine relationships (Fratkin et al., 2013). This enabled breeders to select specific dog breeds that possess certain temperaments and trainability levels suitable for a specific purpose (i.e., working dogs). However, those studies were limited due the lack of diversity of dog breeds (i.e. not all existing dog breeds were included within the studies; Serpell, 2017).

Temperament characteristics of different dog breeds

Tonoike et al., (2015) compared owner-reported behavioural characteristics amongst a variety of Australian dog breeds using the C-BARQ, a standardised survey tool that was designed to measure a variety of domestic dog behaviour and temperament (Serpell & Hsu, 2005). The questionnaire consists of a 101 questions which addresses different response behaviours of domestic dogs to certain situations and events (C-BARQ); (Serpell & Hsu, 2005). The survey takes seven factors into consideration; gender, age, state, neuter status, body weight, owners

20 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs dog-ownership experience, and source where the dog was acquired, to identify characteristics associated with genetically-clustered breed groups. For example, Tonoike et al (2015) found that there are five identified characteristics of temperament in dogs. These included (Radcliff,

2016; Tonoike, Nagasawa, Mogi, Serpell, & Ohtsuki, 2015):

 temperament is a function of the dog’s neurological design;

 temperament is inherited and is determined at moment of conception;

 a dog’s temperament cannot be transformed or eliminated from one type to another.

Therefore, a dog’s temperament cannot be changed in its lifetime, it is a permanent

mental characteristic of that dog;

 it is possible for there to be an overlap of different temperament characteristics within

the same dog, i.e., they could be aggressive and submissive at the same time; and,

 the addition of training, socialisation and changed environmental conditions can modify

the expression of a dog’s temperament.

Temperament in dogs has been classified into two broad categories; sound and unsound temperament. Sound temperament is used to describe a dog that has shown confidence and self- assertiveness towards a stimulus. The dog would be certain of itself and would investigate any new situations or stimuli confidently. Working dogs are good examples of sound temperament, given they recover and learn quickly when startled or frightened by stimuli, often negatively reinforced stimuli. Unsound temperament is used to describe dogs that display frightened, submissive and anxious behaviour. Those dogs do not have the ability to adapt easily to stimuli, often negative reinforcement, within their environments without becoming overly aggressive

(for example terriers) or submissive and extremely frightened (for example toy dogs) (Radcliff,

2016).

Prior to Radcliff’s (2016) research, Svartberg and Forkman (2002) demonstrated that temperament traits such as aggression, fearfulness, playfulness and curiosity varied amongst

Robyn Taylor – October December 2017 21 Temperament and breed influences on learned aversion in domestic dogs dog breeds. Dogs lacking in confidence (i.e., dogs with unsound temperament) often showed signs of fearfulness and anxiety once faced with fear invoking or negative stimuli. Whereas, confident dogs (i.e., dogs with sound temperament) showed signs of curiosity and fearlessness towards fear invoking or negative stimuli, suggesting these dogs have a high level of trainability due to their rapid recovery and behavioural responses. Svartberg and Forkman (2002), demonstrated that it is possible, as well as practical, to predict temperament and behavioural traits within dog breeds, both to identify dogs that could be selected for a specific purpose (i.e., working dogs and pig detection dogs), as well as to minimise any behavioural concerns such as excessive aggression within dog breeds. However, the data collection was perceived to be bias, as dogs were selectively chosen to participate in the study, rather than randomly selected. This limited the variety of breeds that were investigated.

Effect of temperament and/or breed on dog trainability

Dog breeds have been influenced through controlled breeding programs (Gregory, 2011;

Helton, 2010), often selecting traits such as trainability, working intelligence, and problem- solving skills. Dog breeds have previously been ranked in order of working obedience from high to low, where working and toy dogs have been shown to be the most obedient towards their owners when compared to other dog breeds (Helton, 2010). Differences amongst dog breeds is attributed to their underlying trainability as well as their physical capabilities such as agility (Helton, 2010). Temperament has often been characterised by the dogs’ trainability evidenced by their compliance to commands, low level of fearfulness towards their owners, other dogs as well as non-social objects and events (Starling, Branson, Thomson, & McGreevy,

2013).

Temperament influences the rate at which a dog is able to demonstrate changes in their level of trainability, e.g., it might take certain breeds a longer period of time to upgrade from a low trainability level to a moderate level and then to a high level (Batt, Batt, Baguley, & McGreevy,

22 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

2008; Starling et al., 2013). For example, dogs (such as working dogs) who are patient and obedient tend to be focussed, enabling them to learn at a faster rate and to improve their trainability level when compared to dogs who are easily distracted and overly excitable (Batt et al., 2008; Starling et al., 2013). Dog breeds (such as terriers) who demonstrate boldness and easy distractibility are difficult to train due to their loss of focus and high levels of excitement during training sessions (Batt et al., 2008; Starling et al., 2013; Ziv, 2017).

A study by Maejima et al., (2006) on canine temperament examined whether traits and genotypes could predict the successful training of drug detection dogs. Six temperament traits namely; defence drive, ability to cooperate, obedience, hardiness and low to medium level of aggression towards other dogs were the desired traits in domestic dogs during this study

(Maejima et al., 2006). Evaluation of genotypes and behavioural characteristics within 197 participating dogs were recorded in conjunct comparing the first and last training session of breed group percentages in relation to distance from bait, time spent with bait, type of learned aversions demonstrated, body positions and behaviour responses to bait with experiments to test dog obedience, concentration, affection demand, aggression, anxiety and target interest

(Maejima et al., 2006). Results demonstrated that a desire for work was indicated a higher likelihood that an individual dog would complete the training course successfully (Maejima et al., 2006). Mean scores from the study could be used to predict the success rate of canine training within different dog breeds as well as the training of individual dogs for a variety of different jobs (Maejima et al., 2006).

Learned aversion training

Aversion training is a form of conditioning where the subject should respond negatively to an unpleasant stimulus (i.e., a small electrical correction) creating a ‘learned aversion’ towards the stimulus (Peckmezian & Taylor, 2015). The training utilises a negative stimuli that uses electrical cues (Peckmezian & Taylor, 2015), taste cues, pre-feeding, odour cues (Baker et al.,

2007) or inequity aversion (Range et al., 2008).

Robyn Taylor – October December 2017 23 Temperament and breed influences on learned aversion in domestic dogs

Electrical learned aversion training has a number of advantages which are demonstrated in studies done by Peckmezian and Taylor, (2015) and Kreplins et al., (Draft), where aversions were generated in jumping spiders; and in feral pig detection dogs respectively. Both studies demonstrated that aversions can be generated when using an electrical correction with equipment that is easily accessible, simple to use and simple to construct. Timing during learned aversion training is crucial in order to ensure that the aversion is related to the ‘negative’ stimuli rather than the trainer (Peckmezian & Taylor, 2015). Generally, only one exposure to the device is required ensuring minimal stress to the dog being trained (Kreplins et al., Draft).

Conditional taste aversion (CTA) is defined as a form of classical conditioning. Conditional taste aversion followed a form of classical conditioning utilising a sickness-inducing substance, ingested by the subject which elicited a form of nausea); (Welzl, D'Adamo, & Lipp, 2001).

Within the study, the taste stimuli were associated with nausea resulting in avoidance of the substance. (Welzl et al., 2001). Animals that participate within CTA studies were expected to learn to associate a novel sensation with a negative response being the consequence (Welzl et al., 2001). Previous studies have used this form of conditioning on rodents, such as mice and rats (Welzl et al., 2001). For example Welzl’s (2001) study associated a novel taste with the feeling of nausea, which caused the test subjects to avoid drinking any fluid that had that specific flavour (Welzl et al., 2001). The aim of the experiment was to have the test subjects learn avoidance behaviour towards the novel taste, which indicated that the established conditioned reaction was successful and that the animal subject developed a CTA towards the specific flavour (Welzl et al., 2001). CTA’s have quite recently been involved with Levamisole

Hydrochloride such as Cagnacci’s et al., (2004) study where meat baits were injected with

Levamisole Hydrochloride to induce learned aversion in community groups of free-living badgers (Cagnacci, Mossei, Cowan, & Delahay, 2004). The results from their study illustrated that the badgers developed learned aversion towards Levamisole and not to the meat bait itself

(Cagnacci et al., 2004). This, in turn, established that Levamisole could be used to induce bait-

24 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs aversion within animal communities through CTA and that learned aversion through CTA has shown to be successful in past studies (Cagnacci et al., 2004).

Another form of aversion training in animals involves the method of pre-feeding, where species were encouraged to move to the baiting sites through the use on non-toxic baits as a lure and to feed on the non-toxic baits till conditioned to return to the baited area, after which those non- toxic baits were replaced with toxic ones in order to poison the pest animals (Moss, O'Connor,

& Hickling, 1998). Such as in the study investigated by Moss et al., (1998) in which brushtail possums (Trichosurus vulpecula) were used to identify whether there would be any effects that may be caused from pre-feeding. The pre-feeding technique has been used as a well-known management procedure involved in the development of aversion in animals, such as brushtail possums (Moss et al., 1998). The results from the study revealed that the possums had learned aversion to 1080 baits (96% of the non-pre-fed possums showed aversion towards 1080 baits and 90% of the non-pre-fed possums showed aversion towards Brodifacoum baits), which also indicated that the method of prefeeding was a successful technique within learned aversion

(Moss et al., 1998).

Odour cues have been used to protect agricultural land from wild animal foraging (Baker,

Johnson, Slater, Watkins, & Macdonald, 2007). The results from the study by Baker et al.,

(2007) found that the use of Ziram-clove combination induced learned aversion in wild mammals, therefore protecting agricultural crops from damage caused by foraging (Baker et al.,

2007).

Another form of aversion training is inequity aversion training, a study done by Range et al.,

(2008) investigated whether domestic dogs showed sensitivity toward the inequity of a reward received after following a command, e.g., to give their paw to a researcher when in pairs (Range et al., 2008). Results of the study supported their theory that the presence of a rewarded partner

Robyn Taylor – October December 2017 25 Temperament and breed influences on learned aversion in domestic dogs had a significant effect on the performance of the unrewarded dog, wherein the dogs would eventually stop responding appropriately to the command issued by the researchers following

15 to 20 repeated attempts of not being rewarded (Range et al., 2008). Horowitz (2012) investigated a similar study involving the fairness of inequity aversion training on domestic dogs. The study focussed on the advantageous and disadvantageous inequity that was observed amongst the dogs (Horowitz, 2012). Horowitz identified that, during the experiment, the dogs preferred to attend to the over-awarding researcher when they were given the choice to choose which researcher to go to (under-awarding, fair-awarding and over-awarding researchers;

(Horowitz, 2012). Both studies showed the influence of reward as well as quantity of reward within inequity aversion training .(Horowitz, 2012; Range et al., 2008)

The studies reviewed in relation to aversion training (learned aversion, inequity, pre-feeding and odour cue aversion) evidenced the concept of inducing learned aversion within animals, wild mammals, rodents and domestic dogs (Baker et al., 2007; Cagnacci et al., 2004; Horowitz,

2012; Kreplins et al., Draft; Moss et al., 1998; Range et al., 2008). For learned aversion training to be successful in any animal, 100% aversion towards the relevant object (bait, odour and taste) was a requirement within the training program (Baker et al., 2007; Cagnacci et al., 2004;

Horowitz, 2012; Kreplins et al., Draft; Moss et al., 1998; Range et al., 2008). However, learned aversion training in animals hold some limitations such as the identification of individual differences within species being tested in relation to personality traits and trainability (Baker et al., 2007; Cagnacci et al., 2004; Horowitz, 2012; Kreplins et al., Draft; Moss et al., 1998; Range et al., 2008). Potential for additional research has been identified in relation to learned aversion training within a variety of animal species through the testing of differences in animal temperament, age and species, as well as how those factors could relate to their trainability levels (Baker et al., 2007; Cagnacci et al., 2004; Horowitz, 2012; Kreplins et al., Draft; Moss et al., 1998; Range et al., 2008).

26 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Conclusion

Previous studies have evidenced that temperament and behaviour have a role in the trainability of domestic dogs. Some dog breeds possess human-like temperament traits which may have originated from the dog domestication processes where a variety of selective pressures from different environments and time eras have caused the variation in the morphology, personality traits, behaviour and genome sequences within the domestic dog species to date (Svartberg &

Forkman, 2002). Domestic dogs are also viewed as social animals and, as such, may be inclined to be taught through social stimuli rather than through operant conditioning (McKinley &

Young, 2003). Generally, to examine temperament and trainability of domestic dogs standardised testing is used (McKinley & Young, 2003). However, it is also argued that standardised methodology could be restrictive, which could potentially reduce the knowledge gained through the study of dog testing behaviour (Diederich & Giffroy, 2005). Despite this, research to date has shown that domestic dog training is influenced by the temperament and breed of canines. Based on further research opportunities identified in previous research, this thesis aims to investigate how learned aversion training is influenced by domestic dog temperament and breed.

Robyn Taylor – October December 2017 27 Temperament and breed influences on learned aversion in domestic dogs

Draft Manuscript:

How does temperament and breed influence learned aversion training

in domestic dogs?

Authors

, ,

Affiliations

1Murdoch University, School of Veterinary and Life Sciences, 90 South Street, Murdoch, WA,

6150, Australia.

*Email: [email protected],

*Phone: (+61) 409 108 241

†These authors supervised, reviewed and edited the work.

