Assessment of Methods for On-Farm Euthanasia of Layer Chickens

Home , Cervical dislocation

Rathnayaka Mudiyanselage Amila Subhashinie Bandara

A Thesis Presented to The University of Guelph

In partial fulfilment of requirements for the degree of Doctor of Philosophy in Animal Biosciences

Guelph, Ontario, Canada © Rathnayaka Mudiyanselage Amila Subhashinie Bandara, June 2019

ABSTRACT

ASSESSMENT OF METHODS FOR ON-FARM EUTHANASIA OF LAYER CHICKENS

R. M. A. S. Bandara Advisor: University of Guelph, 2019 Professor T. M. Widowski

Animal care guidelines for poultry require that the methods used for routine killing result in rapid and irreversible loss of sensibility and cause minimal pain and distress. This thesis assessed the efficacy of three types of physical on-farm euthanasia methods in different age groups of layer chickens, and the degree of aversion and time to loss of sensibility for different CO2 concentrations (25%, 35%, 50%, and 70%) in laying hens. In Study 1, all three commercially available non-penetrating captive bolt devices tested caused sufficient brain trauma to result in rapid insensibility and brain death in four different age groups (10-11, 20-21, 30-35, 60-70 weeks) of layer chickens. This study also identified and corroborated practical behavioural indicators of death in layer chickens that can be used in field conditions to achieve the animal care guideline requirements of confirming the death before disposing of carcasses; onset of tonic convulsions, last movement, and cloacal relaxation were good indicators of clinical death. A second study assessed efficacy of a commercially available mechanical cervical dislocation device (MCD) in comparison to manual cervical dislocation (CD) in different age groups (12,

27-29, and 65-70 weeks) of layer chickens. Killing methods were assessed in anesthetized chickens to minimize welfare concerns. MCD resulted in a longer time to brain death than CD.

Radiographs revealed that the majority of birds killed by CD had ideal dislocation sites between the skull and atlas (C1) or between cervical vertebrae C1-C2. The MCD resulted in a majority of

dislocations at lower cervical vertebrae. There were few fractures in birds killed by either method. A final study demonstrated that concentrations of 50% and 70% CO2 were significantly more aversive to laying hens than 25% and 35%, based on an approach avoidance test. However, hens demonstrated headshaking and open mouth breathing at all tested CO2 concentrations, and some birds displayed conditioned place avoidance at the low concentrations. Loss of posture, indicating insensibility, occurred in less than 25s in all CO2 concentrations with shorter latencies at higher concentrations. The thesis provides important information for refinement of future euthanasia guidelines for the layer chicken industry.

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my advisor, Dr. Tina M Widowski for her tremendous support and guidance during my graduate career. Her understanding of me as an international student whose family lives thousands of miles away was unbelievable. I am grateful for the opportunities she has provided, and this would not have been possible without her encouragement and belief in my capabilities.

My special sincere thanks should go to Dr. Stephanie Torrey, a member of the advisory committee, whose continued support and dedication has been invaluable. Her frequent encouragement and guidance to develop my research and further my career were remarkable. A big thank you to Dr. Suzanne Millman for her valuable guidance as the co-advisor throughout the completion of my research work at Iowa State University, and for constructive feedback as co- author. I would also like to thank my advisory committee – Dr. Patricia Turner and Dr. Karen

Schwan-Lardner for their expertise, feedback and continuous involvement.

Thank you to Linda Caston for the technical support and wonderful companionship given throughout this project. Your understanding and advice were remarkable for me to develop my personality and to adjust to a different culture. Thank you to Dr. Anna Bolinder and Dr. Alex zur-Linden for your expertise and technical support. Thank you to Dr. Michelle Edwards for your vast knowledge of statistics, and the given support on advising on statistical analysis through this thesis.

Thank you to Kahlee Latreille, Rebbeca Parsons, Alex Hurtado-Terminal, Dilan C. Waters, and

Daniel Rothschild for the given technical support, and wonderful companionship during the research project. Thank you to everyone in the Widowski Lab for helping with data collection.

Thank you to the farm staff at the Arkell Poultry Research Station and ISU LAR technicians for animal care and technical help with this project.

Thank you to Dr. M.A.J.P. Munasinghe, former Head of the Department of Livestock

Production, Sabaragamuwa University of Sri Lanka for assisting to obtain study leave to pursue this PhD degree programme, and arranging staff to cover my duties in the department.

I have no words to thank my loving husband Dhanushka. Your willingness to share all our responsibilities and make sacrifices were unbelievable during this difficult period. You are the one who gave the biggest support in this journey, physically being across the oceans but keeping me in your heart.

Finally, I would like to sincerely thank all the birds who sacrificed for this project, in order to attempt to improve welfare for poultry as a whole.

TABLE OF CONTENTS

ABSTRACT ...... ii ACKNOWLEDGEMENTS ...... iv TABLE OF CONTENTS ...... vi LIST OF TABLES ...... ix LIST OF FIGURES ...... xii CHAPTER 1 ...... 1 Introduction ...... 1 1.1 Euthanasia needs in the layer industry ...... 1 1.2 Welfare concerns associated with euthanasia ...... 1 CHAPTER 2 ...... 4 Literature Review and Thesis Objectives ...... 4 2.1. Avian anatomy that helps to explain the efficacy of different euthanasia techniques ...... 4 2.1.1. Anatomy of the avian skull ...... 4 2.1.2. Neuroanatomy of avian brain and spinal cord ...... 5 2.1.3. Neuroanatomy of the reflex arc...... 10 2.1.4. Traumatic brain injuries (TBI) ...... 11 2.1.5. Anatomy of the avian neck ...... 12 2.1.6. Avian respiratory system ...... 14 2.2. Sensibility, death, and pain ...... 18 2.2.1. Sensibility ...... 18 2.2.2. Death ...... 20 2.2.3. Pain and its mechanism ...... 21 2.3. Assessing insensibility and time of death ...... 26 2.3.1. Brain stem and spinal reflexes ...... 27 2.3.2. Physiologic measures and behaviours ...... 29 2.4. Euthanasia methods and mode of action ...... 33 2.4.1. Physical methods...... 33 2.4.2. Inhaled gas ...... 41 2.5. Motivation, preference and aversion tests ...... 45 2.5.1. Introduction ...... 45 2.5.2. Approach avoidance and conditioned place avoidance paradigms ...... 47

2.6. Thesis objectives ...... 49 CHAPTER 3 ...... 52 Anatomical Pathology, and Behavioural and Physiological Responses Induced by Application of Non-Penetrating Captive Bolt Devices in Layer Chickens ...... 52 3.1. Abstract ...... 53 3.2. Introduction ...... 54 3.3. Materials and methods ...... 56 3.3.1. Animals and facilities...... 57 3.3.2. Non-penetrating captive bolt devices ...... 57 3.3.3. Ante mortem assessments ...... 59 3.3.4. Macroscopic assessment of tissue damage ...... 60 3.3.5. Statistical analyses ...... 61 3.4. Results ...... 62 3.4.1. Ante mortem assessments ...... 62 3.4.2. Pathology evaluations ...... 65 3.5. Discussion ...... 68 3.6. Conclusion ...... 74 CHAPTER 4 ...... 88 Efficacy of a Novel Mechanical Cervical Dislocation Device in Comparison to Manual Cervical Dislocation in Layer Chickens ...... 88 4.1. Abstract ...... 89 4.2. Introduction ...... 90 4.3. Methods...... 92 4.3.1. Animals and facilities...... 93 4.3.2. Koechner Euthanizing Device (KED) ...... 93 4.3.3. Anesthesia and killing procedures ...... 94 4.3.4. Ante mortem assessment ...... 95 4.3.5. Post mortem assessment...... 96 4.3.6. Statistical analyses ...... 97 4.4. Results ...... 99 4.4.1. Assessment of Ante mortem measures ...... 99 4.4.2. Assessment of postmortem measures...... 101 4.5. Discussion ...... 105 4.6. Conclusion ...... 110

vii

CHAPTER 5 ...... 125

Aversion to CO2 Gas in Laying Hens Using Approach-Avoidance and Conditioned Place Avoidance Paradigms ...... 125 5.1. Abstract ...... 126 5.2. Introduction ...... 127 5.3. Methods...... 130 5.3.1 Animals, Housing, and Management ...... 131 5.3.2. Experimental room and equipment ...... 131 5.3.3. Experimental design ...... 133 5.3.4. Training procedures ...... 133 5.3.5. Testing protocol ...... 134 5.3.6. Behavioural data collection ...... 135 5.3.7. Statistical analyses ...... 137 5.4. Results ...... 139 5.4.1. Approach and avoidance behaviour ...... 139 5.4.2. Other observed behaviours ...... 141 5.4.3. Parameters associated with loss of sensibility ...... 143 5.5. Discussion ...... 144 5.6. Conclusions ...... 153 CHAPTER 6 ...... 162 General Discussion ...... 162 6.1. Evaluation of physical on-farm euthanasia methods ...... 162 6.1.1. Humaneness of the euthanasia methods ...... 162 6.1.2. Pathology caused by euthanasia methods and effectiveness ...... 165 6.1.3. Limitations of reflex and behaviour measures ...... 169 6.1.4. Device success and operator safety ...... 172

6.2. Assessment of aversion to CO2 in laying hens ...... 175 6.2.1 Evaluation of the method ...... 179 6.2.2 Implications ...... 181 6.3 In retrospect ...... 183 6.4 Overall conclusion ...... 184 BIBLIOGRAPHY ...... 188

viii

LIST OF TABLES

Table 3. 1 : List of number of birds killed with the different NPCB devices by age group, strain, body weight, and sex. Number of failed birds are indicated in parentheses...... 75

Table 3. 2 : Ante-mortem assessment measures, descriptions, and procedures used, listed in order of observation after application of each killing method ...... 76

Table 3. 3: Gross and microscopic pathology scoring criteria for macroscopic, and microscopic hemorrhage ...... 77

Table 3. 4: Mean time (± SE, s) to onset of specific measures after application of different NPCB devices in different age groups of layer chickens ...... 78

Table 3. 5: Pearson correlation coefficients to assess the relationship between the antemortem measures for different NPCB devices for all ages of layer females and males (n = 94 for Zephyr E, n=92 for Zephyr EXL and n=93 for TED) ...... 79

Table 3. 6: Regression and relative contribution (R2) for response of dependent variable (Y) for independent variables (X) of different NPCB devices ...... 80

Table 3. 7: Summary of gross scores for subcutaneous hemorrhage, skull fractures, and subdural hemorrhage in birds killed by different NPCB devices. Number of birds with each score are indicated...... 81

Table 3. 8: Summary of microscopic scoring of brains for trauma following application of each of the three NPCB devices in layer chickens. Number of birds with each score are indicated. ... 82 Table 3. 9: Overall summary of microscopic scoring of subdural hemorrhage and parenchymal hemorrhage in the brain and spinal cord of layer chickens killed by NPCB device. Number of birds with each score are indicated ...... 83

Table 4. 1: Strain, sex, body weight and sample sizes for the different age classes of birds used in the study ...... 111

Table 4. 2: List of ante-mortem assessment measures, description, and procedure use, recorded in order of observation after application of each killing method (based on Chapter 3) ...... 112

Table 4. 3: Definitions for the terminology used in radiograph evaluation ...... 113

Table 4. 4: Gross and microscopic pathology scoring criteria for macroscopic, and microscopic hemorrhage (Woolcott et al., 2018ab; Chapter 3) ...... 114

Table 4. 5: Number of birds presenting with ante-mortem measures following application of the killing methods...... 115

Table 4. 6: Mean latencies to or durations of (± SE, s) ante-mortem measures in conscious and anesthetized chickens killed by manual cervical dislocation in different age groups. P values are given for effects of age, anesthesia and age by anesthesia interaction...... 116

Table 4.7: Mean latencies to or durations of (± SE s) ante-mortem measures in anesthetized chickens killed by manual or mechanical cervical dislocation in different age groups. P values are given for effects of age, method and age by method interaction...... 117

Table 4. 8: Presence and location of luxation/subluxation (from radiographs) and spinal cord transections (from macroscopic evaluation) in conscious and anesthetized chickens killed by manual cervical dislocation and KED1. Values indicate number of birds...... 118

Table 4. 9: Results of radiographic scoring on number of birds with fractures and the types and locations of fractures in conscious and anesthetized chickens...... 119

Table 4. 10: Macroscopic evaluation of subcutaneous hemorrhage (SCH) at the site of dislocation. Number of birds with each score are indicated ...... 120

Table 4. 11: Summary of microscopic scoring of brains for trauma following application of each of the three killing methods in layer chickens. Number of birds with hemorrhage (any score >0) in each section are indicated...... 121

Table 4. 12: Overall summary of microscopic scoring of subdural hemorrhage and parenchymal hemorrhage in the spinal cord of layer chickens killed by three killing methods. Number of birds with each score are indicated...... 122

Table 5. 1: Definitions of behaviours recorded in the Control Chamber (CC) and Treatment Chamber (TC) ...... 154

Table 5. 2: Test order, assigned CO2 treatment and outcomes in each of the four rounds for the 12 hens enrolled in the study...... 155

Table 5. 3: Number of laying hens that demonstrated different behaviour patterns in the Control (CC) and Treatment Chamber (TC) on the three different test days (Baseline Day, Gas Day, Washout Day)1 ...... 156

Table 5. 4: Number of laying hens that demonstrated different behaviour patterns in the Control (CC) and Treatment Chamber (TC) at the different CO2 concentrations on Gas Days...... 157

Table 5. 5: Frequencies and latencies of approach /avoidance behaviours (mean ± SE) demonstrated by laying hens in different test days, CO2 concentrations, and cycles...... 158

Table 5. 6: Behaviours associated with loss of consciousness on Gas Day in laying hens: Mean time ± SE s are presented for 25%, 35%, and 50% CO2 concentrations. Average time ± SD s are presented for 70% CO2 concentration...... 159

Table 5. 7: Time (s) at onset of behaviour during recovery in laying hens exposed to different CO2 concentrations. Mean time ± SE s are presented for 25%, 35%, and 50% CO2 concentrations ...... 160

LIST OF FIGURES

Figure 2. 1: (A) Schematic diagram of modern consensus view of avian brain according to the conclusions of the Avian Brain Nomenclature Forum (Jarvis et al., 2005, with permission). (B) Dorsal view of a chicken brain: C- cerebrum, D-cerebellum, H- hind brain, P- optic lobes, S- spinal cord...... 7

Figure 2. 2 : Ventral view of the brain of a chicken with cranial nerves (Orosz, 2007) ...... 8

Figure 2. 3 : A radiograph of the cervical vertebrae of a layer chicken (21 weeks old hen. SK- skull, O- occipital lobe, C1- first cervical vertebra, C2- second cervical vertebra, C3- third cervical vertebra, C4- fourth cervical vertebra...... 14

Figure 3. 1: Non-penetrating captive bolt devices. (A) Zephyr-E standard: (B) Conical shape bolt head, (C) Standard subject adapter. (D) Zephyr-E-layer: (E) Round shape bolt head, (F) Chicken subject adapter. (G) Zephyr-EXL: (H) Conical shape bolt head, (I) chicken subject adapter. (J) Turkey Euthanasia Device (TED): (K) Flat bolt head, (L) R-3 subject adapter...... 84

Figure 3. 2: Application of the Zephyr-EXL device on a 30 w.o. hen: The bird was restrained in sternal recumbency with its neck resting ventrally on the ground, and the wings held gently towards the body of the bird. Device was placed perpendicular to the top of the frontal bone just behind the comb and on the mid line between the eyes and ears...... 85

Figure 3. 3 : Figure 3. Gross pathology scoring criteria for skull fractures. Arrows indicate the fracture type [modified from Erasmus et al., (2010b) and Casey-Trott et al., (2013)]. (A) No fracture, intact skull (score 0). (B) Depression fracture (score 1). (C) Penetrating fracture-no imbedded fragments (score 2). (D) Penetrating fracture- with imbedded fragments (score 3). ... 86

Figure 3. 4 : Skin reflected to demonstrate gross subcutaneous hemorrhage. (A) Hemorrhage with less than 25% of area covered (score 2) of a 65 w.o. bird killed by the TED. (B) Hemorrhage completely covering area from the eyes to base of the skull (score 4) of a 10 w.o. bird killed by the TED. (C) Gross subdural dorsal hemorrhage covering less than 25% of the brain surface (score 1) of a 33 w.o. bird killed by the Zephyr-E. (D) Gross subdural dorsal hemorrhage covering 51 -75% of the brain surface (score 3) of a 33 w.o. bird killed by the Zephyr-E...... 87

Figure 4. 1: A- Koechner euthanasia device (KED model-C): S-single side blade; D-double angle blade. B-application of KED model-C in a 65 week old anesthetized rooster...... 123

xii

Figure 4. 2 : Radiographs of chickens showing cervical dislocations. A- Luxation between the skull and C1 vertebra in a 65 week old rooster killed by manual cervical dislocation: Sk-Skull, C1- first cervical vertebra, C2- second cervical vertebra. B - Subluxation between C2 and C3 (letter S shows the site of subluxation) in a 65 week old rooster killed by KED: fractures (F) are present on the articular processes of the C3 vertebra...... 124

Figure 5. 1 : A- Lateral camera view of a laying hen inside control chamber (CC) waiting for the sliding door (S) to open and to enter treatment chamber (TC). Wire mesh (W) can be seen on top of the CC. B- Lateral camera view of hen pushing through the curtain (C) to enter treatment chamber (TC) to access meal worms. Both images were collected on a Baseline Day, with ambient air conditions in CC and TC...... 161

Figure 6. 1 : The fabric chicken restraint device (fabric sleeve made of nylon): A- a toggle at the head end to tighten it, B- Opening at the bottom end ...... 187

xiii

CHAPTER 1

Introduction

1.1 Euthanasia needs in the layer industry

On-farm euthanasia is considered to be a fundamental aspect of management in livestock production. In the poultry industry, some of the most common reasons for killing birds on the farm are (1) to prevent suffering from injury or sickness, (2) disease control, and (3) stock management (e.g., of male layers). In Canada, all layer farms are required to have a euthanasia plan and any injured or diseased birds found in layer flocks need to be identified and euthanized in a timely manner (NFACC, 2017). Therefore, on-farm euthanasia is necessary for eliminating suffering and distress and maintaining flock health and well-being.

1.2 Welfare concerns associated with euthanasia

The term euthanasia is used to describe ending the life of an individual to minimize or eliminate pain and distress (AVMA, 2013). The term euthanasia is derived from the Greek word “eu” meaning good and “thanos” meaning death. When an animal no longer has good welfare (when it no longer has a life worth living; Mellor, 2016), the humane thing to do is to offer it a good death. The desired outcome of euthanasia is relieving suffering (minimize pain, distress, and negative affect to the animal). Therefore, the humaneness of the technique is considered to be an important ethical issue. Many authors define the humaneness of a killing technique based on how quickly the animal loses sensibility (Erasmus et al., 2010a; Martin et al., 2016; Woolcott et al.,

2018 a,b). Therefore, the preferred methods are ones that cause minimal pain and distress and

1 result in rapid loss of sensibility followed by loss of brain function and, ultimately, respiratory and cardiac arrest (AVMA, 2013).

Consideration must also be given to selecting a euthanasia method that is reliable, effective, practicable, safe, easy to use and as esthetically acceptable as possible (AVMA, 2013). It is mandatory for the personnel performing the euthanasia to demonstrate proficiency in the use of the technique (NFACC, 2016). Experience in humane restraint of the animal is also important to ensure that pain and distress are minimized. Moreover, proper physical handling is vital to ensure the safety of the person performing the euthanasia and to protect other persons and animals.

On these grounds, the American Veterinary Medical Association classified euthanasia methods as acceptable, acceptable with conditions, and unacceptable (AVMA, 2013). Acceptable euthanasia methods for poultry include overdose of injectable anesthetic (barbiturates and barbituric acid derivatives). Inhaled gases (CO2, CO, N2, Ar) are acceptable with the condition that they cause rapid loss of sensibility and no or minimal aversiveness to the species. Manual and mechanical cervical dislocation are acceptable, but with the condition of no crushing of cervical bones and spinal cord unless the bird is first rendered insensible. Decapitation is acceptable with the condition that it must be performed by a competent person with a sharp instrument, ensuring rapid and unobstructed severing of the head from the neck. Penetrating and non-penetrating captive bolts are acceptable for euthanasia of large poultry with appropriate restraint and rapid loss of sensibility. Blunt force trauma applied manually to the head is not listed as either acceptable or acceptable with conditions for poultry euthanasia in AVMA (2013) guidelines, as this method could be displeasing to the personnel when performing it and repeated

2 application can result in personnel fatigue, loss of efficacy, and humane concerns. However, in the recent Code of Practice for the Care and Handling of Pullets and Laying Hens, NFACC

(2017) listed blunt force trauma as an on-farm euthanasia option.

Therefore, all euthanasia techniques must meet certain criteria to be considered humane.

However, not all existing methods have been scientifically evaluated to determine whether or not they meet these criteria for laying hens, and new methods are being developed that also require evaluation.

CHAPTER 2

Literature Review and Thesis Objectives

Current on-farm euthanasia techniques for poultry involve separation of the body from brain and central nervous system control (eg., mechanical/manual cervical dislocation or decapitation), direct destruction of the brain (blunt force trauma, penetrative or non-penetrative captive bolts), and use of inhaled gas leading to death by hypoxia. This chapter explains: 1. Basic anatomy and related functions of the avian skull, brain, neck and cardiorespiratory system; 2. Consciousness, sensibility, death and pain; 3. Assessing insensibility and time to death by using different parameters; 4. Euthanasia methods and their modes of action; 5. Motivation, preference and aversion tests, which explains methods for assessing aversiveness to gaseous methods in animals.

All five topics in the chapter help to gain a better insight into scientific assessment of efficacies and welfare implications of poultry on-farm euthanasia techniques.

2.1. Avian anatomy that helps to explain the efficacy of different euthanasia techniques

2.1.1. Anatomy of the avian skull

The avian skull is light weight and strong and is characterized by a beak which replaces the true jaw of mammals (Hogg, 1982). The dorsal view of the skull is cone shaped where the beak creates the point of the cone. However, the skull is flattened on the ventral side. Two large orbital fossa help to accommodate relatively large eyes. The avian skull is divided into two cranial and facial anatomical areas which are formed by collections of several bones. Facial

4 bones are located at the front of the skull (nasal bone) and beak. The beak is formed by the premaxillary, the maxillaries, and the nasal bones (these bones are completely fused in most adult birds, and sometimes pneumatized). Cranial bones are those that form the back of the skull and the cranial cavity, which accommodates the brain. The main cranial bones are the occipital bone, the parietal bones, the temporal bones and the frontal bones. The “foramen magnum” is the large opening located in the back of the skull (within the occipital bone) where the spinal cord passes through to connect with the brain. The suture lines of the cranial bones and facial bones are visible until 4-5 months of age, but fuse at older ages in domestic fowl (Hogg, 1982).

2.1.2. Neuroanatomy of avian brain and spinal cord

The avian brain is a delicate structure accommodated inside the cranial cavity. Anatomically it is divided into three major regions: cerebrum (telencephalon), cerebellum (metencephalon), and brain stem (compromised of medulla oblongata, pons, mesencephalon [mid brain] and diencephalon) (Figure 2.1). The brain is encased in three protective connective tissue layers known as meninges: dura mater, arachnoid mater and pia mater, all developed from the mesoderm of the embryo. The subarachnoid space contains cerebrospinal fluid.

The largest area of the avian brain is the cerebrum. The avian cerebrum is divided into the left and right hemispheres by a fissure, which is similar to the mammalian brain and is involved in the processing of sensory information (visual and auditory) and control of motor activity (Jarvis et al., 2005). Some differences exist in the cerebral cortex of a bird: the cerebral cortex has a smooth surface and lacks gyri and sulci that are present in mammalian brains. In addition, the cortical cells which process information are absent on the surface of avian brains but process the information using subcortical nuclei located deep in the cortex (Orosz, 2007). The olfactory

5 lobes of birds are relatively smaller than in mammals (Husband and Shimizu, 1999). The cerebellum is a well-developed structure in birds and controls essential functions of locomotion and balance (Pearson, 1972). Dorsally, the mid brain is covered by the cerebellum and cerebral hemispheres. The mid brain is mainly responsible for coordinating sensory input, and possesses two channels which connect left and right sides of the brain. It consists of the optic tectum (roof) and tegmentum (floor) (Whittow, 2000). The avian optic tectum, which processes visual stimuli, is comparatively enlarged and highly developed in comparison to mammals who have a tiny superior colliculus instead (Whittow, 2000). The medulla oblongata (brain stem) connects the spinal cord to the brain and is responsible for controlling respiration, blood circulation, motor functions and movement related to food intake (Nickel, 1977). The reticular formation, formed by a bewildering array of nuclei in the brain stem helps to integrate information between the medulla, cerebellum and spinal cord through ascending and descending neural networks (Jones,

1995). It plays a crucial role in sensibility and arousal in both mammals and birds (Revzin,

1965). Therefore, loss of sensibility results from any damage to the reticular formation or to the pathway that connects the reticular formation to the cerebral cortex. The pons is located above the medulla oblongata which helps to integrate information in between cerebellum and cerebrum by forming the nuclei bridge. The diencephalon consists of the epithalamus, thalamus, and hypothalamus (Breazile and Heartwig, 1989). The epithalamus controls some functions in the limbic system while the pineal gland, located in the epithalamus, regulates circadian rhythms.

The thalamus integrates sensory inputs which are received from different sensory nerves. The hypothalamus regulates homeostasis by producing a variety of hormones responsible for various functions in the body.

A B

Birds have 12 cranial nerves (CN) which are similar to those of mammals (Orosz, 1996; Figure

2.2). The olfactory and optic nerves emerge from the cerebrum while the other ten nerves emerge from the brain stem. The olfactory nerve (CN I) is a sensory nerve responsible for smell, while the optic nerve (CN II) is responsible for vision. The oculomotor nerve (CN III) is a motor nerve which controls the dorsal, ventral and medial rectus muscles, and the ventral oblique muscle of the eye. It also innervates the muscle of the eyelid. The dorsal oblique muscle of the eye is controlled by the trochlear nerve (CN IV). The trigeminal nerve (CN V) is a mixed nerve consisting of two sensory branches (the ophthalmic branch and maxillary branches) and one motor branch (mandibular). The sensory branches supply nerves to the upper eye lid, skin of the forehead, nasal cavity, upper beak, lower eyelid, palate and infraorbital sinus. The mandibular branch innervates the muscles for mastication and skin and the mucosa at the commissures of the

7 beak. The abducent nerve (CN VI) controls the lateral rectus muscle and muscle of the third eyelid. The facial nerve (CN VII) controls the hyoid and cutaneous neck muscle and is not involved in the sense of taste in birds as it is in mammals. The vestibulocochlear nerve (CN VIII) is responsible for posture in relation to movement of the head, and eye coordination. In addition,

CN VIII facilitates perception of hearing. The glossopharyngeal nerve (CN IX) receives sensory input from taste fibres and controls larynx and trachea joining with the vagus nerve (CN X).

Superficial muscles of the neck are controlled by the spinal accessory nerve (CN XI) and vagus nerve. The hypoglossal nerve (CN XII) and glossopharyngeal nerve also innervate the tracheal muscles.

Figure 2. 2 : Ventral view of the brain of a chicken with cranial nerves (Orosz, 2007)

The spinal cord is located within the vertebral column and is the primary neural pathway connecting the body and the brain. The spinal cord connects to the caudal end of the brain stem.

The meninges covering the brain continue around the spinal cord and work as a protective casing. Pia matter envelops the spinal cord as the innermost layer, and the middle layer is arachnoid matter (Yew et al., 1996). There are spaces in between the meninges, and cerebrospinal fluid and arteries which supply oxygenated blood to the spinal cord are located inside the subarachnoid space. The outer layer of the spinal cord is formed by the myelinated nerve tracts (white matter) while the inner layer consists of grey matter (cell bodies of inter neurons and motor neurons, unmyelinated axons and neuroglia cells (Butler and Hodos, 1996).

A pair of right and left spinal nerves exit between each pair of vertebrae. Spinal nerves transmit information about the internal and external environment of the body to the brain and vice versa.

The region of the cord from which one set of spinal nerves emerges is known as a spinal segment. Each spinal segment is named for its corresponding vertebrae. These segments are grouped into regions according to main body regions through which the vertebral column passes

(cervical- in the neck, thoracic- in the chest cavity, lumbar- in the abdominal region, sacral- in the pelvic region). The avian spinal cord vertebral column are the same length, and birds do not have a cauda equina as in mammals (Orosz, 1996). Spinal nerves emerge laterally rather than caudally through the vertebral foramin. However, the dura is separated from the periosteal lining, forming an epidural space in the cervical and thoracic regions. This epidural space is filled with a fluid (gelatinous tissue) and acts as a shock absorber, and facilitates the flexibility of the neck in birds (Orosz, 1996).

2.1.3. Neuroanatomy of the reflex arc

Reflexes are automatic responses to internal or external stimuli. The reflex arc is the functional unit of nerve reflexes. In mammals and birds, some sensory neurons synapse with the motor neurons in the spinal cord without passing directly into the brain, and facilitate reflex actions more quickly than signals that pass through the brain. However, the brain can receive sensory impulse from the reflex.

Reflex arcs that consist of only two neurons (a sensory and a motor neuron) are considered monosynaptic and are the simplest skeletal muscle reflexes; an incoming sensory axon from a muscle receptor enters through the dorsal root and terminates on a motor neuron in the ventral horn of the gray matter. Most reflex arcs consist of three neurons: a sensory neuron, an intermediate neuron (a connecting or relay neuron) and a motor neuron. For example, a stimulus activates the pain receptors (nociceptors) of the skin and initiates an impulse in a sensory neuron.

This impulse travels to the spinal cord via the sensory neuron and passes to the relay neuron in the spinal cord. The relay neuron passes the impulse to the brain and the motor neuron that transmits the impulse to the muscles, causing them to contract and/or pull away from the source of pain (Butler and Hodos, 1996).

Reflexes whose arcs pass through the spinal cord are called spinal reflexes. The spinal reflexes are important in providing an automatic muscle reaction in response to a stimulus eg: withdrawal reflex (pedal reflex) in chickens. In the withdrawal reflex, stimuli that are painful or unexpected result in the reflexive withdrawal of a limb or the entire body. This type of withdrawal is caused

10 by positive excitatory reflexes where stimuli activate motor neuron to contract the muscle (Butler and Hodos, 1996).

Brain stem reflexes are controlled by the reflex center in the hind brain. Brain stem reflexes are associated with the cranial nerves. These reflexes help to indicate the source or location of brain stem lesions and neurological disorders (Benett, 1994). Insensibility and complete loss of function in the brain of an animal is assessed based on the absence of brain stem reflexes

(Erasmus et al., 2010; Sandercock et al., 2014; Martin et al., 2016; Terlouw et al., 2016b).

2.1.4. Traumatic brain injuries (TBI)

Brain injuries are categorized in to traumatic brain injuries (TBI) and non-traumatic brain injuries. TBI is defined as an alteration in brain function, or other evidence of brain pathology, caused by an external force (David et al. 2010). When the injury is caused by an internal force

(e.g., stroke, infectious disease, electrical shock, lack of oxygen) it is known as non-traumatic brain injury.

There are two types of head injuries; these are classified as closed head injuries and penetrating head injuries (Gerstenbrand and Stepan, 2001). The closed head injury is an injury to the brain caused by an outside force without any penetration of the skull. When a foreign body penetrates the skull and passes into the meninges, it is known as a penetrating brain injury. There are four main pathoanatomical sequelae of TBI: contusions, subarachnoid hemorrhage, hematomas

(including epidural, subdural, and intra parenchymal lesions), and diffuse axonal injuries

(Gerstenbrand and Stepan, 2001). Other than those four, ischemic brain injury and cerebral

11 edema might be included in a “pathoanatomic” classification scheme. Strong rotation of the head or shaking of the head causes diffuse axonal injury (Shaw, 2002).

Cerebral concussion is one of the most common traumatic brain injuries which is caused by violent physical shaking of the brain and is responsible for a sudden temporary loss of sensibility. Two authors define the cerebral concussion as a temporary disturbance of neural activities due to sudden acceleration or deceleration of the head (Rosenthal, 1993; Label, 1997).

Due to high speed impact to the skull, blood vessels in the brain are torn and cause internal bleeding. Hemorrhages are classified according to the location in the brain by using the name of the meninges as a guide. Hemorrhage above the dura matter is classified as epidural hemorrhage

(EDH). A subdural hematoma (SDH) is a collection of blood below the inner layer of the dura but external to the brain and arachnoid membrane. EDH and SDH are the most common type of traumatic intracranial lesions in humans (Haselsberger et al., 1988). Mortality in acute SDH is reported between 57% to 90% in humans (Haselsberger et al., 1988; Wilberger et al., 1991).

Bleeding into the cerebrospinal fluid which is located below the arachnoid membrane is known as subarachnoid hemorrhage.

2.1.5. Anatomy of the avian neck

The neck of the chicken consists of 14 cervical vertebrae (C1-C14) (McLeod et al., 1964; Figure

2.3). The bodies of the vertebrae consist of freely movable saddle joints making the vertebral column flexible (McLeod et al., 1964). The outer surface of the vertebrae consists of articular and transverse processes (McLeod et al., 1964; Whittow, 2000). The first vertebrae (Atlas) and

12 the second (Axis) are morphologically different from the rest of the vertebrae. The atlas (C1) is a small, ring-like structure with a deep cavity (McLeod et al., 1964). The atlas attaches to the skull via the occipital condyle to form the occipito-atlantal joint and facilitates the head to turn on the neck (McLeod et al., 1964). The Axis (C2) attaches to the Atlas by atlanto-axial joints and helps for cervical rotation (McLeod et al., 1964). The rest of the cervical vertebrae are similar to each other. Each vertebra has a body that is concave on its superior surface and convex on its inferior surface (McLeod et al., 1964). The hollow center of the body is called as vertebral foramen where the spinal cord is located (McLeod et al., 1964; Whittow, 2000). Intervertebral discs are located between the vertebral bodies, and are involved in cervical spine motion, stability, and weight-bearing. (McLeod et al., 1964; Whittow, 2000)

Dislocation or fracture in vertebrae can result from mechanical damage (flexion, rotation, compression, or extension) (Taneichi et al., 2005; Veras et al., 2000). Mechanical damage to the vertebrae commonly results in damage to the spinal cord. Carotid arteries supply the head and neck with oxygenated blood. The chicken has paired carotid arteries that run one on top of the other in ventral side of the neck (MacLelland, 1990). Each carotid artery divides into three separate arteries: the occipital artery, the internal carotid artery and the external carotid artery at the base of the skull, and provide blood to separate areas of the head.

2.1.6. Avian respiratory system

The primary function of the respiratory system is gas exchange by delivering enough oxygen and removing sufficient carbon dioxide for metabolic demand. There are marked differences between the mammalian and avian respiratory systems. In chickens and turkeys, the nostril is closed dorsally by a horny flap, the operculum. There is a vertical lamella of cartilage at the ventral border of the nostril (McLelland, 1990). The avian trachea is typically longer and wider than the trachea of mammals (Gleed et al., 2001), and lies on the right side of the neck. The tracheal cartilage are complete rings and consist of broad and narrow parts (Gleed et al., 2001). The

14 trachea divides into two primary bronchi at the syrinx which is responsible for vocalization in birds (Gleed et al., 2001).

