**Dog Behaviour, Physiology and Welfare**

Editors

**Angelo Gazzano Chiara Mariti**

MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin

*Editors* Angelo Gazzano Universita di Pisa ` Italy

Chiara Mariti Universit`a di Pisa Italy

*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland

This is a reprint of articles from the Special Issue published online in the open access journal *Animals* (ISSN 2076-2615) (available at: https://www.mdpi.com/journal/animals/special issues/ Dog Behavior).

For citation purposes, cite each article independently as indicated on the article page online and as indicated below:

LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Article Number*, Page Range.

**ISBN 978-3-03943-044-4 (Hbk) ISBN 978-3-03943-045-1 (PDF)**

Cover image courtesy of Chiara Mariti.

c 2020 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.

The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.

### **Contents**



### **About the Editors**

**Angelo Gazzano** (Dr.) graduated in Veterinary Medicine, with a PhD in physiology. Angelo Gazzano is currently Associate Professor at the University of Pisa (Italy). His research is mainly focused on veterinary physiology and behavioural medicine. He is a diplomate of the European College of Animal Welfare and Behavioural Medicine.

**Chiara Mariti** (Dr.) graduated in Veterinary Medicine, with a PhD in equine physiology, Chiara Mariti is currently a senior researcher at the University of Pisa (Italy). Her research is mainly focused on pet behaviour and welfare and anthrozoology.

### **Preface to "Dog Behaviour, Physiology and Welfare"**

Dogs were the first animal species to be domesticated. Large amounts of evidence and science-based knowledge support the unique role that domestic dogs play in human life. Dogs can play many roles, ranging from companionship to very specialized roles. Dogs, though descended from the wolf, show noticeable differences when compared to their ancestor, leading to the need for studies focused on the species, as demonstrated by several comparative studies highlighting differences between wolves and domestic dogs. One main difference concerns inter-specific relationships. The kinds of relationship domestic dogs can establish with humans greatly vary and this is a relatively new field of research whose results can provide relevant inputs for both humans and dog welfare. In the last years, research on dog behaviour and welfare is also increasing, with studies ranging from those having an ethological approach to those more focused on applied ethology, physiology, endocrinology, anthrozoology and behavioural medicine. The objective of this Special Issue and book is to publish research papers dealing with dog behaviour, physiology and welfare and their interrelations with the dog–human relationship, to strengthen the knowledge of our "best friend". Thanks to the open access policy, we hope to ensure these data are available to all stakeholders, ranging from people working in the field of dog training, canine behavioural medicine, shelters, etc. to researchers and, possibly, even dog owners.

> **Angelo Gazzano, Chiara Mariti** *Editors*

### *Article* **A Retrospective Analysis of Complaints to RSPCA Queensland, Australia, about Dog Welfare**

**Hao Yu Shih 1,\*, Mandy B. A. Paterson <sup>2</sup> and Clive J. C. Phillips <sup>1</sup>**


Received: 29 April 2019; Accepted: 22 May 2019; Published: 27 May 2019

**Simple Summary:** Animal neglect and cruelty are important welfare and social issues, and dogs are one of the most commonly reported species to have experienced both. Most previous studies related to canine cruelty and welfare focused on animal abuse and dog fighting. However, literature dealing with the milder but more common forms of animal welfare concerns is limited. Therefore, this retrospective study aimed to understand the epidemiology of different types of canine welfare complaints in Queensland in the past decade and also to identify risk factors and their roles in different types of welfare complaints. The number of complaints received each year increased by 6.2% annually. The majority of complaints were neglect-related rather than related to deliberate cruelty, with the most common complaints being that dogs had poor body conformation, insufficient food and/or water, and receiving inadequate exercise. Poor living conditions and leaving dogs in a hot vehicle unattended were more commonly reported in recent years, potentially due to higher public awareness. Adult dogs that were reported were more likely to be alleged to have been poisoned, left unattended in a hot car, abandoned, and to have had inadequate exercise and shelter, compared with puppies. Puppies that were reported were more likely to be alleged to have experienced cruelty, lack of veterinary support, overcrowding, poor living and health conditions, and inappropriate surgery. Recognising which dogs are at most risk of cruelty will inform strategies to address this serious welfare problem.

**Abstract:** Animal neglect and cruelty are important welfare and social issues. We conducted an epidemiological study of dog welfare complaints and identified risk factors. The retrospective study included 107,597 dog welfare complaints received by RSPCA Queensland from July 2008 to June 2018. The risk factors considered were the age of dogs and the year of being reported. The number of complaints received each year increased by 6.2% per year. The most common complaints were poor dog body conformation, insufficient food and/or water, dogs receiving inadequate exercise, and dogs being confined or tethered. Increasing numbers were most evident for poor living conditions and leaving dogs in a hot vehicle unattended, both of which may have resulted from increasing public awareness. The majority of complaints were neglect-related rather than related to deliberate cruelty. Compared with puppies, adult dogs were more likely to be reported to have been poisoned, left unattended in a hot car or abandoned, as well as to have had inadequate exercise and shelter. Reported puppies were more likely to be alleged to have experienced cruelty, lack of veterinary support, overcrowding, poor living and health conditions, and inappropriate surgery. In conclusion, animal neglect was the most commonly reported welfare concern in dogs. Due to an assumed increasing public awareness of some types of cruelty, the trends of reported concerns differed. Adult dogs and puppies were reported to be involved in different types of welfare concerns. Strategies to address cruelty to dogs can be informed by an understanding of risk factors and trends in types of cruelty.

**Keywords:** dog; canine welfare; canine cruelty; neglect; RSPCA; age

#### **1. Introduction**

Animal cruelty involves all human behaviours towards animals that are morally and/or legally unacceptable, causing them to be inflicted with unnecessary and unjustifiable physiological, psychological, and behavioural discomfort or pain [1,2]. It is a complex issue implicating animal welfare, moral concerns, criminal activity, and violence [2–4]. It is regulated by state and territory law in Australia; for example, in Queensland by the Animal Care and Protection Act 2001 (ACPA [1]). This state-based legislation empowers the State to appoint inspectors, some of whom are employed by the Royal Society for the Prevention of Cruelty to Animals, Queensland (RSPCA Qld), to investigate potential breaches of the Act and enforce compliance with the Act [1]. There are two main offences under the ACPA, one is failure to fulfil duty of care responsibilities and the other is cruelty. There are a number of other specified offences. The Act recognises that a person who has charge of an animal owes that animal a duty of care. Failure to provide such care potentially constitutes a 'breach of duty of care' offence. This offence covers such actions as not providing sufficient food, water, exercise, veterinary care, and suitable living conditions. It is not only the owner that has a duty of care towards an animal. Anyone who even temporarily is in charge of an animal has a duty of care. The second major offence is 'animal cruelty' and the Act describes what it sees as cruelty in Section 18. A cruel act towards an animal can be committed by anyone, whether it is their own animal, another domestic animal, or even a wild animal [1]. It is important to note, that under the ACPA, the intention of a person to be cruel is not a prerequisite for committing the offence of cruelty. If an action carried out by a person causes pain and suffering and the action was intentional (that is not accidental), the person may be charged with cruelty. The intention to carry out the action must be proved but not the intention to be cruel. If a lack of action deprives an animal of its fundamental needs then they may be charged with a breach of their duty of care or even cruelty, depending on the circumstances. Motivation may be considered during sentencing [1]. Other offences under the Act include unreasonable abandonment or release, the carrying out of prohibited surgical procedures (e.g., tail docking, ear cropping, debarking, etc.), being involved in, or having items used for, a prohibited event such as dog or cock fighting, and allowing an animal to injure or kill another animal [1].