Abstract

Accidental poisoning of domestic dogs (Canis familiaris) is an unfortunate occurrence when using poison baits to control introduced pests such as feral cats (Felis catus), red foxes (Vulpes vulpes), feral pigs (Sus scrofa), and wild dogs/dingoes (Canis familiaris). Risks to domestic dogs can limit the use of baiting by land managers and hinder invasive species control programs

(Kreplins et al., Draft). A ‘learned aversion device’ and training methods were developed to train domestic dogs to avoid non-toxic FoxOff® baits to reduce the likelihood of accidental poisoning. Domestic dog temperament and breed were assessed through a canine behavioural assessment and research questionnaire (C-BARQ) filled in by each dog’s owner. Following the questionnaire, three tests were conducted to identify the dog’s aggression towards strangers

28 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

(stranger aggression test), obedience (plain biscuit obedience test) and whether they were fearful, anxious, or aggressive towards strange and unusual objects and sounds (novel object test). Using the ‘learned aversion device’, 56 domestic dogs representing five breed categories

(toy, terrier, sporting, domestic working and working dogs) were trained to avoid non-toxic commercially-available FoxOff® baits for four ‘one-on-one’ learned aversion training, sessions spread across 6 weeks. Of the total 56 dogs that underwent the training methodology, 50 were successfully trained to avoid the bait. The results demonstrated that temperament and breed of domestic dogs played a role in their ‘trainability’. Breeds that were highly trainable were working dogs and toy dogs (which is likely related to their submissive nature and fearful temperament), and the breed that had the lowest trainability level was terriers (this was likely due to their bold temperament). This study has demonstrated that learned aversion can be created in a range of breeds and temperaments of domestic dogs over a period of 6 weeks; however, the ability to learn differs amongst those breeds studied.

Keywords

Breed, domestic dogs, FoxOff® bait, learned aversion training, poisoning, temperament, toxic, 1080

Introduction

1080 (sodium monofluoroacetate) poison baiting occurs throughout Australia, including but not limited to Western Australia, Queensland, New South Wales, Victoria and Tasmania to reduce the number of invasive predators for the conservation of native species and livestock protection

(Allen, 2015; Saunders et al., 1995). Beside Australia, a few other countries have been known to use 1080 within their vertebrate pest poisoning programs, including New Zealand, Mexico,

Japan, Korea, Israel and a few non-restricted areas within the United States (Australia, 2016 ).

Introduced pest species such as the Red Fox, feral pigs, feral cats and wild dogs within the

Robyn Taylor – October December 2017 29 Temperament and breed influences on learned aversion in domestic dogs natural Australian environment have caused changes within the ecosystem. Some of these changes have been dramatic and have on occasion resulted in the extinction of native species or there have been drastic changes within the functioning of the ecosystem, Jeschke, 2014 #90}

The table below illustrates the economic, environmental and social impacts of the major introduced species within the Australian environment and agricultural industries, (McLeod,

Norris, & Cooperative Research Centre for the Biological Control of Pest, 2004). The vertebrate pest species were selected based on their relevance to the current and potential Pest Animal

Control CRC research activities, (McLeod et al., 2004). The annual costs values are included for each species which consists of pest control and loss of agricultural production, (McLeod et al.,

2004)

Figure 1. Annual Impact of Pest Species in order of cost, (McLeod et al., 2004)

Native species in Western Australia have a high level of tolerance to 1080, but introduced species including feral cats and the red foxes, as well as, domestic dogs are susceptible to the poison (Twigg et al., 2000). An unfortunate outcome of invasive species control programs

(baiting) in Western Australia as well as other states and territories of Australia is the accidental poisoning of domestic dogs (Meenken & Booth, 2010). Baits that are used in controlling

30 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs vertebrate pest species in Western Australia are made to appear and smell appetising, and as a result, these baits often attract non-target species (such as domestic dogs); (Glen et al., 2007;

Jackson, Moro, Mawson, Lund, & Mellican, 2007; Kinnear, Pentland, Moore, & Krebs, 2016;

Kreplins et al., Draft). Many land owners and managers have been limited in their ability to employ the baiting method due to this accidental poisoning (Kreplins et al., Draft).

Learned aversion training is one of a range of methods employed to train domestic dogs (Baker et al., 2007; Cagnacci et al., 2004; Kreplins et al., Draft). Learned aversion is a form of conditioning where the subject should respond negatively to an unpleasant stimulus (i.e., a small electrical correction delivered by a training device) creating a ‘learned aversion’ towards the stimulus (Peckmezian & Taylor, 2015). The training within the study conducted by Kreplins

(Draft) utilises a negative stimuli that emits electrical cues, giving a dog a correction after having touched an electrified bait.

Dog trainability has been influenced by breed and temperament as a result of controlled selective breeding programs, where characteristics such as working intelligence, problem solving skills, owner obedience and working drive have been bred over several generations in order to develop breeds that would be suitable for a particular purpose (Gregory, 2011; Helton,

2010). Working dogs with a purpose and toy dogs have shown to be more obedient and easier to train when compared to other breed groups, due to their need to please their owners, their submissive nature and certain characteristics (such as working drive) derived from years of selective breeding (Helton, 2010; Holland, 2007; Vandeloo, 2009). Dog breeds (such as terriers) who demonstrate boldness and easy distractibility have been found difficult to train due to their loss of focus during training sessions after becoming too ‘excited’ within the training process (Batt et al., 2008; Starling et al., 2013; Ziv, 2017). To further explain the influence of temperament on trainability, the temperament traits associated with high levels of

Robyn Taylor – October December 2017 31 Temperament and breed influences on learned aversion in domestic dogs trainability are compliance and willingness to obey commands and low to moderate levels of fearfulness (Starling et al., 2013).

The aim of this thesis is to investigate, how learned aversion methods are influenced by breed and temperament of dogs. This project involved:

1. Quantification of individual domestic dog’s temperament.

2. Identifying individual dog responses to the 1080 aversion training.

This project investigated the hypothesis that the learned aversion method is more effective for particular dog breeds and temperament types. It is likely that certain dog breeds possess certain temperaments which will influence the dog’s trainability and behaviour towards the learned aversion. If dogs are trained to avoid poison baits, land owners may be able to increase their bait deployment, thereby reducing invasive predators and enabling successful long-term conservation and livestock production outcomes.

32 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Materials and methods

Study methodology summary

Fifty-six dogs were recruited and classified into five breed groups (i.e., terriers, working, sporting, domestic working and toy dogs). First, all dog owners were required to fill in a dog temperament questionnaire; Canine Behavioural Assessment and Research Questionnaire (i.e.,

C-BARQ). Thereafter dog temperament tests were conducted (i.e., a test of stranger aggression, owner obedience and response to a novel object) followed by learned aversion training (i.e., training dogs to show aversion towards non-toxic FoxOff® bait). The scores from the temperament tests were combined with those that derived from the C-BARQ dog owner survey, which identified temperament traits associated with each individual dog within the study. The learned aversion training results were then compared with dog breed and temperament in order to identify whether there were any significant findings in relation to dog breed trainability, temperaments associated with those breeds and whether those temperament traits influenced dog breed trainability.

Study animal collection and classification

Fifty-six domestic dogs were recruited through word of mouth, flyer advertisement and through

Facebook ‘dog lovers’ and ‘dog trainers’ groups on the internet (see Table 1.). The dogs recruited varied in age; dogs under 6 months were not included in this study. These 56 dogs were assigned into five breed groups: working dogs (n = 12), domestic working dogs (n = 11), sporting dogs (n = 14), toy dogs (n = 11) and terriers (n = 8).

Robyn Taylor – October December 2017 33 Temperament and breed influences on learned aversion in domestic dogs

Table 1. Dogs within the study are classed into one of the following groups, based on both their historical breed classification and their observed breed classification types (Gregory, 2011), (modified and illustrated by Robyn Taylor, 2017).

Dog breeds from Breed type Description study Golden Retriever, Labrador x Retriever, Labrador x Cocker Spaniel, Labrador x Sporting (n= 14) Red cloud, Labrador x Ridgeback, Flat Bred for the purposes of hunting, coated Retriever x retrieving and sporting. They tend to Labrador, Greyhound concentrate on the chase and do not x Ridgeback, figure in placings in advanced obedience Ridgeback x Kelpie, tests. They have loud voices, which they Staghound x are not averse to using Huntaway, Wolfhound x Staghound, German Short-haired pointer and German Short- haired pointer.

America Staffordshire Terrier, Terriers (n= 8) American Staffordshire Terrier x English Variable in size, the Terriers make are Staffordshire Terrier, generally smart dogs, sharp in character English Staffordshire and vocal to a degree. They are alert, Terrier, English playful and affectionate making them Staffordshire Terrier, excellent pets. Jack Russell Terrier x Dingo, Jack Russell Terrier x Kelpie, and Jack Russell Terrier x Maltese Poodle.

Working (n= 12) Kelpie x Staffordshire Terrier, Red Healer, The working group has the largest Kelpie x Border number of breeds and number of dogs Collie, Kelpie x Roo within breeds. Mostly extremely dog, Kelpie x Roo predictable dogs that have been bred for dog, Kelpie x Roo many generations for a single purpose. dog, Kelpie, Kelpie x Selected for trainability and active minds. Whipper, kelpie They can soon become mischievous if left cross, Red cloud to their own devices. Kelpie, and Bull Arab.

Doberman, Husky, The domestic working dog group consists Husky x Dingo, of dogs that would be classified as a Rottweiler, working dog based on their breed, Working (Domestic) (n= 11) Bullmastiff, however these dogs are bred and used as Bullmastiff cross, domestic pets rather than actual working Kelpie x Border dogs. Collie, Kelpie x

34 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Border Collie, Border Collie, Border Collie, Border Collie x Terrier and New Zealand .

Poodle, Pomeranian, Pomeranian, Toys (n= 11) Pomeranian, Miniature Dachshund, A group of small dogs. They share the Chihuahua, ability to be picked up but are Chihuahua x widespread in temperament and Maltese, Maltese behaviour. Aka, companion dogs. poodle x Cocker Spaniel, Sheppadoodle, Frise and Bichon Frise.

Dog Temperament tests and C-BARQ assessment methods

Figure 1 illustrates a flow chart describing the testing and training program conducted for all 56 dogs. On day 0 dog owners were required to fill out a questionnaire (C-BARQ). The questionnaire measured aspects of dog behaviour associated with eight different temperament traits (obedience, aggression, fear and anxiety, excitability, separation related behaviour, attachment related behaviour, bold behaviour and playful behaviour), based on the dog owner’s past experiences of their dogs in regards to certain situations (such as going for walks, or when caught in a thunderstorm). For all the categories associated with the C-BARQ survey see Table

4 and Appendix 2: Temperament categories.

Following the questionnaire, three temperament tests were conducted (stranger aggression test, owner obedience test and response to novel object test), these tests were used to identify whether a dog showed signs of aggression towards strangers, whether they were obedient to their owners and what their responses were towards a strange and unfamiliar novel object.

The main components involved with the temperament tests were:

Robyn Taylor – October December 2017 35 Temperament and breed influences on learned aversion in domestic dogs a). Stranger aggression test: the dog was approached by an unfamiliar person,and their initial behavioural reactions were recorded by the researcher, (Table 4). b). Novel object response test: the dog was shown a ‘novel object’, a smart phone playing a looped car alarm sound track inside a small metal container. The levels of fearfulness, playfulness and boldness were recorded by the researcher based on the dog’s reactions and responses towards the novel objects, (Table 4). c). Owner obedience test: the owner of the dog was instructed to show their dog a plain flavoured . The biscuit was placed on the ground in front of the dog who was commanded by their owner to leave it alone and not to eat it. The levels of obedience was recorded by the researcher based on the dog’s response towards their owner’s command, (Table

4).

All three temperament tests were linked to the scores derived from the C-BARQ (i.e., Stranger aggression was linked to C-BARQ aggression, the novel object responses were linked to C-

BARQ fear and anxiety, boldness and playfulness, and the owner obedience test was linked to

C-BARQ obedience and trainability).

Learned aversion training followed the temperament tests, where all dogs were introduced to one to three presentations of electrified non-toxic FoxOff® bait in order to examine and assess dog breed trainability and behavioural responses towards the training. The temperament tests and C-BARQ was only assessed once on Day 0, whereas the learned aversion training was repeated at three more sessions spread over a 6 week period.

Trainability level measurements (Bait inspectional behaviour and bait activity)

Domestic dog levels of trainability (i.e., bait inspectional behaviour and bait activity) were measured using a Likert Scale (where 0-1: low trainability; 2-3: moderate trainability; and 4: high trainability) in order to identify whether the dogs were able to respond quickly, moderately or slowly to the learned aversion training sessions. A number of criteria quantifying bait inspectional behaviour and bait activity were utilised to categorise the dog responses and

36 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs behavioural observations made during the learned aversion training sessions allowing dog breeds to be classified within three levels of trainability; a). Low level of trainability: Dogs who demonstrate low levels of trainability are described to be difficult to train as repetition of training is required, e.g., four training sessions. b). Moderate level of trainability: Dogs who demonstrate moderate levels of trainability are described to be fairly easy to train as a reasonable amount of repeated training is required, e.g., two to three training sessions. c). High level of trainability: Dogs who demonstrate high levels of trainability are described to be easy to train as little to no repetition of training is required, e.g., one to two training sessions.