Avian lungs are paired and attached firmly to the dorsal ribs (Duncker, 1974). Each lung is triangular or quadrilateral in shape, does not divide into lobes and does not change volume during breathing (Duncker, 1974). Most birds have nine air sacs which are poorly vascularized by the systematic circulation (Duncker, 1974). In the chicken, the nine air sacs include four pairs of air sacs and one unpaired sac: two interclavicular air sacs, two abdominal air sacs, two anterior thoracic air sacs, two posterior thoracic air sacs, and one in the cervical area (Maina,

2003). Air sacs do not directly contribute to gas exchange (Magnussen et al., 1979). However, these air sacs contribute to effective respiration by helping to ventilate the lungs (Farmer, 2006).

The total volume of the respiratory system in birds is greater than in comparably sized mammals due to the air sacs (Magnussen et al., 1979). In contrast to mammals, birds do not have a diaphragm to control the air pressure inside the thoracic cavity, and the avian thoracic cavity is at the atmospheric air pressure (Whittow, 2000).

There are three types of bronchi in birds: primary, secondary and tertiary (called parabronchi)

(Duncker, 1974). Tertiary bronchi are the functional unit of gas exchange (Duncker, 1974).

There are connections between the secondary bronchi and air sacs. Smooth muscle rings that surrounded the parabronchi lumen generates the force to move air in and out of the air sacs and through the parabronchial lung (Duncker, 1974).

2.1.6.1. Pulmonary circulation

The avian cardiovascular system is considered a high-performance system in comparison to mammals, due to a proportionally larger heart, larger stroke volume, greater cardiac output, higher blood pressures, and a lower heart rate (Whittow, 2000). Oxygen and carbon dioxide exchange is also more efficient in birds than in mammals (Altman et al., 1997). The functional anatomy of the pulmonary circulation has been described for the domestic fowl (Duncker, 1974;

Abdalla and King, 1975). Interparabronchial arteries run in between the parabronchi and branch out to the pulmonary blood capillaries near the outside edge of the parabronchial mantle forming a meshwork of capillaries (Duncker, 1974). Pulmonary capillary blood (oxygenated blood) is collected in intraparabronchial veins located near the outside edges of the parabronchus and transport to the heart via pulmonary vein (Duncker, 1974).

2.1.6.2. Welfare perspective on breathlessness

Breathlessness is a negative subjective experience that can impact one’s welfare (Beausoleil and

Mellor, 2015). Breathlessness is well documented in humans and identified as an unpleasant sensation (Evans et al., 2002; von-Leupoldt et al., 2008). Dyspnoea is the most common term used to describe sensations and experiences of respiratory discomfort in humans. According to the definition of the American Thoracic Society “Dyspnoea is a term used to characterize a subjective experience of breathing discomfort that consists of qualitatively distinct sensations that vary in intensity. The experience derives from interactions among multiple physiological, social and environmental factors, and may induce secondary physiological and behavioural responses” (Parshall et al., 2012). However, in the veterinary literature, the term dyspnoea is defined as “difficult, laboured breathing” (Mellema, 2008).

Dyspnea, or breathlessness, is described in terms of its 3 different qualities: respiratory effort, air hunger, and chest tightness (Beausoleil and Mellor, 2015). The degree of unpleasantness varies in different qualities of breathlessness (Banzett et al., 2008). Respiratory effort is an increased effort (depth or rate) to achieve the necessary or desired level of ventilation (Beausoleil and

Mellor, 2015). In respiratory effort, a motor command is sent to the respiratory muscles to increase its activity to achieve the desired level of ventilation (e.g., during exercise). Air hunger is the increased urge or increased need to breathe (Beausoleil and Mellor, 2015) and reported as more unpleasant than respiratory effort (Banzett et al., 2008). Poor coordination between automatic motor command and the degree of lung inflation causes air hunger (Lansing et al.,

2009). In general, increased level of CO2 in blood (hypercapnia), reduced arterial oxygen tension

(hypoxaemia) or metabolic acidosis results in the sensation of air hunger (Beausoleil and Mellor,

2015). Chest tightness occurs due to bronchoconstriction which can be a result of respiratory inflammation or allergic bronchitis (Beausoleil and Mellor, 2015).

When assessing the humaneness of different euthanasia methods, authors utilize different terminology to describe the respiratory responses in animals (gasping in broilers exposed to

Argon or CO2 – Gerritzen et al., 2000; Lambooij et al., 1999; deep breathing (deeper than normal inspiration through the mouth) in laying hens and broilers exposed to CO2 or Argon- Webster and Flethcer, 2001, mouth breathing in layers and broilers- Webster and Fletcher, 2004; respiratory disruption (apparent increased inhalation depth and duration) in broilers exposed to controlled atmosphere stunning with nitrogen, CO2, or argon (McKeegan et al., 2005). Some terminologies operationally describe the behaviour (e.g., open mouth breathing) whereas others describe the association with pain and distress (e.g., respiratory disruption). However, it is most

17 reasonable to utilize terminology that describes the behavioural response, since associated feelings of breathlessness or pain are currently unknown in animals.

In hypoxia or ischaemia, the eupnoea (normal rhythmic breathing) fails and gasping (a second pattern) occurrs (St-John, 2009). In gasping, frequency and tidal volume of respiration initially increase and then decline and cease (St-John, 2009). The functional mechanism of gasping is associated with autoresuscitation and return of eupnoea (St John & Paton, 2004). The mechanisms of eupnoea and gasping are generated in the brainstem respiratory control system

(St-John, 2009).

In this thesis the term “gasping” is used to refer to paroxysmal opening of the beak in in Chapter

3 and Chapter 4. This behaviour was observed in the chickens killed by nonpenetrating captive bolts and cervical dislocation. Chest movement did not occur in these birds during gasping behaviour. The term “open mouth breathing” is used to describe the observed breathing behaviour in chickens exposed to CO2 in the Chapter 5. In open mouth breathing, birds breath through their mouth (beak is opened and chest movements occur). It was difficult to conclude the observed pattern of breathing was normal or abnormal in the CO2 chamber. Thus, it is reasonable to describe the observed breathing behavour as open mouth breathing.

2.2. Sensibility, death, and pain

2.2.1. Sensibility

Sensibility cannot be considered as one single phenomenon. In general, sensibility is described as the state of wakefulness which integrates sensation, sight, vocalization, and feeling (Damasio

18 and Meyer, 2009). Human consciousness is described as incorporating external (behavioural) and internal (cognitive, mental) components (Damasio and Meyer, 2009). The external component includes signs of wakefulness, background emotions, sustained attention towards objects and events in the environment, and the internal component includes more about the mental state representing objects and events in relation to onse’s self (Damasio and Meyer,

2009).The primary level of consciousness in humans is defined as "an awareness of one's surroundings, of the self, and of one's thoughts and feelings" (Sommerhoff and MacDorman,

1994). Because wakefulness is always an inferred state in animals and can only be assessed indirectly, the term sensibility is preferred over consciousness to describe animal wakefulness and responsiveness to external cues. In this thesis, the term of insensibility is used over consciousness throughout.

Sensibility is evaluated in different states from wakefulness to sleeping to being in a coma

(Zeman, 2005; Sandercock et al., 2014). In neurobiological definitions, sensibility refers to mental responsiveness and is associated with the reticular activating system. The Ascending

Reticular Activating System (ARAS), which innervates nerve fibres to the reticular formation of the brain stem, dorsal pons and cerebral cortex, is responsible for the wakefulness (Brown et al.,

2012). A good functioning reticular formation and ARAS are essential for the maintenance of sensibility. The cell bodies of the neurons of the reticular activating system are arranged diffusely throughout the telencephalon. The function of this system is associated with arousal, and its destruction produces a permanent coma (Kinney et al., 1994; Turner and Knapp, 1995).

In general, loss of sensibility is a result of irreversible or reversible dysfunction of the ARAS and/or the cerebral hemispheres. Insensibility is defined as: “a state of unawareness (loss of

19 consciousness) in which there is temporary or permanent disruption to brain function” (Terlouw et al., 2016a).

There is considerable scientific evidence based on avian brain physiology to suggest that birds exhibit sentience and awake states much like mammals (Butler et al., 2005; Rattenborg et al.,

2009). Moreover, awake birds showed similar electrical activity patterns of the brain compared to mammals (Edelman et al., 2005). Sandercock et al. (2014) studied different reflex responses by using electroencephalograph (EEG) activity (as a measure of brain function) in four different clinical states of sensibility in hens and turkeys. The clinical states of sensibility they studied were awake, sedated (drowsy), anesthetized (insensible) and deep hypnotic (insensible). Their results revealed different reflex responses and corresponding EEG activities for the different awake states to death.

2.2.2. Death

The traditional concept of death is defined as the end of a life due to cessation of heart beat and respiration. A new concept of brain death was first described clinically in 1959 by two French physicians who identified the state as "coma depasse" (Mollaret and Goulon, 1959). In literal terms, the state of "coma depasse" is beyond the state of coma. The report of the Ad Hoc

Committee of the Harvard Medical School in 1968 elaborated on the concept of brain death and brought awareness of brain death to a much wider audience via the “Harvard criteria”. The

Harvard criteria to establish brain death included 1. lack of receptivity and responsivity (with the most intensely painful stimuli evoking no vocal or other response); (2) no movements (during observation of one hour); (3) no breathing (apnea was to be confirmed during three minutes off

20 any mechanical respirator); (4) no reflexes (emphasis being on brain stem reflexes); and (5) flat electroencephalogram (EEG). To confirm death by the Harvard criteria, all the above findings should show no change 24 hours later. These criteria have been accepted for confirming the death of an individual. However, three years later two neurosurgeons suggested that irreversible damage to the brain stem was the "point of no return” and the state of brain death in a patient with known irreparable pathology can be established solely on clinical grounds (Mohandas and

Chou, 1971). Furthermore, Mohandas and Chaw (1971) suggested that EEG is not mandatory to detect brain death, and they stated that presence of spinal reflexes had no bearing on the question of brain death. These criteria are known as “Minnesota Criteria” or the “first brain stem criterion”. After the Minnesota Criteria were released, a statement on the diagnosis of brain death was published at the Conference of Medical Royal Colleges and their Faculties in the United

Kingdom in 1976: brain death was defined as “complete, irreversible loss of brain-stem function” (BMJ, 1976; Lancet, 1976).

2.2.3. Pain and its mechanism

Pain often signals injury or disease in the body and produces actions to treat its causes. The

International Association for the Study of Pain defined pain in humans as "an unpleasant sensory and emotional experience associated with potential or actual tissue damage, or described in such terms” (IASP, 1994). Melzack and Wall (1965) proposed the theory of pain. The theory proposed that “injury activates specific pain receptors and fibres which, in turn, project pain impulses through a spinal pain pathway to a pain center in the brain”. International Association for the Study of Pain defined nociception as the “encoding of noxious stimuli” (IASP, 1994). “A stimulus that is damaging or threatens damage to normal tissues” is known as a noxious stimulus

(IASP,1994). Nociception involves detection and quantitation of noxious stimuli. The process of nociception consists of modifying and conveying that information to the brain (transmission, modulation, and projection), and recognizing of the stimulus.

Pain may be physiologic, pathologic or neurogenic; acute or chronic; visceral or somatic

(Goldberg, 2014). Physiologic pain sensation is elicited by noxious stimulation of high-threshold receptors, called nociceptors, in the skin that sense changes in heat, pressure and chemical stimuli (Goldberg, 2014). Nociceptors innervate skin, muscle, fascia, joints, tendons, blood vessels, and visceral organs. Varieties of nociceptors exist based on their response to stimuli

(Goldberg, 2014). Mechanical nociceptors selectively respond to intense mechanical stimuli, and polymodal nociceptors respond non-selectively to noxious mechanical, thermal, and chemical stimuli (Goldberg, 2014). These activated nociceptors send their information to the spinal cord via two different afferent fibres: rapid myelinated A-delta fibres and slower unmyelinated C- fibres (Basbaum et al., 2009). In the spinal cord these fibres synapse with the second-order neurons in the dorsal horn, which either send axons into ascending sensory pathways to transmit pain information to higher centers of the brain or serve as interneurons in segmental reflex pathways, which allows the body to rapidly withdraw from noxious stimuli (Basbaum et al.,

2009). Ascending pathways involved in the transmission of pain include the cerebral cortex, limbic system, and reticular activating system (Basbaum et al., 2009).

Chemically-induced pain is known as chemesthesis. The trigeminal nerve (cranial nerve V) is the main functional component of the chemesthetic system (Bryant and Silver 2000). The trigeminal nerve contains chemoreceptive fibers which can detect chemical irritants (Silver and Maruniak,

1981). The trigeminal nerve consists of three major branches which serve the different mucosal regions of the head: ophthalmic nerve (a sensory nerve), maxillary nerve (a sensory nerve), and mandibular nerve (a mixed nerve, both sensory and motor). In chickens, the ophthalmic branch innervates the frontal region, the parts of the eye, the rostrodorsal part of the nasal cavity, and is also responsible for the motor control of eye and associates with reflexive response to irritating stimuli to the ocular region (Mason and Clark, 2000). The maxillary nerve supplies sensory inputs mainly to the mucosa of the conjunctiva and the palate, and the floor of the medial wall of the nasal cavity (Mason and Clark, 2000). The mandibular branch innervates the rest of the mouth including oral mucosa and wattles (Getty, 1975). Recently, McKeegan found that nasal and buccal trigeminal polymodal nociceptors in chickens respond to ammonia, acetic acid vapor, and carbon dioxide (McKeegan et al., 2003, 2005; McKeegan, 2004). McKeegan (2004) reported that avian nociceptive threshold for CO2 is in the region of 40 - 50% CO2 in air.

It is difficult to understand animals’ experience of pain due to the absence of language. In general, pain is detected by observing animals’ general appearance and “nocifensive” (pain-like) behaviours; most laboratory species react in a manner to avoid and reduce the impact of acute noxious stimuli, suggesting that they experience pain (Carstens and Moberg, 2000). There is both physiological and behavioural evidence that domestic fowl experience pain, but they also appear to have the ability to suppress pain by changes in motivation (Gentle, 2011). Gentle (2001) assessed the cognitive perception of pain by a series of studies on selective attention on pain related behaviours in chickens. The author used a sodium urate model of gouty arthritis to cause acute gouty attacks in a single joint of a chicken, which produced a painful inflammation lasting for 3 hrs. Nesting, feeding, exploration, and social interactions were used as the motivational

23 changes. The author reported that the degree of pain suppression (based on expression of pain related behaviours) ranged from hypoalgesia to complete analgesia. Therefore, shift in attention has an ability to suppress the existing tonic pain in chickens. This phenomenon is important in planning of different killing methods to minimize associated tonic pain.

Animal euthanasia guidelines explained bone crushing prior to insensibility is painful and should be avoided (AVMA, 2013). Bone fractures are reported to be acutely painful in humans (Bove et al., 2009), and other mammals (Waran et al., 2010) and in birds (Nasr et al., 2012). Mechanical nociceptors in the periosteum of the affected bone are activated due to fracture displacement

(Freeman et al., 2008).

2.2.3.1. Anesthesia as a pain management tool

Anesthesia is mainly used during operative procedures so that the animal does not experience pain. Anesthesia is defined as the “state in which, as a result of drug-induced insensibility the patient neither perceives nor recalls noxious stimulation” (Roberts., 1987). There are two main types of delivery routes for anesthetic agents: intravenous agents (generally administered together with sedatives or narcotics), and volatile agents. Anesthetics interact with the ion channels that regulate synaptic transmission and membrane potentials in key regions of the brain and spinal cord (Alkire et al., 2008). Anesthetics decrease or inhibit the excitation mechanism of the neurons (Ries and Puil, 1999). With all these mechanisms, anesthetics produce insensibility by preventing integration (blocking the interactions among specialized brain regions) or by reducing information (shrinking the number of activity patterns available to cortical networks)

(Alkire et al., 2009).

The spinal cord is an important site of anesthetic mechanism. Neurons within the dorsal horn of the spinal cord are involved in the transmission of stimuli to other central nervous system sites.

Anesthetics depress dorsal horn neuronal responses to noxious stimuli. Moreover, anesthetics depress the motor neuron excitability within the spinal cord causing anesthetic-induced immobility in response to a noxious stimulus (Collins et al., 1995).

2.2.3.2. Anesthetic depth and reflex responses

Many of the homeostatic control systems in the body are suppressed by anesthetics drugs.

Different states of anesthesia involve different suppression levels of sensibility, pain perception, muscle tone and reflexes, which are due to the effect of the anesthetic drugs on the central nervous system. Death can occur due to anesthetic overdose and pain can be experienced when too light of a dose is used. For example, anesthetic agents, which have narrow safety margins, result in death in one in 679 cases during anesthesia of healthy dogs and cats (Clarke and Hall,

1990). Therefore, maintaining minimum anesthetic depth for the particular procedure is important, through adjusting the anesthesia. In general, anesthetized patients are insensitive to stimuli and passage of time and do not dream (Hameroff, 2001). In animals, anesthetic drugs are generally administered as a single injection (intraperitoneal or intramuscular) providing analgesia, hypnosis, and muscle relaxation.

Anesthetic depth in animals is primarily assessed based on observational techniques of muscle relaxation, reflex activities, and physiologic responses (Sandercock et al., 2014). The pedal withdrawal reflex is commonly used to assess the depth of anesthesia in animals. If the animal withdraws the limb in response to a toe pinch, it is assumed that the animal still experiences pain

25 and need more anesthesia (Nevarez, 2005). The ocular reflexes are commonly used to assess the anesthetic depth, which includes palpebral response, ocular position and corneal reflex (Nevarez,

2005; Lierz and Korbel, 2012). Sandercock et al. (2014) studied the “readily observable reflexes and behaviours that are reliably associated with different states of sensibility” in two poultry species (turkeys and laying hens). The clinical states of consciousness that they studied were fully awake, semi awake (sedated), insensible–optimal (general anesthesia), and insensible–sub- optimal (deep hypnotic state). According to their results, pupillary and nictitating membrane reflexes and muscle tone (jaw and neck tone) were always present in sedated semi awake birds.

Jaw tone, neck tone and response to nociceptive stimuli were absent, and only pupillary and nictitating membrane reflexes were present both in general anesthetic and deep hypnotic state.

Moreover, there were no differences in respiration rate, blood pressure and rectal temperature between anesthetic states or between species.

2.3. Assessing insensibility and time of death

When animals are euthanized, the intention is to induce immediate, irreversible insensibility to ensure minimum pain and distress (AVMA, 2013). Therefore, it is necessary to have a set of criteria to assess insensibility and brain death. Indicators of sensibility must be absent, and indicators of insensibility must be present to confirm insensibility. However, these indicators may be different for different stunning or euthanasia techniques (physical, electrical and gaseous/ modified atmospheric). The available indicators to assess the state of sensibility in relation to brain functions are discussed below.

2.3.1. Brain stem and spinal reflexes

Brain stem reflexes are regulated by 12 pairs of cranial nerves that are not under cortical control.

The presence of central reflexes indicates functioning of the brain stem and the spinal cord, and thus, sensibility. Insensibility or complete loss of brain function is assessed based on the absence of brain stem reflexes (Sandercock et al., 2014; Martin et al., 2016; Erasmus et al., 2010a;

Verhoeven et al., 2015; Terlouw et al., 2016 a,b). Authors often used pupillary light reflex, corneal reflex, and palpebral reflex to assess insensibility (Erasmus et al., 2010a; Sandercock et al., 2014; Woolcott et al., 2018). Absence of these ocular reflexes indicates loss of sensibility.

The pupillary light reflex is tested by flashing a light into the eye and checking the pupil for constriction. The pupillary reflex is controlled by the cranial nerve II (optic) and cranial nerve III

(oculomortor). The integration center of the pupillary reflex is located in the midbrain, near the reticular formation. The presence of fixed dilated pupils is considered an important criterion to evaluate the prognosis of comatose patients (Thomas, 2000). A well-functioning retina is necessary for the activation of the pupillary reflex. The pupillary reflex is not considered a reliable reflex of assessing sensibility during exsanguination because the blood loss can reduce the functioning of the retina (Blackman et al., 1986). Blood covering the cornea or a damaged eye may also reduce the feasibility of the test. Therefore, some limitations occur in assessing pupillary reflex as an indicator of sensibility.

The other two ocular reflexes used to assess sensibility are the corneal and palpebral reflexes.

Both the corneal and palpebral reflexes are controlled by afferent cranial nerve V (trigeminal) and efferent cranial nerve VII (facial) (Adams and Sheridan, 2008). The corneal reflex is tested by lightly touching the cornea. This sensory information passes through the trigeminal nerve to

27 reach the trigeminal nucleus which is located next to the reticular formation (Cruccu and

Deuschl, 2000).The trigeminal nucleus is connected to the facial nucleus to stimulate the facial motor nerve. Due to motor responses of the facial nerve, which innervates the orbicularis oculi muscle, the eye lid closes. The connections between the trigeminal and facial nuclei pass through the reticular formation (Aramideh and Visser, 2002). Any interruption of this neural circuit will cause the modification or absence of the reflex. If the corneal reflex is absent, there is a probability of dysfunction of the reticular formation, because the neural circuit of the corneal reflex crosses the reticular formation (Zerari-Mailly et al., 2003). Therefore, the absence of corneal reflex is considered as a reliable indicator of insensibility (Terlouw et al., 2016).The palpebral reflex is also tested by tapping the edge of the eye lid, which results in blinking. The neural circuit of the palpebral reflex is similar to that of the corneal reflex. The nictitating membrane (third eyelid) is a transparent membrane present in many birds, reptiles, amphibians, fish, and some mammals such as rabbits. This membrane also closes in response to touching the cornea. The neuronal circuit is similar to that described for the corneal reflex (Desmond et al.,

1983).

These brain stem reflexes provide practical means for assessing insensibility in field conditions.

However, brain stem reflexes are difficult to assess in closed or damaged eyes. In a review,

Erasmus et al. (2010a) reported that loss of pupillary light reflex is an indicator of complete insensibility. However, the relationships between ocular reflexes and brain function are not always clear. For example, Anil (1991) studied the effects of head-to-back electrical stunning on the incidence of post-stunning reflex activity and cortical evoked responses in sheep and reported that the corneal reflex and respiratory gasps were present in 10 out of 12 sheep which did not

28 show concomitant visual and somatosensory evoked responses following a stun. Moreover, one study with poultry showed that EEG activity ceased prior to the loss of brain stem reflexes

(Gregory and Wotton, 1990a). Sandercock et al. (2014) studied the presence or absence of brain stem reflexes in three awake stages induced by sevoflurane, and at death, which was induced by overdose of barbiturate in hens and turkeys. Though many of the reflexes disappeared when birds were insensible, the nictitating membrane reflex were present even after brain death had been confirmed by EEG based on isoelectric wave forms. The authors suggested that this may be an artifact of the birds being anesthetized prior to death. Therefore, positive eye reflexes alone cannot be considered as an indicator of sensibility in some cases.

The pedal or withdrawal reflex involves activation of nociceptors, and is used to assess the state of insensibility (Erasmus et al., 2010a; Verhoeven et al., 2015a) . The pedal reflex is indicated by withdrawal of the foot in response to pressure applied to the toes. Nose and ear prick, which cause head withdrawal, are also used to assess sensibility. However, withdrawal reflex is difficult to assess, especially following physical killing methods where severe body convulsions occur

(Tidswell et al., 1987).

2.3.2. Physiologic measures and behaviours

The animals’ loss of standing posture or inability to remain in an initial standing or sitting position are used as indicators of the potential loss of sensibility (Terlouw et al., 2016). Loss of standing posture is often the first sign of successful stunning, and indicates inability of the cerebral cortex to control posture (Llonch et al., 2012; Raj et al., 1992) and damage to the reticular formation (Scheepens, 2004). Both mechanical and electrical stunning should lead to

29 immediate collapse (AVMA, 2013). However, severing of the upper spinal cord also causes paralysis of the animal without causing brain damage. Therefore, this indicator should be interpreted with caution (Terlouw et al., 2016). Anoxia in the brain following bleeding of non- stunned animals results in progressive loss of standing, due to dysfunction of cortical structures

(Terlouw et al., 2016). Further, loss of posture was suggested as a sign of early stages of insensibility (Gibson et al., 2015a; Gregory et al., 2010).

Cessation of rhythmic breathing is another indicator of loss of sensibility or death (Erasmus et al., 2010a; Verhoeven et al., 2015). The control centers of the respiratory muscles are located in the medulla oblongata which is located in the lower area of the brain stem (Siegel and Sapru,

2006). Inspiration and expirations are controlled by the different groups of neurons, located in the control center (Siegel and Sapru, 2006). These neurons are stimulated by the reticular formation that receives information from the periphery and higher brain centers (Siegel and

Sapru, 2006). It is reported that captive bolt caused axonal injuries to the brain stem and rhythmic breathing was immediately abolished after an effective shot (Finnie et al., 2000).

Grandin (2013) suggested that the presence of rhythmic breathing after stunning is an indication to proceed with a second stunning. The presence of rhythmic breathing after stunning is generally accepted to indicate that an animal may not be fully insensible. Regaining rhythmic breathing is considered as one of the first signs of recovery of sensibility after CO2 and electrical stunning in poultry (Anastasov and Wotton, 2012).

Normal rhythmic breathing is known as eupnea. Eupnea is controlled by cranial and spinal nerves (St-John et al., 2004). If eupnea fails, the second pattern of breathing which is called as

30 gasping occurs. Gasping is characterized by irregular unorganized breathing which can be induced by ischemia or hypoxia (St-John, 2009). Gasping is accompanied by guttural sounds in many cases that can be confused with vocalizations. Progressive changing in rhythmic breathing to irregular breathing is reported following bleeding of non-stunned animals (Verhoeven et al.,

2015). Gasping has been observed in successful nonpenetrating captive bolt stunning of turkeys

(Erasmus et al., 2010a) and calves slaughtered by bilateral severance of the common carotid arteries and jugular veins (Blackmore et al., 1983). Terlouw et al., (2016a) suggested that damage to the reticular formation or the medulla could be the reason for cessation of breathing in mechanical stunning.

Convulsions (neuromuscular spams) are uncontrolled or involuntary muscle contractions of the body which occur in electrical, mechanical and gas-stunned animals (Prinz et al., 2010; Erasmus et al., 2010a; Raj et al., 1992). In general, neuromuscular spasms consist of clonic and tonic phases. In birds, the clonic phase is characterized by severe wing flapping behaviour (Prinz et al.,

2010), and the tonic phase consists of extension or outstretching of legs and wings and an arched neck (Terlouw et al., 2016b). Body movements are controlled by an integration mechanism which involves the cerebral cortex, cerebellum, brain stem, spinal cord and the skeletal muscle

(Vander et al., 2001).

Some studies suggested that neuromuscular spams are incompatible with sensibility due to the absence of higher motor control. Convulsions were reported in insensible birds during electrical stunning (Cook et al., 1995) and gas stunning (Raj et al., 1992). In contrast, neuromuscular spams occurred before insensibility in controlled atmosphere stunned broilers (Coenen et al.,

2009). Gerritzen et al.(2004) observed convulsions just after loss of posture in broilers stunned by 100% CO2, and just before loss of posture when broiler chickens were stunned using two other gas mixtures (50% N2 +50% CO2, and 30% O2 + 40% CO2 + 30% N2). Based on the results, the authors suggested that it is not always possible to distinguish between neuromuscular spasms that are incompatible with sensibility and other involuntary muscle contractions

(myoclonic jerks). Moreover, Erasmus et al. (2010a) reported violent wing-flapping began immediately after cervical crushing and cervical dislocation in turkeys, although palpebral and corneal reflexes persisted for some time. They also reported that brain stem reflexes were abolished immediately after blunt trauma and nonpenetrating captive bolt application while wing flapping persisted. However, reflexes returned in some birds even when the wing flapping was observed. Therefore, onset of convulsions cannot be used as an indicator of detecting insensibility.

Dawson et al. (2007) reported that cessation of convulsions (at the end of tonic phase) as measured by the accelerometer, occurs at about the same time as brain death in broiler chickens killed by manual cervical dislocation. Further, the authors reported that cardiac relaxation typically occurs after the convulsive phase. Based on the study of gas stunning of chickens by

Raj et al. (1992), brain death occurs at or shortly after the end of the convulsive phase. Erasmus et al. (2010a) reported that “no reflexes or breathing returned after tonic spasms had completely ceased” in turkeys killed with nonpenetrating captive bolts and blunt force trauma. Therefore, they suggested that cessation of final tonic convulsions can be used as a criteria of irreversible brain death following physical killing methods.

Jaw tone and neck tone are considered as indicators of sensibility. Jaw tone is the resistance that occurs when a force applies downward towards the lower jaw/beak. Neck tone is the ability of the neck to hold the head upright. Both jaw tone and neck tone are controlled by cranial nerves, and loss of neck tone and jaw tone indicate dysfunction of the brain stem (Erasmus et al., 2010a;

Martin et al., 2016). However, jaw relaxation cannot be used as a single indicator to assess insensibility and can support other indicators of insensibility (Grandin, 2002; Gregory, 2007).

Cloacal reflex is also controlled by cranial nerves, and can be used as an indicator of brain death

(Martin et al., 2016). The external sphincter muscle of the vent contracts causing sporadic opening and closing of the cloaca sphincter.

2.4. Euthanasia methods and mode of action

2.4.1. Physical methods

2.4.1.1. Blunt force trauma to head

Blunt trauma is defined as “the result of the impact of a body against a blunt surface, the impact of an object with a blunt surface against a body, or a combination of both” (Merck, 2008). The amount of force applied to an object depends on its acceleration in the space and the mass, according to Newton’s second law: Force = mass × acceleration. The energy that the object acquires and retains as long as it is moving is defined as kinetic energy. This energy is the force transferred to another body when blunt trauma occurs: Kinetic energy = ½ mass × velocity. Light objects that are accelerated to high velocity or heavy objects that impact slowly can deliver the same kinetic energy. Therefore, the ability of an object to induce mild or severe damage depends on its weight and velocity. The brain is a delicate organ which is surrounded by three meninges and encased in calvarium. The calvarium and the cerebrospinal fluid layer helps to protect the

33 brain from blunt trauma. When a blow hits a skull, most of the tissues that are damaged develop more severe hemorrhages due to shearing and tension applied to the meningeal vessels at the site of direct contact, and there may be contusions of soft tissues and fracture of the skull (Cors et al.,

2015).

A blow to the head is accepted as a humane euthanasia method for neonatal animals with a thin cranium such as piglets (AVMA, 2013). The force delivered to the head should be able to cause immediate insensibility in animals. A blow to the head needs to be performed by a well-trained person who is aware of its esthetic implications (AVMA, 2013). European Commission (2009) stated that this method can be applied to birds up to 5 kg in conjunction with cervical dislocation, should not be used as a routine method, and should not be used by one person on any more than

70 animals per day.

Erasmus et al. (2010a) assessed the effectiveness of blunt force trauma in turkeys and found that the method induced immediate insensibility leading to death. The authors further reported that there was a difference across the age groups in failure rate, where 22% of turkey broilers (3.9 kg) and 3% of turkey toms (13.1 kg) needed reapplication. Whiting et al. (2011) suggested the manual blunt force trauma caused severe and fatal skull fractures in weaned piglets which is sufficient to render them immediately insensible. However, the authors reported that there was a

24% failure rate and was esthetically displeasing to personnel. These results suggest that blunt force trauma has a potential to induce immediate insensibility if correct force and placement on the skull has been used, but there is a high degree of risk of failure.

2.4.1.2. Cervical dislocation

Cervical dislocation is a common method for killing individual birds in the context of commercial or non-commercial practice. EC 1099/2009 (The Council of the European Union,

2009) defines cervical dislocation as stretching and twisting of neck causing cerebral ischemia

(deprivation of glucose and oxygen supply to the brain tissues). In cervical dislocation, the skull is separated from the vertebral column while severing the spinal cord and main blood vessels supplying to the brain (Erasmus et al., 2010; Bader et al., 2014). Based on welfare concerns, crushing of the vertebrae must not occur as a result of cervical dislocation (AVMA, 2013). Ideal cervical dislocation is in between the occipital lobe of the skull and the first cervical bone

(Atlas); this causes extensive damage to the brain stem which renders the birds instantaneously insensible due to the concussive effect that dislocation has on the brain from stretching and severing of the spinal cord (Sparrey et al., 2014; Bader et al., 2014). Manual cervical dislocation is performed using only the hands to cause cervical dislocation. However, cervical dislocation must be performed by a skilled and confident operator (AVMA.2013).

Manual cervical dislocation may be esthetically displeasing to the operator (AVMA, 2007).

Additionally, operator fatigue may lead to compromised bird welfare when a large number of birds are being killed. However, Martin et al. (2018) found no evidence of reduced performance over time when stock people applied manual cervical dislocation to 100 birds.

Manual cervical dislocation is more difficult in mature birds than in young birds due to their increased neck muscle tone. To address the limitations around manual cervical dislocation, different tools have been developed to perform cervical dislocation mechanically. Some devices

(e.g. killing cone and heavy stick) involve stretching, whereas others attempt to dislocate the cervical vertebrae by forcing a blunted edge between two vertebrae (e.g. pliers) without stretching or twisting the neck (Sparrey et al., 2014). The ultimate result should be the same as in manual cervical dislocation (severing of spinal cord and blood vessels). However, some mechanical cervical dislocation devices are being used without being scientifically tested for their welfare impact.

Both manual and mechanical cervical dislocation methods have been questioned for humaneness based on evidence that loss of sensibility is not immediate in poultry species (Gregory and

Wotton, 1990a; Sparrey et al., 2014). Erasmus et al. (2010) found that nictitating membrane reflex persisted in turkey hens killed with cervical crushing using a burdizzo and all broiler turkeys killed with manual cervical dislocation.

Several studies compared the efficacy of manual cervical dislocation to mechanical cervical dislocation in poultry. Gregory and Wotton (1990a) studied cervical dislocation by manual stretching and cervical dislocation by crushing (using Semark neck pliers) on chickens. The authors reported that birds killed by using neck pliers exhibited longer time to loss of visual evoked responses compared to those killed by manual cervical dislocation, and concluded that cervical dislocation by stretching was more effective than cervical dislocation by crushing.

Martin et al. (2017) assessed a modified version of pliers on cadavers and the authors determined that the device had poor potential to be effective given that fewer than half of the birds’ vertebrae were dislocated while more than half had damage to trachea and vertebrae. Martin et al. (2016) evaluated the kill efficacy of two novel mechanical poultry killing devices, the Modified Rabbit

Zinger and a novel mechanical cervical dislocation gloved device (NMCD) in comparison with manual cervical dislocation. The authors reported that NMCD was more effective than manual cervical dislocation at killing layers and broilers of various ages (100% killing rate); birds were rapidly rendered insensible due to extensive trauma to the brainstem and/or spinal cord and died from cerebral ischemia due to severing of carotid arteries. The commercially available Koechner euthanasia device (KED) was evaluated in awake broilers (Jacobs et al., 2019) and in anesthetized turkeys (Woolcott et al., 2018b) and both papers reported shorter latency to loss of brain stem reflexes indicative of insensibility in birds successfully killed with manual cervical dislocation in comparison to birds killed by the KED device.