Potential cases are reported to RSPCA through various means. RSPCA Qld has a 'Cruelty Complaints' telephone number manned 24 h a day, seven days a week; complaints also come in through emails. Complaints can be made by members of the public but also by veterinarians and veterinary nurses, council officers, and other government and non-government employees visiting a location as part of their duties. Animals surrendered to the RSPCA or that come in as strays may be investigated if cruelty is suspected. They are considered by RSPCA Qld inspectors and further investigated if necessary.

According to the annual statistics of RSPCA Qld, there were 15,102 animal welfare complaints reported by the general public in 2011 [5], which had increased to 17,929 by 2017 [6]. Of all species falling victim to animal welfare concerns, dogs (*Canis familiaris*) are one of the most commonly reported species [7].

Various risk factors have been identified as contributing to an unsuccessful dog–owner relationship, which potentially resultsin neglect or abuse. These include the age of the dog [8,9], dog behaviour [8,10–12], physical attributes of the dog [9,13], the owner's motivation to care for the dog [14,15], the owner's attachment to the dog [12,16], costs of keeping the dog [16,17], and the owner's socioeconomic status [18,19]. In relation to actions carried out by third parties, most studies have focused on organised industries such as dog coursing [20] and fighting [21]. There has also been research into the origin of 'noxious abuses', e.g., cruelty involving intentional abuse, such as beating, shooting, and burning, that lead to severe physical injuries to the animals [7,15]. Literature dealing with the milder but more

common forms of animal welfare concerns is limited. One report considers neglect, such as exposing dogs to poor nutrition, keeping dogs in a backyard for hours without a shelter, and failing to meet exercise needs [2]. Most studies [20,22,23] stress the moral, legal, and social aspects of animal cruelty, and few explore the epidemiological dimension of this topic. This study addresses the epidemiology of diverse animal welfare concerns reported by the general public, instead of actual neglect or cruelty cases in a typical Western society. It also aims to identify the age of dogs as a risk factor for different forms of canine welfare complaints. Other risk factors, breed and socioeconomic status of the complainant, will be the subject of future papers.

#### **2. Materials and Methods**

#### *2.1. Materials*

From July 2008 to June 2018, RSPCA Qld received 129,036 canine welfare complaints. Some involving more than one dog were recorded as multiple complaints sharing the same case number, while others were recorded as one complaint with multiple animals. To avoid sample bias due to multiple entries, we only retained the first complaint of case numbers with multiple entries, discarding 21,439 entries as a result. There remained 107,597 canine welfare complaints for this retrospective study. Complaints that fell within the zone of responsibility of RSPCA Qld (determined by a Memorandum of Understanding between RSPCA Qld and Biosecurity Queensland, the Government Department tasked with the administration of ACPA) were investigated by RSPCA Qld inspectors. All other complaints were referred to Biosecurity Queensland to be investigated by their inspectors.

Complaints were recorded in Shelter Buddy®, the RSPCA Qld database. The following information was requested from the reporter of each incident by the inspector at the time of taking the complaint: the number of dogs involved (n = 106,104), their age (n = 107,597), their breed(s) (n = 92,021), the coded complaint type(s) (n = 106,983), the suburb (n = 107,413), and the postcode (n = 107,270); in addition, the date was recorded (n = 107,597). Dogs' ages were dichotomised into adult dog and puppy, based on reporters' interpretation. It is important to recognise that the information recorded from the complainant may have been inaccurate or inaccurately interpreted, e.g., a small dog is commonly referred to as a puppy. Records regarding breed and the number of dogs involved were based on either complainants' initial reports or comments from trained inspectors, again recognising inaccuracies with identification of the breed and the number of dogs involved. The 'complaint code' was selected by the staff member receiving the call or email from a drop-down menu of 18 possible complaints (Appendix A, Table A1). Multiple 'complaint codes' were able to be selected for each case, according to the description of what was alleged to have happened to the dog(s), and each was treated as a separate code for analysis.

#### *2.2. Statistical Analysis*

Data were analysed using the statistical package Minitab® 17.3.1. Descriptive analysis was used to investigate the distribution of complaint codes. Polynomial regression analysis and simple linear regression analyses were used to model the prevalence of different complaint codes from 2008 to 2018. The model chosen was that with the highest R-sq value, after ensuring that all components in the model were significant (*p* < 0.05). In 2008 and 2018, only data from July to December, and January to June were available, respectively. Therefore, data in 2009 and 2017 were used to test for within year variation in code citation rates for 2008 and 2018, respectively. Specifically, chi-squared analyses were conducted to compare whether the reported prevalence of each complaint code from January to June were different from those in July to December in 2009 and 2017. If there was no significant (*p* < 0.05) difference between the two six-month periods in that complaint code in 2009 and/or 2017, then the prevalence of the particular complaint code in the six-month period in 2008 and/or 2018 was/were assumed to be partially representative of the entire year(s). However, if there was a significant difference between the two six-month periods in that complaint code in 2009 and/or 2017, the data of the specific complaint

code in 2008 and/or 2018 were excluded from the polynomial regression analyses of year effects. After that, a Grubbs' test was used to identify outliers of each complaint code, which were excluded from polynomial and simple linear regression analyses. In polynomial regression analyses and simple linear regression analyses, years were entered as input variables and the prevalence of the complaint code as the output. The models were chosen on the basis of significant *p* values and the greatest R-sq values yielded. Three complaint codes, Causing captive animal to be injured/killed by a dog (N = 29), Keeping or using animal for blooding/coursing a dog (N = 18), and Emergency relief (N = 8) were not included in polynomial and simple linear regression analyses because the number of reported cases in the past decade was too few. Eighteen stepwise forward binary logistic regression models were constructed to understand how dogs' ages correlated with each complaint code. To determine the effect of age on complaint codes, age was entered (in dichotomous data form) into a binary logistic regression model as a fixed factor, using a logit function, with an alpha value to enter of 0.15. Complaint codes were entered into the model as outcomes. Separate models were constructed for each complaint code with the same input variable.

#### **3. Results**

#### *3.1. Complaint Codes and Dogs' Ages*

There were 18 complaint codes in total (Appendix A Table A1). On average, each case involved 1.76 (SEM = 0.003) codes. The distribution of complaint codes is presented in Figure 1. The most common codes, listed in declining order, were Poor dog condition (n = 29,982, 27.9%), Insufficient food and/or water (n = 28,265, 26.3%), No exercise/confined/tethered (n = 27,913, 25.9%), and Abandonment (n = 21,626, 20.1%). Overall, 93.67% (N = 100,791) of reported cases involved reported adult dogs and 6.33% (N = 6806) of reported cases involved reported puppies.