Table 2. Trainability levels based off bait inspectional behaviour and bait activity. Training categories 0-1: Low trainability 2-3: Moderate 4: High trainability trainability Number of electrified baits 3 2 0-1 presented Number of non-electrified 3+ 2-3 1-2 baits presented Number of corrections 3 2 0-1 Number of consumed 3+ 2-3 0-1 baits Number of bait touches 3+ 2-3 0-1 Number of bait touches 3+ 2-3 0-1 with nose Number of baits licked 3+ 2-3 0-1 Number of pawed at baits 3+ 2-3 0-1 Number of times bait was 3+ 2-3 0-1 sniffed Number of urinated baits 3+ 2-3 0-1 Number of times bait was 3+ 2-3 0-1 investigated Number of times bait was 3+ 2-3 0-1 barked at Number of training 3+ 2-3 1-2 sessions required to show aversion Time spent with bait prior 5-10min 5min Under 5min touch within 0-1m Time spent with bait after 5-10min 5min Under 5min touch within 0-1m Time spent with bait prior 5-10min 5min Under 5min touch within 1-3m Time spent with bait after 5-10min 5min Under 5min touch within 1-3m Time spent with bait prior 5-10min 5min Under 5min touch within 3-10m Time spent with bait after 5-10min 5min Under 5min touch within 3-10m

Robyn Taylor – October December 2017 37 Temperament and breed influences on learned aversion in domestic dogs

Behavioural cue measurements (Body position and behaviour)

Domestic dog behavioural cues (i.e., body position and behaviour) were measured and scored using a Likert Scale (where 0-1: neutral behaviour; 2-3: alert behaviour; and 4: fearful and anxious behaviour). Behavioural cue measurements were based on dog body positioning and behaviour during all four training sessions (Beerda, Schilder, van Hooff, de Vries, & Mol, 1998;

Vas, Topál, Gácsi, Miklósi, & Csányi, 2005). The scoring aided in the identification of whether dogs altered their body responses and positions during the learned aversion training.

Table 3. Dog behavioural cue measurements. Behavioural cue 0-1: Neutral 2-3: Alert behaviour 4: Fearful and categories behaviour Anxious behaviour Hackles Relaxed. Raised. Lowered. Body response Relaxed. Alert. Submissive Body position Relaxed. Alert stance. Lying still with back on ground. Head position Neutral/Relaxed. Head positioned Lowered or leaning forward or raised. against the ground. Ear position Neutral/Relaxed. Alert. Flat against head. Tail position Neutral/Relaxed. Raised to level of Tucked between spine. hind legs. Vocal cues None. Growling and Whining and barking. moaning. Mouth position Neutral/Relaxed. Clenched jaw, Constantly opening occasional bearing of and closing to let teeth and opening of out whines. mouth. Back leg position Normal. Straight and Held in the air as extended, boarding the dog is lying on on flexed. their back, or tucked against body. Front leg position Normal. Straight and Held in the air as extended, bordering the dog is lying on on flexed. their back or tucked against body.

38 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Figure 2. Flow chart illustrating the testing and training programs conducted throughout four sessions for all 56 dogs.

Robyn Taylor – October December 2017 39 Temperament and breed influences on learned aversion in domestic dogs

Table 4. Methodology key for data collected and analysed for the C-BARQ and Temperament tests Methodology Key Terms Definition Descriptor Scale Categories Measurement scale Analysis Temperament traits derived owner survey (C-BARQ) and Temperament tests (stranger aggression test, owner obedience tests and response to novel object) C-BARQ A dog's compliance with an order, Scoring dog behaviour in See obedience categories 0=Never The scored data for the Obedience request or command when under relation to their obedience in appendix 2, section 1: 1=Seldom temperament categories and the authority of their owner. and trainability based on past Obedience and 2=Sometimes were analysed using the Trainability events. Trainability. 3=Usually PAST v. 3.15 program. Non- 4=Always metric analyses were done Temperament A dog’s compliance with their The dog’s behaviour in An obedient dog would in order to show whether Did the dog consume the test: Owner owners’ command not to consume relation to their obedience have shown the following there was any significant dog biscuit? Obedience the plain dog biscuit once the dog towards their owners’ responses: difference between dog has been told to ‘not eat it’ or to command was scored based Did the dog obey their Dog waited for the ‘okay’ breed and temperament. ‘leave it alone’ and the biscuit has on how the dog responded. owners’ command? from their owner before One-way and Two-way been placed on the ground. they ate the dog biscuit (i.e., PERMANOVA analyses were Was the owner required to This test was linked to that of the dog would have not done on the resulting axis repeat the command more the C-BARQ Obedience and consumed the biscuit and scores (axis 1 and 2) of each than once before the dog Trainability temperament would have obeyed their temperament trait scores, ceased all attempts to category. owner). and then based on dog consume the biscuit? group and their owners who A disobedient dog would scored them. If a significant have disobeyed their owner difference was found, a and attempted or SIMPER analysis was succeeded in consuming the conducted to explain why dog biscuit. that was the case.

C-BARQ Dog behaviours that start off as Scoring dog behaviour in See aggression categories 0=Never Aggression warnings and can accumulate in an relation to their aggression in appendix 2, section 2: 1=Seldom attack, i.e., Growls, bares teeth, based on past events. Aggression. 2=Sometimes warning barks and bites. 3=Usually 4=Always

40 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Methodology Key Terms Definition Descriptor Scale Categories Measurement scale Analysis Temperament A dog’s level of aggressive The dog’s level of aggressive Levels of The scenarios in which test: Stranger behavioural responses towards the behaviour towards the (towards strangers) was stranger aggression was Aggression presence of a stranger on their presence of strangers was measured using a Likert measured against were: owners’ property. scored based off their Scale with the following behavioural responses. descriptors:

Unfamiliar person This test was linked to that of 0-1: represented no (researcher) approaches the C-BARQ Aggression aggression (no visible signs the dog’s owner or family temperament category. of aggression). member.

2-3: represented moderate aggression (growling, Unfamiliar person barking and baring teeth). (researcher) is visiting the

dog owner’s home. 4: represented serious

aggression (snaps, bites or attempts to bite).

See appendix 2, section 2: Stranger aggression observed behavioural cues (these cues were used to identify dog stranger aggression levels between 0 and 4, e.g., no aggression to serious aggression. C-BARQ Fear Unpleasant emotions caused by the Scoring dog behaviour in See fear and anxiety 0=Never and Anxiety threat of danger, pain or harm. relation to their fear and categories in appendix 2, 1=Seldom anxiety based on past events. section 3: Fear and 2=Sometimes Anxiety. 3=Usually 4=Always

Robyn Taylor – October December 2017 41 Temperament and breed influences on learned aversion in domestic dogs

Methodology Key Terms Definition Descriptor Scale Categories Measurement scale Analysis Temperament Fearful and anxious behavioural The dog’s fearful and anxious The scenarios in which fear Levels of dog fear and test: Novel responses caused from the dog behavioural responses were and anxiety was measured anxiety towards the novel Object fearful feeling threatened or uneasy due to scored in relation to their against were: object was measured using responses the presence of a unusual novel level of fear and anxiety a Likert Scale using the object. when introduced to the novel The presence of an following descriptors: object (i.e., metal container unusual and strange novel containing a smart phone object (i.e., metal 0-1: represents no fear and emitting a car alarm container). anxiety (no visible signs of soundtrack). cowering and submission). The presence of an This test was linked to that of unusual and strange novel 2-3: represents moderate the C-BARQ Fear and Anxiety object emitting a car alarm fear and anxiety (some temperament category. sound (i.e., a metal cowering and whining is container containing a present). smart phone playing a looped soundtrack of car 4: represents extreme fear alarms). and anxiety (cowering, whining, and submission).

See appendix 2, section 3: Novel object: Fear and Anxiety observed behavioural cues (these cues were used to identify dog fear and anxiety levels between 0 and 4, e.g., no fear and anxiety to extreme fear and anxiety.

42 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Methodology Key Terms Definition Descriptor Scale Categories Measurement scale Analysis C-BARQ Behavioural signs in the form of Scoring dog behaviour in See separation related 0=Never Separation anxiety that dogs show when left relation to their separation behaviour categories in 1=Seldom related alone over short or long periods of related behaviour based on appendix 2, section 4: 2=Sometimes behaviour time. This can be in the form of past events. Separation related 3=Usually whining, or chewing objects. behaviour. 4=Always C-BARQ Dogs are capable of being readily Scoring dog behaviour in See excitability categories 0=Never Excitability roused into a state of excitement. relation to their excitability in appendix 2, section 5: 1=Seldom based on past events. Excitability. 2=Sometimes 3=Usually 4=Always C-BARQ The level of attachment a dog has Scoring dog behaviour in See attachment related 0=Never Attachment for their owner. The dog could relation to their attachment behaviour in appendix 2, 1=Seldom related show strong attachment towards and attention seeking section 6: Attachment and 2=Sometimes behaviour their owner, whereupon they behaviour based on past attention seeking 3=Usually demand constant attention. events. behaviour. 4=Always C-BARQ Bold Bold dogs tend to demonstrate Scoring dog behaviour in See bold behaviour in 0=Never behaviour behaviour that is described as being relation to their bold appendix 2, section 7: Bold 1=Seldom ‘brave’, ‘courageous’ or ‘fearless’. behaviour based on past behaviour. 2=Sometimes events. 3=Usually 4=Always Temperament Bold behavioural responses from The dog’s bold behavioural The scenarios in which Levels of dog boldness test: Novel the dog demonstrating confident responses were scored in boldness was measured towards the novel object Object bold and fearless behaviour due to the relation to their level of against were: was measured using a Likert responses presence of a unusual novel object. boldness when introduced to Scale using the following the novel object (i.e., metal The presence of an descriptors: container containing a smart unusual and strange novel phone emitting a car alarm object (i.e., metal 0-1: represents no boldness soundtrack). container). (no visible signs of fearlessness). This test was linked to that of The presence of an the C-BARQ Bold unusual and strange novel 2-3: represents moderate temperament category. object emitting a car alarm boldness (dog was

Robyn Taylor – October December 2017 43 Temperament and breed influences on learned aversion in domestic dogs

Methodology Key Terms Definition Descriptor Scale Categories Measurement scale Analysis sound (i.e., a metal demonstrating courage). container containing a smart phone playing a 4: represents extreme looped soundtrack of car boldness (dog was alarms). demonstrating fearlessness).

See appendix 2, section 7: Novel object: Bold observed behavioural cues (these cues were used to identify dog boldness levels between 0 and 4, e.g., no boldness to extreme boldness.

C-BSRQ Dogs that are playful are described Scoring dog behaviour in See appendix 2, section 8: 0=Never Playful as being fond of games and relation to their playful Playful behaviour. 1=Seldom behaviour amusement and are often light- behaviour based on past 2=Sometimes hearted. events. 3=Usually 4=Always Temperament Playful behavioural responses from The dog’s playful behavioural The scenarios in which Levels of dog playfulness test: Novel the dog demonstrating boisterous responses were scored in playfulness was measured towards the novel object Object playful and energetic behaviour due to the relation to their level of against were: was measured using a Likert responses presence of a unusual novel object. playfulness when introduced Scale using the following to the novel object (i.e., The presence of an descriptors: metal container containing a unusual and strange novel smart phone emitting a car object (i.e., metal 0-1: represents no alarm soundtrack). container). playfulness (no visible signs of boisterous behaviour). This test was linked to that of The presence of an the C-BARQ Playful behaviour unusual and strange novel 2-3: represents moderate

44 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Methodology Key Terms Definition Descriptor Scale Categories Measurement scale Analysis temperament category. object emitting a car alarm playfulness (dog was sound (i.e., a metal demonstrating boisterous container containing a behaviour). smart phone playing a looped soundtrack of car 4: represents extreme alarms). boldness (dog was demonstrating energetic behaviour).

See appendix 2, section 8: Novel object: Playful observed behavioural cues (these cues were used to identify dog playful levels between 0 and 4, e.g., no playfulness to extreme playfulness.

Robyn Taylor – October December 2017 45 Temperament and breed influences on learned aversion in domestic dogs

Learned aversion training

All 56 dogs were subjected to four learned aversion training sessions which was repeated from their first training session (Day 0), to their second, third and fourth (aka, Day 1, Day 7 and

Month 1), where each dog was presented with a non-toxic FoxOff® bait that was attached to the

‘learned aversion’ training device (creating an electrified bait). This device provided the dogs with a small electrical ‘correction’ if they attempted to touch or consume the bait. The components of the device (including the box, wires and earthing rod) were all hidden by surrounding foilage so that only the bait residing on a clear plastic lid was visible to the dog.

The dogs were allowed to interact with the bait for up to 10 minutes while their behavioural responses and body positions were monitored on a video camera placed 10m away from the baited area. After the intial training with the electrified bait, there was a 20 minute break, where after the dog was presented with another electrified bait (if the dog had touched the previous bait) or a non-electrified version of the bait (if the dog had avoided the previous bait). Each dog was presented with a maximum of three electrified baits per training session to limit stress any dog may experience whilst undergoing the training. If dogs did not show any response (i.e., dog did not demonstrate aversion) to the electrical corrections emitted from the training device, a more powerful training device was utilised, in addition to wetting the dog’s front and back paws with salt water.

Dogs within the study were classified based on the type of aversion they demonstrated towards non-toxic FoxOff® baits during their learned aversion training. The four categories of aversion,

(Kreplins et al., Draft);

1). Inspectional behaviour: The dog is observed to investigate the bait closely (0-1m) rather than avoiding it.