2.4.1.3. Decapitation

In decapitation, the head of the animal is completely severed from the body by using a sharp blade in one action with rapid blood loss to leading to cerebral schemia resulting insensibility leading to death (Mason et al., 2009). This technique is commonly practiced in small laboratory animals to obtain brain tissues and fluid that are not contaminated with euthanizing chemicals/agents like gases, anesthetics, and electric currents. AVMA (2013) guidelines state that for decapitation to be acceptable, the blade must be a sharp one to ensure clean and rapid cut at once. Derr (1991) suggested decapitation is a powerful arousal stimulus based on the evidence of EEG in rat brains. Cutting of the throat of cattle without stunning results in pain associated with the cut due to stimulation of nociceptors located in the neck tissue (Gibson et al., 2009;

Zulkifli et al., 2012) and distress associated with delays in the time to onset of insensibility and with aspiration of blood into the respiratory track (Gregory et al., 2010). In decapitation,

37 nociceptors in the neck tissue are stimulated as it damages to the skin and other tissues.

Therefore, a rapid cut with a sharp blade is essential in decapitation.

2.4.1.4. Captive bolts

Penetrating captive bolts are designed with a retractable steel bolt which penetrates the cranium and enters the brain while firing. Direct brain damage caused by the bolt and the cerebral concussion resulting from the kinetic energy delivered to the head are sufficient to cause the animal to become insensible (Blackmore, 1983; Gregory et al., 2007). Kinetic energy (KE) is proportional to the mass (m) of the bolt and its velocity (v), where the formula is KE = ½ mv2.

The velocity of the moving object has a greater effect on its kinetic energy than the mass.

Therefore, if the velocity of the bolt is reduced, its ability to effectively stun will be impaired.

2.4.1.4.1. Penetrating captive bolts

Captive bolts are commonly powered by pressurized air, cartridge, spring or elastic. Penetrating captive bolts (PCB) are recommended as a euthanasia method for horses, ruminants, swine and large poultry (turkeys, broiler breeders, ratites, waterfowl) with conditions (AVMA, 2013). The conditions are personnel skill and experience, operator safety, correct position of the device and it is also recommended that the animal is immediately exsanguinated or pithed after applying the

PCB device. However, many authors reported incomplete concussion or return to sensibility in livestock after using PCB (Gregory et al., 2007; Gibson et al., 2012; Atkinson et al., 2013).

Sharp et al. (2014) reported that the wound caused by spring powered PCB was the full length of the brain, from the cerebellum into the brain stem in joey cadavers. Authors suggested that though the spring powered PCB can cause damage to a large area of the brain in joey cadavers, it

38 is not reliable for induction of insensibility and death. Gibson et al. (2015b) studied the mechanical factors related to the performance of PCB in cattle and concluded that appropriate

PCB/cartridge combination, the kinetic energy delivered to the head of the animal, bolt penetration depth, and species/animal type must be considered. Other than these mechanical factors, skull thickness is also an important factor which affects the bolt penetration depth

(Gibson et al., 2015b). Therefore, correct placement of the bolt on the head is essential in order to deliver sufficient force to cause brain damage which helps to initiate effective stun in the animal. The best place for the position of the device is where the skull is at its thinnest and the brain is closest to the surface of the head. Raj and O’Callaghan (2001) reported that the penetrative captive bolt should be fired perpendicularly to the skull to achieve effective stunning in broilers. The authors explained that deviations of more than 20° from the right angle caused failures in stunning.

2.4.1.4.2. Non-penetrating captive bolts

The non-penetrating captive bolts (NPCB) are designed with a flat, round or mushroom shaped head (blunt bolt) which does not penetrate the brain. Therefore, although NPCB may be used to stun large mammals such as cattle, slaughter weight pigs and adult sheep, it is not recommended as a sole method of euthanasia (AVMA, 2013). Nonpenetrating captive bolts have the advantage of consistently administering the same force to the skull independent of the strength of the operator in comparison to manual blunt force trauma. The correct positioning of the device on the head of the animal is important for an effective stunning. Finnie et al. (2003) tested a cartridge powered NPCB (model MKL Karl Schermer and Co, Karlsruhe, Germany) on 4-5 weeks old lambs and 7-8 weeks old pigs. The authors reported minimal axonal injury, mild basal subarachnoid hemorrhage and no contusional injury in pigs, and more severe axonal injuries,

39 complicated contusions and subarachnoid and intraparenchymal hemorrhage in lambs.

Moreover, they explained that shape of the skull and thickness were the critical factors to cause less damage in pigs than in lambs. Widowski et al. (2008) compared a prototype Zephyr

(modified pneumatic nail gun with a round bolt head, 120 psi) with manual blunt trauma on low viability piglets (less than 24 hrs old) and concluded that manual blunt trauma was more effective and humane than the non-penetrating captive bolt because piglets showed signs of returning to sensibility with the Zephyr. However, Casey-Trott et al. (2013) studied the same

NPCB modified with conical nylon bolt head (115-120 psi) on neonatal piglets (less than 72 hr old, < 3 kg) and concluded that the modified design of Zephyr-E reliably caused immediate, sustained insensibility followed by death (in a single step) in 94% of neonatal piglets. Zephyr-E was also effective on 3-9 weeks old piglets killing 99% in single discharge (Casey-Trott et al.,

2014).

Erasmus et al. (2010a) compared the same model of Zephyr-E as used in Casey-Trott et al.

(2013, 2014) to mechanical cervical dislocation, manual cervical dislocation and blunt force trauma on turkeys. Their results revealed that Zephyr-E and blunt force trauma were similarly effective in terms of time to loss of sensibility and the end time of convulsions in large turkeys.

Moreover, they concluded that the prototype Zephyr-E was more consistent than blunt trauma at causing insensibility leading to death in small turkeys, and the Zephyr-E appeared to be a humane on-farm killing method for cull turkeys.

The Zephyr is now commercially available from Bock Industries in several variations (Zephyr-E and Zephyr EXL). The commercially available models of Zephyr-EXL and Turkey Euthanasia

Device (TED) were evaluated in turkeys and shown to be highly effective based on rapid, almost immediate, loss of brain stem reflexes and loss of EEG activity (Gibson et al., 2018; Woolcott et al., 2018a).

2.4.2. Inhaled gas

Carbon dioxide is popular for euthanasia of laboratory rodents because of its relative safety, low cost, ease of use and ability to be used as a mass euthanasia agent (Ambrose et al., 2000). The

AVMA (2013) guidelines state that CO2 should be administered by gradual-fill using a flow rate between 10-30% chamber volume per minute (vol/min), until the animal is insensible, then the flow rate can be increased to reduce the time to death. At room temperature and pressure, CO2 is odorless, colorless, nonﬂammable and heavier than air gas. CO2 is dissolved in water according

+ - to the following chemical reaction: CO2 + H2O ↔ H2CO3 ↔ H + HCO3 . This reaction can

- prevent the acid-base balance in the body: pH = pK + log [HCO3 / CO2] (Henderson-Hasselbach equation). Cellular respiration produces CO2 as a waste in animal’s metabolic tissues. CO2 diffuses into the blood capillaries from the metabolic cells and transported via blood either bound to hemoglobin or dissolved as CO2, carbonic acid or bicarbonate iron (Baggott, 1982). A minor amount of CO2 binds to plasma proteins to form carbamino compounds (Guais et al., 2011).

2.4.2.1. Hypercapnia and its influence

The partial pressure of CO2 (PCO2) inside the pulmonary capillary blood under normal condition

(7% or 46 mmHg) is always higher than that in the alveolar air (6% or 40 mmHg), which creates a concentration gradient between blood and air in the alveoli. Therefore, CO2 freely diffuses via the alveolar membrane and is released from the lungs (Guais et al., 2011). The toxic effect of

CO2 can occur rapidly due to its free diffusion and is mainly associated with blood pH, heart and central nervous system (Guais et al., 2011). Excess CO2 in the blood (hypercapnia) induces the

CO2 tension in arterial blood (PaCO2) causing acute or chronic respiratory acidosis (Guais et al.,

2011). Detection of blood CO2, O2 and pH are critical for maintaining spontaneous breathing

(Feldman et al., 2003). Although the O2 sensitive chemoreceptors are located peripheral to the brain (peripheral chemo receptors), CO2/pH sensitive receptor sites are mainly in the brain

(central chemoreceptors). Central chemoreceptor sites contain chemosensitive neurons, and are highly sensitive to CO2. In humans, a 1 mmHg increase in PCO2 increases ventilation rates by 20 to 30% (Feldman et al., 2003).

CO2 permeates the blood brain barrier and decreases the pH in cerebral spinal fluid (CSF) due to increases in cellular bicarbonate and hydrogen ions (Martoft et al., 2003). CSF contains fewer buffering components than in blood. Therefore, pH is decreases more rapidly in CSF than in arterial blood (Guais et al., 2011). The decreased pH in CSF causes decreased nerve cell function and cerebral electrical activity inducing anesthesia and analgesia (Lee et al., 1996). Some authors reported mild effects to the nervous system in humans due to slightly higher concentrations of

CO2: visual impairment due to 1% of CO2 and headache due to 2% of CO2 (Jiang et al., 2005); decrease in stereo acuity (Sun et al., 1996) and decrease in the ability to detect motion in more than 2.5% of CO2 (Yang et al., 1997). Hypercapnia also affects cardiovascular functions. A level of 5% CO2 in air caused increased heart rate, arterial pressure and peripheral vasodilation in humans (Guais et al., 2011). Ventricular failure in dogs has been reported due to hypercapnia

(Stinson and Mattsson, 1970).

2.4.2.2. Pain and distress associated with CO2

Distress is defined as “a negative state that develops when an organism is unable to adapt to a stressor” (Moberg, 2000). There are many ways that CO2 can cause distress in an animal. CO2 reacts with the water molecule in the nasal mucous membranes and forms carbonic acid, which stimulates trigeminal nociceptors and causes pain ( Leach et al., 2002; McKeegan, 2004). CO2 can also activate the nociceptors in the conjunctiva and cornea causing the sensation of burning and stinging in humans (Feng and Simpson, 2003).

Different species detect different gasses in different concentrations in different ways. In general, the olfactory threshold may not be aversive or cause pain. Humans can detect 10% CO2 by smell and reported that this level was not pungent to inhale (Patterson et al., 1962). Moreover, inhalation of 7% CO2 did not induce respiratory symptoms or irritation in humans (Patterson et al., 1962). When humans are exposed to high CO2 concentrations breathlessness is induced and suggested that it could be pungent (as cited in Gregory and Wotton, 1990b). Hypercapnia and hypoxaemia results in air hunger in humans (as reviewed in Beausoleil and Mellor, 2015) and signs of asphyxia and behavioral excitation have been observed because of both hypercapnia and hypoxia (Forslid et al., 1986). CO2 has been identified to be painful to nasal mucosa when the concentration is more than 65% (Hari et al., 1997). Danneman et al., (1997) also reported that increasing CO2 was highly unpleasant at 50% and painful at 100% based in a human study. Pain threshold level for CO2 in humans has been identified as concentrations of between 40% and

55% (Anton et al., 1992).

Evidence exists that that chickens can also detect CO2 at lower concentrations that are not necessarily aversive to them. For example, Raj and Gregory (1991) assessed hens’ behavioural responses to feeding in chambers at 5%, 7.5%, and 10% CO2 concentrations. Authors observed that when given a free choice, chickens avoided feeding at 7.5% and 10% CO2. However, when a dominant hen occupied the chamber at the lower concentration, subordinate birds fed at the 7.5% level. Dominant birds detected the levels and avoided feeding at these levels. Subordinate birds selected to avoid the dominants birds and feed in CO2. Authors suggested that chickens could detect 7.5% CO2 but that is was not aversive, or at least less aversive than the social pressure.

McKeegan et al., (2005) studied behavioural responses that may indicate detection of and aversion to CO2 in chickens. Hens were exposed to short pulses of gas and their responses were observed to indicate detection or aversion. Authors concluded hens detected CO2 at 10% as indicated by mandibulation, orienting toward and interruption of ongoing behaviours, and displayed head shaking at 20%. But authors further explained that behaviour threshholds were not definitive measures of detection ability. Nasal and oral mucosa of the chicken contains trigeminal polymodal nociceptors that can detect the presence and concentration of potentially noxious chemical stimuli (ammonia, CO2 and acetic acid vapors) (McKeegan et al., 2004).

However, there is no olfactory response to CO2. As mentioned previously, the avian nociceptive threshold is in the region of 40 - 50% CO2 (McKeegan et al., 2004), which is in the same range as the electrophysiological thresholds reported for rats and humans (Anton et al., 1991, 1992).

Several behaviour studies have been conducted to assess the aversion to CO2 in poultry species.

Exposure to CO2 could cause distress to poultry because it may result in a sense of breathlessness

44 as well as causing pain. Raj (1996) reported behaviours of head shaking, gasping and vocalization at 72% CO2 in turkeys. Gerritzen et al. (2000) suggested that gasping and head shaking in broiler chickens exposed to 60% CO2, and gas mixtures containing 30% and 40% CO2 were possible indicators of breathlessness and the pungency of gas mixtures. Webster and

Fletcher (2001) also observed deep breathing (breathing through the mouth) followed by head shaking in layer chickens and broilers exposed to different CO2 concentrations (30%, 45%,

60%). These authors suggested that head shaking was an alerting response during early induction of anesthesia due to CO2. McKeegan et al., (2005) reported respiratory disruption (apparent increased inhalation depth and duration) in broiler chickens exposed CO2 concentrations of 10,

25, 40, 55, 70%) for 10 s. Heistand and Randall (1941) found that 6% CO2 caused slight inhibition of breathing, whereas 10% CO2 caused increased depth of respiration in chickens.

Therefore, low levels of CO2 such as 10 % may result in increased respiratory effort but levels over 40% levels are probably painful, possibly with increased respiratory effort combined.

2.5. Motivation, preference and aversion tests

2.5.1. Introduction

Growing interest in the ethical issues of animal care has led to a focus on the subjective feelings of animals. Thus, during the past 30 years, welfare scientists have investigated numerous scientific methods to understand the subjective states of animals (Dawkins, 1980). Most scientists started to describe the importance of affective state in animal welfare (Fraser et al.,

1997; Broom, 1998) while some scientists stated that affective state is the sole criterion of welfare (Duncan, 1996). Similarities across vertebrate species in their neuroanatomy and physiological response to stimuli suggest a capacity for feelings in non-human animals

(Dawkins, 1980). However, the direct observation of an animal’s feelings is not possible.

Cabanac (1979) stated that the correlation between reported feelings and behavioural responses to stimuli in humans suggests that behavioural and physiological responses reflect the strength of associated feelings.

Animals’ motivation is defined in several ways by animal welfare scientists. Simply,

‘motivation is the strength or willingness with which an animal engages in behaviour’ (Toates,

1986). In general, people associate motivation with feelings. For example, motivaton to drink is often described as the feeling of being thirsty. Fraser and Duncan (1998) introduced the term of motivational affective state (MAS) to refer to subjective affective states. They considered pleasant or unpleasant experiences as the positive or negative affects leading to motivating specific types of behaviours. They included as negative MASs states such as hunger and fear which presume to play a role in the motivation of eating and escape, respectively, and as positive

MASs the pleasure that may accompany such behaviours as eating, playing and mating. Kirkden and Pajor (2006) defined positive motivation as the desire to consume a commodity or perform a behaviour, and negative motivation as the desire to avoid a painful or frightening stimulus.

Motivation is varied in strength. Studying the strength of motivation is important to determine what animals truly want. If animals are unable to do what they are strongly motivated to do, it can cause poor welfare (Dawkins, 1990). The term preference has been described as a difference between the strength of motivation to obtain or avoid one resource or stimulus and the strength of motivation to obtain or avoid another (Kirkden and Pajor, 2006).

2.5.2. Approach avoidance and conditioned place avoidance paradigms

Approach avoidance paradigms are used to assess attraction or aversion to a stimulus. Animals have a high motivation to reach food, and are capable of making the association of a specific stimulus with food. When animals are rewarded with food inside an unconditioned environment, they will start to associate that environment with food and the unconditioned environment will become a conditioned stimulus. Once animals consistently choose to be in the conditioned environment, it is possible to apply a different stimulus in the same environment and determine if exposure to the new stimulus is neutral (the animal still enters the environment) or aversive (the animal refuses to enter the environment).

Scientists have used approach-avoidance paradigms to evaluate the aversiveness to CO2 by incorporating CO2 in an environment which has been conditioned with food. Webster and

Fletcher (2004) studied aversion of hens to different gas atmospheres by using an approach avoidance paradigm. Their test apparatus consisted of two plywood chambers. One chamber was elevated 30 cm above the other and connected by a corridor which was constructed of plywood.

Gas was injected into the lower chamber where hens associated with feed. Hens were feed deprived to motivate them to approach food placed in the atmospheres. The number of hens that approached and entered the lower chamber in different atmospheric conditions was observed to detect the aversiveness of different concentrations of CO2. The authors concluded that there was no welfare disadvantage to the use of up to 60% CO2.

Sandilands et al. (2011) examined the aversion of broiler chicks to three lethal gas mixtures at various concentrations. In this study chicks were allowed to place their heads inside three

47 feeding and drinking stations (FDS) in order to access food and water. Each FDS was then filled with a different gas mixture. Aversion was assessed based on the time birds spent with their head in each FDS (with more time indicating less aversion). The authors concluded that 60% CO2 was highly aversive in comparison to 50% and 55% CO2 concentrations. Other studies also support the finding that chickens given a free choice avoid a feeding chamber containing CO2. Raj and

Gregory (1991) studied 3 different CO2 concentrations (5%, 7.5%, 10%), and reported that hens avoid the feeding chamber when the CO2 concentration was increased up to 5%. McKeegan et al.

(2006) assessed the time to cessation of feeding and withdrawal from the feeding chamber to detect the degree of aversion in broiler chickens exposed to different CO2 concentrations (10%,

25%, 40%, 55%, 70%) for 10 s, and reported that chickens exhibited aversion to 40% and above

CO2 concentrations.

Conditioned place preference (CPP) or conditioned place aversion (CPA) paradigms are potentially useful tools in animal welfare assessment, and mainly based on the principles of classical (Pavlovian) conditioning (Pavlov, 1927). These paradigms measure the reinforcing properties of a stimulus in the absence of the stimulus itself. Conditioning is done by pairing a distinct set of environmental cues (usually in a location or a place) with a particular treatment, and a distinctly different set of environmental cues (usually in an adjacent location or place) is paired with a control treatment. Then, treatment and control pairings with location are repeated several times so that the animal learns to associate the respective reinforcements or feelings elicited by the treatments with the locations. In a post-conditioning test trial, the animal will approach the environmental cues (location) that had positive or rewarding effects or avoid the

48 location when the treatment had negative or aversive effects, resulting in either a conditioned place preference or a conditioned place aversion for the treatment-paired environment.

Conditioned place aversion has been used to study the subjective experiences of animals exposed to different gas concentrations. Ramsaya et al. (2003) studied the rewarding/aversive effects of nitrous oxide (N2O) using the conditioning place paradigm in male Long–Evans rats. After exposure to gas for 8 consecutive days, a conditioned place aversion was found in rats exposed to 30% and 60% N2O. Place aversions were demonstrated during a 20-min test session on day 9 when placebo gas was delivered. Withrock (2015) and KC et al. (2016) studied aversion to three different CO2 concentrations in goats and pigs, respectively, by using approach avoidance and conditioned place avoidance paradigms. However, they concluded that aversion was not sufficient to provoke a conditioned place avoidance in their experiments.

2.6. Thesis objectives

This thesis explores the efficacies of three different non-penetrating captive bolt devices

(Zephyr-E, Zephyr-EXL, and Turkey Euthanasia Device [TED]) (Bock-Industries.com) and a novel mechanical cervical dislocation device (Koechner Euthanasia Device [KED]) (Koechner

MFG. CO., INC) as on-farm euthanasia techniques in layer chickens. Replication of previous studies on assessment of aversiveness to CO2 in chicken using different testing protocols and assessing different CO2 concentrations as possible euthanasia agents are important to refine euthanasia protocols. Thus, aversion and time to loss of sensibility for four different CO2 concentrations (25%, 35%, 50%, and 70%) in layer chickens were studied.

Objective 1: To assess the efficacies of the Zephyr-E, Zephyr-EXL and TED as on-farm euthanasia methods for layer chickens.

-To assess the efficacies of Zephyr-E, Zephyr-EXL and TED for different production

stages (rearing, growing, laying and end of production) of layer chickens based on

measures of time to insensibility and death using eye reflexes and behavioural measures.

-To assess the efficacies of Zephyr-E, Zephyr-EXL and TED for different production

stages (rearing, growing, laying and end of production) of layer chicken based on post-

mortem assessment of traumatic brain injury.

-To identify behavioural and physiological indicators to confirm the death in layer

chickens before carcass disposal.

Objective 2: To assess the efficacies of mechanical cervical dislocation using Koechner

Euthanasia Device (KED model C) for three different age groups (12, 27-29, and 65-70 weeks old) of layer chickens in comparison to manual cervical dislocation based on brain stem reflexes, behavioural responses, physiological parameters, and post mortem pathology.

Due to ethical concerns, to avoid possibility of any pain associated with a novel killing method, the KED was tested in anesthetized birds. Thus, the following objective also arose:

To understand the effect of anesthesia on brain stem reflexes, behavioural responses, and physiological parameters by comparing awake versus anesthetized birds killed by manual cervical dislocation in order to determine which measures would be valid for use in anesthetized birds.

Objective 3: Assessing CO2 as a humane killing method for poultry

- To assess the aversion for different CO2 concentrations (25%, 35%, 50%, and 70%) in

layer chickens (25- 27 weeks old) based on approach avoidance test and conditioned

place avoidance test.

- To assess the time to insensibility for different CO2 concentrations (25%, 35%, 50%, and

70%) in layer chickens (25- 27 weeks old)

CHAPTER 3

Anatomical Pathology, and Behavioural and Physiological

Responses Induced by Application of Non-Penetrating Captive Bolt

Devices in Layer Chickens

This manuscript was published in Frontiers of Veterinary Science:

Bandara RMAS, Torrey S, Turner PV, Schwean-Lardner K and Widowski TM (2019). Anatomical Pathology, Behavioral, and Physiological Responses Induced by Application of Non-penetrating Captive Bolt Devices in Layer Chickens. Front. Vet. Sci. 6:89. doi: 10.3389/fvets.2019.

3.1. Abstract

This chapter evaluated three models of non-penetrating captive bolt devices, Zephyr-E, Zephyr-

EXL and Turkey euthanasia device (TED) for time to loss of sensibility and degree of brain damage during euthanasia in four age groups of male and female layer chickens (10-11, 20-21,

30-35, 60-70 weeks respectively). Latencies to onset of insensibility and cardiac arrest were assessed to detect whether killing birds via these devices was humane and effective. Both gross and microscopic pathology evaluations were conducted to score skull and brain trauma post mortem. All three NPCB devices induced loss of breathing, pupillary reflex and nictitating membrane reflex within 5s after application in most chickens. Latencies to loss of jaw tone and neck muscle tone were longer in 60-70 weeks old roosters (P<0.05). Younger birds (10-21 week- old) demonstrated the longest time (P<0.0001) to onset of tonic convulsions, time at last movement, cloacal relaxation and cessation of heart beat. A positive correlation (P<0.0001) was found for all three devices between time of cardiac arrest and time to onset of tonic convulsions, last movement, and cloacal relaxation. More than 80% of birds had skin lacerations with external bleeding following application of all 3 devices. Device type did not affect the incidence of skull fractures but higher skull fracture scores were noted in 10-11 week-old birds compared to other ages. Regardless of device type and age, microscopic SDH was most apparent in the brain and proximal spinal cord of all birds. In summary, all three devices caused significant trauma to the midbrain and spinal cord. Results demonstrated that all three devices induce rapid insensibility after application and can be used as a single-step method that results in a humane death in all age groups of layer chickens.

3.2. Introduction

In the poultry industry, there are several reasons for killing birds during production: to prevent suffering from sickness or injury, for disease control, and for stock management. Therefore, on- farm killing is a routine procedure on commercial poultry farms. Animal care guidelines for livestock and poultry require that the methods used for routine killing cause minimal pain and distress (NFACC, 2016). Moreover, the killing method should result in rapid and irreversible loss of sensibility (or consciousness) to be considered humane (AVMA, 2013).

The most common method for killing poultry on farms is manual cervical dislocation which involves stretching and separating the cervical vertebrae by hand. However, manual cervical dislocation is considered esthetically displeasing to personnel performing it (AVMA, 2013), and there is evidence that both manual and mechanical cervical dislocation methods may not cause immediate unconsciousness (Erasmus et al., 2010a; Woolcott et al., 2018b). Newly designed euthanasia devices are commercially available and, non-penetrating captive bolt devices (NPCB) have been designed with a blunt bolt head that does not penetrate the brain. A NPCB device is commonly used to stun large mammals such as cattle, slaughter weight pigs and adult sheep, but is not recommended as a sole method of euthanasia for large animals, as it may not cause death as a one-step method of euthanasia and another method is not applied animals may return to sensibility (Finnie et al., 2003; AVMA, 2013).

The efficacy of some NPCB devices has been determined for turkeys. Erasmus et al. (2010b) compared the prototype Zephyr-E to mechanical cervical dislocation, manual cervical dislocation and blunt force trauma in turkeys. That study demonstrated that the prototype Zephyr-E device

54 and blunt force trauma were more effective in terms of time to loss of sensibility in turkeys compared to manual and mechanical cervical dislocation. These authors further suggested that a

NPCB device was more consistent than blunt force trauma at causing insensibility and death in small turkeys. Woolcott et al. (2018a) studied two commercial models of NPCB devices:

Zephyr-EXL and Turkey Euthanasia Device (TED) on turkeys at three stages of production and concluded that both devices were highly effective and reliable at inducing immediate insensibility. Gibson et al. (2018) evaluated electroencephalographic (EEG) and behavioural responses of turkeys stunned with three different NPCB devices (Cash Poultry Killer, TED, and

Zephyr EXL), and concluded that all devices were effective in causing insensibility in turkeys, provided they were positioned correctly with the correct power load. Insensibility is often assessed using brain stem reflexes including pupillary light and nictitating membrane reflexes in poultry (Erasmus et al., 2010a; Sandercock et al., 2014; Martin et al., 2016). NPCB devices are used to cause damage to the regions of the brain which control consciousness and vital functions.

Due to the acceleration force of the physical technique, cerebral contusion results with neuronal damage and internal bleeding. Brain pathological lesions associated with hemorrhage were associated with immediate loss of consciousness and indicative of immediate and irreversible loss of central regulation of breathing and heart function in poultry (Erasmus et al., 2010b; Bader et al., 2014; Woolcott et al., 2018a).

There are fewer published scientific evaluations of NPCB devices in layer chickens. Martin et al.

(2018) evaluated a cartridge-powered NPCB device (Accles and Shelvoke Cash Poultry Killer:

Model .22 CPK 200) in layer chickens, and reported 99.1% kill success with the shortest duration to loss of brain stem reflexes compared to manual and mechanical cervical dislocation.

However, the Cash Poultry Killer is cartridge- based, heavier, and more difficult to use compared to other NPCB devices, a significant disadvantage. An evaluation of other lighter weight commercially available pneumatically-powered NPCB devices is needed for layer chickens as an alternative for on-farm euthanasia.

The objective of the present study was to compare the efficacy of the Zephyr-E, Zephyr-EXL, and TED, all pneumatically powered NPCB devices, for on-farm euthanasia of four different ages of layer chickens. Latencies to onset of insensibility and cardiac arrest were assessed as to determine whether the techniques were humane and effective. Gross and microscopic evaluation and scoring of the skull and brain were used to assess induced trauma.

Animal care guidelines require that death be confirmed before leaving birds and disposing of carcasses (AVMA, 2013; NFACC, 2016). Common criteria include lack of breathing, pulse, and cessation of the heart. Auscultation may be used to monitor heart beat but this can be difficult for stock persons in the field as it requires a stethoscope, skill and practice. Therefore, identification of behavioural and physiological reflexes that are correlated with cardiac arrest is needed to readily confirm death before carcass disposal. Therefore, this study also investigated practical behavioural and physiological indicators to confirm the death in layer chickens.

3.3. Materials and methods

The procedures and protocol for this research were reviewed and approved by the University of

Guelph Animal Care Committee (AUP 3321), which holds a Good Animal Practice certificate issued by the Canadian Council on Animal Care.

3.3.1. Animals and facilities

All chickens enrolled in this study were obtained from different research projects that were conducted at the Arkell Poultry Research Station of the University of Guelph. All the birds had been targeted for euthanasia by researchers or staff, because they had reached end of study or because of routine flock depopulation. The Zephyr-E, Zephyr-EXL and TED devices were assessed for killing efficacies and humaneness on four age groups of layer chickens: 10-11 weeks (1.1 ± 0.2 kg), 19-20 weeks (1.7 ± 0. 2 kg), 30-35 weeks (1.8 ± 0.2 kg), and 60-70 weeks

(2.2 ± 0.2 kg). Approximately 25 birds in each age group were evaluated with each of the three devices, assessing 279 chickens in 4 different strains of White Leghorn, Brown Leghorn,

Columbian Rock, and Plymouth Rock (Table 3.1). Different age groups of birds were available on different days. Birds were randomly assigned to a device, and a random order of application for each device was followed on a trial day. Sex was not balanced among the treatments due to fewer available male layer chickens. All the birds were euthanized at the Arkell Poultry Research

Station, University of Guelph.

3.3.2. Non-penetrating captive bolt devices

All three NPCB devices were commercially manufactured by Bock Industries, Inc., Philipsburg,

PA, USA. The Zephyr-E used in this study is the commercial model weighing 0.75kg. Two models of the Zephyr-E device were used in this study: Zephyr-E-standard (Figure 3.1-A) and

Zephyr-E-layer (Figure 3.1-D). Zephyr-E consists of a modified pneumatic nail gun fitted with a nylon head (diameter: Zephyr-E standard = 25 mm; Zephyr-E layer = 19 mm) attached to a cylindrical metal bolt (diameter: 9.5 mm). The shape of the bolt head is conical in Zephyr-E- standard (Figure 3.1-B) and round in Zephyr-E-layer (Figure 3.1-E). When fully extended, the

57 bolt protrudes (Zephyr-E-standard = 19 mm; Zephyr-E-layer = 11.3 mm) past the barrel. In both

Zephyr-E devices, bolt velocity is 20 m/s, delivering 11 Joules when used at 120 psi (Bock-

Industries.com). The Zephyr-E-standard was used with standard subject adapter (Figure 3.1-C), and the Zephyr-E-layer was used with chicken subject adapter (Figure 3.1-F).

The Zephyr-EXL is another commercially available pneumatic-powered NPCB device (Figure

3.1-G). The Zephyr-EXL uses a modified pneumatic nail gun that was fitted with a nylon head

(diameter: 25 mm) attached to a conical tipped cylindrical metal bolt (diameter: 9.5 mm) (Figure

3.1-H). The Zephyr-EXL bolt velocity is 27 m/s and delivers (Bock-Industries.com). The

Zephyr-EXL weighs 0.91 kg. The Zephyr-EXL was used with the chicken subject adapter

(Figure 3.1-I).

The Turkey Euthanasia Device (TED) (Figure 3.1-J) is a propane-powered NPCB device. The

TED consists of a gas-powered modified nail gun fitted with a flat metal head (diameter: 19.1 mm, length: 4mm) (Figure 3.1-K) attached to a cylindrical metal bolt (9.5mm) weighing 1.8 kg

(including battery and fuel cell). The TED delivers 28 Joules with a bolt velocity of 30 m/s

(Bock-Industries.com). The TED was used with adapter R3 (Figure 3.1-L). When the bolt is fully extended, it protrudes 7.1 mm with the R-3 subject adapter.

3.3.2.1. Application of the NPCB devices

Per manufacturer instructions (Bock Industries Inc., 2016), all devices were applied perpendicular to the top of the frontal bone just behind the comb and on a mid-line between the eyes and ears (Figure 3.2). The Zephyr-E was set to 120 psi by connecting to a compressed air

58 power supply (Hitachi EC 510, Hitachi Koki U.S.A., Ltd: running horse power = 0.8 kW, tank capacity = 22.7 L, maximum pressure = 145 psi). The Zephyr-EXL was set to the 98-100 psi. For consistency, the NPCB device operator was constant throughout all trials. Each bird was positioned in sternal recumbency with its keel on a flat hard surface and restrained by another researcher holding the wings gently towards the body of the bird during the application of the device. Restraint cones were not used in order to facilitate observation of convulsions. A single discharge was administered for each device.

3.3.3. Ante mortem assessments

The measures and procedures used for ante-mortem assessment are presented in Table 3.2. Time to loss of brain stem reflexes (pupillary reflex and nictitating membrane reflex), jaw tone, and neck muscle tone were assessed to detect insensibility. The parameters were assessed immediately after device application and then at 10s intervals until cessation. Pupillary reflex was checked for another 30 s (three times) after noting its absence to confirm brain death. If eye reflexes and/or breathing were present more than 60 s following device application, a second discharge was applied immediately, and the euthanasia trial was deemed a failure. Time at onset of tonic convulsions (rigid extension of legs and neck), time at last movement, first feather erection and cloacal relaxation following sporadic opening and closing were also recorded.

Presence and duration of heartbeat was monitored through stethoscope, and cardiac arrest was determined to have occurred when no discernible heartbeat could be heard. All procedures to and responses of each bird were video recorded after device application. Times to first feather erection and cessation of all convulsions were based on blinded video recordings. Other

59 measures were collected unblinded by live observations, and the time of each event was reconfirmed using the video recordings.

3.3.4. Macroscopic assessment of tissue damage

Gross pathologic assessment was performed immediately after death in all chickens successfully killed on the first attempt. Degree of external injury caused by each device was assessed (0-2 scale system) based on lacerations of the skin and presence of external hemorrhage at the site of device application: 0=no laceration of the skin, 1=laceration of the skin with no external bleeding, 2=laceration of the skin with external hemorrhage. Presence or absence of bleeding from mouth and nose were also recorded as binary (yes/no) responses. Macroscopic scoring of the degree of skull damage was based on a 0-3 scale system (Figure 3.3). Subcutaneous hemorrhage (SCH) and subdural hemorrhage (SDH) were assessed based on a 0-4 scale (Table

3.3) (Walsh et al., 2017; Woolcott et al., 2018a). The degree of SCH was assessed by removing the scalp and examining the amount of hemorrhage under the scalp. Then the dorsal surface of the skull was cautiously cleaned, removing blood and tissues to examine the severity of skull damage. Following this, the skull was lifted, and dura were removed to assess the SDH on the brain.