**Figure 1.** Distribution of dogs by complaint code. PD-Poor dog condition (27.9%, N = 29,982); IF-Insufficient food and/or water (26.3%, N = 28,265); NE-No exercise/confined/tethered (25.9%, N = 27,913); AB-Abandonment (20.1%, N = 21,626); PL-Poor living condition (18.7%, N = 20,162); NT-No treatment (17.6%, N = 18,963); CR-Cruelty (15.5%, N = 16,661); NS-No shelter (12.7%, N = 13,682);

HA-Hot animal in car (7.8%, N = 8384); OV-Overcrowding (0.9%, N = 978); BP-Baiting/poisoning (0.9%, N = 974); KK-Knowingly allowing an animal to kill/injure another (0.6%, N = 600); DF-Dog fighting or other prohibited offence (0.3%, N = 277); TD-Tail docking or other surgical procedure (0.2%, N = 214); PO-Prohibition order breached (0.1%, N = 133); CC-Causing captive animal to be injured/killed by a dog (0.03%, N = 29); BC-Keeping or using animal for blooding/coursing a dog (0.02%, N = 18); ER [a]-Emergency relief (0.01%, N = 8); UN-Unknown (0.6%, N = 614). [a] Emergency relief, as opposed to emergency rescuing which occurred when an animal encountered an urgent situation not related to domestic violence, was provided based on the ACPA, Section 123 [1].

#### *3.2. Trends of Complaint Types*

The number of complaints received annually increased by 6.2% per year, and the incidence of most complaint codes changed over the ten years. Results of the Chi-squared analyses showed that the prevalence from January to June and from July to December was significantly (*p* < 0.05) different for Poor living condition and Baiting/poisoning in 2009, No treatment and Poor dog condition in 2017, and No exercise/confined/tethered and Hot animal in car in both 2009 and 2017 (Appendix A, Table A2). Therefore, the data for Poor living condition and Baiting/poisoning in 2008, No treatment and Poor dog condition in 2018, and No exercise/confined/tethered and Hot animal in car in both 2008 and 2018 were excluded from the analyses of year effects. The prevalence of Poor dog condition in 2008 was an outlier (*p* = 0.029), and therefore was excluded as well. Figure 2 demonstrates the trends and the equations used for polynomial regression or simple linear analysis of each complaint code. These trends can be classified into five patterns: negative linear, positive linear, concave, monotonic, and irregular. Negative linear models included No exercise/confined/tethered, overcrowding, and Tail docking or other surgical procedure. Positive linear models included Poor living conditions, Hot animal in car, and Prohibition order breached. A concave pattern, indicating that the prevalence increased to a peak and then slowly decreased, was observed for No treatment, Abandonment, No shelter, and Knowingly allowing an animal to kill/injure another, for which codes the prevalence reached a peak in 2015, 2014, 2015, and 2011, respectively. In monotonic patterns, the trend was to generally increase, but not consistently, e.g., the prevalence of Poor dog condition generally increased, except for 2011–2016. Finally, some complaint codes had irregular changes over time. Cruelty, Insufficient food and/or water, Baiting/poisoning, and Dog fighting or other prohibited offence could not be modelled as they were reported sporadically over the ten years.

**Figure 2.** *Cont*.

**Figure 2.** *Cont*.

**Figure 2.** *Cont*.

**Figure 2.** Polynomial regression of each complaint code. The X axis represents the year, and the Y axis represents the prevalence of the complaint code.

#### *3.3. Risk Factors for Di*ff*erent Complaint Codes*

We considered age as a risk factor. The relationships between dogs' age and complaint codes are displayed in Table 1 and Figure 3. Compared to adult dogs, puppies were more likely to be reported for alleged Tail docking or other surgical procedure (OR = 9.87, *p* < 0.001), Overcrowding (OR = 4.44, *p* < 0.001), Poor living condition (OR = 1.45, *p* < 0.001), No treatment (OR = 1.33, *p* < 0.001), Cruelty (OR = 1.27, *p* = 0.001), and Poor dog condition (OR = 1.23, *p* < 0.001). Adult dogs were significantly more likely to be reported as an alleged case of a Hot dog in car (OR = 0.41, *p* < 0.001), Baiting/poisoning (OR = 0.42, *p* < 0.001), Abandonment (OR = 0.53, *p* < 0.001), No exercise/confined/tethered (OR = 0.64, *p* < 0.001) and No shelter (OR = 0.91, *p* = 0.037).

**Table 1.** Odds ratio of each variable in the logistic regression model of complaint codes. The outputs of these models were different complaint codes. The input variable was dog age (puppy or dog).


(a) Dog age was only classified as dog or puppy. Odds ratio refers to puppy relative to dog. (b) Age factor was not selected in the logistic regression model.

**Figure 3.** Positive relationships between dog age (adult dog and puppy) and complaint codes. Complaint codes listed under 'Adult dog' and 'Puppy' are complaints commonly involving adult dogs and puppies, respectively.

#### **4. Discussion**

#### *4.1. Complaint Codes*

This study reports the prevalence and progression of canine welfare complaints over the past decade. The complaints came from members of the public and may not represent all animal welfare issues or breaches of the ACPA. It must also be recognized that not all calls were found, on investigation, to represent a breach of the ACPA or even a dog suffering poor welfare. The number of complaints may be a representation of the degree of awareness in the community of animal welfare and the vigilance of many. A small number may be vexatious.

The descriptive analyses of different complaints show that the majority of alleged complaints related to neglect rather than deliberate maleficence. In this study, poor body and living conditions, insufficient food and water, and lack of the provision of exercise were the most commonly reported. This is in line with previous research that animal neglect or cruelty was most likely a result of ignorance due to lack of knowledge or forgetfulness [2,4].

In most cases, but particularly where it was decided not to continue to a prosecution, education of the dog owner was undertaken. To address neglect-related issues, people in charge of the animal were informed about food and water requirements, as well as the need for exercise, human companionship, and what represents good living conditions for a dog. For example, diets were recommended that are complete and balanced to replace homemade and all raw meat diets [24], and dog owners were likely to be informed that there were specialized products designed for specific ages of dogs [25] or dogs with specific health concerns [26]. Regular exercise is essential to promote health and quality of life [13,27,28]. Unfortunately, there are sometimes mismatches between an owner's exercise capability and the needs of their dog which inspectors recognise and advise accordingly [13,28]. The amount and frequency of exercise recommended varies according to the age, breed, size, fitness, and health of the dog. According to the ACPA, owners should ensure their dogs exercise for two hours after being continuously confined (e.g., caged or tethered) for 24 h, or for 1 h after 24 h confinement and another 1 h in the next 24 h [1]. This is the minimal time for most dogs to exercise and the majority benefit from receiving more [29]. The United Kingdom kennel club has published guidelines detailing suitable exercise amounts for each breed [29]. These are general in nature and not prescriptive.

Poor living conditions are negatively associated with a dog's quality of life, and increase the risk of diseases caused by ringworm, Giardia, Cryptosporidium, Toxocara, and Ancylostoma, as well as infestation with ectoparasites such as fleas [30–33]. Some of these agents are zoonotic and can cause health problems in humans.