2). Attentional aversion: The dog is observed to be distracted and/or interested in everything around them within their environment rather than the presence of the bait.

3). Boundary aversion: The dog is observed to form an invisible barrier around the bait, and only approaches the area within 3-5m.

46 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

4). Complete aversion: The dog is observed to avoid the bait absolutely, often not entering the area where the bait is present (dog leaves 5-10m between themselves and the bait).

For further information regarding the training device used in the learned aversion training refer to Appendix 1: Training device information and set up, Figures 7 and 8.

Methods for analysis of data

Utilising multivariate non-metric MDS analyses (Ø. Hammer, 2017, PAST, paleontological statistical data software, version 3.15, Oslo, Norway) the results of the learned aversion training of the 56 domestic and working dogs were compared to their breed group (toy, terrier, working, domestic working and sporting) as well as their temperament traits and behavioural responses towards the learned aversion training. Correlations (Microsoft Excel version 2016) and one-way

PERMANOVAs using the resulting non-metric MDS scores, temperament traits, and learned aversion training results were performed for each breed group. When a significant difference was found, a SIMPER analysis was conducted. The resulting MDS plots, with any identified significant differences in the data, provided clear visual representations of the learned aversion training results and how the training results could be affected by temperament traits and breed group within the study. Multiple Regression analyses (TIBCO Software Inc. 2017, Statistica, data analysis software system, version 13, Palo Alto, USA) were conducted to identify whether any temperament traits influenced dog breed trainability, these results were then subjected to a

Levene’s test to identify whether the variables were of assumed equal variances.

Robyn Taylor – October December 2017 47 Temperament and breed influences on learned aversion in domestic dogs

Results

Temperament results from different breed groups

Obedience differed significantly between breed groups (Figure 3a), where working and toy dogs demonstrated a high level of obedience towards their owners; domestic working and sporting dogs displayed a moderate level of obedience towards their owners, and terriers had a low level of obedience to their owners.

Excitability varied significantly amongst breed groups (Figure 3e), where working dogs were shown to have a low level of excitability; toy, sporting and domestic working dogs exhibited e a moderate level of excitability, and terriers demonstrated an extremely high level of excitability.

Boldness differed significantly between breed groups (Figure 3g), where working and toy dogs displayed a low level of boldness; sporting and domestic working dogs demonstrated a moderate level of boldness, and terriers exhibited an extremely high level of boldness.

Playfulness differed significantly between breed groups (Figure 3h), where working, domestic working, sporting and toy dogs measured in the low to medium level of playfulness, and terriers displayed a high level of playfulness.

a. F(4,54) = 2.42 p = 0.000 b. F (25,54) = 0.70, p = 0.382

48 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

c. F (17,54) = 0.91, p = 0.130 d. F (6,54) = 0.97, p = 0.153

e. F (4,54) = 2.72, p = 0.000 f. F (6,54) = 0.97, p = 0.153

g. F (4,54) = 3.53), p = 0.002 h. F (4,54) = 2.88, p = 0.001

Robyn Taylor – October December 2017 49 Temperament and breed influences on learned aversion in domestic dogs

Figure 3. Non-metric MDS Temperament plots: a). Obedience, b). Aggression, c). Fear and Anxiety, d). Separation related behaviour and breed, e). Excitability, f). Attachment related behaviour, g). Boldness, and h). Playfulness. F values represent the results of two-way PERMANOVA where owner and dog breed are included as independent factors. Colour scale: Domestic working dogs (n= 11), Working dogs (n= 12), Sporting dogs (n= 14), Terriers (n= 8), and Toy dogs (n= 11).

Trainability and behaviour in relation to breed

Based on the results of the study, terriers were identified as the breed group with the lower level of trainability (Figure 4a). Analysis of the observational data indicate terriers spent a longer period of time with the baits, between five to 10 minutes reaching the maximum time allocated for interaction with the baits per session. Terriers displayed high levels of activity in the presence of those baits (i.e., more touches/corrections due to having licked, consumed and pawed at the baits). The breed group barked excessively (for two to three minutes) at the baits after having received a correction, which was interpreted as an indication of bold and aggressive behavioural responses towards the negatively reinforced corrections. Thirty-seven percent of male terriers were observed to urinate (one to two times; Table 10.) on two to three bait samples presented during the training session. Terriers were the main dog breed found to urinate and at the baits presented during the learned aversion training sessions, however, as the terriers’ training progressed, their level of trainability reasonably changed from a low level to a moderate level, although, it is important to note that it took longer (three to four training sessions) for terriers to exhibit this change.

Sporting, working, domestic working and toy dogs measured in the moderate to high levels of trainability (Figure 4a). By the fourth training session, 100% of the sporting dog and 90% domestic dog breeds, distanced themselves from the baited area to (three to ten meters), they spent less time with those baits presented (under five minutes), showed limited activity with and around the baits, and were not vocal to any of the bait presentations (Figure 5 and Table 10.).

One male sporting dog was recorded to urinate once on three presentations of electrified baits upon his first and second training session. In comparison to sporting dogs, domestic working

50 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs dogs and terriers, there were a much higher percentage of working dogs (91%; Figure 5) and toy dogs (100%; Figure 5) that spent less than five minutes with all bait samples presented during their final training sessions. These breed groups were often observed to retreat immediately once the researcher arrived at the property. As sporting, working, domestic working and toy dogs’ learned aversion training increased, their activity and inspectional behaviour towards the baits decreased (e.g., they spent less time with the baits and they often removed themselves from the baited areas, putting as much distance between themselves and the baits), thus the level of trainability for toy and working dogs increased from a moderate to a high level, and remained at a moderate level for sporting and domestic working dogs.

Terriers were persistent in consuming the bait at all training sessions (Figure 4b). Ninety-five percent of terriers initially presented an ‘alert’ body stance (i.e., raised hackles, alert body position and response, head slightly raised, ears alert, tail raised, growling, and clenched jaw, straight and extended back and front legs) and barked towards the bait and the researcher.

Seventy percent of terriers directed this form of bold behaviour (i.e., hackles raised, fearless body response, dominant body position, head raised, ears alert, tail slightly lowered, barking, panting, flexed front legs and crouched back legs) and aggressive responses (i.e., hackles raised, threatening body response and position, head raised, ears alert and forward, tail raised above spine, excessive barking and baring of teeth, flexed front and back legs) towards the baits, rather than showing aversion. In addition to the aforementioned behaviour, the terriers often attacked the electrified baits and receive up to three electrical corrections per bait per session. Despite this breed group’s initial behavioural responses, it was noted that after the last training session there was a change in body stance, where 62% (Figure 5) of the terriers demonstrated behaviour associated with a ‘neutral’ body stance (i.e., relaxed hackles, relaxed body response, normal body position, relaxed ears, tail lowered or relaxed, relaxed mouth and jaw, normal front and back leg position and no vocal cues) and began to show aversion towards the baits presented. In terms of learned aversion, the change in behavioural cues, confirmed that 62% (Figure 5) of the terrier breed group was able to respond to the non-toxic FoxOff® bait in an aversive behaviour

Robyn Taylor – October December 2017 51 Temperament and breed influences on learned aversion in domestic dogs manner employing of inspectional or attentional type aversion. The remaining 37% (Figure 5) of terriers continued to demonstrate aggressive and bold behaviour towards the baits presented and were found to continually consume the maximum assigned number of bait presentations at each training session, including their final training session (i.e., they demonstrated inspectional behaviour which meant that they did not successfully learn aversion towards the non-toxic baits).

Domestic working and sporting dogs initially demonstrated a neutral deterrence towards the bait presentations at their first training sessions (Figure 34b), thereafter, 82% of domestic working and 71% of sporting dogs demonstrated a change in their behaviour after their first electrical correction (Figure 5 and Table 10.). By the last training session 36% of sporting and 45% of domestic working dogs demonstrated fearful and anxious behavioural responses (i.e., hackles lowered, submissive body response and position, head lowered, ears held flat against head, tail tucked between hind legs, whining and moaning, mouth opening and closing or clenched, back and front legs tucked against body) to the bait presentations. Within both breed groups, 100% of domestic working and 93% of sporting dogs were able to respond to the non-toxic FoxOff® bait in an aversive behaviour in the form of boundary or attentional aversion (Figure 5 and Table

10).

Approximately half of the breed sample (54%; Figure 5) demonstrated a fearful and anxious body stance towards the baits presented during the first learned aversion training session, (Figure 4b.). Following the second training session, 82% of toy dogs refused to touch a bait and removed themselves from the baited area and as a result, their initial behavioural response of fear and anxiety only escalated (Figure 5 and Table 10). The 82% of toy dogs were able to respond to the non-toxic FoxOff® bait in an aversive behaviour in the form complete aversion (Figure 5). The remaining 18% of toy dogs never touched a bait (i.e., they never received a correction during the training sessions), therefore they did not show any form of

52 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs aversion, nor did they show any fear or anxiety towards the presence of the bait (i.e., they continually demonstrated a neutral behavioural response; Figure 5 and Table 10.).

a). F(21,54) = 7.45, p = 0.000

b). F(18,54) = 5.82, p =0 0.000

Robyn Taylor – October December 2017 53 Temperament and breed influences on learned aversion in domestic dogs

Figure 4. MDS Non-metric Multivariate graph showing domestic dog a). Trainability and b). Behaviour scores across four breed classifications (Domestic working, Working, Sporting, Terrier and Toy) within their four training sessions (Day0, Day2, Day7 and 1Month).

a). % of time (minutes) spent with bait b). % of distance (meters) spent with bait

c). % of body behaviour types during training d). % of learned aversion types

Figure 5. Scatter plots showing the percentages of a). time (in minutes) and b). distance (in meters) that each dog breed group spent with the baits (bait inspectional behaviour) within their first and last learned aversion training session. Column graphs showing percentages of behaviour and learned aversion types displayed by each dog breed group during their first and last learned aversion training sessions.

54 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Effects of time on dog breed trainability and behaviour

The results indicate the time taken for a dog to demonstrate changes in their trainability levels in relation to bait activity, (Figure 6a) and bait inspectional behaviour, (Figure 6b) was dependent on the breed. Given that terriers were identified as the breed group with the lower level of trainability, it stands to reason that this breed group required a longer period of time to demonstrate attentional aversion towards the non-toxic FoxOff® bait evidenced by the reduction of their bait activity. The results indicated 30% of terriers did not demonstrate an aversive response and showed no reduction in their bait activity, but rather displayed a 20% increase in their bait inspectional behaviour (Figure 5 and Table 10.). The bait aversive group of terriers displayed changes in their behavioural cues, from alert body stance to neutral body stance, however required 65% more time and training repetition than toy and working dogs (Figure 5,

Figure 7 and Table 10.).The remaining breed groups; sporting, working, domestic working and toy dogs, demonstrated a decrease from 24% to 16% in their bait activity and bait inspectional behaviour at their last learned aversion training session (Figure 5 and Table 10.) Sporting, working, domestic working and toy dogs were able to demonstrate an aversive response (in the forms of boundary, attentional or complete aversion) to the non-toxic FoxOff® bait presentations more quickly (i.e., two to three training sessions) than the terrier breed group. Toy dogs required the least amount of time before they demonstrated complete aversion to the baits, displaying frequent anxious responses, followed by working, domestic working and sporting dogs with behavioural cues including fearful and anxious responses (Figure 7).

Robyn Taylor – October December 2017 55 Temperament and breed influences on learned aversion in domestic dogs

a). F(12,54) = 23.8, p = 0.000 b). F(12,54) = 2.63, p = 0.003

Figure 6. Generalised linear repeated measures model showing how the five domestic dog breed group's trainability changed across their four training sessions in relation to their a). Bait activity (bait touches and contact with bait), and b). Bait inspectional behaviour (time and distance spent with bait).

a). F(12,54) = 1.43, p=0.157 b). F(12,54) = 1.39, p=0.176

Figure 7. Generalised linear repeated measures model showing how the five domestic dog breed group's changed in a). Body positioning and b). Behaviour across their four training sessions in relation to their visual body reactions and stance when in the presence of bait.

56 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Effects of temperament on dog breed trainability

Dog temperament influenced the level of dog related bait activity, which consequently impacted dog breed trainability, (Table 6). The results indicated that ‘boldness’ and ‘fear and anxiety’, negatively correlated with levels of dog breed trainability. Fear and anxiety were associated with high levels of trainability within dog breeds (as demonstrated by the aversion results of both working and toy dogs); conversely, boldness was associated with low levels of trainability within dog breeds (as demonstrated by the aversion results of the terrier breed group).

Table 5. Univariate Tests of Significance between trainability (bait activity) and temperament traits; Fear and Anxiety, (F1, 54 = 7.21, p = 0.009) and Boldness (F1, 54 = 7.9, p = 0.007). Univariate Tests of Significance for trainability and temperament. Sigma-restricted parameterization Effective hypothesis decomposition; Std. Error of Estimate: 0.0417 Degr. of SS MS F p

Effect Freedom

Intercept 0.24 1 0.24 138.29 <0.001

Obedience 0.00 1 0.00 2.48 0.121

Aggression 0.00 1 0.00 0.00 0.991

Fear & Anxiety 0.01 1 0.012 7.21 0.009

Attachment 0.00 1 0.00 0.53 0.469

Excitability 0.00 1 0.00 1.23 0.272

Separation 0.00 1 0.00 0.00 0.951

Boldness 0.01 1 0.01 7.90 0.007

Playfulness 0.00 1 0.00 0.10 0.748

Error 0.08 47 0.00

Robyn Taylor – October December 2017 57 Temperament and breed influences on learned aversion in domestic dogs

Discussion

Learned aversion towards non-toxic FoxOff® baits differed between dog breeds as well as dog temperaments. The study indicated that working and toy dogs had the highest and easiest level of trainability, domestic working and sporting dogs were observed to have a moderate level of trainability and terriers were identified to be the most difficult to train.