3.3.4.1. Microscopic assessment of brain trauma

Following macroscopic assessment, brains and the cervical spinal cord (from 1st to 3rd cervical bone) were collected from six randomly selected birds in each age group per device for microscopic evaluation. Tissues were placed in 10% buffered formalin for at least 14 days before trimming. For consistency, all trimming was performed by one individual. Three sections of the

60 brain (A: cerebrum, B: mid brain and thalamus, C: cerebellum) and the spinal cord (C1: portion under the first cervical bone, C2: portion under the second cervical bone, C3: portion under the third cervical bone) were sampled. Tissue sections were embedded in paraffin, cut 4µm and stained with hematoxylin and eosin (Animal Health Laboratory, University of Guelph) prior to assessment. Sections were evaluated by a veterinary pathologist blinded to bird age, breed or treatment to determine the degree of SDH and parenchymal hemorrhage (PCH) using a score from 0 to 4: no hemorrhage (0), minimal (<5%) hemorrhage (1), mild (5-10%) hemorrhage (2), moderate (10-30%) hemorrhage (3) and marked (>30%) hemorrhage (4) (Erasmus et al., 2010b;

Casey-Trott., 2013; Walsh et al., 2017; Woolcott et al., 2018a).

3.3.5. Statistical analyses

Statistical analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC, USA).

Generalized linear mixed models (GLMM) were used to analyze the fixed effects of device, age, and their interactions on ante mortem evaluations. Least significant means separation was conducted by using the Tukey-Kramer test.

Pearson product-moment correlation coefficients were used to determine the relationships between the different antemortem measures for each device. Regression analysis was conducted to establish the functional relationship between the highly correlated variables by generating the estimates for the intercept and the regression coefficient by using REG procedure of SAS. Heart beat end time was considered as the dependent variable (Y) and the independent variables (X) were time to onset of tonic convulsions, time at cessation of convulsions, and time at cloacal relaxation. The estimated intercept and the regression coefficient for each variable were used to

61 generate the relevant fixed-effect equation to predict the value of heart beat end time for the given level of the independent variable (cloaca contractions, cessation of convulsions and time at onset of tonic convulsions).

Generalized linear mixed models (GLMM) with multinomial distribution and cumulative logit link functions were used to analyze the effect of the device, age, and their interaction on macroscopic and microscopic trauma assessments (multinomial ordinary data). Odds ratios were computed to compare differences in the levels of fixed effects. Data from the three sections of brain from each bird were pooled and the highest score of the three sections of each brain was used for SDH and PCH analyses (Walsh et al., 2017, Woolcott et al., 2018a). The same procedure was applied to the SDH and PCH analyses for the spinal cord.

3.4. Results

3.4.1. Ante mortem assessments

All three NPCB devices induced loss of breathing, pupillary reflex and nictitating membrane reflex within 5s after application in most chickens. Overall, 100% successful killing was observed for the TED with immediate and irreversible insensibility in all age groups. One failure

(a 29 w.o. hen) was noted of 93 birds for the Zephyr-EXL for unknown reasons. The Zephyr-E resulted in 7 failures of 101 birds. Two failures occurred in 35 w.o. hens (one with the Zephyr-E- standard and the other with the Zephyr-E-layer) and 5 birds were 65 w.o. (three roosters with the

Zephyr-E-standard and two hens with the Zephyr-E-layer) (Table 3.1). These birds continued to demonstrate a pupillary light reflex, gasping, and rhythmic breathing. After 60s, the TED was

62 applied as the second killing method. Ante mortem and pathology assessments were not conducted for failed trials.

Clonic convulsions, with severe wing flapping and leg paddling, were observed immediately following device application, regardless of device type or age group. Feather erection was observed in all successfully killed birds. First feather erection was commonly observed in the neck region, followed by intermittent feather erections in different areas of the body. Time to onset of first feather erection was remarkably consistent between all devices and age groups

(Zephyr-E- 41±2s, Zephyr-EXL- 41±2s, TED - 43±2s, P=0.773). Gasping was observed in some successfully killed birds for all three devices (Zephyr-E: 9 out of 94, Zephyr-EXL: 6 out of 91,

TED: 3 out of 93). The average duration ± SD duration of gasping for Zephyr-E= 51±28s,

Zephyr-EXL=29±17s and TED= 16±2s. Two birds killed with the Zephyr-EXL started gasping at 50s and 90s after the application of device while others started at 10s after application of the device. None of these birds demonstrated concurrent pupillary or nictitating eye reflexes.

A summary of ante mortem responses is shown in Table 3.4. Both time to loss of neck tone and time to loss of jaw tone showed a difference for device, age, and device by age interaction

(P<0.05). The longest times to loss of jaw tone (31±3s) and neck muscle tone (28±3s) were observed in 60-70 w.o. roosters. The TED had the longest times (P<0.05) to loss of jaw tone

(28±3s) and neck muscle tone (36±6s). Overall, time to loss of jaw tone and neck muscle tone were longer in 60-70 w.o. roosters for TED (39±6s and 53±9s respectively, P<0.05). Younger birds (10-11 week and 20-21 w.o.) showed longest times (P<0.0001) for onset of tonic convulsions and cloacal relaxation (Table 3.4). A device effect was not observed for times to

63 onset of tonic convulsions or last movement. Onset of tonic convulsions was easily observed

(stretched neck and legs with sudden feather erection) in 10-11 w.o. birds compared to the older groups. A longer time to cloacal relaxation (P<0.05) was observed for the TED (186±24s) compared to the Zephyr-E (167±24s), and no difference was found between the TED and the

Zephyr-EXL. Some birds defecated during cloacal relaxation regardless of the device. As the last measure, longest time for cessation of heart beat (P<0.0001) was observed in younger birds (10-

11 w.o.: 235±26s). The shortest times for cloacal relaxation (P=0.033) and cessation of heart beat

(P=0.035) were recorded for the Zephyr-E.

The Pearson product-moment correlation coefficient was computed to assess the relationship between the ante mortem measures for each device. Table 3.5 summarizes the results. Overall, there were strong, positive significant correlations between time to cardiac arrest and time to cloacal relaxation (Zephyr-E: r = 0.935, Zephyr-EXL: r = 0.906, TED: r = 0.906), time to last movement (Zephyr-E: r = 0.951, Zephyr-EXL: r =0.934, TED: r = 0.948), and time to onset of tonic convulsions (Zephyr-E: r = 0.917, Zephyr-EXL: r =0.856, TED: r =0.808). There was a positive relationship between times to cardiac arrest and loss of neck muscle tone (r = 0.34, P=

0.0008) and between cardiac arrest and loss of jaw tone (r = 0.30, P=0.003) for TED. However, correlations between cardiac arrest versus loss of jaw tone or loss of neck muscle tone were not significant for the Zephyr-E and Zephyr-EXL devices. Loss of jaw tone and loss of neck muscle tone showed a positive (P<0.0001) correlation for all three devices. Strong positive correlations

(P=0.0001) were also found between onset of tonic convulsions and cloacal contraction, onset of tonic convulsions and time to last movement, and cloacal relaxation and time to last movement

64 for all three devices (Table 3.5). Feather erection was not associated with any other measure for any of the devices evaluated.

Regression analysis was conducted for the highly correlated variables considering time to cardiac arrest as a dependent variable. Regression equations and relevant coefficient of determinations

(R2) are presented in Table 3.6.

3.4.2. Pathology evaluations

3.4.2.1. Macroscopic evaluation

The degree of external damage (external hemorrhage and skin lacerations) was not different among the devices (P=0.595) or; age groups (P=0.062), nor was there an interaction (P = 0.689).

Skin lesions with external bleeding were noted in >80% of birds for all 3 devices (Zephyr-E:

80.6%, Zephyr-EXL: 85.8%, TED: 82.8%) and more than 70% in all age groups (10-11 w.o.:

85.3%, 20-21 w.o.: 72.2%, 30-35 w.o.; 81.3%, 60-70 w.o.: 90.4%). No external damage was observed in a few birds killed by the different devices (Zephyr-E: 11.8%, Zephyr-EXL: 4.3%, and TED: 12.9%) and in the different age groups (10-11 w.o.: 5.3%, 20-21 w.o.:20.3%, 30-35 w.o.;12 %, 60-65 w.o.:11.1%). Some birds showed both nasal and mouth bleeding (Zephyr-

E=7/94, Zephyr-EXL=3/92, TED=0/93) and protruding eyes with the cornea covered with blood

(Zephyr-E=10/94, Zephyr-EXL=11/92, TED=18/93).

Macroscopic lesion scores are presented in Table 3.7. All birds had subcutaneous hemorrhage on the skull regardless of the device and age. The highest score of 4 was observed in 99% of birds in the 10-11 w.o. group. Figure 3.4 shows two skulls with 25% and 100% macroscopic

65 subcutaneous hemorrhage. Macroscopic subcutaneous hemorrhage found on the skull was different among the devices (P<0.0001), age (P<0.0001) and device by age interaction

(P=0.029). Lower SCH scores were noted to occur 3.6 times more in birds killed with the

Zephyr-E than the TED and 0.1 times more likely for the Zephyr-EXL than the TED. No difference was found between the Zephyr-E and Zephyr-EXL.

Skull fracture scores differed among the age groups (P=0.0001) and there was an age by device interaction (P=0.0017). Skull fractures were common in the 10-11 w.o. group and more than

98% of birds of this age had penetrating fractures from all devices. Birds in the 20-21 and 60-70 w.o. groups had lower fracture scores than 11 w.o. birds. No difference was found between the

10-11 w.o. vs 30-35 w.o., and 20-21w.o. birds vs 60-70 w.o. birds. In addition, there was a 0.4 times greater chance of lower scores in 60-70 w.o. birds than in 30-35 w.o. birds. Some birds in the 60-70 w.o. group did not have any skull fractures (Zephyr-E=4%, Zephyr- EXL=29%, TED=

20%). An age difference was found only for TED and not for the Zephyr devices.

Subdural macroscopic hemorrhage on the brain was substantial and 100% of birds had a score of

1 or more for all three devices (Table 3.7). Over 85% of birds had a score 2 or above for all devices. The highest score of 4 was observed in 12% of birds killed by the Zephyr-E, 21% of birds killed by the Zephyr-EXL and 31% of birds killed by the TED. Figure 3.4 shows one brain covered by less than 25%, and another covered by 50-75% of subdural hemorrhage. Subdural macroscopic dorsal hemorrhage was affected by device (P<0.0001), age (P<0.0001), and device x age interaction (P=0.020, Table 3.7). The Zephyr-E generally resulted in lower scores compared to the TED or Zephyr-EXL. There was no difference in degree of subdural

66 hemorrhage between the TED and Zephyr-EXL. Birds at 10-11 w.o. generally had higher injury scores than all the other age groups. A higher chance of having higher injury scores was observed in birds at 20-21 w.o. than in birds at 30-35 w.o. Birds at 30-35 w.o. had 1.8 times higher chance of having lower trauma scores than birds at 60-70 w.o.

3.4.2.2. Microscopic evaluation

Table 3.8 provides a summary for microscopic scores for subdural (SDH) and parenchymal (PH) hemorrhage in three brain sections (cerebrum, mid brain, and cerebellum) killed with different

NPCB devices. SDH was observed in the cerebrum of all 24 birds killed by the TED, but only some birds killed by Zephyr-E and Zephyr-EXL. Similarly, PCH was absent in the cerebrum of some birds killed by Zephyr-E (n=4), Zephyr-EXL (n=3) and TED (n=1). SDH was substantial in the midbrain and observed in all assessed birds killed by all three devices. PCH was also found in the mid brains of all birds killed by Zephyr-EXL and TED except one bird killed by

Zephyr-E. All birds had SDH in the hind brain except one bird killed by Zephyr-EXL and TED.

Nearly 50% of birds did not have PCH in the hind brain when killed by the Zephyr-E (n=11),

Zephyr-EXL (n=7) and TED (n=12). PCH was absent in the hind brain of all 6 birds in 10-11 w.o. group. Four of six birds killed by the TED and one of six killed by the Zephyr-EXL in 10 -

11 w.o. group did not have hind brain PCH. Overall all three devices caused the highest degree of trauma to the mid brain.

There were no device, age or their interaction effects on SDH in the brain or spinal cord (Table

3.9). However, SDH was observed in both the brain and spinal cord for all three devices in all age groups. Similarly, PCH was observed in the brain in all ages for all three devices, but there

67 was a device effect (P=0.024). Birds killed with the Zephyr-EXL had a greater chance of having higher scores for PCH in the brain than those killed with the Zephyr-E or TED. Overall, 46% of birds received a score of 4 for PCH when killed with Zephyr-EXL, compared to 8% for the TED and 29% for the Zephyr-E. There was no difference between the TED and Zephyr-E. In contrast to brain PCH, a device effect was not observed for PCH in the spinal cord (P=0.182). However,

PCH in the spinal cord was different (P=0.044) among the age groups and showed an age by device interaction (P=0.021). Birds in the 60 w.o. group had a greater chance of having higher scores for PCH in the spinal cord than birds in the 10-11or 30-35- w.o. groups.

3.5. Discussion

To the author’s knowledge, this is the first study to evaluate the commercially available pneumatic NPCB devices (Zephyr-E, Zephyr-EXL, TED) for on-farm killing of layer chickens.

Our results demonstrated that all three devices were similarly effective at inducing insensibility and causing death in all age groups.

All devices caused loss of pupillary light reflex, nictitating membrane reflex, and breathing within 5s after application. Direct or indirect (contrecoup) trauma to the brain can result in impairment of brain stem reflexes (White and Krause, 1993; Drew and Drew, 2004). The area that controls the pupillary light reflex (cranial nerve II and III) is located in the mid brain. The nictitating membrane reflex is controlled by both cranial nerve III (located in the mid brain) and cranial nerve V (located in the pons). It is possible that direct damage to the corresponding cranial nerves caused loss of these reflexes. However, the substantial microscopic parenchymal hemorrhage (PCH) demonstrated that all three devices caused severe trauma to the mid brain.

The medulla oblongata in the hind brain contains important regions that regulate the cardiovascular and respiratory systems. The control centers of jaw tone and neck muscle tone are also located in the hind brain. The majority of birds had microscopic SDH (99%) and PCH (more than 50%) in the hind brain. The damage caused by the devices to the hind brain was enough to cause impairment of breathing. Jaw tone and neck muscle tone disappeared in less than 20s and

50s respectively, in all birds indicating that all three devices caused hind brain damage. Overall, all three devices disrupted brain function and caused rapid brain death.

Sandercock et al. (2004) reported that jaw tone and neck muscle tone were the most reliable reflexes distinguishing between sensible and insensible states in poultry based on EEG studies.

Loss of jaw tone has been used as an indicator of loss of sensibility in poultry under field conditions (Erasmus et al., 2010a; Martin et al., 2016). All birds in our study showed jaw tone and neck muscle tone for a few seconds following application of all three devices, and muscle tone did not disappear as quickly as did the eye reflexes. Martin et al. (2016) also reported longer time to loss of jaw tone (21.7s) than pupillary reflex (11.6s) in chickens killed with a penetrating captive bolt (Modified rabbit zinger). Results of the present study revealed longer time to loss of jaw tone and neck muscle tone for TED compared to the Zephyr devices, which corresponded to differences in PCH scores in the hind brain between devices. Moreover, the shortest time for cloacal relaxation and cardiac arrest, two responses controlled by the hind brain, were recorded for Zephyr-E. We suggest that the force caused to the hind brain by the TED is different compared to the Zephyr devices. Differences between the TED and Zephyr devices could be due to the shape of the bolt heads which deliver force differentially across and through the skull; the

TED has a flat bolt head and the Zephyr devices have either round or conical bolt head.

Erasmus et al. (2010a) did not report any gasping in turkeys effectively stunned with the

Zephyr-E or blunt force trauma. In contrast, a few successfully killed birds (~10%) showed gasping following application all three devices nearly for 60s. However, their eye reflexes were completely absent. In the current study, paroxysmal opening of the beak without any chest movement associated with breathing was recorded as gasping. Gasping is not indicative of sensibility and can be present in the absence of auditory evoked potentials (Cors et al., 2015).

Therefore, gasping can be observed both in awake and insensible birds, and while unpleasant to watch, is not necessarily indicative of device failure.

Seven birds failed to lose brain stem reflexes and breathing within 60s for Zephyr-E. Four of them were killed with the Zephyr-E standard with a common subject adapter and 3 with the

Zephyr-E layer with the chicken subject adapter. As the failures with the common subject adapter came first, we considered that the adapter was the cause of the failures. Then the device was switched to Zephyr-E-layer with a chicken subject adapter. The chicken subject adapter allows for a better alignment of the device around the comb of the bird. For the Zephyr-E and

Zephyr EXL the bolt velocity is adjustable as a function of air pressure. The pressure in the air compressor was slightly lower than 120 psi when killing these seven birds, and we suspect this could be a cause of the failed euthanasia. Based on this finding, we recommend that the Zephyr-

E be used exactly at 120 psi in layer chickens of all ages and weights.

Determining reliable indicators of irreversible brain injury is important for detecting clinical death under field conditions since animal care guidelines require confirmation of death before birds are disposed (NFACC, 2016). We studied times to onset of tonic convulsions, last

70 movement, and cloacal relaxation since these are readily observed indicators that can be used to confirm the irreversible brain injury. Tonic convulsions were observed in all successfully killed birds regardless of the device and thus tonic convulsions can be used to indicate a successful euthanasia. Sudden feather erection was observed first around 60s after device application and also during the tonic phase in all birds killed successfully. Gerritzen et al. (2007) confirmed the death of poultry killed with CO2 based on sudden feather erection along with occurrence of tonic convulsions followed by complete muscle relaxation. Heard (2000) stated that sudden feather erection during anesthesia of birds is indicative of cardiac arrest. However, in the current study, time at first feather erection was not correlated with cessation of heart beat after killing with two of the three devices, and cannot be considered a reliable indicator of time at cardiac arrest.

Similar results were determined by Hernandez et al. (2018) who also found that feather erection was not associated with cardiac arrest or isoelectric point using EEG. Cessation of movement has been used to estimate irreversible brain death (Erasmus et al., 2010a; Dawson et al., 2007;

Gerritzen et al., 2007). Time to last movement in the present study was highly correlated with time of cardiac arrest, thereby serving as a reliable on-farm indicator of cardiac arrest. Cloacal relaxation was the last reflex observed before cardiac arrest in all birds, which agrees with

Martin et al. (2016), and demonstrates its utility as a conservative indicator of death. Cardiac arrest typically occurs after all motion has ceased (Dawson et al., 2007; Turner et al., 2012). This coincides with the results in present study as all the birds ceased heart beat after cessation of all movements and reflexes. The presence of a heart beat does not indicate sensibility, and is, itself, a conservative measure. Turner et al. (2012) also reported the presence of a heartbeat in poultry several minutes after brain death, as confirmed by use of an EEG. Auscultation of the heart with a stethoscope can be difficult under the field conditions so it can be difficult to be certain of the

71 exact moment of cardiac arrest. Additionally, the heart may continue to beat irregularly for some time (Hernandez, 2018). Onset of tonic convulsions, last movement, and cloacal contractions can be visually observed in the field. The relationship analysis in this study suggest that cardiac arrest was highly positively correlated with onset of tonic convulsions, time to last movement, and cloacal relaxation for all three NPCB devices and thus, onset of tonic convulsions, last movement, and cloacal relaxation can be used by stock persons to make accurate decisions of successful euthanasia under field conditions.

Younger birds (10-11 w.o. and 20-21 w.o.) had a longer latency to onset of tonic convulsions, last movement, cloacal relaxation, and cardiac arrest. In addition, one, four and six of the six 10-

11 weeks old birds killed by the Zephyr-EXL, TED and Zephyr-E, respectively, had no PCH in the hind brain, indicating less trauma to the hind brain. This might explain the longer latencies to onset of tonic convulsions, last movement, cloacal relaxation, and cessation of heart beat in the

10-11 w.o. group. The oldest birds in the 60-70 w.o. group, with more mature anatomy (fused and larger size skull, large comb) experienced a longer time to loss of jaw tone and neck muscle tone. Therefore, the placement of the device on the head (the place of the skull where the bolt hit), device configurations and anatomic structure of the head may have affected the degree of damage cause to different regions of the brain.

Overall more than 80% birds showed external damage for all three devices. Some birds had mouth and nose bleeding, and damaged eyes. External bleeding is important for biosecurity measures and aesthetic concerns. The fine balance between effectiveness and aesthetics is the key to selection of an appropriate method of euthanasia. Higher external damage caused by the

NPCB devices may indicate a need for lower air pressure levels for the Zephyr-EXL based on the age group. However, the Zephyr-E should only be used with the manufacturer-recommended

120 psi to minimize the chance of failures.

Results of the macroscopic pathology assessment indicated that lower subcutaneous and subdural hemorrhage scores were more likely for the Zephyr-E. This may be due to lower force generated by the Zephyr-E than for the Zephyr-EXL and TED. All three devices caused penetrating fractures with no embedded fragments in most birds. Skull fractures are highly associated with severe traumatic brain injuries leading to death in humans (Tseng et al., 2011). Studies in other animal species have reported that skull fractures are often present in animals that were effectively stunned with non-penetrating captive bolt devices (Finnie et al., 2000; Casey-Trott et al., 2013). However, Erasmus et al. (2010b) reported that presence or absence of skull fractures did not affect onset of insensibility in turkeys killed by NPCB or blunt trauma. Our results also showed that presence or absence of skull fractures did not influence the effectiveness of inducing insensibility and irreversible brain death. Overall, 15 birds in our study had an intact skull, and all of them showed rapid loss of eye reflexes and irreversible brain death.

There was no difference found for the microscopic assessment of subdural hemorrhage in brain sections among the devices. A device effect was found for PCH in that the Zephyr-EXL (at 98-

100 psi) caused higher scores of PCH in the brain than that seen with the TED or Zephyr-E. PCH indicates traumatic brain injuries (TBI). Researchers have suggested that immediate insensibility and irreversible loss of vital functions are associated with subdural and parenchymal hemorrhage in poultry (Erasmus et al., 2010b; Bader et al., 2014). Results from all three devices evaluated in

73 this study confirm this. It is important to note that all three devices also caused SDH and PCH in the cervical area of the spinal cord. Therefore, the force generated by all three devices is sufficient to cause intensive traumatic damage to the brain and to the cervical portion of the spinal cord.

An unexpected problem was encountered with the Zephyr devices for Plymouth Barred Rock chickens. In two chickens of this strain, the feathers became stuck between the bolt and the adapter for both Zephyr-E and Zephyr-EXL. Despite this, the birds were both killed effectively with these devices. We suggest that plumage type should be considered when using Zephyr devices with chicken subject adapter.

3.6. Conclusion

This study demonstrated that brain trauma cause by all three NPCB devices was sufficient to rapidly render the birds insensible, leading to irreversible brain death in all age groups of layer chickens. The Zephyr-E, Zephyr-EXL, and TED devices can be used as a humane single-step euthanasia method for layer chickens. Additionally, we suggest onset of tonic convulsions, last movement, and final cloacal relaxation are good indicators of clinical death in layer chickens in field conditions.

Table 3. 1 : List of number of birds killed with the different NPCB devices by age group, strain, body weight, and sex. Number of failed birds are indicated in parentheses.

Age Device Body wt Strain Male Female Total (weeks) (kg) 10-11 Z-E-std 1.1 ± 0.2 White Leghorn 13 12 25 Z-EXL 1.1 ± 0.2 White Leghorn 13 12 25 TED 1.1 ± 0.2 White Leghorn 13 25 25 20-21 Z-E-std 1.5 ± 0.2 White Leghorn 1 1 2 Z-E-layer 1.7 ± 0.2 White Leghorn 2 12 1.9 ± 0.3 Plymouth Rock 1 2 17 Z-EXL 1.6 ± 0.1 White Leghorn 3 8 1.9 ± 0.3 Plymouth Rock 3 4 18 TED 1.7 ± 0.2 White Leghorn 3 12 2.2 ± 0.9 Brown Leghorn 2 0 1.8 Plymouth Rock 0 1 18 30-35 Z-E-std 1.8 ± 0.2 White Leghorn 0 12+(1) 13 Z-E-layer 1.9 ± 0.2 Brown Leghorn 0 10+(1) 2.0 ± 0.1 Columbian Rock 0 3 14 Zephyr-EXL 1.7 ± 0.2 White Leghorn 0 12 1.9 ± 0.1 Brown Leghorn 0 4 1.8 ± 0.1 Columbian Rock 0 9 25 TED 1.6 ± 0.2 White Leghorn 0 12 1.9 ± 0.1 Brown Leghorn 0 6 1.8 ± 0.2 Columbian Rock 0 7 25 60-75 Z-E-std 2.2 ± 0.1 White Leghorn 2+(3) 0 2.4 ± 0.4 Brown Leghorn 0 4 1.8 ± 0.1 Columbian Rock 0 3 2.2 ± 0.1 Plymouth Rock 0 2 14 Z-E-layer 2.4 ± 0.3 Brown Leghorn 0 8+(2) 2.1 Columbian Rock 0 1 2.3 ± 0.1 Plymouth Rock 0 5 16 Zephyr-EXL 2.2 ± 0.1 White Leghorn 14+(1) 0 2.4 ± 0.7 Brown Leghorn 0 2 2.2 ± 0.2 Columbian Rock 0 7 1.9 Plymouth Rock 0 1 25 TED 2.0 ± 0.1 White Leghorn 14 0 2.2 ± 0.2 Brown Leghorn 0 6 2.1 ± 0.7 Columbian Rock 0 3 2.4 ± 0.1 Plymouth Rock 0 2 25

Z-E-std=Zephyr-E with standard subject adapter and conical shape bolt head. Z-E-layer= Zephyr-E with chicken subject adapter and round shape bolt head. TED= Turkey Euthanasia Device.

Table 3. 2 : Ante-mortem assessment measures, descriptions, and procedures used, listed in order of observation after application of each killing method

Measures Description Procedure

Pupillary light reflex Constriction of the pupil in response to Light from a medical penlight light was directed into the eye and pupil constriction was examined Nictitating Transient closure of the nictitating The medial canthus of the eye membrane reflex membrane in response to mechanical or the cornea was lightly stimulation touched with a fingertip

Jaw tone Resistance to downward pressure Gentle pressure was applied applied to the jaw to the lower jaw with a finger

Neck muscle tension Change in neck muscle tone or The neck was lifted with the movement of the head when the neck fingers of one hand is lifted Gasping Paroxysmal opening of the beak Visual observation for paroxysmal opening of the beak

Feather erection Sudden erection of feathers, not in Visual observation of sudden response to external stimuli feather erection on some part of the body

Tonic convulsions Muscle rigidity with the legs and Visual observation of the wings outstretched time of onset of legs and neck outstretched

Cloacal relaxation Cloaca opening following Visual observation for cloaca contractions of cloaca opening following contractions

Cardiac arrest Cessation of heart beat Auscultation by using a stethoscope

Breathing Rhythmic inhalation and exhalation Visual observation for rhythmic movement of the chest area

Table 3. 3: Gross and microscopic pathology scoring criteria for macroscopic, and microscopic hemorrhage

Score Macroscopic Microscopic Subcutaneous or subdural hemorrhage Subdural or parenchymal hemorrhage

0 None None 1 <25% of surface area Minimal (<5% of section) 2 26–50% of surface area Mild (5–10% of section) 3 51–75% of surface area Moderate (11–30% of section) 4 76–100% of surface area Marked (>30% of section)

Table 3. 4: Mean time (± SE, s) to onset of specific measures after application of different NPCB devices in different age groups of layer chickens

Measure Age Device1 All P value (weeks) devices Z-E Z-EXL TED Device Age Device*Age Loss of jaw <0.0001 <0.0001 0.0063 tone 10-11 13±2bc 22±2bc 25±6b 19±3b 20-21 21±3bc 21±2bc 29±6b 23±3b 30-35 20±3bc 19±2bc 24±6bc 21±3b 60-70 28±3b 27±2b 39±6a 31±3a All ages 21±3b 22±3b 28±3a Loss of neck <.0.0001 <0.0001 0.0297 muscle tone 10-11 18±4e 23±5de 32±9dc 24±6c 20-21 25±4dce 23±5de 33±10dc 26±6b 30-35 27±4dc 29±5dc 33±10dc 30±6b 60-70 35±4bc 44±5ab 53±9a 46±6a All ages 27±6b 30±6b 36±6a Time at first 0.7733 0.2154 0.0546 feather 10-11 38±2 erection 20-21 45±2 30-35 39±2 60-70 44±2 All ages 41±2 41±2 43±2 Onset of 0.2909 <0.0001 0.8839 tonic 10-11 153±22a 20-21 141±23a 30-35 111±23b 60-70 103±22b All ages 122±22 126±22 132±22 Last 0.1042 <0.0001 0.9026 movement 10-11 206±25a 20-21 189±25a 30-35 146±25b 60-70 147±25b All ages 164±25 171±25 181±25 Cloacal 0.0335 <0.0001 0.8349 relaxation 10-11 203±24a 20-21 199±25a 30-35 153±25b 60-70 150±24b All ages 167±24b 175±24ab 186±24a Cessation of 0.0354 <0.0001 0.8866 heart beat 10-11 235±26a 20-21 209±27b 30-35 173±27c 60-70 178±26c All ages 189±26b 198±26ab 209±26a

Different letters indicate statistically significant (P<0.05) differences between the comparisons Bold numbers indicate significance results (P<0.05) 1Z-E = Zephyr-E, Z-EXL = Zephyr-EXL, TED = Turkey Euthanasia Device

Variable Device1 Loss of Loss of jaw First Onset of Cloacal Last neck tone feather Tonic relaxations movement muscle erection convulsions tone Cessation of Z-E -0.1082 -0.1056 0.1088 0.9174 0.9352 0.9519 heart p=0.2991 p=0.3111 p=0.296 p<0.0001 p<0.0001 p<0.0001 beat Z-EXL 0.0917 0.0723 0.0703 0.8566 0.9067 0.9342 p=0.387 p=0.495 p=0.5078 p<0.0001 p<0.0001 p<0.0001 TED 0.3422 0.3003 0.2065 0.8081 0.9068 0.9482 p=0.0008 p=0.0034 p=0.047 p<0.0001 p<0.0001 p<0.0001 Loss of neck Z-E 0.7948 -0.0744 -0.2007 -0.1003 -0.1622 muscle tone p<0.001 p=0.475 p=0.052 p=0.3383 p=0.1183 Z-EXL - 0.7089 0.2021 -0.026 0.058 0.0462 p<0.0001 p=0.054 p=0.8047 p=0.5912 p=0.663 TED 0.7660 0.2971 0.2371 0.2134 0.2832 p<0.0001 p=0.0038 p=0.0221 p=0.0399 p=0.0059 Loss of jaw Z-E -0.1461 -0.1467 -0.0639 -0.1299 tone p=0.1599 p=0.1583 p=0.5427 p=0.2119 Z-EXL - -0.1128 -0.036 0.0259 -0.0105 p=0.286 p=0.7289 p=0.8102 p=0.9212 TED 0.1880 0.2371 0.2138 0.2763 p=0.0711 p=0.0221 p=0.0396 p=0.0073 First Z-E 0.1148 0.1903 0.1235 feather p=0.2705 p=0.067 p=0.2374 erection Z-EXL - 0.0251 0.0487 0.0894 p=0.8128 p=0.6522 p=0.3989 TED 0.0588 0.1013 0.1914 p=0.5755 p=0.333 p=0.066 Onset of Z-E 0.9341 0.9519 Tonic p<0.0001 p=0.0001 Z-EXL - 0.9049 0.9056 p<0.0001 P<0.0001 TED 0.7915 0.8195 p<0.0001 p<0.0001 Cloacal Z-E 0.960 relaxation p<0.0001 Z-EXL - 0.94877 p=<0.0001 TED 0.9428 p<0.0001 1Z-E = Zephyr-E, Z-EXL = Zephyr-EXL, TED = Turkey Euthanasia Device. Bolded numbers indicate significant results (P < 0.05)

Table 3. 6: Regression and relative contribution (R2) for response of dependent variable (Y) for independent variables (X) of different NPCB devices

Independent Dependent Device1 Regression Coefficient of variable (X) equation determination (R2) variable (Y) Time at onset Heart beat Z-E Y= 35.77 + 1.1X 0.8416 of tonic end time Z-EXL Y= 57.15 + 1.1X 0.7339 convulsions TED Y= 84.52 + 0.9X 0.6532 Time at cloacal Heart beat Z-E Y= 19.1 + 1X 0.8746 relaxation end time Z-EXL Y= 33.6 + 1X 0.8222 TED Y= 26.2 + 1X 0.8224 Time at last Heart beat Z-E Y= 24.6 + 1X 0.9062 movement end time Z-EXL Y= 40.5 + 0.9 X 0.8728 TED Y= 34.6 + 1X 0.8991

All regression coefficients were significant (P < 0.05). 1Z-E = Zephyr-E, Z-EXL = Zephyr-EXL, TED = Turkey Euthanasia Device.

Table 3. 7: Summary of gross scores for subcutaneous hemorrhage, skull fractures, and subdural hemorrhage in birds killed by different NPCB devices. Number of birds with each score are indicated.

Device1 Age Number of birds with gross score P-value 0 1 2 3 4 Total Device Age Device* Age SCH <0.0001 <0.0001 0.0295 Z-E 10-11 0 0 0 2 23 25 20-21 0 3 5 4 7 19 30-35 0 5 12 6 2 25 60-70 0 3 2 10 10 25 Z-EXL 10-11 0 0 0 4 21 25 20-21 0 2 2 7 7 18 30-35 0 2 15 5 3 25 60-70 0 7 7 4 6 24 TED 10-11 0 0 1 2 22 25 20-21 0 0 0 1 17 18 30-35 0 2 3 9 11 25 60-70 0 1 4 6 14 25 Skull 0.1485 0.0001 0.0017 fractures Z-E 10-11 0 0 20 5 N/A 25 20-21 1 4 11 3 N/A 19 30-35 0 3 19 3 N/A 25 60-70 1 2 13 9 N/A 25 Z-EXL 10-11 0 2 14 9 N/A 25 20-21 0 2 11 5 N/A 18 30-35 0 0 15 10 N/A 25 60-70 7 4 5 8 N/A 24 TED 10-11 0 0 13 12 N/A 25 20-21 0 7 11 0 N/A 18 30-35 1 4 13 7 N/A 25 60-70 5 5 13 2 N/A 25 Brain <0.0001 <0.0001 0.0203 SDH Z-E 10-11 0 2 13 8 2 25 20-21 0 5 7 4 3 19 30-35 0 7 10 7 1 25 60-70 0 1 12 7 5 25 Z-EXL 10-11 0 0 6 10 9 25 20-21 0 2 4 5 7 18 30-35 0 4 9 10 2 25 60-70 0 4 11 7 2 24 TED 10-11 0 0 4 6 15 25 20-21 0 1 4 10 3 18 30-35 0 3 11 9 2 25 60-70 0 4 5 7 9 25 Bold numbers indicate significance (P<0.05). 1Z-E = Zephyr-E, Z-EXL = Zephyr-EXL, TED = Turkey Euthanasia Device. SCH= subcutaneous hemorrhage, SDH= Subdural hemorrhage

Table 3. 8: Summary of microscopic scoring of brains for trauma following application of each of the three NPCB devices in layer chickens. Number of birds with each score are indicated.