#### *4.2. Trends Over Time*

Overall, the number of complaints received each year increased by 6.2% per year in the past decade, which may be contributed by the increasing population in Queensland in parallel with growing dog ownership in Queensland, and people's rising awareness of animal welfare. The population growth rate in Queensland was around 2% per year from 2008 to 2017 [34,35], and the pet population breakdown across the states and territories mirrored the country's population distribution [36]; it is reasonable that dog ownership in Queensland has also been increasing at a similar rate. Given the mismatch between the growth rate of canine welfare complaints (around 6% annually) and Queensland population (around 2% annually) for the same period of time, we believe that increasing public awareness and propensity to report animal cruelty may be another contributing factor [37]. The issue of animal welfare is becoming popular with the Australian public, in terms of their knowledge, concerns, willingness to participate, and legislation [38,39]. This is reflected by the reported increased concern for animal welfare among the general public [40–42], animal protection movements from advocacy groups [43], and public media [44], as well as the development of a closer relationship between humans and animals [45] and people's inclination to report cases involving animal harm to the police and RSPCA [37]. Such an increase in public awareness may be associated with the connections between animal abuse and human violence [23,46], and animal and human health [30,47], as well as more general dog ethical issues, such as dog consumption in Asia [48]. Moreover, people increasingly acknowledge the importance of good nutrition [49], canine behaviour [50,51] and emotions [52], and the goal of achieving a no-kill policy and improving the live release rate for roaming or sheltering dogs [39,53]. At RSPCA Qld, the number of volunteers has also grown from 2000 in 2011 to over 5000 in 2018 [6], suggesting that more people are actively concerned about the welfare of dogs.

The trends seen with the different complaints are also important to consider. Dogs were commonly reported as being in poor condition, living in poor environments, or being left in a heated vehicle. The increasing frequency of these three complaints may also be the result of a growing awareness of animal welfare among the public. However, the importance of animal welfare law enforcement and public education should not be underestimated. Humans' changing lifestyle may also influence the prevalence of certain kinds of animal welfare concerns. For instance, more people nowadays own a car and travel with their dogs frequently [54]. Consequently, dogs are at greater risk of being left unattended in a vehicle. This is particularly hazardous in Queensland, where median summer maximum temperatures of around 30 ◦C can lead to dangerously high temperatures inside cars [55]. Apart from the increasing trends for the types of complaints mentioned above, some complaint categories, particularly those involving food or water insufficiency, cruelty, dog poisoning, and dog fighting demonstrated irregular patterns, suggesting that they have occurred, or have been reported, inconsistently over the ten years [56]. These complaints also have the potential to jeopardize the dogs' life, and may also be related to violence and crimes [4,21,57]; therefore, these should be closely monitored. Finally, in Queensland, tail docking and other inappropriate surgical procedures (e.g., ear cropping and declawing) have been banned under the ACPA since 2001, unless they are undertaken by and under the recommendation of a veterinary surgeon [1]. Our results demonstrate a steady decline in the number of such cases, which indicates that the law is being adhered to.

#### *4.3. Adult Dogs and Puppies*

Apart from gaining an overall understanding of the prevalence and trends of different complaints, we also identified dog age as a risk factor in our dataset. Figure 3 summarizes the positively significant correlations between dog age and complaint codes. Adult dogs were more likely to be alleged to be subjected to activity-related welfare issues—for example being confined. They were also more likely to be reported as abandoned or left in a heated vehicle. The higher abandonment potential is

supported by an anthropomorphic theory that states that as adult dogs possess fewer infant-like traits, and thus are regarded as less attractive, they are less likely to evoke our nurturing instinct, creating a weaker bond with humans [12,58,59]. Adult dogs may have a greater chance of being left in a hot car because they are more difficult to carry and are usually not allowed to enter public places. On the other hand, puppies need more care and are easier to carry with a portable confinement box, therefore, it is less likely that they will be transported in a car and left when the driver is absent. Drivers may also imagine adult dogs to be more robust than puppies and more able to cope with high temperatures. This problem may be solved by promoting more pet-friendly environments (e.g., shops and restaurants) where dogs can stay with their owners and not be left in a vehicle alone. Increased information and signage in car parks reminding drivers of the danger to their dog of hot cars should also be encouraged. As for puppies, tail docking is generally conducted on three to five-day old puppies [60,61]; therefore, it is not surprising that this complaint category was more commonly reported with puppies than with adults. As the ban on tail docking was introduced in Queensland in 2001 [1], it is likely that some adult dogs were not reported because they were docked before the ban was introduced.

Additionally, in this study, puppies were alleged to be more likely to suffer from cruelty, which may result from the fact that puppies are submissive, and therefore tend to satisfy the controlling motivation for animal cruelty [15]. This finding is partially supported by a previous study that focused on non-accidental injuries of dogs and cats, and which revealed that dogs less than two years of age were more susceptible to intentional abuse [62]. In addition to the submissive nature of puppies, the authors suggest that young dogs are less manageable and thus may provoke owners with aggressive potential [62]. Another inconclusive result was found in a research trial investigating animal cruelty and domestic violence; in that study, authors did not find dogs' age to be a predisposing factor of being targeted for abuse [46]. Finally, access to appropriate medical support and suitable living conditions such as a good environment and enough space is another concern for puppies. These welfare concerns are probably indicators for animal hoarding or puppy farms [63,64].

#### *4.4. Strengths and Limitations*

One strength of this study is that it describes the trends in dog welfare complaints in Queensland over the last ten years and represents a large database, which allowed trends to be determined. It also correlates dog age with specific complaints. However, the current study also has its limitations. First, dogs were only classified as puppy or dog by complaints, which may hide important details related to age [9]. Second, this dataset covers only coastal, highly populated parts of Queensland, and thus generalization should be made with caution. Finally, the complaint code choices made by Call Centre staff were made on the basis of information obtained from the public, which indicates the potential of inaccurate reports. However, the study presents an analysis of what was reported and reflects changes in public awareness and motivation to act. Future studies could assess the accuracy of what people report and include more risk factors; for instance, breeds, behavioural issues of dogs, socioeconomic levels, and the history of an unsuccessful ownership have been reported to negatively influence dog ownership [8]. Dealing with behavioural problems and preventing people with a history of poor dog ownership from acquiring a new dog could reduce the risk of similar incidents being repeated. These factors are useful for addressing canine welfare issues and associated crimes and thus should be considered in future studies.

#### **5. Conclusions**

This study identified prevalence, trends, and the age of dogs as a risks factor for different types of complaints. Breed of the dog and socioeconomic status of the complainant will be the subject of future papers. Some neglect-related complaints, such as offering insufficient food and water, providing poor living conditions, and leaving a dog unattended in a heated vehicle apparently became more prevalent in recent years, probably indicating greater public awareness rather than an increase in neglectful behaviour. However, some serious complaints have been consistently reported over the past decade, including those involving animal abuse or severe injuries, and consequently should be closely monitored. The age of dogs was correlated with complaints about abandonment, neglect-related mistreatment, cruelty, and inappropriate surgery. Adult dogs were more likely to be reported as receiving inadequate exercising and shelter, having been abandoned, and having been left unattended in a hot vehicle; puppies were more likely to be reported as having poor living and health conditions, having undergone inappropriate surgery, and having suffered abuse. Recognising which dogs are at most risk of cruelty will inform strategies to address this serious welfare problem. Furthermore, the local or state government can direct specific attention to the most common and growing types of neglect and cruelty.