Why learned aversion over positive reinforcement methods

Learned aversion training is likely to be more successful than positive reinforcement methods when training domestic dogs to avoid 1080 baits. Positive reinforcement methods work by presenting a motivating/conditioning stimulus to the dog after they have shown a desirable behaviour to their owner (i.e., the owner is required to give praise to their dog in the form of food and approval); (Peckmezian & Taylor, 2015). The advantage of using learned aversion training with domestic dogs to avoid 1080 baits, is that the training can occur in the owner’s absence (i.e., dogs are not reliant on owner’s positive reinforcement methods in order to avoid baits). For example, dogs tend to find baits when they are alone and if they are reliant on their owner for positive reinforcement to avoid it, the dog would more than likely take a bait, whereas if they were trained through learned aversion, they would be able to recognise the smell of the bait and avoid it regardless of their owner’s presence in the area.

Why some dog breeds are easier or more difficult to train than others

Based on the outcome of the study, working dogs had a high trainability level; this indicated that they were easier to train compared to the other breed groups (i.e., domestic working dogs, sporting dogs, toy dogs and terriers). As a result of their level of obligation towards their human demonstrator, highly trainable dogs, like working dogs, have been known to possess more skill in associating a command (such as ‘sit’), with the desired action (such as ‘sitting’), more quickly and efficiently compared to other dog breeds, (Helton, 2009; King et al., 2012;

Young, Olson, Reading, Amgalanbaatar, & Berger, 2011). The high trainability factor

58 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs associated with working dogs is contributed to the constant breeding selection pressures focussing on six desirable traits and characteristics, namely obedience, bite and grip, prey drive, neutral temperament, tracking and retrieving instincts over several generations in order to obtain a suitable working dog breed (such as pig detection dogs); (Turcsán et al., 2011). A working dog that has been bred with those desirable traits and characteristics is considered to be easily and highly trainable, compared to dogs that do not possess those six characteristics (Turcsán et al., 2011). The aforementioned traits selected for the breeding of working dogs have significantly impacted on how they react to certain experiences, what their training capabilities are as well as how they learn and train (Turcsán et al., 2011). Equally, Vandeloo (2009) had found that working dog breeds such as Border Collies demonstrated an easy and high trainability factor as a result of selective breeding programs. Additionally, Craigon et al.,

(2017), highlighted that high trainability levels in working dogs was related to their high levels of obedience, low aggression, fearfulness, low stress levels and low energy levels..

Domestic working dogs and sporting dogs are considered to have moderate levels of trainability compared to working dogs. Both domestic working and sporting dogs, were historically bred to be obedient towards their owners and to have a sufficient drive for a working (i.e., vertebrate pest detector dogs and guide dogs) and hunting (i.e., hunting and retrieval dogs) purpose (King et al., 2012; Turcsán et al., 2011). However, over a period of several generations a few working dog breeds have been domesticated for the purpose of human companionship rather than for their bred purpose, in the case of this study, 100% of both domestic working and sporting dogs were used as companion animals and did not participate in any working or activities (Dundas, Adams, & Fleming, 2014; King et al., 2012; Trut et al., 2004). As a result, those domestic working and sporting dog breeds within the study have lost some, if not all, their working drive and obedience towards their owners, therefore they have become difficult to train. Thus it is likely their level of trainability has been impacted by companionship domestication (Jackson et al., 2007; King et al., 2012). Additionally, Trut et al., (2004) explored the concept that dog breeds that were originally bred for a working purpose had lost their

Robyn Taylor – October December 2017 59 Temperament and breed influences on learned aversion in domestic dogs hunting drives and instincts, due to suppression as a result of domestication where selective pressures were placed on preferred behavioural and temperament traits within those dog breeds.

The concept of loss of working/hunting drives and instincts by working dog breeds due to selective pressures and domestication could explain the difficulties experienced when training both domestic working and sporting dogs to avoid non-toxic FoxOff® baits (Dundas et al., 2014;

King et al., 2012; Trut et al., 2004).

The study identified that terriers exhibited a low level of trainability in comparison to the other breed groups with 37% of terriers observed to demonstrate aggressive and bold behaviour towards the training device once presented with an electrified bait whilst not displaying any form of aversion. Previous research conducted by Radcliff, (2016) regarding dogs with unsound temperament could explain this outcome, where it was demonstrated that dogs that overreact to a situation or a stimuli tend to become excited, bold and aggressive. These temperaments often led to the dog’s attention becoming unfocussed and distracted, which impacts on their ability to learn a new task or skill during their training. Overall, Radcliff,

(2016) argued that dogs who expressed those temperaments during training were considered to be difficult to train. Additionally, Ziv, (2017), Deldalle and Gaunet, (2014) implied that negative reinforcement training methods, tend to elicit aggressive, bold, stubborn, and defensive reactions in dogs, which influences their levels of trainability, whereas positive reinforcement has shown to encourage obedience within breed groups. When considering the above theory, the sample terrier group in this study were considered difficult to train, possibly due to a preference of training, e.g., studies by Ziv, (2017) had demonstrated that dogs responded better towards positively reinforced methods over negatively reinforced ones. As negative reinforced methods of learned aversion were used within this study, research by Ziv, (2017) and Radcliff,

(2016) could explain the behavioural and trainability difficulties that were faced when training terriers.

60 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Aside from working dogs, the toy dogs breed group were often observed to be highly trainable in terms of learned aversion training. Considering that toy dogs were not bred as working dogs, but rather to be companions for their owners, their high trainability factor was unexpected

(Turcsán et al., 2011). The underlying factor that could explain the ease of toy dog trainability is their submissive nature towards negative or aggressive behaviour and/or punishment (Holland,

2007; Vandeloo, 2009). Submissive dogs tend to lack confidence (i.e., they often hesitated when approaching a bait placed within the area), and they tend to display behaviour associated primarily with fear and subordinance in an attempt to stop any negative punishments they were receiving (Holland, 2007; Vandeloo, 2009). Sixty-five percent of toy dogs within the study were observed to show submissive behaviour (extreme fear, hesitation and subordinance) soon after they received their first correction from the training device. This led to the dogs avoiding the area where the bait was present (i.e., they would show forms of complete aversion after having received one correction). Consequently, it is considered that toy dogs were highly trainable during their learned aversion training primarily due to their submissive nature which is substantiated by Hart (1995) and Vandeloo (2009) where both researchers observed that in regards to trainability and temperament, toy dogs were highly trainable as a result of their submissive behaviour and their need to please their owners.

Unique responses to the training

Four male dogs (each with a female partner) from both the sporting and terrier breed groups were observed to urinate on and/or around the electrified non-toxic FoxOff® bait. The behaviour could be explained by examining two techniques of urination used by dogs; ‘cache marking’

(marking food sources that no longer had nutrients left on them or food sources that were considered to be ‘bad’, thus labelling the food source as ‘there is no food left here’ or ‘this food is bad’) and ‘token marking’(the marking of objects or areas for the sole purpose of scent marking where little urine was expelled from the dog’s system to conserve enough urine to complete acts of scent marking). Sporting dogs within the study such as the Golden Retriever cross Labrador displayed possible signs of ‘cache marking’ by urinating on every bait sample

Robyn Taylor – October December 2017 61 Temperament and breed influences on learned aversion in domestic dogs presented. It is likely that the dog did not consider the baits to be food compared to the human- like-food diet provided by his owner, alternately he perceived the bait as ‘bad’ and wanted to communicate this to his female partner (Harrington, 1982). The other three dogs (terriers breed group) showed possible signs of ‘token marking’, all three males seemed to acknowledge the presence of the researcher and assistant, and as a result urinated within the areas where the researcher had been, which occasionally included urinating on the bait and equipment to rid the objects of the strange and unfamiliar scents (Harrington, 1982). These dogs may be considered to be territorial, the backyard was known to be their domain and any unfamiliar scent was removed through the form of ‘token marking’. Research has revealed that studies reported foxes and coyotes urinating on baits, dead rabbits and hamburgers following an encounter of feeling nauseous or sick after consuming a poisoned and/or sickness-inducing substances (Gustavson,

Garcia, Hankins, & Rusiniak, 1974; Gustavson, Kelly, Sweeney, & Garcia, 1976; Kinnear et al.,

2016). However, the main differences between this study in comparison with studies involving foxes and coyotes, was that both coyotes and foxes initially consumed the poisoned and/or sickness- inducing bait/food source and were observed to urinate on the bait/food source only after they were reported to vomit up the bait/food source they initially consumed (Gustavson et al., 1974; Gustavson et al., 1976; Kinnear et al., 2016). Sickness induced aversion may have influenced the coyotes and foxes’ behaviour in terms of avoiding poisoned baits/food sources, thus triggering the urination onto the bait/food source. Whereas, dogs within this study were observed to urinate on the non-toxic bait samples prior to any consumption of bait, therefore no sickness or pain induced aversion was observed that may have led to this particular behavioural response to the non-toxic FoxOff® bait. It is proposed that the male dog (sporting breed) within this study was using his urination as a signal to their female partners to avoid the bait as it was considered to be “bad” or “not food”.

Could the diet of dogs limit their interest towards non-toxic FoxOff® bait

Several dogs were fed unique diets that may have affected the outcome of the learned aversion training. Dogs must be attracted to the poison baits in order to interact with them appropriately,

62 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs facilitating learned aversion training. Dogs that never touch the non-toxic forms of the poison bait cannot be trained but are also unlikely to consume the baits in a scenario where they are free to locate the baits on a ‘baited’ property. Two sporting dogs, a Golden Retriever (female) and a Golden Retriever crossbreed Labrador (male), showed limited interest towards the non- toxic FoxOff® baits. The female retriever was observed to sniff then either sit on or walk away from the bait. The male was observed to sniff then urinate on the bait, thus he acknowledged the presence of non-toxic FoxOff® bait; however, made no attempt to lick, bite or consume it. These responses towards the presence of the baits were consistently observed from both during their four learned aversion training sessions. The owner reported that the dogs, living in the same household, received two meals per day, biscuits/snacks in the morning and a meal for dinner with a dental snack. The meals were freshly prepared, e.g., roast chicken, rice and doggie biscuits and not from a stereotypical can. This would suggest that the two retrievers have been subjected to ‘human-like’ food rather than the stereotypical dog food which is akin to the manufactured non-toxic FoxOff® bait. Dogs with this type of diet could not create a learn aversion towards the non-toxic FoxOff® bait as they did not recognise the bait as a food substance due to their ‘human- like’ food diet. These dogs are unlikely to be susceptible to poisoning from baits as they do not recognise or associate any poison bait they come across as food.

Additionally, an English Staffordshire terrier, from the terrier group, was accustomed to a frozen, raw food diet (raw meat rolled in rice and then frozen). The frozen meals were left outside to defrost, which may have conditioned the terrier to be less food orientated than other terriers, he was required to wait for the meal to defrost and thus was only able to consume parts of the meal over a period of time. In contrast to the sample dogs in the terrier group, the

Staffordshire terrier showed little to no interest in the electrified and non-electrified baits. Given that he was less food orientated towards the baits compared to the other terriers, the frozen food he was conditioned to eat may explain both his disinterest and limited response to the baits, as well as his higher level of trainability during the learned aversion training. Based on the

Robyn Taylor – October December 2017 63 Temperament and breed influences on learned aversion in domestic dogs observations, this dog is unlikely to consume poison bait he may encounter as he is not food orientated.

Could pain tolerance influence learned aversion in domestic dogs

Thirty percent of terriers showed no form of aversion after having been subjected to the four learned aversion training sessions. They were observed to totally consume three electrified baits during all of their training session. Previous research has reported that dog breeds such as terriers and huntaways had low sensitivity to pain (Bowden, Beausoleil, Stafford, Gieseg, &

Bridges, Submitted). The low sensitivity to pain could theoretically imply that terriers have the ability to withstand a certain amount of pain from the electrical corrections emitted by the learned aversion device. This may enable the terriers to ignore the corrections, thus allowing them to consume the bait rather than learn aversion from the pain caused by the corrections.

Consequently, a more powerful learned aversion device was required for the study, in conjunction with the addition of salt water to increase the flow of the electrical current from the device to the dog. It is likely that chronic health conditions with long term pain may be another factor influencing pain sensitivity levels which reduce the effectiveness of aversion training methods.

Genetics and breed: are modern day breed groups accurate in terms of their genetic DNA

It is questionable whether or not the dog breeds that were analysed within this study, did in fact, hold the genetic aspects associated with their breed. Genetic tests have been conducted to determine whether a dog from a specified breed actually did posses the genetic coding associated with that breed, or if the dog’s DNA was a result of a series of genetic coding derived from other breeds not normally associated with the breed that dog was classified as (Parker,

2012; Parker et al., 2004). Additionally, modern day DNA tests used microsatellites (repeating sequences of DNA) from signatures derived from particular breeds, rather than the actual genes derived from a genetic sample (Parker, 2012; Parker et al., 2004). Therefore, the genetic

64 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs identification of dog breeds was not based on their actual genes, and the results did not relate to behavioural or physical characteristics of a particular breed (Parker, 2012; Parker et al., 2004).