Device1 Brain SDH Score PCH Score section 0 1 2 3 4 0 1 2 3 4

Z-E Cerebrum 5 1 6 6 6 4 6 7 6 1 Mid brain 0 0 8 5 11 1 9 7 2 5 Hind brain 0 1 4 17 2 11 6 2 0 5

Z-EXL Cerebrum 3 0 2 10 9 3 3 4 8 6 Mid brain 0 2 2 7 13 0 4 4 7 9 Hind brain 1 4 5 7 7 7 7 7 2 1

TED Cerebrum 0 0 5 10 9 1 7 6 10 0 Mid brain 0 2 3 11 8 0 2 5 15 2 Hind brain 1 1 6 9 8 12 8 4 0 0 1Z-E = Zephyr-E, Z-EXL = Zephyr-EXL, TED = Turkey Euthanasia Device. SDH = Subdural hemorrhage. PCH = Parenchymal hemorrhage

Table 3. 9: Overall summary of microscopic scoring of subdural hemorrhage and parenchymal hemorrhage in the brain and spinal cord of layer chickens killed by NPCB device. Number of birds with each score are indicated

Device1 Age Score Total P Value 0 1 2 3 4 Device Age Device*Age SDH in 1.000 0.5388 0.8972 brain Z-E 0 0 0 12 12 24 Z-EXL 0 0 1 6 17 24 TED 0 0 0 10 14 24 SDH in 0.999 0.995 0.9136 spinal Z-E 0 0 3 9 12 24 cord Z-EXL 0 1 1 7 15 24 TED 0 0 1 7 16 24 PCH in 0.0242 0.1694 0.4794 brain Z-E 0 7 6 4 07 24 Z-EXL 0 1 3 9 11 24 TED 0 2 5 15 02 24 PCH in 0.1822 0.0443 0.0215 spinal Z-E 10-11 0 3 2 1 0 6 cord 20-21 1 3 2 0 0 6 30-35 1 1 2 2 0 6 60-70 0 5 1 0 0 6

Z-EXL 10-11 3 3 0 0 0 6 20-21 1 2 1 1 1 6 30-35 3 2 1 0 0 6 60-70 0 1 4 1 0 6

TED 10-11 1 3 1 1 0 6 20-21 0 1 4 1 0 6 30-35 1 4 1 0 0 6 60-70 0 1 3 2 0 6

1Z-E = Zephyr-E, Z-EXL = Zephyr-EXL, TED = Turkey Euthanasia Device. SDH = Subdural hemorrhage. PCH = Parenchymal hemorrhage. Bolded numbers indicate significant results (P<0.05).

Figure 3. 1: Non-penetrating captive bolt devices. (A) Zephyr-E standard: (B) Conical shape bolt head, (C) Standard subject adapter. (D) Zephyr-E-layer: (E) Round shape bolt head, (F) Chicken subject adapter. (G) Zephyr-EXL: (H) Conical shape bolt head, (I) Chicken subject adapter. (J) Turkey Euthanasia Device (TED): (K) Flat bolt head, (L) R-3 subject adapter.

Figure 3. 2: Application of the Zephyr-EXL device on a 30 w.o. hen. The bird was restrained in sternal recumbency with its neck resting ventrally on the ground, and the wings held gently towards the body of the bird. Device was placed perpendicular to the top of the frontal bone just behind the comb and on the mid line between the eyes and ears.

Figure 3. 3 : Figure 3. Gross pathology scoring criteria for skull fractures. Arrows indicate the fracture location [modified from Erasmus et al., (2010b) and Casey-Trott et al., (2013)]. (A) No fracture, intact skull (score 0). (B) Depression fracture (score 1). (C) Penetrating fracture-no imbedded fragments (score 2). (D) Penetrating fracture- with imbedded fragments (score 3).

CHAPTER 4

Efficacy of a Novel Mechanical Cervical Dislocation Device in

Comparison to Manual Cervical Dislocation in Layer Chickens

This manuscript was accepted to journal of Animals with the following authors: R.M.A.S. Bandara, S. Torrey, P.V. Turner, A. zur Linden, A. Bolinder, K. Schwean-Lardner and T. M. Widowski

4.1. Abstract

The main objective of this study was to assess the efficacy of mechanical cervical dislocation using the Koechner Euthanasia Device Model C (KED) in comparison to manual cervical dislocation in layer chickens. Laying hens and/or roosters in three different age groups (12, 27-

29, and 65-70 weeks old) were randomly assigned to one of three experimental groups: manual cervical dislocation in conscious birds (CD), manual cervical dislocation in anesthetized birds

(aCD), or mechanical cervical dislocation by KED in anesthetized birds (aMCD). Anesthetized birds received an intramuscular dose of 0.3 mg/kg medetomidine and 30 mg/kg of ketamine to achieve a clinical anesthesia. A comparison of CD vs. aCD responses confirmed that the anesthetic plane abolished or reduced clonic convulsions, nictitating membrane reflex, tonic convulsions, and cloacal relaxation. Time to loss of the pupillary light reflex (~123 s), and time to cardiac arrest (~172 s) were longer (P < 0.001) in the birds in aMCD group than aCD (~71 and

~137s, respectively). Radiographs revealed that the majority of the birds killed by manual cervical dislocation (CD + aCD) had dislocations between the skull and atlas (C1) or between cervical vertebrae C1-C2. The KED resulted in a majority of dislocations at C2-C3. Birds killed by manual cervical dislocation presented more subdural and parenchymal hemorrhage in the brain stem compared to birds killed by KED. Radiographs indicated the presence of fractures in a few birds killed by either method (CD + aCD vs aMCD). Compared to manual CD, KED resulted in less brain trauma and a longer latency to brain death indicating lower efficacy of KED as an on-farm killing method.

4.2. Introduction

Over the last decade, techniques used to kill poultry on farms have come under public and scientific scrutiny due to concern for animal welfare. For the welfare of the animal, the goal of any euthanasia technique is to achieve insensibility as quickly as possible, for the process to cause minimal pain, and for death to follow quickly. Manual cervical dislocation is the most common method for killing poultry on farms. Manual cervical dislocation involves stretching and separating the cervical vertebrae by hand, rupturing the blood vessels and causing death by cerebral ischaemia and extensive damage to the spinal cord and brainstem (Gregory and Wotton

1990a; Mason et al., 2009; Bader et al., 2014). Optimal application of cervical dislocation separates the cervical vertebrae between the skull and the first cervical vertebra (C1) and completely transects the spinal cord (HSA, 2004). However, cervical dislocation may be esthetically displeasing to personnel performing it. Additionally, use of manual cervical dislocation to kill poultry on-farm has been restricted to birds weighing less than 3 kg and to 70 birds per person per day through European Union (EU) legislation (EC, 2009).

To address the limitations around manual cervical dislocation, different tools have been developed to perform cervical dislocation mechanically. Some devices (e.g. killing cone and heavy stick) involve stretching, whereas others attempt to dislocate the cervical vertebrae by forcing a blunted edge between two vertebrae (e.g. pliers, and Burdizzo) without stretching or twisting the neck (Sparrey et al., 2014). Both manual and mechanical cervical dislocation methods have been questioned for humaneness based on evidence that loss of sensibility is not immediate in poultry species (Gregory and Wotton, 1990a; Sparrey et al., 2014). In their guidelines, the AVMA (2013) specify that cervical dislocation “must result in luxation of the

90 cervical vertebrae without primary crushing of the vertebrae and spinal cord”. However, there is evidence that vertebral separation with the aid of a tool can cause fractures to the cervical vertebrae in chickens (Bader et al., 2014) and turkeys (Bader et al., 2014; Woolcott et al., 2018b;

Widowski et al., 2018). At present, different devices are manufactured for cervical dislocation and marketed to provide poultry producers with options for on-farm euthanasia. However, there is a lack of evidence on the different devices’ efficacy in causing a humane death. The Koechner

Euthanizing Device (KED; Koechner Mfg. Co., Inc. Tipton, MO) is commercially available for poultry as a mechanical cervical dislocation device. Woolcott et al. (2018b) assessed the efficacy of the KED (Model S designed for chicks and poults) in anesthetized poults and young turkeys.

Their results revealed longer times to cessation of brain stem reflexes and more fractures in cervical vertebrae in the birds killed by KED-S in comparison to manual cervical dislocation.

Jacobs et al. (2019) also found longer times to loss of brain stem reflexes when broiler chickens were killed by KED-model-C compared to manual CD. Therefore, there is a need to evaluate the

KED in layer chickens before recommending its use on farms.

Ethical concerns arise in the evaluation of novel killing devices in conscious animals. Therefore, it is necessary to initially evaluate novel killing devices either using cadavers (Martin et al.,

2017) or anesthetized animals (Woolcott et al., 2018b; Martin et al., 2019). The use of anesthetic agents reduces the distress and pain associated with killing technique by abolishing the awareness of the animal. Anesthetics produce insensibility by preventing integration (blocking the interactions among specialized brain regions) or by reducing information (reducing the number of activity patterns available to cortical networks) (Alkire et al., 2008). Efficacy and humaneness of physical killing methods in poultry are often assessed based on time to loss of

91 brain stem reflexes and degree of physiological trauma that is likely to result in rapid death (e.g.

Erasmus et al., 2010ab; Woolcott et al., 2018ab; Chapter 3). Previous studies have reported some brain stem reflexes present under anesthesia in poultry (Woolcott et al., 2018b; Martin et al.,

2019), whereas others are not (Sandercock et al., 2014), thus the need to compare conscious and anesthetized manual CD to understand the effect of anesthesia on the measures used to determine efficacy.

The main objective of the current study was to assess the efficacy of mechanical cervical dislocation using the Koechner Euthanasia Device (KED model-C) in comparison to manual cervical dislocation based on behavioural responses, brain stem reflexes and postmortem analysis of the physiological damage produced in three different age groups of layer chickens. For ethical reasons, the KED was tested on anesthetized birds and compared with anesthetized birds killed by manual cervical dislocation. In order to determine which measures would be valid for use in anesthetized birds, our secondary objective was to determine the effect of the anesthetic agents used in the study on behavioural and reflex responses by comparing anesthetized and conscious birds killed by manual cervical dislocation.

4.3. Methods

The procedures and protocol for this research were reviewed and approved by the University of

Guelph Animal Care Committee (AUP 3321) which holds a Good Animal Practice certificate

101 issued by the Canadian Council on Animal Care.

4.3.1. Animals and facilities

The birds were obtained from different research projects that were conducted at the Arkell poultry research station of the University of Guelph. All the birds were targeted for euthanasia as they had reached scheduled end of study, or because of routine flock depopulation. Different strains and age groups of birds were available on different days. A health assessment was performed on the birds prior to the experiment because we used anesthesia model, and only healthy birds were used to reduce variable outcomes. Laying hens and/or roosters in three different age groups (12, 27-29, and 65-70 weeks old) were randomly assigned to either manual cervical dislocation (CD), anesthetized manual cervical dislocation (aCD) or anesthetized mechanical cervical dislocation by KED (aMCD). Order of application of assigned treatments on a trial day was determined by using the random number generator in Excel. Sex was not balanced among the treatments due to the disproportionate availability of female and male layer chickens in the different age groups. A total of 72 chickens were enrolled in this study: eight birds in each age group were evaluated with each of the three killing methods. Strain, sex, body weight and sample sizes for the different age classes of birds used in the study are given in Table

4.1.

4.3.2. Koechner Euthanizing Device (KED)

The Koechner Euthanizing Device (KED) is manufactured as a mechanical cervical dislocation device for poultry by Koechner MFG. CO., INC (2016) (U.S. Patent No. 8,152,605). Four different commercial models of the KED are manufactured for different weight classes (KED).

The model-S is designed for birds weighing up to 1.8 kg and the model-T is for birds weighing up to 20.5 kg. The model-Txl is designed for birds weighing over 29.5 kg. The KED model

93 evaluated in the current study was the model-C consisting of a 69 cm handle, marketed for birds weighing up to 13.5 kg (Figure 4.1A).

4.3.3. Anesthesia and killing procedures

Anesthesia was induced in the birds by using a combination of medetomidine (1 mg/ml)

(CepetorTM, DIN:02337177, Modern Veterinary Therapeutics, LLC, Miami, FL, USA) and ketamine (100 mg/ml) (Ketaset®, DIN: 02173239, Pfizer Animal Health, Kirkland, QC, CAN).

Appropriate drug doses to induce surgical anesthesia in chickens were established in a pilot study. The chosen doses (medetomidine=0.3 mg/kg body weight; ketamine= 30 mg/kg body weight) were administered as separate injections into the breast muscle by a veterinarian monitoring the anesthesia.

After injecting the drugs, the birds were kept in a crate in a quiet and dark room for ten minutes for the anesthesia to take effect. After another 5 minutes without interference, breathing, heartbeat and pedal reflex were assessed. Fifteen minutes after the drug application, birds were assessed for breathing, pupillary reflex, pedal reflex, neck muscle tone, jaw tone, and heartbeat.

The same responses were checked directly prior to application of the killing method. Birds were determined to be ready to apply the killing method when they were non-responsive to handling, and when assessment of breathing pattern, heart auscultation, jaw tone, and pedal reflex by the monitoring veterinarian indicated surgical anesthesia.

Manual cervical dislocation was performed by a trained and experienced technician. The bird’s head was held in the operator’s palm, with the neck between the index finger and thumb. Manual

94 cervical dislocation was performed in one swift movement with the operator pulling down on the bird’s head, stretching the neck, while rotating the bird’s head upwards into the back of the neck.

A trained and experienced technician applied the KED device. The KED was applied according to the manufacturer’s instructions (Koechner MFG. Co., INC. 2016). Prior to the experiment, the

KED device was applied to cadavers in different age groups and radiographs were used to confirm correct placement and pressure. All of the birds were manually restrained on a table in a sternal recumbent position when applying the KED. The double angle blade was placed under the neck of the bird while the single side blade was placed dorsally above the top of the neck at the base of the skull. The handles were brought together quickly and firmly to cause dislocation

(Figure 4.1B). Then the device was removed from the neck of the bird.

4.3.4. Ante mortem assessment

The measures used in ante-mortem assessment are described in Table 4.2. Each measure was checked immediately after application of the killing method and then at 10s intervals until cessation. Presence of heartbeat was monitored through a stethoscope, and cardiac arrest was determined to have occurred when no discernible heartbeat could be heard. All of the procedures and responses of each bird were video recorded. Time to first feather erection was based on blinded video recordings. The other measures were collected unblinded by live observations, and the time at each event was confirmed with the video records.

4.3.5. Post mortem assessment

4.3.5.1. Assessment of radiographs

Radiographs were performed on all birds immediately following euthanasia in an adjacent room.

Extreme care was taken with minimum handling to prevent further damage to the cervical dislocation site. Four views were performed of each bird (dorsal ventral, ventral dorsal, right lateral, left lateral) using a portable x-ray unit (Poskom VET-20BT X-ray unit with 80 kVp, 2.5 mAs power). The birds were placed directly on the wireless direct digital imaging receptor panel

(ULTRAMAXX, Promark imaging; 19.65 x 23.6 cm image matrix, a-SI TFT-PIN (thin filament transistors), 77-micron pixel pitch, 6.0 to 7.0 lp/mm resolution) for radiograph acquisition.

Radiographs were assessed for site of luxation/subluxation, and for presence, type, and site(s) of fractures (Table 4.3) by a board-certified veterinary radiologist blinded to the treatment. Digital

Imaging and Communications in Medicine (DICOM) formatted radiographs were interpreted after importing the images into a DICOM viewer (Osirx MD, version 8.0.2, Pixmeo SARL,

Bernex, Switzerland).

4.3.5.2. Macroscopic assessment of tissue damage

Degree of external damage and bleeding caused by each killing method was assessed (0-2 scale) based on laceration of the skin and presence of external hemorrhage on the neck: 0=no laceration of the skin, 1=laceration of the skin with no external bleeding, 2=laceration of the skin with external bleeding.

Dissection was performed on all birds to assess the degree of subcutaneous hemorrhage (SCH) at the site of cervical dislocation, damage to the trachea, transection of the spinal cord, and degree

96 of subdural hemorrhage (SDH) on the brain. Scoring criteria for macroscopic SCH and SDH are presented in Table 4.4 (Woolcott et al., 2018ab; Chapter 3). SCH was scored by excising the skin around the neck of the bird to measure the degree of hemorrhage at the site of cervical dislocation. The skull was lifted, and dura were removed to assess the SDH on the brain.

4.3.5.3. Microscopic assessment of brain trauma

Once the macroscopic assessment was completed, brains and a sample of spinal cord from the site of cervical dislocation were collected from six randomly selected birds in each age group per killing method for microscopic evaluation. The brains and spinal cord samples were placed in

10% buffered formalin for at least 14 days before trimming. For consistency, all trimming was performed by one individual. Three sections of the brain (cerebrum, mid brain and thalamus, cerebellum/hind brain) and a section of the spinal cord at the site of cervical dislocation was sampled (Woolcott et al., 2018b). The tissue sections were embedded in paraffin, cut 4µm and stained with hematoxylin and eosin using standard techniques to make microscopic slides

(Animal health laboratory, University of Guelph). The sections on the slides were assessed microscopically by a veterinary pathologist blinded to the treatments to determine the degree of

SDH and parenchymal hemorrhage (PCH). The degree of SDH and PCH were scored for each microscopic slide using the same procedure as Chapter 3 (Table 4.4).

4.3.6. Statistical analyses

Statistical analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC, USA). Two separate statistical analyses were used. The first analysis was used to compare the behavioural responses and reflex variables of anesthetized birds killed by manual cervical dislocation (aCD)

97 to conscious birds killed by manual cervical dislocation (CD) in order to determine the effects of anesthesia on each outcome variable. Then the response variables of the aCD were statistically compared to those of the anesthetized birds killed by KED (aMCD).

Fisher’s exact tests were performed on the numbers of birds pooled across age groups that showed pupillary light reflex, nictitating membrane reflex, gasping, feather erection, clonic convulsions, tonic convulsions, and cloacal relaxation to find the effect of anesthesia (CD vs aCD) and killing method (aCD vs aMCD). The numbers of birds presenting with luxation, fractures, macroscopic SDH, and transection of spinal cord were similarly analyzed by Fisher’s exact tests.

Generalized linear mixed models (GLMM) were used to analyze the fixed effects of killing method, age, and killing method by age interaction of selected antemortem measurements (time to loss of pupillary light reflex, duration of gasping, time at first feather erection, time to cessation of heart beat). Nictitating membrane reflex, clonic convulsion, tonic convulsion, and cloacal relaxation were excluded from the analysis due to low numbers of birds exhibiting these measures in the anesthetized groups. Least significant means separation was conducted by using the Tukey-Kramer test.

Generalized linear mixed models (GLMM) with multinomial distribution and cumulative logit link functions were used to analyze the effect of the killing method, age, and their interaction on postmortem macroscopic SCH at the site of dislocation, microscopic SDH and PCH of the brain and spinal cord, and dislocation site in the neck (multinomial ordinal data). Odds ratios were

98 computed to compare differences in the levels of fixed effects. Data from the three sections of brain from each bird were pooled and the highest score of the three sections of each brain was used for SDH and PCH analyses (Woolcott et al., 2018 ab; Chapter 3). The same procedure was applied to the SDH and PCH analyses for the spinal cord. Dislocation site was analyzed by conversion to a numerical category (skull–C1 = 1; C1-C2= 2; C2–C3 = 3; etc.); luxation and subluxation were pooled for the analyses.

4.4. Results

4.4.1. Assessment of Ante mortem measures

4.4.1.1. Effects of anesthesia on birds killed by manual cervical dislocation

The proportion of birds presenting with ante-mortem measures following application of the killing methods are given in Table 4.5. All conscious and anesthetized birds demonstrated pupillary light reflex before application of the killing method. Pupillary light reflex was observed in all conscious and anesthetized birds after application of the killing method indicating that there was no anesthesia effect on pupillary light reflex. All conscious birds had nictitating membrane reflex before application of manual cervical dislocation but nictitating membrane reflex was present in only one anesthetized bird before application of manual cervical dislocation

(P=0.001).

The proportion of birds observed gasping was not different between conscious birds killed by manual CD (75%) and anesthetized birds killed by manual CD (45%, P=1). The anesthetic did not affect occurrence of feather erection in conscious birds killed by manual CD (conscious manual CD = 87%, and anesthetized manual CD = 83%, P=1). Some anesthetized birds killed by

99 manual CD did not exhibit clonic convulsions, tonic convulsions, or cloacal relaxation indicating that there was anesthetic effect on these measures (Table 4.5). The number of birds exhibiting clonic convulsions was significantly lower (P=0.001) in the anesthetized manual CD group than the conscious manual CD group. Clonic convulsions with severe wing flapping and paddling were observed in all conscious birds killed by manual CD. However, in 50% of anesthetized birds killed by manual CD, clonic convulsions were absent except at the time of application of the killing method (considered as no clonic convulsions in the analysis, only 2-3 wing flaps were observed at the time of killing). An effect of anesthesia was found in the number of birds that showed tonic convulsions (P<0.001) and cloacal relaxation (P<0.001).

Comparison of times to cessation of pupillary light reflex and heartbeat, time at first feather erection, and duration of gasping in conscious and anesthetized birds killed by manual CD are presented in Table 4.6. Durations were calculated for only the birds that showed the responses.

Overall, time to cessation of pupillary light reflex was longer (P=0.016) in conscious birds than in anesthetized birds. There was a significant age by treatment interaction (P=0.036) with a longer duration for the cessation of pupillary light reflex in conscious birds killed by manual CD compared to the anesthetized birds killed by manual CD birds in the 65-70 week group. Duration of gasping did not differ between the conscious and anesthetized groups (P=0.349). Anesthesia did not affect time to first feather erection (P=0.226). There was no difference in time to cessation of heart beat in conscious versus anesthetized birds killed by manual CD (P=0.137).

Time to cessation of heart beat took longer in the younger birds compared to older ones

(P<0.001), but there was no anesthesia by age interaction (P=0.596).

100

4.4.1.2. Effect of manual CD versus mechanical CD in anesthetized birds

The number of anesthetized birds killed by manual CD or mechanical CD and their responses are shown in Table 4.5. All anesthetized birds, regardless of treatment, presented pupillary light reflex. The number of birds that demonstrated gasping was significantly higher (P<0.001) in anesthetized birds killed by mechanical CD than anesthetized birds killed by manual CD. Feather erection occurred in 83% of anesthetized birds killed by manual CD and in 75% of birds killed by mechanical CD (P=0.723). There was no difference in the number of birds that showed clonic convulsions in anesthetized manual CD and anesthetized mechanical CD groups (P=1). Number of birds exhibiting tonic convulsion (P=1) and cloaca relaxations (P=0.740) was not different in anesthetized birds killed by manual CD and mechanical CD.

Table 4.7 shows mean latencies to loss of antemortem measures or their durations in anesthetized birds killed by manual CD versus mechanical CD in different age groups. Time to cessation of pupillary reflex (P<0.001), duration of gasping (P=0.011), time to first feather erection (P=0.042) and time to cessation of heart beat (P=0.012) were all longer in the anesthetized birds killed by mechanical CD compared to anesthetized birds killed by manual CD in all age groups.

4.4.2. Assessment of postmortem measures

4.4.2.1. Assessment of cervical vertebrae radiographs

Results of radiographic scoring for presence and site of cervical dislocation are tabulated in

Table 4.8. The cervical vertebrae were assessed for any sites of luxation and/or subluxation. All birds killed by manual CD had luxation (complete dislocation) irrespective of whether they were conscious or anesthetized, whereas only 54% of birds killed by mechanical CD had luxation

101

(P<0.001). The site of dislocation was significantly different for treatment (P<0.001) but not for age (P=0.655) or age by treatment interaction (P=0.915) in birds pooled across luxation and subluxation. There was a greater chance of causing dislocation further down the vertebrae than the recommended skull to C1 in chickens killed with mechanical CD than manual CD. Overall

58% of the birds (conscious and anesthetized) killed by manual cervical dislocation had luxation in between the skull and C1. In contrast, the majority of birds (75%) killed with mechanical CD had luxation or subluxation between the C2 and C3 vertebrae.

The number of birds scored with fractures and the different types of fractures observed in the cervical vertebrae are presented in Table 4.9. Fractures were observed in less than 10% of the birds regardless of treatment or whether they presented with luxation or subluxation. There was no difference between manual cervical dislocation and MCD in number of birds with fractures (P

= 0.195). Overall, 4 birds out of 24 (17%) killed with MCD and 3 out of 48 (6%) killed by manual cervical dislocation had fractures of the cervical vertebrae. Figure 5.2 shows cervical luxation (5.2-A), and subluxation with a fracture (5.2-B) in a layer chicken.

4.4.2.2. Macroscopic tissue damage assessment

Spinal cords were transected in all conscious and anesthetized birds killed by manual cervical dislocation in all age groups (Table 4.8). The number of birds that had transected spinal cords was significantly different between manual CD group (both conscious and anesthetized) and the mechanical CD (P<0.001). Spinal cords were intact in 25% of birds killed by mechanical CD;

3/8 in both the 12 and 27-29 week groups, respectively. All spinal cords were transected in birds killed by mechanical CD in the 60-75 week group.

102

Trachea damage was not observed in any of the birds irrespective of the killing method. External skin damage at the site of cervical dislocation was greater in the birds killed with mechanical CD

(12 week: 25%; 27-29 week:75%; 60-65 week:50%) than the birds killed by manual CD

(P<0.001). External skin damage at the site of cervical dislocation was observed in 8% of birds killed by manual CD (conscious and anesthetized combined). Mechanical CD resulted in external bleeding in 33% of the birds pooled across age groups whereas 4% of the pooled manual

CD birds had external bleeding (P=0.002).

Subcutaneous hemorrhage (SCH) at the site of dislocation was different among the treatments

(P<0.001) but there was no age effect (P=0.76) or age by treatment interaction (P=0.23) (Table

4.10). Except for one, all conscious and anesthetized birds killed by manual cervical dislocation had a score of 2 or more for SCH. The highest score of 4 was found in over half of the birds killed by manual cervical dislocation (conscious= 54%, anesthetized= 56%). Score 4 was absent in all birds killed by the mechanical CD device. Macroscopic subdural hemorrhage (SDH) (score

1) in the brain was observed in only one bird out of 24 (12 w.o.) killed with MCD. Overall, 21% of the birds killed by manual CD (10/48) had macroscopic SDH on the brain (P=0.086).

4.4.2.3. Microscopic evaluation

Table 4.11 gives descriptive statistics for SDH and PCH (all scores > 0) in the brain of birds killed by the different methods in the three age groups. None of the birds in any group had SDH of the cerebrum. Birds killed by manual CD (conscious and anesthetized combined) had SDH both of the mid brain (30%) and the hind brain (63%) whereas only one bird out of 24 killed with mechanical CD had SDH of the hind brain. Parenchymal hemorrhage (PCH) was observed in

103 few birds overall. Six of 36 killed by manual CD had PCH in the hindbrain. No PCH was observed in any of the birds killed by MCD. Although more birds killed by manual CD showed microscopic damage compared to those killed by MCD, killing method, age, and method ˟ age interaction did not affect brain SDH or PCH scores.

The number of birds with different scores for subdural hemorrhage (SDH) and parenchymal hemorrhage (PCH) in spinal cord are presented in Table 4.12. Over 80% of all birds sampled had

SDH of the spinal cord (method P=0.168), but there was an effect of age (P=0.006), and tendency for an age by killing method interaction (P=0.057). Birds at 12 weeks had 22 times of a greater chance of having lower scores than birds at 65-70 weeks and 6 times greater chance of having lower scores than birds at 27-29 weeks. Birds at 27-29 weeks had 4 times greater chance of having lower scores for SDH of the spinal cord than birds at 65-70 weeks. There was a trend of having lower scores of SDH of the spinal cord in the birds killed by mechanical CD in younger age (12-29 weeks) groups. All the birds killed by mechanical CD had a score of 3 or 4 at

65-70 weeks. PCH in the spinal cord was significantly different among the killing methods

(P=0.041). Age and age by killing method interaction tended to be significant for the PCH

(P=0.088 and P=0.071 respectively). Conscious and anesthetized birds killed by manual CD had a greater chance of having lower scores for PCH than birds killed by mechanical CD. Of the 18 anesthetized birds killed by mechanical CD, 13 birds (72%) had mild (score 2) to marked (score

4) PCH. Mild to marked PCH in the spinal cord was found in 50% of birds in conscious manual

CD group and 27% in anesthetized manual CD group.

104

4.5. Discussion

Due to ethical concerns about testing the mechanical device on conscious birds, general anesthesia was induced by using a combination of medetomidine and ketamine in order to abolish any pain associated with the killing technique. Brain stem reflexes, behavioural responses, and physiological parameters could be affected by the anesthetics used, which is why we chose to study both the effect of the anesthesia and the effect of the method.

In the current study medetomidine was used in combination with ketamine. Medetomidine is rapidly absorbed causing reliable sedation, analgesia, muscle relaxation, and anxiolysis (Sinclair,

2003). However, medetomidine decreases cerebral blood flow (Zornow et al., 1990; Keegan et al., 1995) and heart rate (Cullen, 1996). Thus, medetomidine is combined with other anesthetics to reduce the side effects (Sinclair, 2003). Ketamine is often used in combination with medetomidine for avian anesthesia (Paul-Murphy et al., 2001). The major effects of ketamine administration involve the central nervous system. Ketamine can affect eye reflexes and degree of muscle tone and poses a problem when assessing the level of sedation or anesthesia (Hass et al., 1992). Therefore, our anesthesia protocol had the potential to affect both antemortem measures (eye reflexes and heart rate) and postmortem measures (brain hemorrhage), and needed to be evaluated.

The anesthetic protocol used did not abolish the pupillary light reflex, which allowed us to use this as an indicator of brain stem death in all birds. However, latency to loss of pupillary light reflex was longer when birds were anesthetized. This is different than Woolcott et al., (2018b) who found that pupillary light reflex was no longer in turkeys using the same anesthesia protocol

105 as in this study. The anesthetics used in the current study also abolished or reduced four of the other observed behavioural responses and reflexes (nictitating membrane reflex, clonic convulsions, tonic convulsions, and cloacal relaxation) precluding their use for comparing killing methods. Woolcott et al. (2018b) reported that presence of the nictitating membrane reflex, presence of clonic convulsions, and duration of gasping were affected by the same anesthetic protocol in turkeys. Sandercock et al. (2014) observed nictitating membrane reflex until after respiratory arrest and brain death in laying hens anesthetized with sevoflurane, and suggested that the longer time to cessation of nictitating membrane reflex may have been due to combined effects of the anesthetic and sedative they used. In the current study, nictitating membrane reflex was not present in many of the chickens as a result of the anesthetic protocol. We previously found that onset of tonic convulsion and cloacal relaxation correlated with cessation of heart beat in laying hens and can be used as a practical measure of clinical death (Chapter 3). However, our current results revealed that both tonic convulsions and cloacal relaxation were reduced in many of the anesthetized birds. Thus, latency to pupillary reflex, nictitating membrane reflex, clonic convulsions, tonic convulsions, and cloacal relaxation were affected by our anesthetic protocol.

The anesthetics did not affect time to cessation of heart beat in chickens. Since latency to loss of pupillary reflex was shortened by the anesthetic, our estimate for time of brain death using the anesthetized bird as a model is conservative.

The main objective of the current study was to assess the efficacy of mechanical cervical dislocation using the KED (mechanical CD) in comparison to manual cervical dislocation

(manual CD). The results of the current study reveal that time to loss of pupillary light reflex, an indicator of brain stem death, was longer in the anesthetized birds killed with mechanical CD

106 than by manual CD. Previous studies also reported shorter latency to loss of eye reflexes in poultry successfully killed with manual cervical dislocation in comparison to birds killed by mechanical CD using the KED (conscious broilers: Jacobs et al., 2019; anesthetized turkeys:

Woolcott et al., 2018b). Gregory and Wotton (1990a) reported that birds killed by mechanical

CD (using Semark neck pliers) exhibited longer time to loss of visual evoked responses compared to manual CD, and concluded that cervical dislocation by stretching was more effective than cervical dislocation by crushing. Hernandez (2018) also reported a longer time to onset of isoelectric EEG indicating brain death in layers (60 weeks of age; anesthesia was induced by isoflurane at 5% concentration with oxygen via a face mask) killed by mechanical

CD by KED in comparison to manual CD. Results from the current study reveal that time to cessation of heart beat and time at feather erection were also longer in the anesthetized birds killed by mechanical CD compared to manual CD. Heard (2000) suggested sudden feather erection in anesthetized birds as an indicator of cardiac arrest or reduced blood flow to the heart.

In the current study, >75% of birds exhibited feather erection irrespective of the killing method.

Recent studies reported that there is no association between time at feather erection and cardiac arrest in chickens (Hernandez 2018; Chapter 3). Thus, the delayed feather erection in the birds killed by mechanical CD is difficult to explain.

Ideally, cervical dislocation separates the cervical vertebrae in between the skull, and the atlas

(C1), completely transecting the spinal cord and disrupting blood vessels. In the current study, the vast majority of the birds killed by manual CD had dislocations in between the skull and atlas

(C1) or between C1-C2. However, the KED resulted in a majority of dislocations between C2 and C3. Thus, manual CD caused more cranial dislocations than the KED. Previous studies also

107 reported similar results indicating that manual cervical dislocation resulted in the majority of dislocations between the skull and atlas (C1) in chickens (Bader et al., 2014; Martin et al., 2016,

2017, 2018a) and turkeys (Woolcott et al., 2018b). Martin et al. (2018b) reported that more cranial dislocation was associated with improved kill success and more rapid loss of reflexes compared to more caudal dislocations for cervical dislocation.

Anatomical damage to the top of the spinal cord and the base of the brain stem is possible with more cranial dislocation, which is associated with spinal cord concussion, neurogenic shock, and loss of consciousness in humans (Dumont et al., 2001; Harrop et al., 2001). Manual CD resulted in all birds having cervical luxation, more cranial luxation locations, more spinal cord transections, greater SDH and PCH in the brain stem, and shorter time to loss of brain stem death and cardiac arrest. Except for one bird, none of the birds killed with mechanical CD using the

KED had SDH or PCH in the brain, indicating that KED did not cause severe brain trauma.

Previous studies suggested that immediate insensibility and irreversible loss of vital functions are associated with SDH and PCH in the brain (Erasmus et al., 2010b; Bader et al., 2014). We suggest the absence of SDH and PCH may be the reason for the observed longer time to loss of pupillary light reflex and clinical death in the birds killed by mechanical CD.

The majority of birds killed by either manual CD or mechanical CD showed transected spinal cords and SDH and PCH in the spinal cord, indicating both methods were sufficient to cause traumatic damage to the spinal cord. The spinal cord was transected in all 65-70 week old birds killed by mechanical CD irrespective of whether there was luxation or subluxation. This indicates that KED model-C caused severe damage to the cervical area of the spinal cord of 65-

108

70 week old roosters. Some muscles responsible for inspiration are innervated by cervical nerves and the accessory cranial nerve (CN X1) (DeTroyer and Estenne, 1984). Some authors reported the evidence that damage to the cervical area of the spinal cord impaired the functioning of these respiratory muscles resulting in hypoxia due to failures in respiration process, and finally cause death in humans (Winslow and Rozovsky, 1983). A higher number of birds killed by mechanical

CD exhibited longer duration of gasping in comparison to manual CD. Therefore, we suggest that hypoxia resulting from spinal cord trauma and leading to cerebral ischemia could be a main reason for the observed gasping reflex and death in mechanically cervical dislocated birds; they also had minimal levels of brain trauma.