**Author Contributions:** Conceptualization, C.J.C.P. and M.B.A.P.; methodology, H.Y.S., C.J.C.P. and M.B.A.P.; software, H.Y.S. and C.J.C.P.; validation, H.Y.S., M.B.A.P. and C.J.C.P.; formal analysis, H.Y.S.; investigation, H.Y.S.; resources, M.B.A.P. and C.J.C.P.; data curation, H.Y.S., M.B.A.P. and C.J.C.P.; writing—Original draft preparation, H.Y.S.; writing—Review and editing, C.J.C.P. and M.B.A.P.; visualization, H.Y.S.; supervision, C.J.C.P. and M.B.A.P.; project administration, C.J.C.P.

**Funding:** This research received no external funding.

**Acknowledgments:** We thank RSPCA, QLD for providing the database, and RSPCA inspectors for consultation.

**Conflicts of Interest:** Mandy B.A. Paterson is employed as the principal scientist by RSPCA, Qld. None of the authors receive any interest or financial support from people or organizations who can bias the research.

#### **Appendix A**

**Table A1.** Description of each complaint code alleging a welfare issue.


(a) Prohibition order—A prohibition order is given by the court when a person convicted of an animal welfare offense must not possess any or a specific animal for a prescribed period of time [1].



*Animals* **2019**, *9*, 282

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*

### **Feeding Enrichment in a Captive Pack of European Wolves (***Canis Lupus Lupus***): Assessing the E**ff**ects on Welfare and on a Zoo's Recreational, Educational and Conservational Role**

### **Giacomo Riggio 1, Chiara Mariti 2, Chiara Boncompagni 1, Simone Corosaniti 1, Massimiliano Di Giovanni 3, Asahi Ogi 2, Angelo Gazzano 2,\* and Robert Thomas <sup>4</sup>**


Received: 11 April 2019; Accepted: 4 June 2019; Published: 8 June 2019

**Simple Summary:** Feeding enrichment is widely used to improve the welfare of zoo animals, but it may also affect zoo visitors' experience and perception of the animals. The objective of this study was to assess the effects of a naturalistic and a non-naturalistic feeding enrichment program, on both wolf behaviour and visitors' interest in the exhibit. A questionnaire was administered to visitors with the aim of assessing whether our feeding enrichment programs might affect their perception of captive wolf welfare as well as their attitude towards wolf conservation issues. Our findings suggest that, although wolves seemed to benefit from enrichment, their behavioural responses were highly variable among individuals. Visitors' interest in the exhibit and perception of captive wolf welfare improved by observing the wolves interacting with food, especially when novel feeding objects were provided. Finally, their attitude towards wolf conservation issues did not change in relation to enrichment, but improved when they observed the wolves performing feeding-related behaviours. These findings may help zoos implement enrichment programs that are effective for enhancing their wolves' welfare as well as their recreational and educational role.

**Abstract:** This study investigated the effects of two feeding enrichment programs on the behaviour of a captive pack of European wolves (*Canis lupus lupus*) and their correlation with both zoo visitors' interest towards the exhibit and their overall perception of the species. Behavioural data (exploration, stereotypies, social interactions, activity/inactivity rates) were collected on four male wolves during four two-week long phases: initial control, hidden food, novel object, final control. Three observation sessions were performed daily: before, during and after feeding. Number of visitors and their permanence in front of the exhibit were recorded. After watching the wolves, visitors were asked to fill out a brief questionnaire in order to investigate their perception of captive wolf welfare, as well as their attitude towards wolf conservation issues. Despite the high inter-individual variability in their behavioural response, all wolves seemed to benefit from feeding enrichment. With regard to visitors, interest in the exhibit increased when enrichment was provided. Visitors' perception of the level of welfare of wolves improved if they attended a feeding session, especially during the novel object phase. Visitors' attitude towards wolf conservation issues also improved during feeding sessions, regardless of enrichment provision.

**Keywords:** zoo; wolf behaviour; animal welfare; visitor; conservation; education

#### **1. Introduction**

Wolf conservation has always been a very controversial topic because of the impact of wolf predatory activity on livestock [1]. If on one hand, wolves held in zoos may help raise public awareness on the need to preserve the species, on the other they may have their welfare compromised by the impossibility to perform natural behaviours, including predation itself [2,3].

In the Carnivora, predatory motivation influences the individual's behaviour from the early stages of its psycho-physical development, even in apparently extraneous contexts than those related to food consumption (i.e., play) [4]. Furthermore, in social species like the wolf, group hunting may also have a cohesive function among members of the pack [5,6].

For management and ethical reasons, captive carnivores are rarely given the opportunity to perform predatory behaviour, as food is often served ready to be consumed [7–9]. Constantly preventing animals from appeasing their predatory motivation may lead to chronic frustration and stress [10]. Abnormal behaviours may develop as coping strategies in response to the stressful environment [11].

Feeding enrichment programs aim to promote the expression of natural behaviours and the interactions with a more adequately stimulating environment [12,13]. In captive carnivores, such programs are often aimed to stimulate at least some phases of predatory behaviour [14]. The ultimate goal is, of course, that of enhancing the animals' level of welfare [15,16]. Several behavioural indicators of welfare have been used to assess the effects of feeding enrichment programs. In carnivores, feeding enrichment has been shown to decrease stereotypies [14,17], improve behavioural diversity [18,19], increase exploration [20] and activity levels [21].

Improving animal welfare may be beneficial not just for the animals, but for zoo visitors [22,23]. For instance, animals that interact with their environment have been shown to increase visitor interest in the exhibit [22,24–26]. Furthermore, sleeping or pacing animals objectively provide less information about the behavioural repertoire of the species and may fail to meet visitors' expectations of the animals' behaviours [27]. Visitors who have a negative experience may leave the zoo with wrong or little, if at all, new knowledge on the species and a sense of frustration for not having their expectations met [8,28]. On the opposite, more active and naturally behaving animals have been shown to improve visitors' perception of animal welfare and, as a consequence, their perception of the educational importance of zoos [8,29,30].

This study aimed to implement two distinct feeding enrichment procedures in a captive pack of European wolves (*Canis lupus lupus*) in order to (1) simultaneously assess their effects on both captive wolves' behaviour and visitors' interest towards the exhibit, (2) assess the possible correlation between enrichment and visitors' perception of captive wolf welfare, as well as their attitude towards wolf conservation.

#### **2. Materials and Methods**

#### *2.1. Time and Setting*

The study was carried out at the Bioparco of Rome (Italy) for two consecutive months, between March and May 2017. The wolf enclosure is one of the largest exhibits of the park. It is a 200 m2 area, surrounded bya3m high wooden palisade on one long side (the visitor side). A central 10 m<sup>2</sup> trapezoidal recess virtually divides the enclosure in a right and a left area. Windows in the two diagonal sides of the recess allow visitors to observe the animals in both areas. When necessary, metal sliding doors are used to physically divide the enclosure in three smaller areas (Figure 1).

**Figure 1.** Simplified map of the wolf exhibit. LW = left wing, RW = right wing, IA = isolation area, TR = trapezoidal recess. Red lines = sliding doors, blue lines = windows, yellow lines = wooden palisade.

#### *2.2. Animals*

Experimental subjects were four European wolves (*Canis lupus lupus*). All of them were un-neutered male siblings: wolf 1, 2 and 3 were 8 years old at the time of the study and belonged to the same litter, while wolf 4 was 1 year younger. All of them had always lived together and they were moved together from another zoo where they were born and raised. They were fed once a day for 6 days a week and their diet comprised either chunks of buffalo or cow meat, or entire rabbits or quails.