Thus, breeds that were considered to be different in characteristics could possibly have the same or similar genetic coding, thus the similarity would be based on DNA rather than physical appearance (Parker, 2012; Parker et al., 2004). Consequently, the dogs sampled in this study could potentially have possessed the genetic codes associated with their breed type, or they could have genetic codes more similar to other breeds rather than their assigned breed (Parker,

2012; Parker et al., 2004). Introducing a DNA testing aspect (using the Wisdom panel) to future similar studies could potentially reduce the uncertainty of whether or not the dogs are genetically associated with their assigned breeds (Parker, 2012; Parker et al., 2004). This could strengthen the connections and dissimilarities identified between different breed groups, temperament traits and their reactions towards learned aversion training (Parker, 2012; Parker et al., 2004).

Further research opportunities

Ideally, based on the variety of accidental poisoning occurrences to the domestic dog population, training programs should not only be restricted to train learned avoidance towards

1080 poison baits, but expanded to other bait forms such as insecticides, rodenticide and slug baits, as well as towards other toxic substances such as human medications and chocolate.

Study limitations

A larger sample of each breed group would have improved the accuracy of the results obtained from the analyses. The sample of dogs within each breed group remained consistent until the last training session (month 1), where approximately 12 dogs were withdrawn from the study, either due to the owners moving, the death of a dog or no data having been collected for that particular training session. The majority of the working dog group lacked the necessary data for their final training session as they were originally trained using FoxOff®, but then were trained using Probait® as the bait was more appropriate to their specific training requirements. The

Robyn Taylor – October December 2017 65 Temperament and breed influences on learned aversion in domestic dogs sampling method was limited to social media, no responses from pamphlet requests or posters in community centres generated participants. An additional limitation was the lack of pure bred dog breed groups, the majority participants were mixed breeds and the classification into dog breeds was reliant on the owner’s statement of .

Conclusion

Accidental poisoning of domestic dogs within Western Australia occurs due to 1080 baiting management plans, established within agricultural and conservational areas, to control vertebrate pest species (such as wild dogs, feral pigs, feral cats, and foxes). As domestic dogs are highly susceptible to the poison they are often at risk of a fatality. Learned aversion training in domestic dogs has shown that chances of accidental poisoning scenarios can be reduced, and thus baiting boundaries can be expanded.

Generally, the hypothesis that differences in temperament and breed in domestic dogs affects their reactions and behaviours when undergoing learned aversion training, was supported by the results of this study. The study identified that learned aversion training was more successful with particular dog breed groups (working, domestic working, sporting and toy dogs) than others (terriers). Significant differences between certain temperament traits (obedience, excitability, bold and playful behaviour) and different breed groups, as well as significant differences in the trainability and behaviour between those different breed groups were identified to support the hypothesis.

Dog breeds that were easier to train (working dogs) could either be attributed to selective breeding programs focussed on enhancing the dogs’ obedience and working purpose drive or the result of a naturally submissive nature; for example toy dogs. These factors made working and toy dog breed groups easier to train through learned aversion than the other breed groups selected for this study. The breeds that were fairly difficult to train, e.g., sporting and domestic

66 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs working dogs could be credited to their loss of working and hunting drive due to companionship domestication as a result of selective pressures placed on preferred behavioural traits by their owners. Lastly, terriers were the most difficult to train as they became aggressive and bold towards the negatively reinforced training method that was used within the study.

There is potential to conduct further research on the use of Canid Pest Ejectors (CPEs) using pepper as the substance that is expelled from the device into the mouth of the domestic dog

(Government of Western Australia, 2017; Fleming et al., 2006). This form of CPE could be combined with taste aversion, where, once lured to the CPE with bait, pepper would be ejected from the shaft of the CPE and into the dog’s open mouth during their attempt to consume the bait (Government of Western Australia, 2017; Fleming et al., 2006). This could potentially discourage the dog from attempting to consume bait, due to the burning sensation and discomfort caused to the dog by the pepper (i.e., taste aversion). This method could potentially result in learned aversion within domestic dogs, including those who display no reaction corrections delivered from the learned aversion device.

The outcomes from this study may contribute towards land managers being more confident in terms of bait deployment, as well as domestic dog owners feeling reassured about the welfare of their companions. By meeting both parties’ needs, there is potential for 1080 control programs to continue reducing invasive predators within areas with long-term conservation and livestock production outcomes, as well as reducing accidental poisoning.

Robyn Taylor – October December 2017 67 Temperament and breed influences on learned aversion in domestic dogs

Acknowledgements

Many thanks to John Snowball, Murdoch University’s electrical technical officer who created the self-training device that was used in this project. Thanks also goes to Peter Adams for assistance with the training methodology.

To Damon van der Linde for keeping me company on the long drives and the training sessions, you were the best assistant I ever had.

A special thank you to my parents, Lynda and Johann Taylor for their help and support during the editing and analysis process.

This study has approval of the Animal Research Ethics Committee (R2809/16).

68 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Appendix

APPENDIX 1: TRAINING DEVICE INFORMATION AND SET UP

The device consists of an energizer which is attached to a non-toxic Fox-off ® bait (Animal

Control Technologies Australia) an earthing rod which is placed in the ground (Figure 7). The bait was placed on a piece of clear plastic so that bait did not earth out on the ground and an electric current was created. A single electrical pulse (cycle rate of 1 Hz) is given out every second allowing the dog time to investigate the area and to possibly make contact with the bait before the dog receives a correction (Kreplins et al., Draft). This is because the dog may be alerted too early if a continual or faster pulse rate was used, which would in turn only train the dog to avoid baits in which they sense to be electrified (Kreplins et al., Draft). The maximum output that the device has consists of 550V and 10mA, which sits well below the lethal level of

100mA (Kreplins et al., Draft). The device system was designed so the dog undergoing the training would only see the bait (Kreplins et al., Draft). All cords connecting the earthing rod and learned aversion device to the bait are hidden from view using vegetation, sand, wood and other items from nearby (Figure 8) (Kreplins et al., Draft).

Robyn Taylor – October December 2017 69 Temperament and breed influences on learned aversion in domestic dogs

a). b).

c).

Figure 8. Self- training device. (a) The unit itself, (b) the self-training device including the energizer, cords, earthing rod and non-toxic version of the FoxOff Bait and, (c) the dogs undergoing the training will only see the FoxOff® Bait (sourced from Animal Control Technologies, contains green scat markers) (Kreplins et al., Draft).

70 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Figure 9. The two drawings depict the set-up of the training device, bait and wires. The first image above showcases the device before it is hidden, and the image below it shows how the device is to be hidden by bushes, leaves or other objects within the area from the participating dog, (images by Robyn Taylor, 2017).

Robyn Taylor – October December 2017 71 Temperament and breed influences on learned aversion in domestic dogs

APPENDIX 2: TEMPERAMENT CATEGORIES DERIVED FROM C-BARQ AND TEMPERAMENT

TEST

C-BARQ Obedience and Trainability scoring categories

1. When off the leash, returns immediately when called.

2. Obeys the "sit" command immediately.

3. Obeys the "stay" command immediately.

4. Seems to attend/listen closely to everything you say or do.

5. Slow to respond to correction or punishment; "thick-skinned".

6. Slow to learn new tricks or tasks.

7. Easily distracted by interesting sights, sounds, or smells.

8. Will "fetch" or attempt to fetch sticks, balls, or objects.

C-BARQ Aggression scoring categories

1. When verbally corrected or punished (scolded, shouted at, etc.) by you or a household

member.

2. When approached directly by an unfamiliar adult while being walked/exercised on a

leash.

3. When approached directly by an unfamiliar child while being walked/exercised on a

leash.

4. Toward unfamiliar persons approaching the dog while s/he is in your car (at the gas

station, for example).

5. When toys, bones or other objects are taken away by a household member.

6. When bathed or groomed by a household member.

7. When an unfamiliar person approaches you or another member of your family at home.

8. When unfamiliar persons approach you or another member of your family away from

your home.

72 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

9. When approached directly by a household member while s/he is eating.

10. When mailmen or other delivery workers approach your home.

11. When his/her food is taken away by a household member.

12. When strangers walk past your home while your dog is outside or in the yard.

13. When an unfamiliar person tries to touch or pet the dog.

14. When joggers, cyclists, rollerbladers or skateboarders pass your home while your dog is

outside or in the yard.

15. When approached directly by an unfamiliar male dog while being walked/exercised on

a leash.

16. When approached directly by an unfamiliar female dog while being walked/exercised

on a leash.

17. When stared at directly by a member of the household.

18. Toward unfamiliar dogs visiting your home.

19. Toward cats, squirrels or other animals entering your yard.

20. Toward unfamiliar persons visiting your home.

21. When barked, growled, or lunged at by another (unfamiliar) dog.

22. When stepped over by a member of the household.

23. When you or a household member retrieves food or objects stolen by the dog.

24. Towards another (familiar) dog in your household.

25. When approached at a favourite resting/sleeping place by another (familiar) household

dog.

26. When approached while eating by another (familiar) household dog.

27. When approached while playing with/chewing a favourite toy, bone, object, etc., by

another (familiar) household dog.

Robyn Taylor – October December 2017 73 Temperament and breed influences on learned aversion in domestic dogs

Temperament test: Stranger Aggression behavioural cues

The behavioural cues that were measured to distinguish whether a dog showed any stranger aggression was based on a scale of 0-4 where 0-1 represented no aggression (no visible signs of aggression), 2-3 represented moderate aggression (growling, barking and baring teeth), and 4 represented serious aggression (snaps, bites or attempts to bite). Table 6.

Table 6. Dog Stranger Aggression levels based on behavioural cues that are measured on a Likert Scale, adapted from; (Vas et al., 2005). 2-3: Moderate Behavioural cues 0-1: No aggression Aggression 4: Serious Aggression Hackles Neutral. Lowered. Raised. Body response Neutral. Alert. Threatening. Body position Neutral. Alert stance. Threatening stance. Head positioned Head positioned high forward in line with with nose directed Head position Neutral/Relaxed. spine and directed upwards, towards towards stranger. stranger. Ear position Neutral/Relaxed. Alert. Alert and forward. Raised high above Tail position Neutral/Relaxed. Raised to level of spine. spine. Excessive Vocal cues None. Growling at stranger. barking/growling at stranger. Clenched jaw, occasional bearing of Mouth position Neutral/Relaxed. Baring teeth. teeth and opening of mouth. Straight and extended, Back leg position Normal. Flexed. boarding on flexed. Straight and extended, Front leg position Normal. Flexed. bordering on flexed.

C-BARQ Fear and Anxiety scoring categories

1. When approached directly by an unfamiliar adult while away from your home.

2. When approached directly by an unfamiliar child while away from your home.

3. In response to sudden or loud noises (e.g., vacuum cleaner, car backfire, road drills,

objects being dropped, etc.).

4. When unfamiliar persons visit your home.

5. When an unfamiliar person tries to touch or pet the dog.

6. In heavy traffic.

74 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

7. In response to strange or unfamiliar objects on or near the sidewalk (e.g., plastic trash

bags, leaves, litter, flags flapping, etc.).

8. When examined/treated by a veterinarian.

9. During thunderstorms, firework displays, or similar events.

10. When approached directly by an unfamiliar dog of the same or larger size.

11. When approached directly by an unfamiliar dog of smaller size.

12. When first exposed to unfamiliar situations (e.g., first car trip, first time in elevator, first

visit to veterinarian, etc.).

13. In response to wind or wind-blown objects.

14. When having nails clipped by a household member.

15. When groomed or bathed by a household member.

16. When stepped over by a member of the household.

17. When having his/her feet toweled by a member of the household.

18. When unfamiliar dogs visit your home.

19. When barked, growled, or lunged at by an unfamiliar dog.

Robyn Taylor – October December 2017 75 Temperament and breed influences on learned aversion in domestic dogs

Temperament test: Novel object Fear and Anxiety observed behavioural cues

The behavioural cues that were measured to distinguish whether a dog showed any fear and anxiety towards the novel object was based on a scale of 0-4 where 0-1 represented no fear and anxiety (no visible signs of cowering and submission), 2-3 represented moderate fear and anxiety (some whining and cowering is present), and 4 represented extreme fear and anxiety

(whining, cowering and submission is present). Table 7.

Table 7. Dog Fear and Anxiety levels based on behavioural cues that are measured on a Likert Scale, adapted from; (Beerda et al., 1998). 0-1: No fear and 2-3: Moderate fear 4: Extreme fear Behavioural cues anxiety and anxiety and anxiety Hackles Neutral. Lowered. Flat against body. Body response Neutral. Cowering. Submissive Sitting or lying on Lying still with back Body position Neutral. stomach in a on ground. cowering stance. Lowered or leaning Head position Neutral/Relaxed. Lowered. against the ground. Lowered against Ear position Neutral/Relaxed. Flat against head. head. Lowered from the Tucked between Tail position Neutral/Relaxed. spine, but not tucked hind legs. between hind legs. Whining and Vocal cues None. Yapping. moaning. Clenched and Constantly opening Mouth position Neutral/Relaxed. occasionally opening and closing to let to let moans escape. out whines. Held in the air as Tucked in, under or Back leg position Normal. the dog is lying on against body. their back. Held in the air as Tucked in, under or Front leg position Normal. the dog is lying on against body. their back.