There were few fractures in cervical vertebrae of the chickens irrespective of the killing method.

This is in contrast to previous studies that reported prevalence of fractures in poultry killed by manual CD and mechanical CD. Woolcott et al. (2018b) found considerably more fractures in 3 week old turkeys killed by mechanical CD (KED model-S), with 78% having transverse or comminuted fractures or fragmentation. In their study, turkeys killed via manual CD also had fractures, although fewer (44%) than with KED. Bader et al. (2014) also reported different types of fractures (transverse, comminuted, longitudinal) in chickens by manual cervical dislocation and turkeys killed by mechanical cervical dislocation using a nonpenetrating forcep with three blunt shear arms.

Less external blood loss and fewer skin tears are preferred in the commercial environments due to biosecurity and esthetic concerns. Similar to the findings of Woolcott et al. (2018b) in turkeys and Jacobs et al. (2019) in broiler chickens, the KED resulted in more external skin damage in

109 layer chickens than manual CD. Other important factors when assessing killing methods for practical on-farm use, are safety for stock people and ease of training. Cervical dislocation is easy to perform manually, practical in any emergency killing, and can be applied immediately during barn walkthroughs. However, manual CD is associated with variability in operators’ performance (Martin et al., 2018b). We assume a person can be easily trained on operating the

KED device, and there is little risk associated with mis-use of the device for the operator or technical difficulties.

4.6. Conclusion

Compared to manual cervical dislocation, mechanical CD using the KED caused more external skin trauma but less trauma to the brain with more incomplete dislocations located in the lower regions of the cervical vertebrae. Birds killed with the KED also showed longer times to loss of pupillary light reflex and cessation of heart beat. Overall, mechanical cervical dislocation by

KED resulted in lower efficacy in comparison to manual cervical dislocation as on-farm killing method for layer chickens. The anesthetic protocol abolished or reduced clonic convulsions, nictitating membrane reflex, tonic convulsions, and cloacal relaxation in the current study suggesting that these behavioural responses or reflexes are not useful as an approximate measure of brain death in cases where ketamine and medetomidine have been administered in chickens.

110

Table 4. 1: Strain, sex, body weight and sample sizes for the different age classes of birds used in the study

Age Killing Number Strain Sex Body weight (week) method of birds (Kg) (Avg wt ± SD) 12 CD 8 White Leghorn All females 1.00 ± 0.10 aCD 8 White Leghorn All females 0.94 ± 0.09 aMCD 8 White Leghorn 2 Males + 6 Females 1.02 ± 0.62

27-29 CD 8 White Leghorn 1 Male + 1 Females 1.77 ± 0.03 Columbian Rock 1 Male + 5 Females 1.75 ± 0.36 aCD 8 White Leghorn 3 Males + 3 Females 1.79 ± 0.29 Columbian Rock 2 females 2.03 ± 0.29 aMCD 8 White Leghorn 3 Males + 4 Females 1.86 ± 0.35 Columbian Rock 1 female 1.98

65-70 CD 8 White Leghorn All males 2.16 ± 0.18 aCD 8 White Leghorn All males 2.10 ± 0.19 aMCD 8 White Leghorn All males 2.29 ± 0.29 CD = Conscious manual cervical dislocation aCD = Anesthetized manual cervical dislocation aMCD = Anesthetized mechanical cervical dislocation by KED

111

Table 4. 2: List of ante-mortem assessment measures, description, and procedure use, recorded in order of observation after application of each killing method (based on Chapter 3)

Measures Description Procedure

Pupillary light reflex Constriction of the pupil in Light from a medical penlight response to light was directed into the eye and pupil constriction was examined

Nictitating membrane reflex Transient closure of the The medial canthus of the eye or nictitating membrane in the cornea was lightly touched response to mechanical with a fingertip stimulation

Gasping Paroxysmal opening of the beak Visual observation of paroxysmal opening of the beak

Feather erection Sudden erection of feathers, not Visual observation of first in response to external stimuli occurrence of feather erection

Clonic convulsions Rapid, uncoordinated movement Visual observation of rapid wing of the body and wings flapping and foot paddling

Tonic convulsions Muscle rigidity with the legs Visual observation of the time and wings outstretched of onset of legs and neck outstretched

Cloacal relaxation Cloacal opening following Visual observation for cloacal contractions of cloaca opening following contractions

Cardiac arrest Cessation of heart beat Auscultation by using a stethoscope

112

Table 4. 3: Definitions for the terminology used in radiograph evaluation

Terminology Definition Luxation Complete dislocation of two vertebrae at the articular process joints. Subluxation Partial (incomplete) dislocation of two vertebrae at the articular process joints. Fracture type Sagittal Parallel to the long axis of the vertebra on midline. Trans (transverse) Perpendicular to the long axis of the vertebra. Articular processes Fractures of the articular processes of a vertebra. Dens Fracture of the dens (tooth-like process that projects from the cranial aspect of the centrum of the axis (C2) to articulate with the atlas (C1). Crushed vertebral bodies Multiple fractures of a vertebra that cannot be classified as a fragment, articular process or dens fracture.

113

Table 4. 4: Gross and microscopic pathology scoring criteria for macroscopic, and microscopic hemorrhage (Woolcott et al., 2018ab; Chapter 3)

Score Macroscopic Microscopic Subcutaneous or subdural hemorrhage Subdural or parenchymal hemorrhage

114

Table 4. 5: Number of birds presenting with ante-mortem measures following application of the killing methods.

Method CD aCD aMCD N=24 N=24 N=24

Pupillary light reflex 24 24 24

Nictitating membrane reflex 16a 1b, d 9c

Gasping 18 11d 24c

Feather erection 21 20 18

Clonic convulsions 24a 12b 11

Tonic convulsions 15a 1b 1

Cloacal relaxation 24a 7b 5

a,b Different superscripts indicate numbers of birds presenting the measure is different between CD and aCD according to Fishers’ exact tests (P<0.005) c,d Different superscripts indicate numbers of birds presenting the measure is different between aCD and aMCD according to Fishers’ exact tests (P<0.005) CD = Conscious manual cervical dislocation aCD = Anesthetized manual cervical dislocation aMCD = Anesthetized mechanical cervical dislocation by KED

115

P value

Age CD aCD (week) Age Anesthesia Age × Anesthesia Time to 12 81 ± 7 70 ± 7 0.8507 0.0166 0.0369 loss of 27-29 77 ± 5 80 ± 5 pupillary 65-70 96 ± 7a 62 ± 7b reflex Overall 85 ± 3 71 ± 3

Gasping 12 22 ± 7 50 0.1497 0.3490 0.7706 duration 27-29 57 ± 8 68 ± 8 65-70 73 ± 13 74 ± 17 Overall 51 ± 7 64 ± 11

Time to 12 82 ± 15 104 ± 16 0.8942 0.2264 0.3785 first feather 27-29 78 ± 19 114 ± 17 erection 65-70 93 ± 15 84 ± 11 Overall 85 ± 10 101 ± 10

Time to 12 191 ± 20 173 ± 20 0.0005 0.1377 0.5960 cessation 27-29 122 ± 9 119 ± 9 of heart 65-70 152 ± 12 119 ± 12 beat Overall 155 ± 3 137 ± 8

Durations calculated for only the birds that showed the responses Superscript letters show the significant difference (P<0.05) CD = Conscious manual cervical dislocation aCD = Anesthetized manual cervical dislocation Bolded numbers indicate significant results (P<0.05)

116

P value Age aCD aMCD Age Method Age × Method Time to 12 70 ± 14 142 ± 14 0.2014 <0.0001 0.3091 loss of 27-29 80 ± 8 118 ± 8 pupillary 65-70 62 ± 10 107 ± 10 reflex Overall 71 ± 7b 123 ± 7a

Gasping 12 50 117 ± 18 0.5367 0.0109 0.7338 duration 27-29 63 ± 13 102 ± 10 65-70 70 ± 20 132 ± 16 Overall 64 ±16b 117 ± 9a

Time to first 12 104 ±25 146 ± 30 0.7083 0.0423 0.5599 feather 27-29 114 ± 10 125 ± 11 erection 65-70 87 ± 15 135 ± 14 Overall 101 ± 12b 135 ± 11a

Time to 12 173 ± 23 172 ± 23 0.0998 0.0125 0.0945 cessation of 27-29 119 ± 10 153 ± 10 heart beat 65-70 119 ± 13 193 ± 13 Overall 137 ± 9b 172 ± 9a

aCD = Anesthetized manual cervical dislocation aMCD = Anesthetized mechanical cervical dislocation by KED Durations calculated for only the birds that showed the responses Superscript letters show the significant difference (P<0.05) Bolded numbers indicate significant results (P<0.05)

117

Luxation Sub-Luxation Spinal cord transection

Age Method Total # birds Sk-C1 C1-C2 C2-C3 C3-C4 # birds Sk-C1 C1-C2 C2-C3 C3-C4 Number number of with with sub- of birds birds luxation luxation assessed

12 CD 8 8 8 0 0 0 0 0 0 0 0 8 aCD 8 8 7 0 0 1 0 0 0 0 0 8 aMCD 8 5 2 1 2 0 3* 1 1 3 0 5

27-29 CD 8 8 1 7 0 0 0 0 0 0 0 8 aCD 8 8 3 5 0 0 0 0 0 0 0 8 aMCD 8 5 0 0 5 0 3 0 0 1 2 5

65-70 CD 8 8 4 3 1 0 0 0 0 0 0 8 aCD 8 8 5 3 0 0 0 0 0 0 0 8 aMCD 8 3 0 0 3 0 5 0 0 4 1 8

1Number of birds presenting luxation and subluxation were pooled for statistical analysis. Results of statistical analyses are given in the text. CD = Conscious manual cervical dislocation aCD = Anesthetized manual cervical dislocation aMCD = Anesthetized mechanical cervical dislocation by KED *Two birds had sub-luxation in multiple places

118

Table 4. 9: Results of radiographic scoring on number of birds with fractures and the types and locations of fractures in conscious and anesthetized chickens.

Age (weeks) Method1 Number of birds with Fracture type Location fractures 12 CD (n=8) 0 - aCD (n=8) 0 - aMCD (n=8) 1 Trans C3 1 Crushed vertebral bodies C3 and C4

27-29 CD (n=8) 1 Small fractures Dens of C2 aCD (n=8) 1 Trans Dens of C2 1 Small fractures Dens of C2 aMCD (n=8) 0 -

65-70 CD (n=8) 0 - aCD (n=8) 0 - aMCD (n=8) 2 Articular processes C3

1CD = Conscious manual cervical dislocation aCD = Anesthetized manual cervical dislocation aMCD = Anesthetized mechanical cervical dislocation by KED

119

Table 4. 10: Macroscopic evaluation of subcutaneous hemorrhage (SCH) at the site of dislocation. Number of birds with each score are indicated

Method1 Age Score P value 0 1 2 3 4 Method Age Method*Age SCH CD 12 0 0 1 2 5 0.0002 0.7603 0.2330 27-29 0 0 1 3 4 65-70 0 1 3 0 4

aCD 12 0 0 2 2 4 27-29 0 0 0 0 8 65-70 0 0 2 5 1

aMCD 12 0 5 2 1 0 27-29 0 2 0 6 0 65-70 0 2 5 1 0

1CD = Conscious manual cervical dislocation aCD = Anesthetized manual cervical dislocation aMCD = Anesthetized mechanical cervical dislocation by KED

120

Age Method1 Subdural hemorrhage Parenchymal hemorrhage (SDH) (PCH) (Week) Cerebrum Mid Hind Cerebrum Mid Hind brain brain brain brain 12 CD 0 0 5 0 0 1 aCD 0 0 4 0 0 0 aMCD 0 0 0 0 0 0

27-29 CD 0 5 6 0 0 2 aCD 0 1 3 0 0 2 aMCD 0 0 1 0 0 0

65-70 CD 0 1 4 0 0 0 aCD 0 4 0 0 0 1 aMCD 0 0 0 0 0 0

Six randomly selected brains were assessed for each device in each age group. 1CD = Conscious manual cervical dislocation aCD = Anesthetized manual cervical dislocation aMCD = Anesthetized mechanical cervical dislocation by KED

121

Method1 Age Score P value 0 1 2 3 4 Method Age Method*Age SDH of CD 12 4 2 0 0 0 0.1680 0.0006 0.0572 Spinal 27-29 1 1 2 2 0 cord 65-70 0 0 2 2 2 aCD 12 2 0 2 1 1 27-29 0 0 2 3 1 65-70 0 2 1 2 1 aMCD 12 1 3 1 1 0 27-29 2 0 4 0 0 65-70 0 0 0 4 2 PCH in CD 12 3 1 1 0 1 0.0407 0.0885 0.0712 spinal 27-29 2 1 2 2 0 cord 65-70 2 0 2 1 0 aCD 12 2 0 1 3 0 27-29 3 3 0 0 0 65-70 2 3 0 1 0 aMCD 12 0 2 0 3 0 27-29 3 0 4 0 0 65-70 0 0 1 3 2

Randomly selected 6 brains with spinal cords were assessed for each killing method in each age group. 1CD = Conscious manual cervical dislocation aCD = Anesthetized manual cervical dislocation aMCD = Anesthetized mechanical cervical dislocation

122

Figure 4. 1: A- Koechner Euthanasia Device (KED model-C): S-single side blade; D-double angle blade. B-application of KED model-C in a 65 week old anesthetized rooster.

123

124

CHAPTER 5

Aversion to CO2 Gas in Laying Hens Using Approach-Avoidance

and Conditioned Place Avoidance Paradigms

This manuscript is in preparation for submission to Poultry Science with the following authors: R. M.A.S. Bandara, T. M. Widowski, S. Torrey, R. L. Parsons, and S. T. Millman

125

5.1. Abstract

Carbon dioxide (CO2) has been evaluated as an inhalant euthanasia agent in poultry but animal welfare concerns have been raised on associated pain and aversion. This study assessed the aversion to four different CO2 concentrations (25%, 35%, 50%, 70%) in laying hens based on approach avoidance test and conditioned place avoidance test. An apparatus consisting of two identical chambers separated by a sliding door and a curtain was used. The control chamber (CC) maintained ambient air conditions; the treatment chamber (TC) maintained predetermined CO2 concentrations. Twelve laying hens (1.2 ± 0.1 kg; 24 week-old) were individually trained to access rewards in the TC. Hens were randomly assigned to the first CO2 treatment and then assigned to the remaining rounds to balance treatments. Each round of testing consisted of one

Gas Day, preceded by one Baseline Day (1% CO2) and followed by one Washout Day (1% CO2).

On Gas Days, the TC was maintained at one of four CO2 levels. Hens were considered to show aversion if they pushed the curtain and/or inserted their heads into the TC without fully entering.

Hens were considered to display CPA if they did not enter the TC on Washout Day after entering the TC on the previous Gas Day. Eight of 9 hens entered the TC at 25% and 35% CO2, 6 of 11 entered at 50%, and 2 of 7 entered at 70% CO2. Head shaking, open mouth breathing, ataxia and loss of posture were observed in all hens at all concentrations. Latency to onset of ataxia (less than 15s; P=0.09) tended to decrease and loss of posture (less than 25s; P=0.02) decreased with increased CO2 concentrations. Some hens displayed conditioned place avoidance following loss of posture at 25% and 35%, and more hens at 50% CO2.We suggest 50% and above CO2 concentrations are more aversive to hens than 25% and 35%. Therefore, euthanasia with CO2 should target concentrations less than 50% to achieve welfare advantages in layer hens.

126

5.2. Introduction

Over the last three decades, carbon dioxide (CO2) has been evaluated by various researchers as a method of gas stunning and as an inhalant euthanasia agent for poultry using either static

(prefilled) or gradual fill techniques. Animal welfare concerns have been raised regarding pain, distress and associated aversion experienced by animals exposed to high concentrations of CO2 from the time of gas introduction to the onset of insensibility (Raj, 1996; McKeegan, 2005;

Sandilands et al., 2011). To be considered acceptable as an inhalant euthanasia agent, established guidelines require that the concentration not be aversive to the species, and that the agent induce rapid loss of sensibility without recovery (AVMA, 2013; NFACC, 2017).

Carbon dioxide is an anesthetic gas (Kohler et al., 1999) that produces insensibility by suppressing afferent transmission of sensory stimuli to the brain (Raj et al., 1992). Carbon dioxide causes death by inducing hypercapnia (excessive CO2 in the bloodstream) and hypoxia.

Carbonic acid is formed when CO2 contacts moisture on respiratory and ocular membranes, which can cause pain and distress in animals due to stimulation of nociceptors in respiratory and ocular membranes (Anton et al., 1992; Chen et al., 1995; Feng and Simpson, 2003; McKeegan et al., 2004).

Different species can detect different gas levels through gustation, olfaction and nociception

(McKeegan et al., 2005). Humans can detect CO2 at the 10% level even though it is not pungent to inhale (Patterson et al., 1962). Pain threshold level for CO2 in humans has been identified at concentrations between 40% and 55% (Anton et al., 1992). Laying hens appear to be able to detect CO2 concentrations as low as 7.5% (Raj and Gregory, 1991). Different behaviourl

127 responses to CO2 have been suggested as indicators of distress, aversion and pain perception.

Gerritzen et al. (2000) suggested that gasping and head shaking in broiler chickens exposed to

60% CO2 and gas mixtures containing 30% and 40% CO2 as possible indicators of breathlessness and the pungency of gas mixtures. The avian nociception threshold for CO2 is in the 40-50% which is in the same range as reported in humans.

Cognitive paradigms of approach avoidance (AA) have been used to measure the aversion to a negative experience associated with different gas types (CO2 in ewes: Grandin et al., 1986; NH3 in laying hens: Kristensen et al., 2000; CO2 and Ar in laying hens: Webster and Fletcher, 2004).

These tests directly ask the animal whether they are willing to enter an environment filled with an inhalant agent. In an AA test, CO2 (potentially causing a negative experience) is paired with a positive stimulus (reward) to elicit motivational conflict, and the motivation to approach the reward (e.g. feed and enrichment) is compared against the motivation to avoid gas exposure

(layer chickens: Raj and Gregory, 1991, Webster and Fletcher, 2004; broilers: Sandilands et al.,

2011).

Previous studies that assessed aversion in chickens to concentrations of 30% CO2 and higher by using AA tests have produced conflicting results. Webster and Fletcher (2004) tested aversion to

30%, 45% and 60% CO2 and concluded that there is no welfare disadvantage to the use of up to

60% CO2. They assessed avoidance of CO2 in a test apparatus consisting of a raised entry chamber connected by a descending chute to a lower chamber into which gas was injected. Hens had to go down a chute to the test chamber and were able to detect gas before reaching the chamber. Hens made a number of stops and retreats during the approach to the lower chamber,

128 so the level of gas they avoided was uncertain. Gerritzen et al. (2000) assessed aversion to 60%

CO2 in broiler chickens by using an AA test. In this study, chickens were isolated from the flock and then allowed to reinstate social contact by walking through a CO2-filled tunnel towards the companions placed behind a glass panel. No difference was found in numbers of birds that entered the tunnel for different CO2 concentrations, and the authors concluded that the birds could not detect CO2 concentrations. However, social isolation is highly distressing for chickens, and the motivation for social reinstatement may have confounded the results. Additionally, very few control birds entered the tunnel limiting the value of this test.

Previous studies also evaluated the aversion to CO2 in chickens based on interruption of ongoing feeding behaviour when increasing CO2 concentration. Only sub-stunning levels of CO2

(5%, 7.5%, 10%) were assessed by Raj and Gregory (1991) who reported that hens spent less time in a feeding chamber when the concentration was above 5%. McKeegan et al. (2006) studied the immediate aversion to CO2 (10%, 25%, 40%, 55%, 70%) in broiler chickens. The chickens were exposed to gas for 10 s and the authors concluded that the birds demonstrated a moderate immediate aversion to 40% or above CO2 based on the time at cessation of feeding and withdrawal from the food dish. Sandilands et al. (2011) assessed feeding withdrawal time of broiler chickens in higher CO2 concentrations of 50%, 55%, and 60%, and concluded that all

CO2 concentrations were aversive.

Conditioned place avoidance (CPA) tests measure animal responses when they are re-exposed to an environment previously associated with a negative experience. CPA tests have been used to evaluate the responses in goats (Withrock, 2015), swine (KC et al., 2016), and turkeys (Bandara

129 et al., unpublished) exposed to different CO2 concentrations. While they did not measure CPA directly, Webster and Fletcher (2004) observed that laying hens exposed to different gas concentrations had significantly longer latencies to leave the control chamber on the day following testing (retraining day) indicating that their previous experience affected subsequent behavioural responses.

Replication of previous studies on assessment of aversiveness to CO2 in chickens using different testing protocols, and assessing different CO2 concentrations as possible euthanasia agents are important to refine euthanasia protocols. Therefore, our first objective was to assess the degree of aversion in laying hens exposed to four different CO2 concentrations (25%, 35%, 50%, 75%) based on AA test and distress behaviours. We hypothesized that birds would avoid entering the treatment chamber when CO2 concentrations were aversive. The second objective was to assess the degree of aversion associated with previous exposure and loss of sensibility based on CPA and distress behaviours. We hypothesized laying hens would avoid the environment previously associated with aversive CO2 concentrations and where loss of sensibility was experienced. The third objective was to estimate the time to insensibility by assessing time to loss of posture in laying hens exposed to different CO2 concentrations.

5.3. Methods

The protocol for this experiment was approved by the Institutional Animal Care and Use

Committee (IACUC) at Iowa State University.

130

5.3.1 Animals, Housing, and Management

Twenty 36-week-old laying hens (White Leghorn) were obtained from a research flock at Iowa

State University. Upon arrival at Laboratory Animal Research (LAR) facility at Iowa State

University, the birds were housed in a single pen in a temperature-controlled room (2.5 m x 3.65 m) in the LAR facility. Room temperature was 24°C and relative humidity was maintained between 60 – 65% with a photoperiod of 12 hrs (6.00 a.m to 6.00 p.m). The concrete floor of the room was covered with wood shavings. Five nest boxes were provided inside the pen.

Environmental enrichment included two metal chains, two hanging compact discs, and a perch.

Birds were provided with ad libitum access to a commercial layer diet (Heartland Co-op, Prairie

City, IA) and water. Birds were individually identified by using colored leg bands. During a 4- week acclimation period, pullets were habituated to the researcher’s presence and handling, and birds were familiarized with features of the test apparatus (e.g. the same plastic curtain made of transparent PVC strips, which was used in the test apparatus) and food rewards (live meal worms) within the home pen.

5.3.2. Experimental room and equipment

The experiment was conducted in a room located in the same building where the hens were housed. Tests were performed using a custom designed apparatus, as described by Withrock

(2015) and KC (2018). In brief, the apparatus consisted of two identical chambers (61cm x 61cm x 76 cm) separated by a sliding door and a plastic curtain made of transparent PVC strips (Figure

5.1). Birds were able to see through the curtain and could easily push through it to enter the treatment chamber. Each chamber was outfitted with a hinged door through which birds could be placed or removed from the chamber. The doors and top panels were made of Plexiglass to

131 facilitate behavioural observations. The side panels of the box were made of opaque, hard plastic to avoid the researchers being seen by the birds. Chambers floors were covered with black rubber mats.

The test apparatus was engineered for independent atmospheric control in each chamber, using recirculating positive pressure systems to maintain separate designated CO2 concentrations. This included two small square axial fans (Model 4WD47, Dayton Electric Mfg. Co. IL) that drew air into the right (double) sidewall of the chamber where CO2 gas was injected and mixed into the air stream. The CO2 enriched air was then delivered to the chamber through perforations evenly distributed in the left side wall of the treatment chamber (TC). A continuous flow of ambient air was similarly introduced into the control chamber (CC). To maintain the separate CC and TC environments, a negative pressure exhaust sink was located between the two chambers where any excess CO2 was evacuated.

To further facilitate maintaining CO2 concentration at or near ambient levels (<1%) within the

CC during all treatments, the top panel was opened to release air and was covered with a plastic mesh (5.08 x 5.08 cm) to prevent birds escaping. The opened top panel also facilitated recording of vocalizations inside CC.

To ensure that hens would be highly motivated to enter TC, enrichments were provided to stimulate foraging and exploration. The rubber mat inside the treatment chamber (TC) was covered by a waterproof disposable underpad (Pet All Star Training Pad, Pet All Stars, ASIN,

B014RDSMFU) to facilitate cleaning between birds. Wood shavings were scattered over the

132 underpad. Enrichments included a compact disc, a plastic lid (diameter = 12.7 cm), some plastic strings (green and white color, 30 cm in length, diameter = 2.5 mm), and a gray plastic carpet (12 x 50 cm). Green vegetables (broccoli and spinach, approximately 10g together) were scattered over the plastic carpet to facilitate foraging. Commercial feed and live meal worms in plastic cups were provided inside the TC to attract the birds. We assumed that hens would prefer the enriched environment in the TC in the absence of an aversive stimuli.

5.3.3. Experimental design

Individual birds were initially assigned to receive each of four concentrations of CO2 (25%, 35%,

50%, 70%) in four rounds, and each bird acted as its own control. Birds received one gas treatment per day, and were randomly assigned to gas treatment in the first round, then systematically assigned to remaining gas treatments in subsequent rounds (Table 5.2). A treatment round consisted of three consecutive days. On Baseline Day, both CC and TC were maintained at ambient air conditions (<1% CO2). On Gas Day, TC was maintained at the predetermined CO2 concentration, while CC was maintained at near ambient levels (<1%). On

Washout Day, TC and CC were again, both maintained at ambient air conditions (<1% CO2).

5.3.4. Training procedures

Home pen training was conducted during the 4-week acclimation period to habituate the birds to researchers, handling, and to the act of acquiring meal worms by pushing through a curtain. A covered area of 45 x 45 cm was established inside the home pen and the entrance to this area was covered with the same curtain material used in the test apparatus. Meal worms were provided

133 inside this covered area and hens were trained to enter the covered area by pushing through the curtain.

The hens were individually trained to the experimental procedures within the apparatus. In all tests and training sessions, feed was withheld from birds for one hour prior to training to increase their motivation to access the food rewards in TC. As the first step, the bird was introduced into the CC. The sliding door separating the two chambers was kept opened, after nearly 60 s the bird was gently pushed through the slit in the middle of the curtain, and then allowed five minutes to access the feed rewards and enrichments in the TC. On the second day, the bird was placed in

CC with the sliding door closed. After 5 minutes, the sliding door was opened, allowing access to TC. If the bird did not enter TC within five minutes, a researcher gently guided it through the curtain. Upon entry to the TC, the bird was again, provided five minutes during which it could access the enrichments in the TC and freely move about the chambers. After the 5 minutes, the bird was returned to the home pen.

All hens were trained individually in the apparatus for 5-6 consecutive days to achieve the enrollment criteria for the testing phase. Training was concluded when the bird learned to independently enter the TC on two consecutive days.

5.3.5. Testing protocol

Testing order of the birds was assigned based on their level of motivation to access the feed and enrichment in treatment chamber during the training sessions. Hens who accessed the enrichments fastest were placed first in the testing order in order to ensure longer feed

134 deprivation for the less motivated hens. A health assessment was performed on each bird to identify any wounds on the body prior to the test, and birds who were in good health were eligible for the testing phase.

The testing procedures on Baseline Day and Washout Day were identical to those described for final training with the exception that birds that chose to not enter TC after five minutes were not guided through the sliding door; the test was concluded, and the bird was returned to the home pen. On Gas Day, the TC was prefilled and stabilized at the designated CO2 treatment concentration prior to placing each bird in the CC. If the bird did not enter the TC within the given five minutes, the test was concluded, and the bird was returned to the home pen. When loss of posture (LOP) occurred inside the TC, the bird was immediately removed and placed into a recovery pen located within the testing room. During recovery, affected birds were observed until they regained full sensibility and standing posture. Feed and live meal worms were provided inside the recovery pen, and the bird was returned to the home pen after normal foraging behaviour occurred. The apparatus was cleaned and disinfected with Accel

(ACCDISC1G-US, Virox Technologies Inc, Ontario, Canada) between tests, and the underpad and feed enrichments were replaced in the TC.

5.3.6. Behavioural data collection

Behaviour data were collected via live observations and video recordings (Table 5.1). Frequency of vocalizations in CC and TC, latency to fully enter TC, latency to loss of posture, and latency to loss of neck tone were recorded during live observation. Live observations were collected by two researchers, with Observer 1 sitting approximately 0.5 m from the hinged door of the TC

135 from which the bird was visible. A black fabric curtain (2.1 m x 0.9 m) and lighting placement were used to obscure Observer 1 from the bird’s view. Observer 1 recorded latencies using a digital timer (National Presto IND. Inc., Eau Claire, WI). On Gas days, Observer 1 determined time at loss of neck tone, and was responsible for moving the bird as quickly as possible from

TC the recovery pen. Observer 2 was positioned beside the chambers, out of the test bird’s view, and operated the sliding door. Observer 2 listened for and counted all vocalizations using a manual tally counter (Great Star Tally counter, Model 30665, Great Star Industrial USA LLC,

Huntersville, NC). All live observations were confirmed with video recordings, except for vocalizations which were not audible by the digital recordings.

Continuous video was recorded using four color digital video cameras (Panasonic, Model WV-

CP-484, Matsushita Co. Ltd., Kadoma, Japan), which were positioned to provide views from the top and side of the CC and TC. Cameras were fed into a multiplexer using a Noldus Portable Lab

(Noldus Information Technology, Wageningen, The Netherlands) that enabled capture of a dual recording at 30 frames/s onto a computer using Handy AVI software (version 4.3 D, Anderson’s

AZcendant Software, Tempe, AZ). Data were collected from videos by a trained observer, blinded to the animal ID, date and CO2 concentrations, using The Observer software (The

Observer version 10.1.548; Noldus Information Technology, Wageningen, The Netherlands). A neutral individual performed the blinding procedures for the video recordings from all tests. The blinding procedures involved cutting the video recordings to remove identification presented at the beginning of each video, assigning a random number to each video segment and sorting for the purpose of providing a random sequence in which videos were to be scored. Seven videos were selected at random and duplicated within this sequence for the purpose of determining

136 intra-observer reliability. Prior to data collection, the observer was trained to use the Observer

XT program by repeatedly scoring 2 videos and an ethogram from an unrelated CO2 study until reaching an intra-observer reliability score of k ≥ 0.90 as calculated by the Observer program.

After reaching the desired level of competence, data collection began using the blinded videos and ethogram shown in Table 5.1.

Latencies to first curtain pushing, first head insert, and full entry were calculated relative to the time the sliding door was opened. Latencies for onset of ataxia and loss of posture were calculated relative to the time of head insert at full body entry to TC. Duration was calculated as total time recorded over all bouts during testing. Hens spent different amounts of time in CC before entering the TC, and amounts of time spent in TC until loss of posture was also differed.

Therefore, frequency of vocalizations was calculated as a rate (number of vocalizations/minute) for the analysis.

5.3.7. Statistical analyses

Fisher’s exact tests (SAS 9.4) were used to determine whether the number of hens that entered

TC and the number of hens that presented avoidance behaviours was independent of test day

(Baseline Day, Gas Day, Washout Day) using 2 × 3 tests. If the results showed significant difference among the days, then pairwise comparisons between Baseline Day vs Gas Day (to find any proximate effect), and Baseline Day vs Washout Day (to find any conditioned place avoidance/ any learning behaviour) were performed for number of birds that entered TC

(yes/no), and presented behaviour (yes/no) using Fisher’s exact tests. The same analysis was performed to determine whether entry to TC was independent of CO2 concentration (25%, 35%,

137

50%,70%) using 2 × 4 tests. If the results showed significant difference among the CO2 concentrations, then pairwise comparisons between every combination of concentration (25% vs

35%, 25% vs 50%, 25% vs 70%, 35% vs 50%, 35% vs 70%, 50% vs 70%) were performed for number of birds that entered TC (yes/no) using Fisher’s exact tests (Zar, 1998).

Frequencies and durations of behaviour during observations in CC and TC were analyzed by using PROC GLIMMIX (SAS 9.4, SAS Inst. Inc., Cary, NC) with identity link function and

Gaussian distribution. Degrees of freedom and fixed effects standard errors were adjusted by using Kenward-Roger method. Bird was included as a random effect.

The model for analyzing all behaviours observed in Control Chamber and Treatment Chamber included the fixed effects of day (BD, GD, WD), CO2 concentration (25%, 35%, 50%, 70%,), and round (1st, 2nd, 3rd, 4th). Interactions were not tested due to small sample size and too few degrees of freedom available among the model factors.

The model for analyzing recovery behaviour inside the recovery pen included the fixed effects of

CO2 concentration (25%, 35%, 50%, 70% CO2).

Least square means estimates and standard errors (SE) are reported in the results. Least significant means separation was conducted by using the Tukey-Kramer test.

138

5.4. Results

Eighteen hens met the performance criteria for enrollment in the training phase within the home pen. Among them, eight hens successfully completed apparatus training within 5 days. Four other hens needed 6 days to complete the apparatus training. Therefore, 12 hens were enrolled in the testing phase (1.2 ± 0.1 kg; 24 week-old).

Some birds died or failed to enter TC following gas exposure, resulting in uneven treatment allocations (Table 5.2). One (F) hen never re-entered TC after her first gas experience. Another

(G) never entered TC after her 2nd gas experience, while 2 other hens (B and K) stopped entering TC after their 3rd gas experience. Two birds (L and H) died during exposure to 50% and

70% CO2, respectively, during the 3rd round. Therefore, only 3 hens completed all 4 treatment rounds.

5.4.1. Approach and avoidance behaviour

The numbers of hens that entered TC differed across days (p= 0.001; Table 5.3). According to pairwise comparisons, fewer hens entered TC on Gas Days than Baseline Days (P<0.001) and on

Washout Days than Baseline Days (P<0.001). The numbers of hens that entered TC at the different CO2 concentrations are presented in Table 5.4. The number of hens that entered TC on

Gas Day differed for different CO2 concentrations (P=0.028) with fewer birds entering TC as

CO2 concentration increased. Pairwise comparisons showed that the number of hens that entered

TC on Gas Day was significantly different between 25% and 70% CO2 (P=0.035), and between

35% and 70% CO2 (P=0.035). None of the birds that entered TC on Gas Day returned to CC although they were able to freely move between the two chambers.

139

The hens who did not fully enter TC on Gas Day always pushed at the curtain and/or inserted their head into the TC (Table 5.3). In this way, they were exposed to CO2 concentrations but chose not to enter. The proportion of hens who demonstrated head insert into TC without entering into TC was significantly higher for Gas Days than Baseline Days (P=0.025). The proportion of hens who showed curtain pushes without head insert and entering into TC tended to be higher on Gas Days than baseline Days (P<0.054). Some hens also performed curtain pushes and/or head inserts into TC on Washout Day without entering into TC (Table 5.3). More hens pushed the curtain without head insert and entering into TC on Washout Day than Baseline

Days (P<0.001).