#### *2.3. Experimental Design*

The experimental protocol was scheduled into four different phases: initial control (ICp), hidden food (HFp), novel object (NOp), final control (FCp). Each phase lasted 2 weeks. ICp, in which no enrichment was administered, served as a baseline control. Food was administered as usual, in the small area at the edge of the right wing of the enclosure. During HFp, food was divided into 12 pieces, regardless of its nature. Four of them were singly buried 10 to 15 cm deep in holes purposely dug by the keepers. Four were suspended 2–3 m-high on tree branches. The remaining pieces were singly inserted deep in the middle of four woodpiles previously built in each wing of the enclosure. These piles had been introduced into the enclosure 2 months before the beginning of the study in order to allow the wolves to habituate to the new elements. During NOp food had to be wrapped in eight canvas bags in order to make it more difficult for the wolves to reach for it. In addition, a hard-plastic feeding ball (Aussiedog®) filled with smaller chunks of meat had to be randomly left around the area. Again, during FCp, no enrichment was provided. During this phase, which served as final double control, food was administered using the same procedure as in ICp. Feeding procedures were standardized across the four phases. Keepers would alternately lock the wolves in the left or right area, by using the central metal sliding door. Hence, they would enter the other area in order to clean it and place food. At the end of these procedures they would step out, lock the cage, and free the wolves. Once the sliding door was opened, keepers had 2 min to walk out of sight to the wolves. The study was observational in nature. The animals involved in this study were housed at the zoo and were part of a program of enrichment promoted by the zoo itself, i.e., the procedures were thus not carried out for research purposes. No animal care license nor approval of ethical committees were therefore needed.

#### *2.4. Wolf Data Collection*

Behavioural observations were carried out daily, 5 days a week, from Tuesday to Saturday. Every day, three observation sessions were performed before, during and after feeding. Each session lasted 1 h, plus a 1-h interval between consecutive sessions, for a total of 3 h per day. In order to meet keepers' working and break hours and zoo opening hours, but still avoid predictability by wolves on feeding times, observations could start at three different times of the day: 9.30, 11.00, or 12.15. Daily observation starting times were randomly set. Behavioural data were collected through instantaneous scan sampling technique [31]. During each hour of observation, the behaviour of each wolf was recorded every 1 min. The ethogram comprised 16 behavioural categories (Table 1) and was obtained by integrating behaviours described in scientific literature [5,18,32–34] with direct behavioural

observations over a 7-day period, prior to the study. Since behavioural data from the first observation session on 29 March 2017 had not been recorded, data from the entire day were eliminated.

**Table 1.** Description of wolf behavioural categories grouped in macrocategories used for statistical purposes. Social interactions were categorized as either positive or negative. All behaviours, except stereotypies, were categorized as either active or inactive.


#### *2.5. Visitor Data Collection*

Data from visitors were collected by an additional observer using the same observation schedule used for the wolves. Both the number of visitors and the duration of their permanence in front of the exhibit were recorded [35]. The number of visitors in front of the enclosure was recorded by instantaneous scan sampling with 1-min intervals for the entire hour of observation for each session. Children up to 16 years were excluded from the count. The amount of time (seconds) spent in front of the exhibit by every third visitor (excluding children) was also recorded [28,29,36,37]. Visitor permanence recording began when the selected subject walked across the virtual line that delimited the wolf exhibit observation zone (trapezoidal recess) and ended when they crossed it the opposite way.

#### *2.6. Questionnaire*

When visitors exited the observation zone, a researcher asked them (not children) to fill in a brief questionnaire on wolves. Visitors selected for filling the questionnaire and those selected for recording their permanence in front of the exhibit were not necessarily the same. All visitors were asked to fill the questionnaire although not all of them would accept (respondents *n* = 630). The questionnaire consisted of two sections. The first section investigated on demographic information, which are summarized in Table 2. The second section consisted of 11, 1–5 Likert-scale items partially based on surveys used in previous studies [28,38] (Table 3). Visitors could respond by marking with an X a number from 1 to 5 depending respectively on the degree of disagreement/agreement with the corresponding statement, with 1 meaning "strongly disagree" and 5 meaning "strongly agree." Items' positivity and negativity was intended in relation to the quality of visitors' perception towards wolves.


**Table 2.** Demographic information for questionnaire respondents.

Nationalities below 1% were not reported.

**Table 3.** Likert-scale items and scoring system.


S.D. = strongly disagree, S.A. = strongly agree. Negative items are marked with an \*.

#### *2.7. Statistical Analysis*

All the statistical analyses were run with the software SPSS Statistic 17.0 (Chicago, IL, USA).

Due to the small sample size and the possible individual differences in behavioural responses to enrichment, data were analysed separately for each wolf, which acted as its own control. For analysis purposes, all behaviours except stereotypies were grouped into broader macro-categories (Table 1). Exploratory behaviour was also analysed singularly. In order to assess potential differences in each wolf behaviour among phases and among observation sessions, the Kruskal Wallis test and then Wilcoxon signed rank test were applied (*p* < 0.05 multiple comparison corrections were performed using the Benjamini–Hochberg procedure). The Wilcoxon signed rank test was also used to compare data obtained from feeding sessions in the first and second week for both hidden food and novel objects phases, with the aim of assessing a possible effect of the decreased novelty of the enrichment. Potential correlations between number of visitors in front of the exhibit and respectively phase, session and number of active wolves, were analysed using the Spearman test (*p* < 0.05). The Kruskal Wallis test was used to investigate potential differences in visitor stay time in front of the exhibit across observation sessions and experimental phases. When appropriate, pair-wise comparisons were carried out using the Mann–Whitney U test (*p* < 0.05) for independent variables.

As for the questionnaire, descriptive statistics was used for demographic information. Furthermore, all the negative items on the Likert-scale questionnaire were reverse-scaled to match with the score of positive items. Principal Component Analysis (PCA) was used to identify a possible underlying structure across the items; promax rotation with a correlation matrix was used.

Items were included in a component if their loading in that component was >0.50 and their loading in the other components was <0.25. The three principal components found were further analysed by using a Kruskal Wallis Test and then a Mann–Whitney U test (*p* < 0.05) in order to assess differences among phases and sessions.

#### **3. Results**

#### *3.1. Wolf Behavioural Results*

Within each enrichment phase, none of the comparisons between the first and second week of observation during feeding session revealed a statistically significant difference. Activity and inactivity rates did not significantly differ across phases, for any of the wolves. Significant results were obtained for stereotypies, social behaviour and exploration, as reported below.

A significant difference was found in the rate of stereotypic behaviours for wolf 2 and wolf 4 across phases. For wolf 2, stereotypy rates were significantly lower in the NOp when compared with the HFp (*p* = 0.012) and the FCp (*p* = 0.001). Similarly, wolf 4 stereotypy rates were significantly lower during the NOp when compared with the HFp (*p* = 0.003) and the FCp (*p* = 0.001). In addition, wolf 4 stereotypy rates tended to be higher in FCp compared to ICp (*p* = 0.054) (Figure 2).