C-BARQ Separation related behaviour scoring categories

1. Shaking, shivering, or trembling.

2. Excessive salivation.

3. Restlessness, agitation, or pacing.

4. Whining.

5. Barking.

6. Howling.

76 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

7. Chewing or scratching at doors, floor, windows, curtains, etc.

8. Loss of appetite.

C-BARQ Excitability scoring categories

1. When you or other members of the household come home after a brief absence.

2. When playing with you or other members of your household.

3. When doorbell rings.

4. Just before being taken for a walk.

5. Just before being taken on a car trip.

6. When visitors arrive at your home.

C-BARQ Attachment and Attention seeking scoring categories

1. Displays a strong attachment for one particular member of the household.

2. Tends to follow you (or other members of the household) about the house, from room to

room.

3. Tends to sit close to, or in contact with, you (or others) when you are sitting down.

4. Tends to nudge, nuzzle or paw you (or others) for attention when you are sitting down.

5. Becomes agitated (whines, jumps up, tries to intervene) when you (or others) show

affection for another person.

6. Becomes agitated (whines, jumps up, tries to intervene) when you (or others) show

affection for another dog or animal.

C-BARQ Bold behaviour scoring categories

1. Chases or would chase cats given the opportunity.

2. Escapes or would escape from home or yard given the chance.

3. Steals food.

Robyn Taylor – October December 2017 77 Temperament and breed influences on learned aversion in domestic dogs

4. Urinates against objects/furnishings in your home.

5. Investigates unknown objects with confidence.

6. Not fearful.

Temperament test: Novel object Bold observed behavioural cues

The behavioural cues that were measured to distinguish whether a dog showed any boldness towards the novel object was based on a scale of 0-4 where 0-1 represented no boldness (no visible signs of fearlessness), 2-3 represented moderate boldness (dog was demonstrating courage), and 4 represented extreme boldness (dog was demonstrating fearlessness). Table 8.

Table 8. Dog bold levels based on behavioural cues that are measured on a Likert Scale, adapted from;(Beerda et al., 1998). 2-3: Moderate 4: Extreme Behavioural cues 0-1: No boldness boldness boldness Hackles Neutral. Lowered. Raised. Body response Neutral. Courageous. Fearless. Body position Neutral. Alert. Dominant. Held along spine Head position Neutral/Relaxed. Held upwards. level. Ear position Neutral/Relaxed. Alert. Alert and Forward. Tucked slightly between hind legs Held along length (cautious boldness) Tail position Neutral/Relaxed. and height of spine. or raised high above spine (fearless boldness). Vocal cues None. Yapping. Excessive barking. Open, often Mouth position Neutral/Relaxed. Clenched and closed. panting. Back leg position Normal. Flexed. Crouched. Front leg position Normal. Flexed. Flexed.

C-BARQ Playful behaviour scoring categories

1. Barks persistently when excited.

2. Licks people persistently.

3. Begs constantly for attention.

4. Playful, puppyish and boisterous.

5. Chases or would chase anything.

6. Chases own tail.

78 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Temperament test: Novel object Playful observed behavioural cues

The behavioural cues that were measured to distinguish whether a dog showed any playfulness towards the novel object was based on a scale of 0-4 where 0-1 represented no playfulness (no visible signs of boisterous behaviour), 2-3 represented moderate playfulness (dog was demonstrating boisterous behaviour), and 4 represented extreme playfulness (dog was demonstrating energetic behaviour). Table 9.

Table 9. Dog playfulness levels based on behavioural cues that are measured on a Likert Scale, adapted from; (Beerda et al., 1998). 2-3: Moderate 4: Extreme Behavioural cues 0-1: No playfulness playfulness playfulness Hackles Neutral. Lowered. Raised. Body response Neutral. Boisterous. Energetic. Backside positioned Relaxed with slight up in the air with Body position Neutral. alert crouch. body at an alert stance. Head is lowered as Head is lowered Head position Neutral/Relaxed. the backside is along the spine. lifted into the air. Ear position Neutral/Relaxed. Relaxed. Raised. High in the air, Tail position Neutral/Relaxed. Raised above spine. often wagging. Vocal cues None. Slight yapping. Barking. Open, tongue may Opening and closing be exposed and Mouth position Neutral/Relaxed. on occasion. dog may be panting. Flexed in order to Back leg position Normal. Slightly flexed. raise backside into the air. Crouched low to the ground in order Front leg position Normal. Slightly crouched. to raise backside into the air.

Robyn Taylor – October December 2017 79 Temperament and breed influences on learned aversion in domestic dogs

APPENDIX 3: LEARNED AVERSION TRAININ CERTIFICATE

80 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Figure 10. Learned aversion training certificate for dog owners, congratulating them on their dog’s successful performance in the training program, with the additional information of their dog’s temperament levels situated across six different temperament traits, (designed by Robyn Taylor, 2017).

Robyn Taylor – October December 2017 81 Temperament and breed influences on learned aversion in domestic dogs

APPENDIX 4: TABLE OF BREED GROUP PERCENTAGES

Table 10. Table comparing the first and last training session of breed group percentages in relation to distance from bait, time spent with bait, type of learned aversions demonstrated, body positions and behaviour responses to bait. a). Distance from bait (meters) Dog classification Training Sessions Descriptor 0 1 2 3 4 5 Working dogs (n = 12) Day 0 Count 12 0 0 0 0 0 Percentage 100.00% 0.00% 0.00% 0.00% 0.00% 0.00% Month 1 Count 0 0 1 4 0 7 Percentage 0.00% 0.00% 8.33% 33.33% 0.00% 58.33% Domestic working (n = 11) Day 0 Count 11 0 0 0 0 0 Percentage 100.00% 0.00% 0.00% 0.00% 0.00% 0.00% Month 1 Count 2 2 2 2 2 1 Percentage 18.18% 18.18% 18.18% 18.18% 18.18% 9.09% Sporting dogs (n = 14) Day 0 Count 13 0 0 0 0 1 Percentage 92.86% 0.00% 0.00% 0.00% 0.00% 7.14% Month 1 Count 1 2 4 6 0 1 Percentage 7.14% 14.29% 28.57% 42.86% 0.00% 7.14% Toy dogs (n = 11) Day 0 Count 10 0 1 0 0 0 Percentage 90.91% 0.00% 9.09% 0.00% 0.00% 0.00% Month 1 Count 1 1 2 2 5 0 Percentage 9.09% 9.09% 18.18% 18.18% 45.45% 0.00% Terriers (n = 8) Day 0 Count 8 0 0 0 0 0 Percentage 100.00% 0.00% 0.00% 0.00% 0.00% 0.00% Month 1 Count 3 2 1 2 0 0 Percentage 37.50% 25.00% 12.50% 25.00% 0.00% 0.00%

82 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Dog Training b). Time with bait (minutes) classification Sessions Descriptor 0 1 2 3 4 5 6 7 8 9 10 Working dogs Day 0 Count 2 3 3 1 1 1 0 1 0 0 0 (n = 12) Percentage 16.67% 25.00% 25.00% 8.33% 8.33% 8.33% 0.00% 8.33% 0.00% 0.00% 0.00%

Month 1 Count 12 0 0 0 0 0 0 0 0 0 0 Percentage 100.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% Domestic Day 0 Count 6 2 1 1 0 1 0 0 0 0 0 working Percentage 54.55% 18.18% 9.09% 9.09% 0.00% 9.09% 0.00% 0.00% 0.00% 0.00% 0.00% (n = 11) Month 1 Count 10 1 0 0 0 0 0 0 0 0 0

Percentage 90.91% 9.09% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% Sporting dogs Day 0 Count 10 2 1 0 1 0 0 0 0 0 0 (n = 14) Percentage 71.43% 14.29% 7.14% 0.00% 7.14% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

Month 1 Count 14 0 0 0 0 0 0 0 0 0 0 Percentage 100.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% Toy dogs Day 0 Count 8 2 1 0 0 0 0 0 0 0 0 (n = 11) Percentage 72.73% 18.18% 9.09% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

Month 1 Count 9 2 0 0 0 0 0 0 0 0 0 Percentage 81.82% 18.18% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% Terriers Day 0 Count 5 0 1 1 0 1 0 0 0 0 0 (n = 8) Percentage 62.50% 0.00% 12.50% 12.50% 0.00% 12.50% 0.00% 0.00% 0.00% 0.00% 0.00%

Month 1 Count 0 0 0 0 6 2 0 0 0 0 0 Percentage 0.00% 0.00% 0.00% 0.00% 75.00% 25.00% 0.00% 0.00% 0.00% 0.00% 0.00%

Robyn Taylor – October December 2017 83 Temperament and breed influences on learned aversion in domestic dogs

Dog classification Training Sessions Descriptors c). Urination on bait d). Demonstrated learned aversion Working dogs (n = 12) Day 0 Count 0 10 Percentage 0.00% 83.33% Month 1 Count 0 12 Percentage 0.00% 100.00% Domestic working (n = 11) Day 0 Count 0 8 Percentage 0.00% 72.73% Month 1 Count 0 11 Percentage 0.00% 100.00% Sporting dogs (n = 14) Day 0 Count 1 12 Percentage 7.14% 85.71% Month 1 Count 0 13 Percentage 0.00% 92.86% Toy dogs (n = 11) Day 0 Count 0 7 Percentage 0.00% 63.64% Month 1 Count 0 9 Percentage 0.00% 81.82% Terriers (n = 8) Day 0 Count 3 4 Percentage 37.50% 50.00% Month 1 Count 3 5 Percentage 37.50% 62.50%

84 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

e). Type of aversion Dog classification Training Sessions Descriptors Inspectional Boundary Attentional Complete None Working dogs Day 0 Count 1 6 4 1 0 (n = 12) Percentage 8.33% 50.00% 33.33% 8.33% 0.00%

Month 1 Count 0 6 0 6 0 Percentage 0.00% 50.00% 0.00% 50.00% 0.00% Day 0 Count 3 2 4 0 2 Domestic working (n = 11) Percentage 27.27% 18.18% 36.36% 0.00% 18.18%

Month 1 Count 0 4 7 0 0 Percentage 0.00% 36.36% 63.64% 0.00% 0.00% Sporting dogs Day 0 Count 4 3 6 0 1 (n = 14) Percentage 28.57% 21.43% 42.86% 0.00% 7.14%

Month 1 Count 0 5 8 0 1 Percentage 0.00% 35.71% 57.14% 0.00% 7.14% Toy dogs Day 0 Count 3 0 4 2 2 (n = 11) Percentage 27.27% 0.00% 36.36% 18.18% 18.18%

Month 1 Count 0 0 1 8 2 Percentage 0.00% 0.00% 9.09% 72.73% 18.18% Terriers Day 0 Count 3 0 2 0 3 (n = 8) Percentage 37.50% 0.00% 25.00% 0.00% 37.50% Month 1 Count 0 1 4 0 3 Percentage 0.00% 12.50% 50.00% 0.00% 37.50%

Robyn Taylor – October December 2017 85 Temperament and breed influences on learned aversion in domestic dogs

f). Body position g). Behaviour to bait Dog classification Training Sessions Descriptor Neutral Fearful Alert Neutral Bold Fear Aggressive Working dogs Day 0 Count 9 2 1 9 0 3 0 (n = 12) Percentage 75.00% 16.67% 8.33% 75.00% 0.00% 25.00% 0.00%

Month 1 Count 1 10 1 2 0 10 0 Percentage 8.33% 83.33% 8.33% 16.67% 0.00% 83.33% 0.00% Domestic working Day 0 Count 8 1 2 9 1 1 0 (n = 11) Percentage 72.73% 9.09% 18.18% 81.82% 9.09% 9.09% 0.00%

Month 1 Count 5 4 2 6 1 5 0 Percentage 45.45% 36.36% 18.18% 54.55% 9.09% 45.45% 0.00% Sporting dogs Day 0 Count 9 2 3 10 2 2 0 (n = 14) Percentage 64.29% 14.29% 21.43% 71.43% 14.29% 14.29% 0.00%

Month 1 Count 6 5 3 7 2 5 0 Percentage 42.86% 35.71% 21.43% 50.00% 14.29% 35.71% 0.00% Day 0 Count 3 5 1 5 0 6 0 Toy dogs (n = 11) Percentage 27.27% 45.45% 9.09% 45.45% 0.00% 54.55% 0.00%

Month 1 Count 1 9 1 2 0 9 0 Percentage 9.09% 81.82% 9.09% 18.18% 0.00% 81.82% 0.00% Day 0 Count 1 0 7 5 3 0 0 Terriers (n = 8) Percentage 12.50% 0.00% 87.50% 62.50% 37.50% 0.00% 0.00%

Month 1 Count 5 0 3 1 3 0 3 Percentage 62.50% 0.00% 37.50% 12.50% 37.50% 0.00% 37.50%