The numbers of hens that displayed approach/avoidance behaviours in the Control Chamber at the different CO2 concentrations are presented in Table 5.4. The proportion of hens that entered the TC on Gas Day was significantly lower for 70% CO2 compared to 25% and 35% CO2

(P=0.035). The proportion of hens that displayed other approach/ avoidance behaviours (e.g. curtain pushes, head inserts, head withdrawals), and other behaviours (e.g. defecation, vocalization, head shaking, open mouth breathing) did not differ between the CO2 concentrations

(Table 5.4).

Some hens who entered to TC on Gas Day did not enter in to TC on Washout Day. Number of birds entered to TC on Washout Day was same for 25%, 35%, and 50 % CO2. Proportion of hens

(number of hens entered to TC on Washout Day out of number of hens entered to TC on Gas

Day) who showed conditioned place avoidance did not differ (P>0.05) among the CO2

140 concentrations (3/8 for 25%; 3/8 for 35%; 3/6 for 50% CO2). Two birds entered to 70% CO2 and one hen died. The remaining hen entered the TC on Washout Day.

Assessment of the motivation to enter the TC was based on latency to first push the curtain, the frequency of curtain pushes, latency to first instance of the inserting head into the TC, frequency of inserting head into the TC, and latency to full entry into the TC. Frequencies and latencies of approach/avoidance behaviours are presented in Table 5.5. The frequency of curtain pushes was higher on Washout Days (9.0 ±1.3) than Baseline Days (2.8 ± 1.2; P=0.0046). Hens demonstrated more curtain pushes in the 4th round in comparison to 1st and 2nd rounds and this tended to be significant (P=0.0562). Latency to first push the curtain and latency to first head insert were significantly longer in the 4th round compared to the 2nd round (P=0.0286 and

P=0.0336 respectively). Latency to full entry to TC tended to be different among the rounds

(P=0.0543) with longer latencies in 3rd and 4th rounds.

5.4.2. Other observed behaviours

Open mouth breathing was observed only on Gas Day (Table 5.3) and by all the birds that entered the TC (Table 5.4). The average frequency ± SD of open mouth breathing (the number of gasps for the entire time) was 6 ± 2 in 25% CO2 (n=8), 5 ± 2 in 35% CO2 (n=8), 3 ± 1 in 50%

CO2 (n=6), and 2 ± 0.7 in 70% CO2 (n=2). Hens demonstrated head shaking on all 3 test days.

More hens demonstrated head shaking inside the TC on Gas Days than Baseline Days (P=0.001).

All hens that entered the gas showed head shaking, and this was not different between concentrations (Table 5.4). The average frequency ± SD of head shaking until loss of posture

141 inside the TC was 9.5 ± 3 in 25% CO2 (n=8), 9 ± 3 in 35% CO2 (n=8), 7.7 ± 3.9 in 50% CO2

(n=6), and 6 ± 2.8 in 70% CO2 (n=2).

The number of hens that vocalized on different days, and different CO2 concentrations are presented in Table 5.3 and Table 5.4, respectively. The proportion of hens that vocalized inside the CC did not differ for all 3 tests days and CO2 concentrations. The proportion of hens that vocalized inside the TC was higher for the Gas Day in comparison to Base line Day (P=0.004) but no difference was found between Washout Day and Baseline Day (P=0.42). Rates of vocalizations (mean vocalizations per minute) before the sliding door was opened did not differ among the 3 different test days (P=0.48), CO2 concentrations (P=0.12), and rounds (P=0.30). The rate of vocalizations after the sliding door was opened among the 3 different test days tended to be different (P=0.08), and a higher rate was observed on Washout Days (5.6 ± 1.3 per minute) than Baseline Days (2.1 ± 1.2 per minute). However, no differences were found in the rate of vocalizations after the sliding door was opened for CO2 concentration (P=0.76) or round

(P=0.46). Hens demonstrated higher rates of vocalizations inside TC on Gas Days (4.1 ± 0.7 per minute) than Baseline Days (0.2 ± 0.5 per minute) and Washout Days (0.5 ± 0.8 per minute;

P=0.0001). Rate of vocalizations inside the TC did not differ between CO2 concentrations

(P=0.79) or rounds (P=0.19).

Duration of standing active in the CC was significantly longer (P=0.0011) on Gas Days (76±16 s) and Washout Days (98±15 s) than Baseline Days (20±14 s). Duration of standing active in the

CC tended to be significant for CO2 concentrations (P=0.0990). Hens performed standing active in the CC for longer durations at 50% CO2 (95±16 s) than 25% CO2 (37±18 s) on Gas Days.

142

Duration of standing active in CC did not differ among the rounds (P=0.41). Duration of standing inactive in the CC was not different among the CO2 concentrations (P=0.2476), or rounds (P=0.13). Duration of standing inactive in the CC was higher in 4th round (122 ± 27 s) in comparison to the 1st round (36 ± 16 s; P=0.04). Except for two hens, all the other hens did not sit inside the CC on any of the 3 test days. However, some hens demonstrated sitting inside CC during the retraining period. Escape behaviour was absent in all hens.

5.4.3. Parameters associated with loss of sensibility

All the birds that entered TC on Gas Day showed ataxia and loss of posture irrespective of CO2 concentration. Mean time ± (SE) s for latencies (ataxia, loss of posture) and durations (ataxia) are presented for 25%, 35%, and 50% in Table 5.6. Latency to onset of ataxia was tended to decrease as concentration increased (P=0.09). Longer latency to onset of loss of posture was observed at 25% CO2 than 50% CO2 (P<0.05). Table 5.7 shows behaviour observed during recovery from exposure to CO2. Hens that lost sensibility at 50 % CO2 demonstrated the longest average time to raise their head (175 ±11 s, P<0.0001), stand up (510 ± 77, P=0.0118), walk

(1077 ± 145; P=0.0006), and begin feeding (1346 ± 58, P<0.0001) compared to the birds exposed to 25% and 35% CO2.

Only one hen who entered 70% CO2 was recovered and the relevant data inside the recovery pen are shown in the Table 5.7. One hen that entered at 50% CO2 and another that entered at 70%

CO2 in their 3rd round died inside the recovery pen without regaining sensibility.

143

5.5. Discussion

In approach avoidance paradigms, the motivation to approach a desirable stimulus is gauged against the motivation to avoid an unpleasant stimulus. Results are expected to vary according to strengths of competing motivations.

In the current study, all hens who were eligible for the apparatus training were able to learn the assigned task within 5-6 days. Therefore, the assigned task to the hens (entering the TC pushing through a curtain) is within hens’ cognitive capacity, and hens were sufficiently rewarded inside the TC. Hens were able to see through the curtain which was between the CC and TC, and could easily push through it to enter the chamber. In the current study, all the hens were interested in meal worms. Upon entering the chamber, they first approached the meal worms and spent time eating them. They were also interested in the commercial feed and spent the second longest time on feeding. This suggests that the one hour feed was withheld induced feeding motivation of the hens. None of the hens went back to CC on Baseline Days during the given 5 minutes, indicating their willingness to stay inside the TC, and a preference for it compared the CC. All hens entered

TC in less than 60s on Baseline Days in the first round. We suggest that all hens were motivated to enter the TC to acquire the provided feed and enrichments.

In the current study, refusal to enter the TC on Gas Day reveals that the motivation of a bird to avoid the aversive environment inside TC was higher than the motivation to acquire the rewards inside the TC. Except for one bird, all birds voluntarily entered to 25% and 35% CO2. The same bird avoided 35% and 25% CO2 in its 1st and 2nd rounds respectively, and this might be due to the difference in individual sensitivity of the hens. However, this bird entered into 50% CO2 in

144 its 3rd round and died on Gas Day without regaining consciousness. The reason that this bird avoided lower CO2 concentrations but entered at higher CO2 concentration is difficult to explain.

Overall, we observed that the number of hens entering into the TC decreased with increasing

CO2 concentrations. Webster and Fletcher (2004) observed similar results in layer hens in an approach avoidance test, which used 30%, 45%, and 60% CO2 in a feeding chamber. The authors reported that the percentage of birds that entered the feeding chamber decreased at higher concentrations of CO2, and further suggested that hens could detect the presence of CO2 in the corridor leading to the feeding chamber and tended to hesitate to push through the entrance of the gas chamber as a result. In the current study, nearly half of the birds tested at 50% CO2 avoided entering the TC on Gas Day indicating that half of the hens were able to detect 50% CO2 and avoided it. Sandilands et al. (2011) assessed 50%, 55%, and 60% CO2 in broilers and reported that all gas levels were aversive to some extent in broiler chickens based on time spent with head in a feeding station filled with CO2 gas. Further the authors explained that 60% CO2 was evidently the most aversive option among the gas mixtures tested, but even this concentration of

CO2 was not totally refused by broilers. In the current study, only two of 7 hens entered into 70%

CO2. Therefore, we suggest that 70% CO2 is relatively more aversive to layer chickens based on avoidance. In the current study, 25% was the lowest CO2 concentration tested, and this is slightly higher than the proposed aversion threshold level of 25% (McKeegan et al. 2003) for the chickens. Except one, all hens chose to enter into 25% and 35% CO2 levels. Therefore, we suggest that hens entered 25% and 35% CO2 with minimal or no aversion and that their motivation to access the rewards inside the TC was higher than the motivation to avoid 25% and

35% CO2.

145

The number of curtain pushes and head withdrawals before entering or avoiding TC on Gas Day could be considered to reflect conflict or competing motivation between entering TC to acquire the resources and avoidance of gas. All hens that did not enter into 50% CO2 showed a number of curtain pushes and head withdrawals. Except for two hens, all other hens who did not enter into 70% CO2 demonstrated curtain pushes without inserting their head into TC. The absence of head insert into TC could be due to higher aversiveness of 70% CO2. All birds did push the curtain several times, and many birds made head withdrawals prior to avoiding 50% and 70%

CO2 concentrations on the Gas Day, indicative of several attempts to enter the TC. Head withdrawals on Baseline Day were only observed in 4 hens in their later rounds of testing.

Except for two hens at 25%, and two hens at 35% CO2, all other hens who voluntarily entered into either 25% or 35% CO2 did not demonstrate head withdrawals prior to entering TC, supporting our suggestion that these two concentrations were only minimally aversive to the hens. However, it is difficult to say whether layer hens can detect 25% and 35% CO2 or not based on the current results of the approach-avoidance test. Raj and Gregory (1991) reported concentrations above 7.5% CO2 were detectable to laying hens and given a free choice, the hens learned to avoid such atmospheres. In contrast, all except for one hen entered into 25% and 35%

CO2 in the current study.

In the current study, we hypothesized that a bird that entered into the gas and experienced loss of consciousness would demonstrate conditioned place avoidance (CPA). Although the majority of hens entered at 25% and 35% CO2, some of those hens did not enter TC on Washout Day.

Among them, one hen that entered 25% needed 4 days of retraining, and another hen that entered

35% CO2 needed 8 days of retraining to resume entering the TC voluntarily, providing evidence

146 that they demonstrated CPA. None of the hens who entered and lost consciousness at 50% CO2 entered TC on Washout day, and all needed 4-5 days of retraining before re-entering TC voluntarily. Moreover, all hens who demonstrated CPA did not insert their head into TC on

Washout Day and some did not even push the curtain. Many of the birds who demonstrated CPA vocalized 50-90 times (within the given 5 minutes) after the sliding door was opened on

Washout Days showing evidence of either fearfulness or conflict to reenter the TC. Therefore, I observed a trend of higher rate of vocalizations after the sliding door was opened on Washout

Days in comparison to Baseline Days.

The majority of hens who lost consciousness at 25% and 35% CO2 did not demonstrate CPA, and this could be due to less aversion experienced at these two CO2 concentrations prior to loss of consciousness. In this study, three birds who experienced 50% CO2 (by pushing the curtain and head inserting) without full entry to the TC on Gas Day never reentered the TC, either on the

Washout Day or the next 4 days of retraining. Three other birds tested at 70% CO2 pushed the curtain several times (two birds inserted their heads once into TC) but did not enter the TC on

Gas Day, and they did not enter the TC on Washout Day. Therefore, we suggest that exposure to

50% or higher CO2 concentration by pushing the curtain and/or inserting their heads into the TC without experiencing loss of consciousness was enough to create CPA in the laying hens.

McKeegan et al. (2005) reported that chickens detected CO2 at 10% while Raj and Gregory

(1991) reported that hens can detect CO2 concentrations at 7.5%. Gerritzen et al. (2007) compared the behaviour of ducks, broilers, laying hens and turkeys to assess differences in susceptibility among species following exposure to CO2. increasing from 0% to 45% CO2 at a

147 rate of 14 l/min. The authors reported that the first signs that birds noticed a change in their environment was recorded from as little as 2% CO2 in the air within the test box for broilers, ducks and turkeys which was at a significantly (P<0.05) lower concentrations than for laying hens (6.6% CO2). The used least concentration of 25% CO2 in the current study is higher than the detection levels in chickens. Thus, it is suggested that all tested concentrations were detected by the hens.

Breathlessness is induced in humans when they exposed to high CO2 concentrations (Gregory and Wotton, 1990b). Breathlessness is a negative experience and can activates the same regions of the brain as other negative sensory experiences such as pain, hunger, and thirst in mammals

(Beausoleil and Mellor, 2015). Breathlessness is described in terms of its 3 different qualities: respiratory effort, air hunger, and chest tightness (Beausoleil and Mellor, 2015). Respiratory effort and air hunger are induced in CO2 euthanasia. Lack of available O2 resulted in tissue hypoxia in the respiratory muscle, thus decreasing the muscle activity. Therefore, respiratory effort (an increased conscious effort for respiration) is induced to achieve the necessary or desired level of ventilation for the respiratory muscular functioning and respiratory ability

(Beausoleil and Mellor, 2015). Air hunger is the increased urge to breathe. Air hunger is more unpleasant than respiratory effort (Banzett et al., 2008) and induced by hypercapnia and hypoxaemia when there is no proper coordination between increased automatic command to breath and degree of lung inflation (Beausoleil and Mellor, 2015). In humans, signs of asphyxia and behavioral excitation have been observed because of both hypercapnia and hypoxia (Forslid et al., 1986). McKeegan et al. (2005) conducted a behaviour study, which hens were exposed to

148 short pulses of gas and their responses were observed to indicate aversion, and concluded that aversion to CO2 was seen at 24%.

Exposure to CO2 could be distress to poultry because it may result a sense of breathlessness.

Different behaviour studies have been conducted to assess the aversion to CO2 in poultry species.

Gerritzen et al. (2000) suggested gasping and head shaking in broiler chickens exposed to 30% and 60% CO2 as indicators of breathlessness and the pungency of gas mixtures. Webster and

Fletcher (2001) also observed head shaking and deep breathing (breathing through the mouth) in layer chickens and broilers exposed to different CO2 concentrations (30%, 45%, 60%).

McKeegan et al., (2005) reported respiratory disruption (apparent increased inhalation depth and duration) in broiler chickens exposed CO2 concentrations (10, 25, 40, 55, 70%) for 10 s. It is reasonable to say all tested concentrations in the current study were aversion to the hens up to some extent. Hens demonstrated open mouth breathing in all CO2 concentrations, suggesting that they were experienced air hunger and may be with respiratory efforts in lower concentrations.

McKeegan et al. (2004) reported that the nasal and oral mucosa of the chicken contain trigeminal polymodal nociceptors that can detect and transmit information relating to the presence and concentration of potentially noxious chemical stimuli (ammonia, CO2 and acetic acid vapours).

The authors further suggested that the avian nociceptive threshold is in the region of 40 - 50%

CO2, which is in the same range as the electrophysiological thresholds reported for rats and humans (Anton et al., 1991, 1992). Therefore, low levels of CO2 such as 25 % seems to be probably air hunger but high levels are probably pain, possibly with air hunger combined.

149

All hens in the current study entered the TC on Gas Day and showed head shaking and open mouth breathing irrespective of CO2 concentration. Raj (1996) studied the degree of aversiveness to 72% CO2 in air in turkeys by using 1-5 scale system based on behaviours of head shaking, gasping and vocalization. He stated that all three behaviours were clearly observed in all turkeys in the CO2 environment. However, interpretation of head shaking on animal welfare is controversial. Head shaking has been described as indicative of an aversive reaction to CO2 and respiratory distress in poultry (Raj and Gregory, 1994). However, other researchers suggested head shaking as an alerting response to unexpected or novel events in poultry (Hughes, 1983) or an attempt to regain an alert state when birds are feeling dizzy or start to lose consciousness during the exposure to CO2 (Webster and Fletcher, 2001; Gerritzen et al., 2007). We observed head shaking in some hens on Baseline Days and Washout Days when the CO2 concentration was the same as ambient air. Therefore, the argument that head shaking is an aversive reaction to

CO2 or that it indicates respiratory distress is not supported. McKeegan et al. (2005) reported that gustatory mandibulation was associated with head shaking, and suggested that the acidity of

CO2 may provide a gustatory stimulus which causes head shaking. Significantly more hens showed head shaking on Gas Days than Baseline Days for all CO2 concentrations. Therefore, hens are more likely to perform head shaking when exposed to CO2.

By rotating the same hens through different CO2 concentrations, there is a possibility that experience of one CO2 concentration could carry over to influence the results of a test with another CO2 concentration. Webster and Fletcher (2004), who also used repeated tests at different concentrations of gas in laying hens, explained the possibility of carry-over effect due to hens being exposed to more than one stunning gas atmosphere, and the suggested reasons

150 were inability of complete physiological recovery between tests, or learned responses associated with recovery from loss of consciousness or from exposure to the stunning atmosphere itself. In the current study, there was a significant effect of round in some observed behaviours. Hens tested in 50% and 70% CO2 in their previous rounds demonstrated longer latency to insert head into TC and fully enter into TC for even lower concentrations of 35% and 25% CO2 in 3rd and

4th rounds. Overall, latency to first push the curtain was longer in 3rd and 4th rounds and this could be also due to previous aversive experience in the TC. A higher number of curtain pushes were also observed in the 4th round. This may be due to hens testing the air in TC due to their previous experience with higher CO2 concentrations. Head withdrawals (2-3 times) on Baseline

Day were only observed in 4 hens in their 3rd round and one hen in its 2nd round. The hens who demonstrated head withdrawals for 25% and 35% CO2 were tested for 50% and 70% CO2 in their previous rounds, and this previous experience may be the reason for head withdrawals prior to entering the lower concentrations of 25% and 35% CO2. More hens demonstrated CPA in the

3rd round for different CO2 concentrations compared to the first and second rounds, and needed retraining, indicating that more hens learned to avoid TC after the 3rd gas experience.

For optimal welfare of the animal, the goal is to achieve insensibility and loss of consciousness as quickly as possible, for the process to be pain-free, and for death to follow quickly. CO2 causes hypercapnia, suppressing afferent transmission of sensory stimuli to the cortex of the brain, producing anesthesia (Forslid et al., 1986; Raj et al., 1992; Raj and Gregory, 1993). Loss of posture (LOP) has been proven to be the reliable indicator of onset of unconsciousness in poultry based on electroencephalography (EEG) studies in broiler chickens (Coenen et al., 2000;

Gerritzen et al., 2004; Benson et al., 2012). In the current study, onset of ataxia occurred

151 regardless of CO2 concentration. However, there was a tendency of decreasing latency to onset of ataxia with increased CO2 concentrations. Latency to onset of LOP was significantly different among the CO2 concentrations. Longer latency to onset of LOP was observed in 25% CO2 than

50% CO2. Similar to the current study, previous research reported shorter times to LOP in higher

CO2 concentrations. Webster and Fletcher (2001) reported that higher CO2 concentrations reduced the time to loss of consciousness based on time to LOP in layer hens and broilers.

Gerritzen et al. (2007) reported slight and occasionally significant differences in time to LOP between different poultry species following exposure to CO2 treatments that increased from 0% to 45% at a rate of 14 l/min. Broilers lost posture at 19.0% CO2, which was significantly different than in ducks (23.8% CO2), while loss of posture occurred in layers and turkeys at

19.9% and 19.3% CO2, respectively. Results of our first study with turkeys (Bandara et al., unpublished data) are in accordance with these findings, where laying hens showed faster LOP in comparison to turkeys. In the current study we could not observe loss of neck tone, an indicator of loss of posture (Sandercock et al., 2014) in hens because the time between LOP and loss of neck tone was short (~2 s) but in the turkeys, recording of loss of neck tone was possible due to longer time (> 25 s) between LOP and loss of neck tone. Therefore, these findings revealed the need to set different criteria for different poultry species using CO2 as a humane killing agent.

Recovery pen data revealed that birds that lost consciousness at 50% CO2 demonstrated the longest time to raise their head, stand up, walk, and start feeding compared to the birds exposed to 25% and 35% CO2. Overall, all hens regained consciousness within 6 minutes. We suggest at least ~12 minutes to observe hens to confirm that they are dead following CO2 euthanasia if using 25% or above CO2 concentration.

152

5.6. Conclusions

The majority of hens avoided entering into 50% and 70% CO2. Except for one, all hens entered into 25% and 35% CO2. Therefore, hens found 50% and 70% CO2 concentrations more aversive than 25% and 35% CO2 concentrations. Hens demonstrated conditioned place avoidance for all

CO2 concentrations and needed several days of retraining after gas experience to start voluntarily entering into TC irrespective of CO2 concentration. Therefore, all concentrations tested in the current study were aversive to a certain extent. Loss of posture indicating loss of consciousness occurred in less than 20s in all CO2 concentrations. Latency to LOP was shorter in 50% than

20% CO2.

153

Table 5. 1: Definitions of behaviours recorded in the Control Chamber (CC) and Treatment Chamber (TC)

Measure Variable type Definition States Standing inactive v Duration Bearing weight on feet without locomotion Standing active v Duration Bearing weight on feet with locomotion Sitting v Duration Breast is touching the chamber ground Exploring v Duration Pecking floor or walls (curtains excluded). Can be done while sitting, standing or in locomotion Events Vocalization D Frequency Vocalizations occur in the CC and TC Push curtain VD Latency, First time beak contacts with curtain; this may Frequency include pushing into the TC but head does not cross plane of curtain Insert head VD Latency, Head crosses plane of curtain into TC with or Frequency without neck Withdraw head V Latency, Head and beak return to CC after curtain contact Frequency Full entry VD Latency Bird enters TC with both feet Head shaking V Frequency Rapid lateral movement of the head Open mouth Latency, Bird breath through their mouth (beak is opened breathing V Frequency and chest movements occur) Wing flapping V Duration Coordinated movement with rising and lowering wings without feet losing contact with the ground. Ataxia V Latency, Uncoordinated movement, while bearing weight Duration on one or more legs Loss of posture VD Latency Non-weight bearing on feet; may occur rapidly (collapse) after ataxia. Head righting response present Escape V Frequency Leaping or climbing; feet left floor with wing flapping

V Measured via video recordings D Measured via direct observations. Vocalizations were recorded separately before the sliding door open, after the sliding door open, and after entry to the Treatment Chamber

154

Table 5. 2: Test order, assigned CO2 treatment and outcomes in each of the four rounds for the 12 hens enrolled in the study.

Order of Bird ID CO2 concentration (%) testing 1st Round 2nd Round 3rd Round 4th Round

1 A 70r 50r 35e 25e 2 B 25e 70e 50r* - 3 C 35e 25e 50ec* - 4 D 25e 35ec 70r 50r 5 E 50r* 35e 25ec* - 6 F 35ec* - - - 7 G 70r 50ec* - - 8 H 25e 50e 70ed - 9 I 70r 35ec 50ec* - 10 J 50e 70r 35e 25ec 11 K 50r* 35e 25ec* - 12 L 35r 25r 50ed - e Birds entered TC on Gas Day. r Birds did not enter TC on Gas Day. c Birds did not enter TC on Washout Day d Birds did not recover from CO2 exposure on Gas Day (died) *Birds required retraining (did not enter on subsequent Baseline Day)

155

Table 5. 3: Number of times laying hens demonstrated different behaviour patterns in the Control (CC) and Treatment Chamber (TC) on the three different test days (Baseline Day, Gas Day, Washout Day)1

Baseline Day Gas Day Washout Day (n=36) (n=36) (n=34) Approach/avoidance Behaviours Fully entered into TC 36ac 24b 18d Head insert only 0b 6a 2 Pushed curtain only 0bd 5a 14c Neither head insert nor curtain 0 1 1 pushes Head withdrawals 4 9 4

Other behaviours Defecation in CC 5 11 10 Vocalization in CC-before door open 11 9 10 Vocalization in CC- after door open 8 5 13

Inside Treatment Chamber (n=36) (n=24) (n=18) Vocalization 5b 12a 1 Defecation 0 0 1 Open mouth breathing 0b 24a 0 Head shaking 4b 24a 4

1 Pairwise comparisons of Baseline Day vs Gas Day and Baseline Day vs Washout Day were analyzed by Fishers’ exact test (P<0.005). a,b Different superscripts indicate proportion of birds presenting the measure is different between Baseline Day and Gas Day according to Fishers’ exact tests (P<0.005) c,d Different superscripts indicate proportion of birds presenting the measure is different between Baseline Day and Washout Day according to Fishers’ exact tests (P<0.005)

156

Table 5. 4: Number of laying hens that demonstrated different behaviour patterns in the Control (CC) and Treatment Chamber (TC) at the different CO2 concentrations on Gas Days.

CO2 concentration 25% 35% 50% 70%

(n=9) (n=9) (n=11) (n=7) Approach/avoidance Behaviours a a ab b Fully entered into TC 8 8 6 2 Head insert only 0 0 4 2 Pushed curtain only 0 1 1 3 Neither head insert nor curtain pushes 1 0 0 0 Head withdrawals 2 1 4 2

Other behaviours Defecation in CC 2 2 4 3 Vocalization in CC- before door open 3 2 4 0 Vocalization in CC- after door open 1 0 3 1

Inside Treatment Chamber (n=8) (n=8) (n=6) (n=2) Open mouth breathing 8 8 6 2 Head shaking 8 8 6 2 Vocalization 4 4 4 0

Numbers within a row with different superscripts indicate proportion of birds presenting the measure is different among the concentrations according to Fishers’ exact tests (P<0.005)

157

Table 5. 5: Frequencies and latencies of approach /avoidance behaviours (mean ± SE) demonstrated by laying hens in different test days, CO2 concentrations, and cycles.

Latency to first Frequency of Latency to first Frequency of Latency to full push the curtain curtain pushes head insert TC head insert TC entry to TC (s) (s) (s) Day Baseline Day 13.3 ± 3.6 2.8 ± 1.2b 17 ± 8 1.3 ± 0.1 26 ± 10 Gas Day 9.3 ± 3.8 6.6 ± 1.3ab 42 ± 10 1.2 ± 0.2 46 ± 13 Washout Day 17.2 ± 3.9 9.0 ± 1.3a 37 ± 12 1.0 ± 0.2 47 ± 14 P value 0.3491 0.0046 0.1450 0.5578 0.3485 CO2 concentration 25% 8.6 ± 4.3 4.3 ± 1.6 13 ± 11 1.5 ± 0.2 25 ± 13 35% 11.9 ± 4.4 6.2 ± 1.5 42 ± 11 1.1 ± 0.2 46 ± 13 50% 15.3 ± 3.9 7.9 ± 1.4 45 ± 10 1.3 ± 0.2 47 ± 14 70% 17.3 ± 4.7 4.9 ± 1.7 14 ± 13 0.8 ± 0.3 25 ± 18 P value 0.5303 0.3597 0.0736 0.2030 0.5106 Round 1st 14 ± 3ab 5.1 ± 1.3 23 ± 10ab 1.2 ± 0.1 29 ± 12 2nd 6 ± 4b 4.5 ± 1.4 17 ± 10b 1.0 ± 0.1 19 ± 12 3rd 16 ± 5ab 6.4 ± 1.5 43 ± 11ab 1.3 ± 0.2 62 ± 13 4th 29 ± 7a 12.4 ± 2.5 75 ± 20a 1.4 ± 0.4 72 ± 27 P value 0.0286 0.0562 0.0336 0.6531 0.0543 Different letters indicate statistically significant (P<0.05) differences between means within the group of comparisons. Bold numbers indicate significance results or tend to be significance results (P<0.05)

158

CO2 concentration 25% 35% 50% P value 70% (n=8) (n=8) (n=6) (n=2) Latency to ataxia 13 ± 0.8 11 ± 0.8 10 ± 0.9 0.096 11 ± 1* Latency to loss of posture 19 ± 1a 17 ± 1ab 13 ± 1b 0.022 17 ± 3*

Different letters indicate statistically significant (P<0.05) differences between concentrations within a row *Did not statistically analyze due to only two data points

159

Table 5. 7: Time (s) at onset of behaviour during recovery in laying hens exposed to different CO2 concentrations. Mean time ± SE s are presented for 25%, 35%, and 50% CO2 concentrations

CO2 concentration Behaviour 25% 35% 50% P value 70% (n=8) (n=8) (n=5) (n=1) Raised head 35±9b 37±9b 175±11a <0.0001 327* Stand up 273±60b 192±60b 510±77a 0.0118 1703* Start walking 267±115b 179±115b 1077±145a 0.0006 1803* Feeding 295±45b 312±58b 1346±58a <0.0001 1860*

*Did not statistically analyze due to only one data point available Different letters indicate statistically significant (P<0.05) differences between concentrations within a row

160

W C

A B

161

CHAPTER 6

General Discussion

On-farm euthanasia is a common task within the poultry industry. The traditional methods by which poultry are killed on-farm need to be evaluated for humaneness due to the increasing concern for animal welfare. Additionally, new euthanasia methods continue to be developed, often without scientific assessment for their effectiveness and humaneness. Therefore, there is an urgent need to develop appropriate euthanasia methods that cause rapid, irreversible insensibility and minimum pain or distress (AVMA, 2013; NFACC, 2016).

In this thesis, I assessed the efficacy of three physical euthanasia methods (non-penetrating captive bolts; manual cervical dislocation; mechanical cervical dislocation), and the aversion to different CO2 concentrations in layer chickens. The methods were compared for effectiveness and humaneness allowing for recommendations to be made.

6.1. Evaluation of physical on-farm euthanasia methods

6.1.1. Humaneness of the euthanasia methods

In Chapters 3 and 4, I assessed three different commercially available non-penetrating captive bolt (NPCB) devices and a mechanical cervical dislocation device (KED) in comparison to manual cervical dislocation, respectively. Overall, all three non-penetrating captive bolts caused rapid loss of sensibility in all age groups. All eye reflexes and rhythmic breathing were abolished in ~5 s of application for all NPCB devices. Loss of jaw tone and neck muscle tone were observed at ~ 30 s, while time to last movement and cardiac arrest occurred at less than 250 s.

162

Previous research also reported rapid insensibility in poultry killed by similar NPCB devices

(Erasmus et al., 2010b; Woolcott et al., 2018a; Gibson et al., 2018).

Chapter 4 revealed that awake chickens killed by manual cervical dislocation demonstrated a loss of pupillary light reflex at ~85 s, indicating that they may be sensible for a longer time in comparison to birds killed by NPCB devices. Anesthetized birds killed by KED demonstrated loss of pupillary light reflex at ~140 s, and this value is longer than for conscious birds killed via

NPCB devices and manual cervical dislocation. I found an anesthetic effect on eye reflexes and other behavioural responses. Time to loss of pupillary light reflex was significantly reduced in anesthetized birds compared to awake birds killed by manual cervical dislocation. The results revealed a shorter latency to loss of pupillary light reflex indicative of brain death in the anesthetized birds killed by manual cervical dislocation than in the anesthetized birds killed by

KED. Previous studies also revealed shorter latencies to insensibility and brain death in poultry killed with manual cervical dislocation compared to other different mechanical cervical dislocation devices (chickens killed by semark neck pliers: Gregory and Wotton, 1990a; turkeys killed by burdizzo: Erasmus et al., 2010b; turkeys killed by KED: Woolcott et al., 2018b and

Jacobs et al., 2019; layer chickens killed by KED: Hernandez, 2018).Overall, the ability to cause rapid loss of sensibility, shown through loss of reflexes and behaviours, was faster in birds killed with NPCB devices than the cervical dislocation methods. Erasmus et al., (2010b) also reported shorter latencies to insensibility in turkeys killed by a NPCB than by manual cervical dislocation.

However, a few birds failed to lose sensibility within 60 s for the Zephyr-E device in the current study, and the cause was suspected to be due to a mechanical failure of the air compressor.

Results of the current study revealed that device success was 100% for TED and 99% for the

163

Zephyr-EXL. Therefore, NPCB devices were shown to be relatively consistent in killing layer chickens in an on-farm context. I suggest NPCB devices are more humane for killing of layer chickens than the cervical dislocation methods.

AVMA (2013) guidelines indicate that cervical dislocation must result in luxation of the cervical vertebrae without primary crushing of the vertebrae and spinal cord. Previous research reported prevalence of vertebral fractures during mechanical cervical dislocation. Bader et al. (2014) reported cervical fractures (in C1 and C2) in 43% of turkeys killed by using a non-penetrating custom-made forceps with three blunt shear arms, and in 33% of turkeys killed by manual cervical dislocation. Further, the authors reported cervical fractures (in C1 and C2) in 48% of chickens killed by manual cervical dislocation. However, all the birds were stunned by a sharp blow to the head from a wooden pole prior to cervical dislocation. Thus, the observed cervical fractures may not be solely due to cervical dislocation and some may have resulted from the blunt force trauma to the head. Woolcott et al. (2018b) found cervical fractures in 78% of 3 weeks old turkeys killed by mechanical CD (KED model-S), compared to 44% of turkeys killed by manual CD. In Chapter 4, I found comparatively less vertebral fractures both in the layer chickens killed by manual CD (6%) and KED (16%), which indicates differences across specific cervical dislocation tools and across poultry species. When AVMA guidelines were established in 2013, no evidence was available on cervical fractures in poultry killed by manual cervical dislocation. In my study, cervical fractures were found in layer chickens killed by manual cervical dislocation for the first time. These cervical fractures were found in 12-29 weeks old birds and not in 60-75 weeks old birds. Thus, further studies need to assess the fractures in different age groups of layer chickens killed by manual cervical dislocation.

164

For humane reasons, we decided to assess the KED device in awake chickens only after assessing it on anesthetized birds. However, based on the results of the trials using anesthetized chickens (Chapter 4), we decided not to apply the KED device in awake chickens due to the prolonged time for brain death, prevalence of sub-luxations versus full luxations and resultant cervical fractures.

6.1.2. Pathology caused by euthanasia methods and effectiveness

The degree of tissue trauma caused by different euthanasia methods can be used to corroborate the antemortem measures of loss of sensibility and brain death. The preferred killing methods are ones that cause direct trauma to the brain disturbing normal electrical activity (Alexander, 1995;

Claassen et al., 2002) and result in rapid loss of sensibility and brain death. NPCB are manufactured to cause direct damage to the brain and to result in loss of sensibility. However, cervical dislocation methods aim to transect the spinal cord and destroy the blood vessels which supply blood into the brain, causing death by cerebral ischemia (White and Krause, 1993;

Claassen et al., 2002). It has been suggested that the stretching and twisting motion in manual cervical dislocation results in a concussive effect and brain stem damage (Shi and Whitebone,

2006; Cartner et al., 2007).

Skin laceration and external bleeding, skull fractures, and macroscopic and microscopic hemorrhages and trauma of the brain and spinal cord were assessed for chickens killed with

NPCB. For birds killed by manual and mechanical cervical dislocation by KED, skin laceration

165 and external bleeding, macroscopic and microscopic hemorrhage were also assessed, in addition to dislocation site and fracture of the cervical vertebrae, and transection of the spinal cord.