**Figure 2.** Number of times stereotypic behaviour was observed for each wolf across the four experimental phases.

Social behaviour rates significantly differed for wolf 1, wolf 2 and wolf 4 across phases, whilst no difference was observed for wolf 3. More specifically, for wolf 1, negative social behaviour rates were significantly lower during NOp compared to ICp (*p* = 0.013). However, they were also lower during FCp compared to ICp (*p* = 0.025). Similarly, for wolf 2, negative social behaviour rates were significantly lower during NOp, if compared with ICp (*p* = 0.012). However, they were also different between control phases, being significantly lower during the FCp (*p* = 0.042) (Figure 3). As for wolf 4, statistical differences were instead found for positive social behaviour rates, that resulted higher during HFp than ICp (*p* = 0.011) and NOp (*p* = 0.024) (Figure 4).

**Figure 3.** Number of times negative social interactions were observed for each wolf across the four experimental phases.

**Figure 4.** Number of times positive social interactions were observed for each wolf across the four experimental phases.

For all wolves there were no statistically significant differences in exploratory behaviour rates between initial and final control phases. Wolf 1 exploratory behaviour rates were significantly higher during the NOp when compared with both ICp (*p* = 0.008) and FCp (*p* = 0.029). For wolf 2, they were significantly higher in NOp when compared with any other phases (NOp versus ICp: *p* = 0.001; NOp versus HFp: *p* = 0.001; NOp versus FCp: *p* = 0.001), as well as in HFp when compared with ICp (*p* = 0.043). Additionally, for wolf 3 exploration was significantly higher in NOp when compared with any other phases (NOp versus ICp: *p* = 0.001; NOp versus HFp: *p* = 0.002; NOp versus FCp: *p* = 0.001). Finally, for wolf 4, exploratory behaviour rates were significantly higher in both the enrichment phases when compared with the control phases (HFp versus ICp: *p* = 0.001; HFp versus FCp: *p* = 0.001; NOp versus ICp: *p* = 0.005; NOp versus FCp: *p* = 0.003) (Figure 5).

**Figure 5.** Number of times exploratory behaviour was observed for each wolf across the four experimental phases.

#### *3.2. Visitors' Interest Results*

A positive correlation was found between mean number of visitors in front of the exhibit and experimental phase (Rho value > 0.001), as well as mean number of visitors and observation session (Rho value > 0.001). The mean number of visitors was the highest during NOp (*p* < 0.001). Unexpectedly, it was also higher during FCp phase than HFp (*p* < 0.001) (Figure 6). As for the sessions, the highest mean number of visitors was found during feeding sessions (*p* < 0.001), regardless of the use of enrichment (Figure 7). Moreover, a positive correlation between the number of active wolves and the number of visitors in front of the exhibit was found (*p* < 0.001).

**Figure 6.** Mean number of visitors in front of the exhibit across the four experimental phases.

**Figure 7.** Mean number of visitors in front of the exhibit across the three daily observation sessions.

Time spent by visitors in front of the exhibit differed significantly across phases. It was longer during NOp than during any other phase (*p* < 0.001). It was also longer during HFp when compared with both control phases (*p* < 0.001). No significant difference was found between control phases. Furthermore, visitors' permanence in front of the exhibit also differed across observation sessions, being the longest during feeding sessions (*p* < 0.001) and longer before feeding than after feeding (*p* = 0.016).

#### *3.3. Questionnaire Results*

The 11, 1–5 Likert-scale items from the questionnaire were analysed using a Principal Component Analysis (PCA) with promax rotation (*p* < 0.05). The component correlation matrix (Table 4) did not show a correlation between the three components. The PCA model was adequate as verified by the Keiser Meyer Olkin (KMO) test (KMO = 0.694; *p* < 0.001). Three principal components (PC1, PC2 and PC3) were identified that explained 49.895% of the variation (Table 5). PC1 was characterized by those items relating to the respondent's perception of wolf welfare in zoos. PC2 was characterized by those items relating to the respondent's perception of wolves as a threat to humans and human activities. Finally, PC3 comprised those items relating to respondent's attitude towards wild wolf population management measures. No significant differences in questionnaire responses were found when data from the three daily sessions were compared across phases. On the contrary, differences were found for PC1 and PC3 across sessions. In particular, respondents scored significantly higher for PC1 and PC3 when questionnaires were administered during feeding when compared with sessions before feeding (PC1: *p* = 0.016; PC3: *p* = 0.010), suggesting that their perception of captive wolves' welfare and their attitude towards wild population management measures improved by watching the animals interacting with food, regardless of the use of enrichment. However, by comparing questionnaire responses during the sole feeding sessions across the four experimental phases, we found that respondents scored higher for PC1 when they observed the wolves get fed during NOp and HFp rather than during ICp.


**Table 4.** Component correlation matrix obtained for principal component analysis with promax rotation on questionnaire items.

**Table 5.** Results of the principal component analysis and promax rotation carried out on items of the questionnaire. For each item, the load on each component has been reported.


Loadings with an absolute value greater than 0.5 (and lower than 0.25 for another component) are in bold.

#### **4. Discussion**

#### *4.1. Wolves*

In this study, none of our significant results regarding wolf behaviour were consistent across all four experimental subjects. This suggests individual factors played a major role in determining the wolves' responses to feeding enrichment. Three behavioural categories were used to assess wolf responses to enrichment: stereotypies, quality of social interactions and exploration.

Stereotypic behaviours are probably the most used behavioural indicator of welfare in captive animal studies [39,40]. They have been defined as "repetitive, invariant behavioural patterns with no apparent goal" [41]. In this study, two distinct behaviours fit such definition: pacing and jumping against the fence. The latter was performed by only one individual, namely wolf 2. Both of them fall in the category of "locomotor stereotypies," which are the most common type of stereotypy in captive carnivores [42,43]. In order to meet the criteria of "repetitive, invariant behavioural pattern" we reported such behaviours only when they were performed for at least three consecutive times [27,42]. In this study, two out of four wolves showed a significantly lower rate of stereotypic behaviours when novel artificial feeding objects were provided. Although the underlying processes of stereotypic behaviour in zoo animals are not yet clearly understood, one of the most widely accepted theories is that it represents a copying strategy performed by an individual in response to a chronic stress condition generated by a sub-optimal and restrictive environment [2,39,42,44,45]. Therefore, a reduction in the rate of stereotypic behaviours may be suggestive of increased welfare, especially when other behavioural indicators of welfare improve concurrently [39].

Quality of social interactions was assessed classifying social behaviour in either "positive" or "negative" [32]. Social play and affiliative interactions were considered to be positive, whereas agonistic and aggressive interactions were considered to be negative [32]. Again, in the context of our study, a quite high degree of individual variability in wolves' behavioural responses to enrichment was observed. While wolf 4 increased positive social interactions when food was hidden around the enclosure, wolf 1 and wolf 2 showed a decrease in negative interactions when artificial feeding objects were provided, whereas wolf 3 showed no significant change in any enrichment programs. As by nature, social interactions occur among more individuals, the interpretation of results needs to take into consideration the possible reciprocal influence of the behaviour of the various parties involved. In this case, the decrease of negative social interactions observed in wolf 1 and 2 are likely to be the result of the decrease of negative social behaviours in one wolf; the parallel trend in the second wolf probably mirrors the first decrease, as a consequence of receiving fewer negative interactions. However, looking at the whole unit, both the direct and indirect impact of enrichment can be regarded as beneficial.