86 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

References

Allen, L. R. (2015). Demographic and functional responses of wild dogs to poison baiting. Ecological Management & Restoration, 16(1), 58-66. Baker, S. E., Johnson, P. J., Slater, D., Watkins, R. W., & Macdonald, D. W. (2007). Learned Aversion with and without an odour cue for protecting untreated baits from wild mammal foraging. Applied Animal Behaviour Science, 102(3), 410- 428. Bates, N. S., Sutton, N. M., & Campbell, A. (2012). Suspected metaldehyde slug bait poisoning in dogs: a retrospective analysis of cases reported to the Veterinary Poisons Information Service. The Veterinary record, 171(13), 324-324. doi: 10.1136/vr.100734 Batt, L. S., Batt, M. S., Baguley, J. A., & McGreevy, P. D. (2008). Factors associated with success in training. Journal of Veterinary Behavior: Clinical Applications and Research, 3(4), 143-151. doi: https://doi.org/10.1016/j.jveb.2008.04.003 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 to different types of stimuli in dogs. Applied Animal Behaviour Science, 58(3), 365-381. doi: 10.1016/S0168-1591(97)00145-7 Bowden, J., Beausoleil, N. J., Stafford, K. J., Gieseg, M. A., & Bridges, J. (Submitted). A prospective study of breed differences in the thermal pain sensitivity of dogs. Veterinary Anaesthesia and Analgesia. doi: 10.1016/j.vaa.2017.07.003 Cagnacci, F., Mossei, G., Cowan, D. P., & Delahay, R. J. (2004). Can learned aversion be used to control bait uptake by Eurasian badgers? Applied Animal Behaviour Science, 92(1-2), 159-168. Craigon, P. J., West, P. H., Gary, C. W. E., Whelan, C., Lethbridge, E., & Asher, L. (2017). “She’s a dog at the end of the day”: Guide dog owners’ perspectives on the behaviour of their guide dog. PLoS One U6 Journal Article, 12(4). doi: 10.1371/journal.pone.0176018 Deldalle, S., & Gaunet, F. (2014). Effects of 2 training methods on stress-related behaviors of the dog and on the dog-owner relationship. Journal of Veterinary Behavior: Clinical Applications and Research, 9(2), 58-65. doi: 10.1016/j.jveb.2013.11.004 Diederich, C., & Giffroy, J. M. (2005). Behavioural testing in dogs: a review of methodology in search for standardisation. Applied Animal Behaviour Science, 97, 57-72. Dundas, S. J., Adams, P. J., & Fleming, P. A. (2014). First in, first served: uptake of 1080 poison fox baits in south-west Western Australia. Ewing, T. (2006). Poisoning: an ever-present threat: deadly toxic substances are everywhere in the typical dog's environment. Dog Watch, 10(9), 1. Fleming, P. J. S., Allen, L. R., Lapidge, S. J., Robley, A., Saunders, G. R., & Thomson, P. C. (2006). A strategic approach to mitigating the impacts of wild canids: proposed activities of the Invasive Animals Cooperative Research Centre.

Robyn Taylor – October December 2017 87 Temperament and breed influences on learned aversion in domestic dogs

Australian Journal of Experimental Agriculture, 46(7), 753. doi: 10.1071/EA06009 Fratkin, J. L., Sinn, D. L., Patall, E. A., & Gosling, S. D. (2013). Personality consistency in dogs: A meta-analysis. PLOS ONE, 8(1). Glen, A. S., Gentle, M. N., & Dickman, C. R. (2007). Non‐target impacts of poison baiting for predator control in Australia. Mammal Review, 37(3), 191-205. doi: 10.1111/j.1365-2907.2007.00108.x Goh, C. S. S., Hodgson, D. R., Fearnside, S. M., Heller, J., & Malikides, N. (2005). Sodium monofluoroacetate (Compound 1080) poisoning in dogs. Australian Veterinary Journal, 83(8), 474-479. doi: 10.1111/j.1751-0813.2005.tb13296.x Government of Western Australia (2016 ). 1080: Characteristics and use.: Government of Western Australia Retrieved 19/06, 2017 from https://www.agric.wa.gov.au/1080/1080-characteristics-and-use. Government of Western Australia (2017). Canid Pest Ejectors. Department of Primary Industries and Regional Development Retrieved 20/06, 2017 from https://www.agric.wa.gov.au/invasive-species/canid-pest-ejectors. Gregory, T. (2011). Qualitative behavioural assessment on shelter dogs: Do enriched hides or human interaction affect dog behaviour? (Bsc. Honours thesis in Veterinary Biology), Murdoch University, Murdoch, Western Australia. Gustavson, C. R., Garcia, J., Hankins, W. G., & Rusiniak, K. W. (1974). Coyote Predation Control by Aversive Conditioning. Science, 184(4136), 581-583. doi: 10.1126/science.184.4136.581 Gustavson, C. R., Kelly, D. J., Sweeney, M., & Garcia, J. (1976). Prey-lithium aversions. I: coyotes and . Behavioral Biology, 17(1), 61-72. doi: 10.1016/S0091-6773(76)90272-8 Harrington, F. H. (1982). Urine marking at food and caches in captive coyotes. Canadian Journal of Zoology, 60(5), 776-782. doi: 10.1139/z82-107 Hart, B. L. (1995). Analysing breed and gender differences in behaviour. The Domestic Dog: It’s Evolution, Behaviour and Interactions with People. (Vol. 5, pp. 66- 78): Serpell, J. Helton, W. S. (2009). Cephalic index and perceived dog trainability. Behavioural Processes, 82(3), 355-358. doi: 10.1016/j.beproc.2009.08.004 Helton, W. S. (2010). Does perceived trainability of dog ( Canis lupus familiaris) breeds reflect differences in learning or differences in physical ability? Behavioural Processes, 83(3), 315-323. doi: 10.1016/j.beproc.2010.01.016 Helton, W. S. (2010). Does perceived trainability of dog (Canis lupus familiaris)breeds reflect differences in learning or differences in physical ability? Behavioural Processes, 83, 315-323. Holland, C. C. (2007). Leaking with excitement: if your dog urinates during greetings and other times, behavioral remedies can help. Dog Watch, 11, 6+. Horowitz, A. (2012). Fair is fine, but more is better: Limits to inequity aversion in domestic dog. Social Justice Research, 25, 195-212. Jackson, J., Moro, D., Mawson, P., Lund, M., & Mellican, A. (2007). Bait Uptake and Caching by Red Foxes and Nontarget Species in Urban Reserves. The Journal of Wildlife Management, 71(4), 1134-1140. doi: 10.2193/2006-236

88 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

Jones, A. C., & Gosling, S. D. (2005). Temperament and Personality in dogs (Canis familiaris). A review and evaluation of past research. Applied Animal Behaviour Science, 95, 1-53. King, T., Marston, L. C., & Bennett, P. C. (2012). Breeding dogs for beauty and behaviour: Why scientists need to do more to develop valid and reliable behaviour assessments for dogs kept as companions. Applied Animal Behaviour Science, 137(1-2), 1-12. doi: 10.1016/j.applanim.2011.11.016 Kinnear, J. E., Pentland, C., Moore, N., & Krebs, C. J. (2016). Fox control and 1080 baiting conundrums: time to prepare for a CRISPR solution. Australian Mammalogy. doi: https://doi.org/10.1071/AM16020 Kreplins, T. L., Adams, P., Bateman, P. W., Dundas, S. J., Kennedy, M. S., & Fleming, P. A. (Unpublished data). Training domestic dogs to avoid poison baits using a 'learned aversion' device. Murdoch University. Ley, J. M., & Benett, P. C. (2007). Understanding personality by understanding companion dogs. Anthrozoos, 20(2). Maejima, M., Inoue-Murayama, M., Tonosaki, K., Matsuura, N., Kato, S., Saito, Y., . . . Ito, S. (2006). Traits and genotypes may predict the successful training of drug detection dogs. . Applied Animal Behaviour Science, 19, 572-579. Marshall-Pescini, S., Frazzi, C., & Valecchi, P. (2016). The effect of training and breed group on problem-solving behaviours in dogs. Animal Cognition, 19(3), 572- 579. McKinley, S., & Young, R. J. (2003). The efficacy of the model-rival method when compared with operant condition for training domestic dogs to perform a retrieval-selection task. Applied Animal Behaviour, 81, 357-365. McLeod, R., Norris, A., & Cooperative Research Centre for the Biological Control of Pest, A. (2004). Counting the cost: impact of invasive animals in Australia, 2004. Canberra: Cooperative Research Centre for Pest Animal Control. Meenken, D., & Booth, L. H. (2010). The risk of dogs to poisoning from sodium monofluoroacetate (1080) residues in possums (Trichosurus vulpecula). New Zealand Journal of Agricultural Research, 40(4), 573-576. Moss, Z. N., O'Connor, C. E., & Hickling, G. J. (1998). Implications of prefeeding for the development of bait aversions in brush tail possums. Wildlife Research, 25, 133-138. Most popular dog breeds. (2003). Weekly Reader, Senior Edition (including Science Spin), 82(11), S3. Parker, H. G. (2012). Genomic analyses of modern dog breeds. Mammalian Genome, 23(1), 19-27. doi: 10.1007/s00335-011-9387-6 Parker, H. G., Kim, L. V., Sutter, N. B., Carlson, S., Lorentzen, T. D., Malek, T. B., . . . Kruglyak, L. (2004). Genetic Structure of the Purebred Domestic Dog. Science, 304(5674), 1160-1164. doi: 10.1126/science.1097406 Peckmezian, T., & Taylor, P. W. (2015). Electric shock for aversion training of jumping spiders: towards an arachnid model of avoidance learning. Behavioural Processes, 113, 99–104. Popular breeds of 2011. (2012). Dog Watch, 16(5), 2.

Robyn Taylor – October December 2017 89 Temperament and breed influences on learned aversion in domestic dogs

Radcliff, R. (2016). Understanding Dogs: Temperament in Dogs- Its role in decision making. Humane Society International, 1-5. Range, F., Horn, L., Virany, Z., & Huber, L. (2008). The absence of rewards induces inequity aversion in dogs. Proceedings of The National Academy of Sciences, 106, 340-345. Saunders, G., Coman, B., Kinnear, J., & Braysher, M. (1995). Managing Vertebrate pests: Foxes. Canberra: Australian Government Publishing Service. Serpell, J. A. (2017). Canine Behavioural Assessment and Research Questionnaire. CBARQ. Retrieved 20/05, 2017, from http://vetapps.vet.upenn.edu/cbarq/about.cfm Serpell, J. A., & Hsu, Y. (2005). Effect of breed, sex, and neuter status on trainability in dogs. Journal of the Interactions of People and Animals, 13, 197-207. Starling, M. J., Branson, P., Thomson, P. C., & McGreevy, P. D. (2013). "Boldness" in the domestic dog differs among breeds and breed groups. Behavioural Processes, 97, 53-62. Svartberg, K., & Forkman, B. (2002). Personality traits in the domestic dog (Canis familiaris). Applied Animal Behaviour Science, 79, 133-155. Taylor, K. D., & Mills, D. S. (2006). The Development and assessment of temperament tests for adult companion dogs. Journal of Veterinary Behaviour: Clinical Applications and Research, 1(3), 94-108. Tonoike, A., Nagasawa, M., Mogi, K., Serpell, J. A., & Ohtsuki, H. (2015). Comparison of owner-reported behavioural characteristics among genetically clustered breeds of dogs (Canis familiaris). Scientific Reports, 5(117710). Trut, L. N., Plyusnina, I. Z., & Oskina, I. N. (2004). An Experiment on Fox Domestication and Debatable Issues of Evolution of the Dog. Russian Journal of Genetics, 40(6), 644-655. doi: 10.1023/B:RUGE.0000033312.92773.c1 Turcsán, B., Kubinyi, E., & Miklósi, Á. (2011). Trainability and boldness traits differ between dog breed clusters based on conventional breed categories and genetic relatedness. Applied Animal Behaviour Science, 132(1), 61-70. doi: 10.1016/j.applanim.2011.03.006 Twigg, L. E., Eldridge, S. R., Edwards, G. P., Shakeshaft, B. J., dePreu, N. D., & Adams, N. (2000). The longetivity and efficiancy of 1080 meat baits used for dingo control in central Australia. Wildlife Research, 27, 473-481. Vandeloo, J. (2009). The importance of intergrating natural selection within a detector dog selective breeding model. Journal of Veterinary Behavior: Clinical Applications and Research, 4(6), 245-245. doi: 10.1016/j.jveb.2009.05.005 Vas, J., Topál, J., Gácsi, M., Miklósi, Á., & Csányi, V. (2005). A friend or an enemy? Dogs’ reaction to an unfamiliar person showing behavioural cues of threat and friendliness at different times. Applied Animal Behaviour Science, 94(1), 99- 115. doi: 10.1016/j.applanim.2005.02.001 Veterinary Poisons Information Service (2015). Veterinary Poisons Information Service Annual Report, 2015. Victoria State Government (2017). Information on 1080 and PAPP pest animal bait products. Victoria State Government: Economic Development, Jobs, Transport and Resources Retrieved 21/05, 2017 from

90 Robyn Taylor - December 2017 Temperament and breed influences on learned aversion in domestic dogs

http://agriculture.vic.gov.au/agriculture/farm-management/chemical- use/agricultural-chemical-use/bait-use-and-1080. Welzl, H., D'Adamo, P., & Lipp, H. (2001). Conditioned taste aversion as a learning and memory paradigm. Behavioural Brain Research, 125(1-2), 205-213. Young, J. K., Olson, K. A., Reading, R. P., Amgalanbaatar, S., & Berger, J. (2011). Is Wildlife Going to the Dogs? Impacts of Feral and Free-roaming Dogs on Wildlife Populations. BioScience, 61(2), 125-132. doi: 10.1525/bio.2011.61.2.7 Ziv, G. (2017). The effects of using aversive training methods in dogs—A review. Journal of Veterinary Behavior: Clinical Applications and Research, 19, 50-60. doi: https://doi.org/10.1016/j.jveb.2017.02.004

Robyn Taylor – October December 2017 91