An increase in exploration when novel feeding objects were provided was the only behavioural response common to the four wolves. Nevertheless, for wolf 2 and wolf 4 exploration also increased when food was hidden around the enclosure. In accordance with previous studies [18–20,46,47], this finding suggests that the relationship between feeding enrichment and exploratory behaviour is more linear and less affected by individual variables than the other behavioural indicators we assessed.

Overall, both feeding enrichment programs seemed to be effective at positively modifying our wolves' behaviour. However, inconsistency of behavioural responses among wolves indicates that individual variables might have qualitatively and quantitatively affected the beneficial effects of enrichment. Previous studies also found a high inter-individual variability in behavioural responses to feeding enrichment [18,48]. Since all of our experimental subjects were males with equal rearing conditions, this finding cannot be attributed to gender or rearing differences, as suggested by Cummings et al. [18] in their study on maned wolves. In our case, it is most likely explained by individual temperament differences and pack social dynamics [16,48–50].

A major limitation of this study is the small size of the sample, which prevents us from drawing any conclusion at the species level. Studies that investigate enrichment programs in zoo animals are often performed on small samples [18,34,51–53] and in specific environments that are likely not representative of the entire zoo population. Being opportunistic, our study presents these limitations. Our findings should be regarded as preliminary and further studies should be performed on larger samples and different captive environments.

In addition, each enrichment program was provided for 2 weeks only. The lack of differences between behaviours displayed in the first and second week of both enrichment phases suggest that, although the novelty of the enrichment decreased, it maintained beneficial effects in the analysed period. This is relevant for zoo animal caretakers, as their investment seems to have persistent benefits. However, 2 weeks may be a sufficient amount of time to determine short-term effects, but not enough to assess a possible long-term impact. This should be investigated in future studies by extending periods of enrichment.

Lastly, for practical reasons we could not assess any physiological indicator of acute or chronic stress, which could have helped us draw a clearer picture of the internal changes that each animal underwent during the study and their link with the behaviours observed [54–56].

#### *4.2. Visitors*

The number of visitors and the duration of permanence in front of the exhibit have often been used as indicators of their interest in zoo exhibits [22,23,35]. While the latter may be a more reliable parameter, the former may be more strongly affected by other variables such as season, day of the week, time of the day and weather [57]. This may explain why visitor number was higher in the final control phase than during the first enrichment phase. Nonetheless, our overall findings suggest that visitors were more interested in the exhibit during feeding sessions and during enrichment phases, especially when novel non-naturalistic objects were involved. In the latter case both number and time of permanence in front of the exhibit increased significantly. In previous studies, greater visitor interest in the exhibit has been linked to higher levels of animals activity [24,26,58]. However, although we found visitor number to be positively correlated with the number of active wolves at a given time, no significant difference in animal activity levels across experimental phases was detected. Other variables, such as the type of enrichment used [59] or the type of behaviour elicited [28,38,59] may also affect visitors' interest in the exhibit.

Interestingly, those experimental conditions that increased visitors' interest also enhanced their perception of captive wolf welfare. In fact, among daily observation sessions, the highest scores for animal welfare-related questions were obtained from those visitors who observed the wolves during feeding times, regardless of the use of enrichment. This result reflects those of previous studies in which visitors' perception of captive wolf welfare improved by observing the animals perform active natural behaviours, such as feeding [28,29]. Furthermore, when we compared responses obtained from visitors who observed the wolves during feeding times, across all four experimental conditions, we found that higher scores for welfare-related questions were obtained when enrichment was provided, especially the non-naturalistic type. Previous studies [30,60,61] did not find any differences in the effects of naturalistic versus non-naturalistic enrichment on visitors' perception of the animals' well-being. However, some methodological differences, such as type of enrichment used [30,60,61], species involved [30,60,61], partial or total lack of complementary data on the animals' behaviour [30,60] may explain the different results.

More importantly, visitors seemed to be aware of the implications that enrichment might have had on the wolves' welfare. Price et al. [29] suggest that visitors' perception of animal welfare may in turn be influenced by their own perception of zoos' commitment at caring for their animals. In our study, artificial enrichment objects may have rendered the zoo's commitment more evident in the eyes of the visitors.

Although this assumption should be further investigated with more specific questions in future studies, it may also explain why visitors who had higher perception of the wolves' welfare also had a more positive perception of the educational role of zoos [59]. On one hand, negative emotions elicited by perceiving animals suffering may generate a deep sense of distrust in the zoo as an animal preservation institution, thus diminishing the value of its educational role [8]. On the other hand, animals stimulated to engage in natural behaviours, such as foraging and feeding, may not only be perceived as happier by visitors [8], but they may objectively provide more information about the species' behavioural repertoire, if compared with inactive or stereotyping animals [29].

Findings from previous studies indicate that visitors who observe animals perform active natural behaviours show greater appreciation for the species' biological significance and greater conservation intent [8,38,62]. According to that, our visitors showed a more positive attitude towards wolf conservation issues when they observed the animals during feeding times, regardless of the use of enrichment. However, feeding enrichment, whether naturalistic or non-naturalistic, failed to further improve visitors' attitude towards wolf conservation. Due to the increasing wild wolf population, Italy is currently going through a period of intense political debate over wolf conservation measures. Actual or spurious episodes of attacks to livestock and even citizens reported by the media are likely to affect people's attitude on the matter. This may explain why, in our study, visitors did not seem to change their perception of wolves as a threat to humans and human activities. Finally, it should be taken into account that the wolf is a historically controversial species. The cross-generational "evil wolf" myth embedded in the occidental culture may represent a significant obstacle for a mind shift on attitude towards wolf conservation [63]. Similar studies on other species or conducted in other countries may lead to different results.

#### **5. Conclusions**

Overall, both hiding food within the enclosure and providing animals with novel artificial feeding objects appeared effective at modifying the behaviour of our wolves in a way that suggests an increased level of welfare. However, inconsistency in results across experimental subjects indicates that individual variables play an important role in determining the degree and the type of behavioural response to feeding enrichment. Among the behavioural indicators assessed, an increase in exploration activity when novel feeding objects where provided was the only change in behaviour common to all the wolves.

Feeding enrichment, especially when artificial objects were used seemed to be effective at increasing visitors' interest in the exhibit. Visitors' perception of captive wolf welfare and the educational role of zoos, as well as their attitude towards wolf conservation issues were more positive when they observed the wolves during feeding times, regardless of the use of enrichment. Feeding enrichment, especially the non-naturalistic type, further improved visitors' perception of captive wolf welfare. On the contrary, it failed to modify visitors' attitude towards wolf conservation issues. In order to confirm this study's findings future research on the effects feeding enrichment on wolves and visitors should be conducted on larger animal samples and in different countries.

**Author Contributions:** Conceptualization, G.R., C.B. and A.O.; methodology, G.R., C.M., A.O.; formal analysis, C.M.; investigation, G.R., C.B., S.C.; resources, M.D.G.; data curation, G.R., C.M.; writing—original draft preparation, G.R.; writing—review and editing, G.R., C.M., A.G. and R.T.; supervision, A.G., R.T.; project administration, M.D., R.T.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors would like to thank the Biorparco di Roma curator Yitzhak Yadid and the carnivore keepers Marco Subrizi, Mauro Bedin, Massimiliano Pellicciotta, Barbara Leoni, Francesco Pompei and Nino Pinsone for their active participation in the study; Aussiedog® for providing them with its enrichment products.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