Brain hemorrhage and skull fractures are frequently used to score the damage caused by physical methods of euthanasia (Erasmus et al., 2010c; Casey-Trott et al., 2013; Bader et al., 2014;

Woolcott et al., 2018a,b). Researchers have suggested that immediate insensibility and irreversible loss of vital functions are associated with evidence of subdural and parenchymal hemorrhage in poultry (Erasmus et al., 2010c; Bader et al., 2014). Chapter 3 revealed that all three NPCB devices caused substantial subdural and parenchymal hemorrhage in the brain tissues. All birds sampled for microscopic scoring showed both parenchymal and subdural hemorrhage in the mid brain for all NPCB devices except one bird for Zephyr-E. Subdural hemorrhage (SDH) was substantial in the hind brain for all three NPCB devices with the vast majority given the highest scores (moderate to marked levels). Nearly 50% of birds had parenchymal hemorrhage (PCH) in the hind brain for all three NPCB devices. In contrast,

Chapter 4 revealed that brain damage was minimal for mechanical cervical dislocation method.

None of the birds killed by KED had PCH in the brain whereas only 1 of 18 birds sampled for microscopic evaluation had minimal level of SDH in the hind brain. Brain damage was also comparatively less in birds killed by manual cervical dislocation in comparison to NPCB. In the birds killed by manual CD, minimal levels (<1%) of PCH were observed only in the hind brain in 14% birds sampled whereas 76% of birds had SDH in the hind brain. A few birds (30%) killed by manual CD had SDH also in the mid brain. Thus, in contrast to mechanical CD by

KED, stretching and twisting the neck during manual CD caused brain stem damage as suggested in previous studies (Shi and Whitebone, 2006; Cartner et al., 2007). The neuronal

166 network of the reticular formation which is responsible for consciousness (Seth et al., 2005;

Erasmus et al, 2010a) is located in the brain stem, and its ascending pathway innervates the fore brain (Starzl et al., 1951) to process information of consciousness. Thus, trauma to the brain stem region was suggested to disconnect the neuronal pathway that is responsible for sensibility

(Bader et al., 2014). In the current study, brain stem trauma was minimal in the birds killed with

KED, and less than in the birds killed with manual CD. Brain stem damage was higher in the birds killed by NPCB in comparison to cervical dislocation methods, suggesting that this could be the reason for rapid loss of sensibility in the birds killed with NPCB. The longest time to brain death was observed in the birds killed with KED, corresponding to the minimal brain stem trauma.

Gregory and Wotton (1990a) reported that cervical dislocation by pliers did not result in immediate loss of sensibility in broiler chickens based on visual evoked responses, and the spinal cord remained intact in 21% of birds. Carotid arteries also remained intact but left and right arteries were damaged. The authors concluded that the birds killed by the pliers died from asphyxia. In Chapter 4, subcutaneous hemorrhage in the neck was observed in the birds killed by

KED but brain trauma was absent in all except one bird. Thus, these findings support the suggestions that death results from cerebral ischemia in the birds killed by mechanical cervical dislocation and not because of the brain trauma. Overall, the minimal brain trauma could be the reason I observed longer latencies to insensibility and clinical death in chickens killed with mechanical cervical dislocation than the chickens killed with manual CD. The highest scores (3 and 4) for subcutaneous hemorrhage in the neck was found in more than 70% of the birds killed by manual cervical dislocation. Therefore, I suggest that loss of sensibility and death results from

167 both cerebral ischemia and brain trauma in the birds killed by manual CD. The killing methods which do not stretch and twist the neck, resulting in cerebral ischemia alone (e.g. burdizzo or pliers), have been prohibited since 2013 under the Council Regulation 1099/2009 (European

Commission, 2009).

Anatomical damage to the top of the spinal cord and the base of brain stem is associated with spinal cord concussion, neurogenic shock, and loss of sensibility (Dumont et al., 2001; Harrop et al., 2001). In all of the birds killed by manual cervical dislocation, the spinal cord was transected at the top (between the skull and C1 or C1-C2). For the mechanical cervical dislocation, 75% of birds had a transected spinal cord but further down the vertebrae than the recommended skull to

C1. I suggest that lower level transected spinal cord or dislocation could be the reason for minimal brain stem trauma in the birds killed by KED. Martin et al. (2018) reported that high neck dislocation was associated with improved kill success and more rapid loss of reflexes compared to lower level dislocations for cervical dislocation. Overall 58% of the birds (awake and anesthetized) killed by manual cervical dislocation had luxation in between the skull and C1.

In contrast, birds killed with mechanical CD presented dislocation with the majority (71%) of sites occurring caudal to C2.

Subdural and parenchymal hemorrhage in the spinal cord were observed, indicating traumatic damage to the spinal cord in birds after both manual and mechanical cervical dislocation. Some muscles that are responsible for inspiration are innervated by cervical nerves and the accessory cranial nerve (CN X1) (DeTroyer and Estenne, 1984). Some authors reported that damage to the cervical area of the spinal cord impaired the functioning of these respiratory muscles resulting in

168 hypoxia due to respiratory failure, and death in humans (Winslow and Rozovsky, 1983). The observed traumatic damage to the spinal cord resulted in hypoxia causing death in the birds killed by cervical dislocation. I suggest that hypoxia could be a reason for the observed open mouth breathing in birds killed via cervical dislocation. This is also supported by the longer duration of open mouth breathing in the birds killed by KED. Therefore, trauma to the spinal cord could be the main reason for death in the birds killed by KED as they had minimal brain trauma.

Erasmus et al. (2010c) suggested that the presence or absence of skull fractures did not affect the onset of insensibility in turkeys killed by NPCB or blunt trauma. Chapter 3 revealed that some birds in the 60-70 week old group did not have any skull fractures (Zephyr-E=4%, Zephyr-

EXL=29%, TED= 20%). However, all these birds were successfully killed and had a rapid loss of sensibility, similar to those with skull fractures. Therefore, my results agree with Erasmus et al. (2010c) that there is no association with degree of skull damage and time to insensibility or brain death.

6.1.3. Limitations of reflex and behaviour measures

In this thesis, euthanasia methods were assessed based on how quickly insensibility and subsequent brain death was induced (AVMA, 2013; NFACC, 2016). However, determining the exact onset of brain death is highly debated in both human and veterinary medicine. Verhoeven et al., 2015 stated that insensibility and brain death are best measured through electroencephalogram (EEG), and it is the most objective measure. But EEG is not practical for on-farm assessment of brain death. Animal care guidelines require that clinical death be

169 confirmed before leaving birds and disposing of carcasses (AVMA, 2013; NFACC, 2016). Some authors reported birds regaining sensibility or brain stem reflexes after application of a NPCB or cervical dislocation (turkeys killed with NPCB- Erasmus et al., 2010b; broiler chickens killed by

KED and manual cervical dislocation- Jacobs et al., 2019). Thus, loss of brain stem reflexes is not enough to confirm clinical death. If a bird regains sensibility, it may suffer until its clinical death. Stockpeople need to monitor birds until clinical death to minimize the birds’ suffering; if there is a return to sensibility, they need to apply a second killing method immediately.

Therefore, there is the need for behavioural and physiological reflexes, measures other than

EEG, to be identified as reliable indicators of clinical death, to be applied on birds prior to carcass disposal.

Brainstem reflexes and behavioural measures have been validated as conservative measures of sensibility and brain death (Gregory and Wotton, 1990a; Sandercock et al., 2014). I assessed the relationships between time to loss of brain stem reflexes, some behavioural observations and time of cessation of heart beat following application of the NPCB. My study suggests that cardiac arrest was positively correlated with onset of tonic convulsions, time to last movement, and cloacal relaxation for all three NPCB devices and thus, onset of tonic convulsions, last movement, and cloacal relaxation can be used by stockpeople to make accurate decisions of successful euthanasia under field conditions. Cessation of movement has been suggested to estimate irreversible brain death (Dawson et al., 2007; Erasmus et al., 2010a). Raj et al. (1990) studied the effect of CO2 stunning of hens on electroencephalogram (EEG) suppression and found that cessation of convulsions happened (at 93 s) just before brain death (at 101s).

Moreover, Raj et al. (1992) reported brain death in hens exposed to CO2 and Ar mixture at 58 s

170 based on EEG, and cessation of visible movements at 60s. My results agree with these findings, that the last visible movement occurred immediately after the brain death. In more a recent study,

Jacobs et al. (2019) assessed the time at cessation of musculoskeletal movements (vent, legs, tail) to interpret onset of brain death in broiler chickens. They reported that cessation of musculoskeletal movements were the last visible measures to confirm brain death.

During this project, reflexes were assessed every 10 s, which is more frequent than previous studies that used 15 s intervals to measure the reflexes (Martin et al., 2016, Woolcott et al, 2018 a,b). A 15s interval may have reduced the accuracy in determining when reflexes were lost, and may have led to their overestimation. The 10 s interval used in this study was chosen to give better accuracy, and still allow for multiple measures to be made sequentially within each interval, providing a large amount of information.

In Chapter 4, we used anesthetics (ketamine and medetomidine) to reduce distress or pain associated with the killing technique by abolishing the awareness of the animal. However, there was a significant effect of the anesthetics on several brain stem reflexes and behaviour measures.

The anesthetic protocol used allowed measurement of pupillary reflex as a conservative indicator of insensibility (due to brain death) in all birds. However, the anesthetics abolished or reduced nictitating membrane reflex, clonic convulsions, tonic convulsions, and cloacal relaxation, precluding their use for comparing killing methods. Woolcott et al. (2018b) also reported that the nictitating membrane reflex, clonic convulsions, and gasping were affected by the same anesthetic protocol in turkeys, although jaw tone was unaffected. In contrast, Sandercock et al.

(2014) reported observing nictitating membrane reflex even after respiratory arrest and brain

171 death in laying hens anesthetized with sevoflurane. The authors suggested that the long time to cessation of nictitating membrane reflex may have been due to combined effects of the used anesthetic and sedative. Combined, these studies suggest that, nictitating membrane reflex, clonic convulsions, tonic convulsions, and cloacal relaxation are more sensitive to the injectable anesthetics and may not be useful as an approximate measure of brain death in cases where ketamine and medetomidine have been administered in chickens.

In the current study the anesthetic protocol resulted in faster brain death in chickens. Thus, the observed time for brain death in chickens killed by KED (while under anesthesia) is conservative. Since anesthesia resulted in a faster time to brain death in birds killed via manual

CD, the time to death for birds killed via KED would likely be longer in awake chickens.

Sandercock et al. (2014) suggested that semi-awake (sedated) chickens demonstrated jaw tone and neck muscle tone but these two measures were absent when birds were under general anesthesia. Most of the birds in the current study were under general anesthesia and did not have jaw tone. Therefore, we did not assess jaw tone as a measure of insensibility in anesthetized birds. However, we found that individual variation occurred in depth of anesthesia in the chickens for the same anesthetic protocol.

6.1.4. Device success and operator safety

All three NPCB devices used in this study caused rapid, irreversible insensibility and subsequent death in layer chickens. Kill success was 100% for the TED device whereas only one failure was observed for Zephyr-EXL device (one bird out of 93 birds). However, a few device failures occurred for Zephyr -E-standard (4 birds out of 54 did not lose sensibility within 60 s). At the

172 beginning of the experiment, birds were killed by using Zephyr-E-standard with a common subject adapter and the conical shape bolt head. This common subject adapter is mainly designed for swine and rabbits. Three birds killed by Zephyr-E-standard failed to lose brain stem reflexes within 60s. Therefore, we switched to using the Zephyr-E-layer with a chicken subject adapter and the round shape bolt head. The Zephyr-E-layer differs from the standard Zephyr-E in that round shape bolt head and the chicken subject adapter. The chicken subject adapter allows for a better alignment of the device around the comb of the bird. Selection of the correct subject adapter for the species is an important factor for the killing success of Zephyr devices. However, another three birds out of 47 killed by the Zephyr-E-layer also failed to lose brain stem reflexes within 60s. The bolt velocity of Zephyr devices is adjustable as a function of air pressure.

Therefore, any pressure drop will result a lower bolt velocity and generate lower kinetic energy from the device. The pressure in the air compressor was decreased below the recommended level of 120 psi when killing these 7 birds by using Zephyr-E (4 birds by Zephyr-E standard + 3 birds with Zephyr-E- layer). Therefore, I suspect this could be a cause for the observed failures. The bolt velocity of the Zephyr-E is 20 m/s and generates 11 J of kinetic energy. The bolt velocity of

Zephyr-EXL is 27 m/s and generates 26 J kinetic energy. Therefore, the energy generated by

Zephyr-E is comparatively low. For the Zephyr-E, any air pressure level below 120 psi is not enough to generate sufficient force to successfully kill a chicken. Therefore, Zephyr-E should be used with the manufacturer-recommended exact air pressure level of 120 psi.

Another problem was encountered with Zephyr devices for one strain of chickens, the Plymouth

Barred Rock. After device application, the feathers of the Plymouth Barred Rock became lodged between the bolt and the adapter for both Zephyr-E and Zephyr-EXL. Despite this, the birds

173 were killed effectively with these devices. The Plymouth Barred Rock, a heritage strain of chicken, differs from the other tested strains in feathering with easily removable feathers.

Seventeen of the birds used in this study were Plymouth Barred Rock, and two of them resulted in this problem, when killed using the chicken subject adapter. Therefore, extra care needs to be taken when using Zephyr devices with chicken adapters on chickens in different plumage types.

When evaluating killing methods for practical on-farm use, the methods need to be safe for stockpeople. Sufficient training and skill are required for the method to be effective. Cervical dislocation is widely used as an on-farm euthanasia technique by poultry producers because it is easy to do, and practical and can be applied immediately during barn walk-throughs. People can be easily trained on how to do cervical dislocation and there is minimal risk to the safety of the operator. Operating the KED device is also easy, with minimal risk associated for the operator.

Comparatively, the potential for injury may be higher with operating NPCB devices. However, all three devices are designed with a safety pin to minimize risk, and the safety pin needs to be removed for the devices to function. The Zephyr devices additionally have a secondary safety switch as an extra precaution, and the TED requires depression of the activator for the device to function. Every individual using these devices should follow the steps of manufacturer’s operator safety guide and should receive proper training in order to ensure the device is used in an appropriate manner.

Additionally, proper restraint of the bird during device application is important, both for operator safety and for the welfare of the bird. We did not use restraining devices in the current study

174 because we needed to observe the reflexes and behavioural measures. Two people were involved during the killing process: one person applied the device while the other one restrained the bird.

However, one person can kill a bird with NPCB devices or KED using a restraint device without having assistance from a second person. A cone-shaped fabric (made of nylon) sleeve with an opening at the top and bottom ends was designed for the purpose of our study, but also to facilitate producers in restraint during euthanasia (Figure 6.1). This sleeve restrains the wings towards the body of the bird during convulsions, and one person can easily apply the restraining device on a bird.

Overall, more that 80% of birds showed external damage after application of the three NPCB devices. Some birds had bleeding from the mouth and nose, and damaged eyes. Minimizing external bleeding is important for biosecurity measures and aesthetic concerns. Serpa et al.

(2017) evaluated the efficacy of the Zephyr-EXL, powered at different air pressures levels (80,

90, 100, 110 and 120 psi) in killing broiler breeder roosters. They found that pressures between

80-120 psi are were equally effective at humanely killing broiler breeder roosters, and pressures from 80-110 psi reduced the external damage compared to 120 psi.

6.2. Assessment of aversion to CO2 in laying hens

In Chapter 5, I presented an assessment of the degree of aversion in laying hens exposed to 4 different CO2 concentrations (25%, 35%, 50%, 70%) based on approach avoidance and conditioned place avoidance paradigms with addition of a distress behaviour assessment. Results revealed that approach avoidance and conditioned place avoidance paradigms were successful in

175 assessing aversion to CO2 in layer chickens. Some hens avoided entering into some CO2 concentrations. Some hens displayed conditioned place avoidance to some CO2 concentrations.

I hypothesized that birds would avoid entering the treatment chamber when CO2 concentrations were aversive. To support this hypothesis, I observed some proximate effects by comparing the behaviours on Baseline Day versus Gas Day. Fewer hens entered the treatment chamber (TC) on

Gas Days than on Baseline Days. Avoidance of entering the TC on Gas Day revealed that the motivation of a bird to avoid the aversive environment inside the TC was higher than the motivation for acquiring the rewards inside the TC. More birds showed a higher number of curtain pushes and head insertions into TC before entering or during avoidance on Gas Days than

Baseline Days. The number of curtain pushes and head withdrawals before entering / avoiding

TC on Gas day reflect conflict or competing motivation between entering the TC to acquire the rewards and avoidance of gas. Overall, more hens demonstrated higher rates of vocalizations inside the control chamber (CC) on Gas Day than Baseline Day, showing evidence of either fearfulness or conflict to enter the TC.

All hens in the current study that entered the TC on Gas Day showed head shaking and open mouth breathing irrespective of CO2 concentration. Head shaking and open mouth breathing, previously observed in poultry for varied CO2 concentrations, have been suggested to be indicators of aversion (Webster and Fletcher, 2001) and respiratory distress (turkeys: Raj, 1996; broilers: Gerritzen et al., 2000; laying hens: Webster and Fletcher, 2001). Interpretation of head shaking in this regard is difficult. Some researchers suggested head shaking served as an alerting response to unexpected or novel events in poultry (Hughes, 1983) or an attempt to regain an alert

176 state when birds are feeling dizzy or start to lose sensibility during the exposure to CO2

(Gerritzen et al., 2007). In the current study, some hens demonstrated head shaking on Baseline

Days and Washout Days when the CO2 concentration was the same as ambient air. Therefore, the arguments that head shaking is an aversive reaction to CO2 or that it indicates respiratory distress is not fully supported. McKeegan et al. (2005) suggested that the acidity of CO2 may provide a gustatory stimulus which causes head shaking. Significantly more hens showed head shaking on Gas Days than Baseline Days for all CO2 concentrations. Therefore, hens are more likely to perform head shaking when exposed to CO2, but it is not necessarily indicative of respiratory distress.

In the current study, all except one bird voluntarily entered to 25% and 35% CO2. Overall, the number of hens entering into the TC decreased with increasing CO2 concentrations. Webster and

Fletcher (2004) suggested that hens could detect the presence of CO2 in the chute leading to a chamber filled with 30%, 45%, and 60%. In the current study, the purpose-built chamber allowed for greater control over the concentration of CO2 both within and outside the test chamber, so that birds were exposed to the appropriate concentrations. The TC was monitored in real-time with a wide span CO2 sensor and a real-time O2 sensor connected to an automatic data recording system. The CC was monitored with a wide span CO2 sensor and an ambient range CO2 sensor.

Further, a negative pressure exhaust sink was located between the two chambers to evacuate excess CO2 and maintain separate TC and CC environments. Some hens who did not enter into the TC demonstrated a number of curtain pushes and head insertions indicating several attempts to enter TC for 50% and 70% CO2 concentrations. Hens were exposed to the CO2 concentration inside the TC when they performed head inserts and curtain pushes, and were able to detect the

177 particular concentration inside the TC. Nearly half of the birds tested for 50% CO2 avoided entering the TC on Gas Day suggesting that half of the hens found 50% CO2 as an aversive stimulus. Only two hens entered into 70% CO2, and all the other hens tested avoided 70% CO2.

Therefore, the vast majority of hens were able to detect 70% CO2 and avoided entering into it.

McKeegan et al. (2003) observed an aversion threshold (based on an electrophysiological assessment) to CO2 of 24% in air for chickens and further reported that chickens can detect 11%

CO2 based on a behaviour study. The thresholds of hen nasal and buccal trigeminal nociceptors to CO2 have been investigated (McKeegan, 2004) and the authors suggest thresholds of 40–50%.

Therefore, in principle, hens should be able to detect the 25% and 35% CO2 concentrations, and should experience pain and/or discomfort when inhaling 40% or above CO2 concentrations. The hens who that did not enter into 50% CO2 and 70% CO2 showed a number of curtain pushes and head withdrawals indicating that hens found these CO2 concentration aversive, and quite possibly painful. Except for two, all other hens who voluntarily entered into either 25% or 35%

CO2 did not demonstrate head withdrawals prior to entering TC suggesting that hens entered

25% and 35% CO2 with minimal or no aversion, or that their motivation to access the rewards inside the TC was higher than the motivation to avoid 25% and 35% CO2.

Loss of posture (LOP) has been proven to be the reliable indicator of onset of insensibility in poultry based on electroencephalography (EEG) studies in broiler chickens (Coenen et al., 2000;

Gerritzen et al., 2004; Benson et al., 2012). In the current study, onset of ataxia occurred regardless of CO2 concentration. Latency to LOP was significant among the CO2 concentrations.

Overall latency to LOP decreased with increasing CO2 concentration. Similar results were

178 observed by Webster and Fletcher (2001) who reported that higher CO2 concentrations reduced the time to loss of sensibility based on time to LOP in layer hens.

6.2.1 Evaluation of the method

Hens who were in the control chamber were given a free choice to enter or avoid the Treatment

Chamber (TC) filled with a CO2 concentration. Prior to the testing phase hens were trained to enter TC pushing through a curtain. Overall, the vast majority of laying hens were able to learn the task by the fifth day of training. This indicates that the assigned task was within the cognitive capacity of laying hens.

In this experiment 12 birds were used for the study. Our plan was to use the same birds to check the aversiveness for four gas concentrations. We randomly assigned the birds for different CO2 concentrations. But some hens never entered the TC after their first, second or third gas treatments. Because of this, the number of birds tested for each concentration was different. Most of the hens avoided entering TC after their third gas experience. It seems that hens were able to learn the adverse effect inside the TC, with increasing exposure to any concentration. By rotating the same hens through different CO2 concentrations, there is a possibility that experience of a previous CO2 concentration could carry over to influence the results of a test with another

CO2 concentration. Birds that have experienced exposure to CO2 may not be representative of naïve birds. When the same bird was exposed to the next gas level, their sensitivity or tolerance to a gas concentration may have changed. Two hens who were exposed to 50% and 70% CO2 in their third round died without regaining of consciousness. Previously they had been exposed to

179 two other CO2 concentrations in their first and second rounds. The deaths could have been due to lack of physiological recovery or anatomical damage due to previous gas exposures.

In this experiment I assumed that the strength of motivation to get food (commercial food and meal worms) and motivation to avoid CO2 were additive. But this assumption does not always hold true. In the current trial, I assigned the order of birds based on their motivation to enter the treatment chamber during the training period. The most motivated birds were tested first. Food was withheld 1 hour prior to testing the first bird. All the birds were tested within 5 hours of feed deprivation. Kirkden et al. (2008) found that tolerance to the aversion of CO2 in rats was increased with the deprivation time and then started to decrease after a certain point. They found the least level of aversiveness at the intermediate feed deprivation. I provided both commercial feed and meal worms in the treatment chamber, with meal worms only provided in the treatment chamber except during the training. I observed that hens had a higher motivation to enter the TC to eat meal worms. I also assumed that providing other enrichments for foraging helped to increase their motivation to enter TC.

During the testing phase, we allowed 5 minutes for acclimatization in the control chamber prior to opening the sliding door which was in between the two compartments. We allowed another five minutes for the bird to enter the treatment chamber after the door was opened. The vast majority of the birds entered TC within 3 minutes from opening the sliding door in their training sessions. Therefore, I conclude that the allocated 5 minutes time was sufficient for the hens to make the decision on whether to enter the TC or not.

180

Time at egg laying was not considered in the current trial. However, most hens laid before 1.00 p.m., and the test began at 1.30 p.m. every day. We assumed all hens completed egg laying by that time, but we did not manage hens individually for their egg laying. This confounding effect could have interfered with the results of distress behaviour or avoidance. Every day we performed a health assessment for all the birds before the test. Therefore, it can be confirmed that health issues did not interfere with the results of the current study.

6.2.2 Implications

This thesis assessed available on-farm euthanasia methods for laying hens. However, in Chapter

5, aversion to different CO2 concentrations was assessed with their possibility to induce rapid insensibility in layer chicken. Thus, it is important to discuss how does the results of Chapter 5 apply to on-farm euthanasia context.

Carbon dioxide is used in two different techniques to euthanise chickens. One method is gradual filling the euthanasia box with CO2 and the other one is immersed birds in pre-filled CO2 chambers (Close et al., 1996). Gradual filling of 100% CO2 is accepted to achieve minimum pain and distress (AVMA, 2013). Geritzen et al. (2004) suggested that anesthesia is induced in a bird before CO2 level reach painful level. Immersion in CO2 is acceptable at low CO2 concentrations to avoid associate distress (AVMA, 2013). Gentle death in adult bird with minimal pain that take longer is preferred than quick death with distress (AVMA, 2013).

I studied 25% and 35 % CO2 concentrations, and found that these levels were relatively less aversive to the hens. Onset of insensibility occurred less than 20 s at 25% and 35% CO2

181 concentrations. Thus, suggested that there are welfare advantages to the hens if use 20% and

30% CO2 concentrations as prefilled technique in euthanasia. However, practical issues arise in on-farm context to use lower level of CO2. Many farms used gradual filling technique and it is not sure whether a lower CO2 concentration is possible to hold for a particular time period inside the CO2 chamber. The commercially available euthanasia boxes are mainly designed to gradual increase up to 60% CO2 or above levels.

Gerriteze et al. (2013) reported that multistage gas stunning has distinct advantages for bird welfare. Authors started with a lower concentration (30% or 40% CO2) and increased to 65%

CO2. Birds were exposed to 45 or 60 s at the particular CO2 concentration. Authors concluded that CO2 concentration did not exceed 40% before all the birds expressed loss of posture indication of unconsciousness. Moreover, it was reliable in both the experimental and the semicommercial setup. I did not find difference in latency to onset of LOP at 35% and 50% CO2.

Therefore, it is reasonable to start with 30% CO2 concentration and hold the level until bird loss its sensibility, and then increase to a higher concentration to achieve a rapid death in layer chicken euthanasia to achieve more welfare benefits.

However, these lower CO2 concentrations cannot recommend for chick euthanasia. The atrial pCO2 level is comparatively higher (60 mm Hg) in the embryo (Freeman and Vince, 1974).

Thus, day-old chicks are more resistant to hypercapnia and anoxia (Jaksch, 1981), Therefore, higher CO2 concentrations need to be targeted in day old chick euthanasia and the levels recommended for adult chickens are not sufficient to achieve rapid insensibility or death (Jaksch,

1981). Bethany et al., (2019) concluded that shortest time of distress and fastest time of death

182 occurred in old chicks immersed in 100% CO2 concentration in comparison to gradual filling technique. Therefore, results of the current study can only apply in on-farm context by targeting the welfare advantages of adult birds and not the chicks.

6.3 In retrospect

The results of the current study demonstrate that non-penetrating captive bolt (NPCB) devices caused rapid, irreversible insensibility and subsequent death in layer chickens. However, external skin damage and bleeding was higher for NPCB devices in comparison to cervical dislocation methods. Some birds killed by NPCB devices had mouth and nose bleeding, and damaged eyes.

External bleeding is important when considering biosecurity and aesthetic concerns. I would recommend doing a pilot study to identify the degree of external damage caused by different

NPCB devices using different pressures. In that way, it will be possible to come up with a more suitable scoring system to evaluate the external damage. In the current study we did not score the degree of external bleeding. Nose bleeding and mouth bleeding should also be included in the external damage and bleeding scoring system. Future studies to evaluate suitable air pressure levels for Zephyr-EXL, which cause minimal external damage and rapid insensibility leading to death should be based on a comprehensive external damage scoring system.

I would have liked to use pentobarbital sodium, which is considered the gold standard in euthanasia to induce rapid insensibility and brain death (Hernandez, 2018) as the secondary killing method instead of having to apply a second shot with the TED in the failed birds. This would preserve brain architecture better for macroscopic and microscopic assessment. In that

183 way, I would have been able to get a better idea about the degree (or lack) of pathology of the brains of the birds that failed to be killed with the tested method.

Chapter 5 revealed that layer hens showed learned behaviours going through the treatment rounds during testing for aversion to CO2 concentrations. Some birds might not have fully recovered physically when they were tested in the next immediate round. Therefore, I would have preferred to have 12 naïve birds for each concentration rather than rotating the same 12 hens through different concentrations. However, that would increase the number of animals necessary for the trial, and this may not fit with the 3Rs concept. This would also add considerable time and expense for training the birds. An alternative would be to permit additional days of rest to the hens between testing rounds. The other issue of repeating the hens through the rounds was that a hen may learn that she will be returned to her home pen if she does not enter TC. She also learnt that the TC is an unpredictable environment. This could be the reason that a greater number of hens avoided entering the TC in the 3rd round. However, I did not provide food in the home pen until the last hen completed the test in order to decrease the motivation of a hen to return to the home pen Meal worms were never provided inside the home pen after the habituation period, and a hen’s only chance to get meal worms was by entering the

TC.

6.4 Overall conclusion

The death of an animal at human hands, is an important element of an animal’s welfare.

Therefore, to achieve a good death in an animal, all possible measures need to be taken. People argue that euthanasia means ending of one’s life to eliminate pain and distress. However, many

184 of the available killing methods cause pain and distress. This thesis addresses how painful the available killing methods and their degree of suitability as a humane method in layer chickens, thus giving a chance to think and argue on whether we can achieve a good death without a humane killing method. It is our responsibility to ensure that the existing killing methods truly result in a good death to animals without inducing unnecessary pain or distress.

Available euthanasia guidelines and codes of practice listed several conditions which should be achieved during killing an animal in order to offer a humane death. These euthanasia protocols are based on previous research studies conducted on few species. A killing method assessed for one species or one age group cannot be generalized to the other species or other age groups.

Therefore, available killing methods need to be assessed on target species and target age groups.

Euthanasia guidelines should focus on species and age categories to achieve accurate recommendations.

The series of experiments conducted in this thesis have identified the potential of new physical killing devices (NPCB and KED) in layer chicken industry. Assessments were done in different age groups facilitating more accurate recommendations. Three different types of commercially available NPCB devices were assessed in layer chickens assuming that device differences could cause a different degree of welfare advantages in birds. This facilitates the most accurate recommendation of devices to use on layer chickens rather than recommending all NPCB devices in general. KED-model-C was assessed in different age groups as there were no recommended minimum body weight for application of the device. Manufacturers should clearly indicate age or weight limits (both minimum and maximum) for different poultry species.

185

Different poultry species have anatomical differences, thus using only the weight category may not be enough to select the correct model of the device.

Existing euthanasia protocols or guidelines have not recommended specific CO2 concentrations for layer hens to achieve welfare advantages. In Chapter 5, I identified relatively less aversive

CO2 concentration (25% and 35%) which resulted in insensibility less than 20s in layer chickens.

Results of approach avoidance tests are always relative and depend on the cost and rewards to the animal. The design of the current study provided an advantage in the terms of welfare as birds were not highly motivated to enter the TC. I provided luxury treat (meal worms) and some environmental enrichments which resulted in a positive motivation. Since the bird does not need them they can easily avoid entering TC. On the other hand, I withheld the feed, and this is a negative motivation. However, the feed withdrawal period was short and may not resulted in strong negative motivation in birds. Overall, the cost was not great for the bird and suggest of that 25% and 35% levels were not that aversive.

This thesis provides important information for euthanasia guidelines to be considered in the future with special attention to laying hen industry. Comprehensive trauma analysis of each physical killing method gives information for manufacturers to consider in further improvement of the existing devices to be more efficient as euthanasia methods. Every method had limitations, thus selecting of the most appropriate method for the existing situation is important in an on- farm context.

186

Figure 6. 1 : The fabric chicken restraint device (fabric sleeve made of nylon): A- a toggle at the head end to tighten it, B- Opening at the bottom end Images curtesy of Institute for Applied Poultry Technologies - euthanasia techniques http://www.iaptwest.org/welfare-and-pain-management/

187

BIBLIOGRAPHY

Abdalla M. A. and King A. S. (1975). The functional anatomy of the pulmonary circulation of the domestic fowl. Respiration Physiology 23(3): 267–290.doi: 10.1016/0034- 5687(75)90078-X. Adams D. B. and Sheridan A. D. (2008). Specifying the Risks to Animal Welfare Associated with Livestock Slaughter without Induced Insensibility. Australian Government Department of Agriculture, Fisheries and Forestry, (November), pp. 1–100. Available at: http://www.daff.gov.au/animal-plant----eealth/ welfare/aaws/a_scientific_comment_on_the_welfare_of_sheep_slaughtered_without_stu nning/specifying_the_risks_to_animal_welfare_associated_with_livestock_slaughter_wit hout_induced_insensibility. Alexander M.P. (1995). Mild traumatic brain injury: pathophysiological, natural history, and clinical management. Neurology 45, 1253-1260. Alkire M. T., Hudetz A. G. and Tononi G. (2008). Consciousness and Anesthesia. Science 322(5903): 876–880. doi:10.1126/science.1149213. Altman R.B. (1997). Avian Medicine and Surgery. Clubb SL, Dorrestein GM, et al (eds) Philadelphia, PA, Saunders Ambrose N, Wadham J, Morton D (2000) In: Refinement of Euthanasia. Progress in the Reduction, Refinement and Replacement of Animal Experimentation (Balls M, van Zeller AM, Halder ME, eds). Amsterdam: Elsevier Science. pp.1159-1169. American Veterinary Medical Association AVMA Guidelines for the Euthanasia of Animals: 2013 Edition (2013). Available online: https://www.avma.org/KB/Policies/Documents/euthanasia.pdf (accessed on 28 March 2016). Anastasov M. I. and Wotton S. B. (2012). Survey of the incidence of post-stun behavioural reflexes in electrically stunned broilers in commercial conditions and the relationship of their incidence with the applied water-bath electrical parameters. Animal Welfare, 21(2): 247–256. doi: 10.7120/09627286.21.2.247. Anil M. H. (1991). Studies on the return of physical reflexes in pigs following electrical stunning. Meat Science, 30(1): 13–21. doi: 10.1016/0309-1740(91)90030-T. Animal Production Science, 54: 187–193.http://dx.doi.org/10.1071/AN12128 Anton F., Euchner I., and Handwerker H. O. (1992). Psychophysical examination of pain induced by defined CO2 pulses applied to the nasal mucosa. Pain, 49: 53-60.

188

Anton F., Peppel P., Euchner I., and Handwerker H.O. (1991). Controlled noxious chemical stimulation: responses of rat trigeminal brainstem neurons to CO2 pulses applied to the nasal mucosa. Neuroscience Letters, 123: 208-211. Anton F., Euchne, I., Handwerker H.O., (1992). Psychophysical examination of the pain induced by defined CO2 pulses applied to the nasal mucosa. Pain 49 (1):53–60.

Aramideh M. and Ongerboer De Visser B. W. (2002). Brainstem reflexes: Electrodiagnostic techniques, physiology, normative data, and clinical applications. Muscle and Nerve, 26(1): 14–30. doi: 10.1002/mus.10120.

Atkinson S., Velarde A., and Algers B. (2013). Assessment of stun quality at commercial slaughter in cattle shot with captive bolt. Animal Welfare, 22: 473–481 Bader S., Meyer-Kuhling B., Gunther R., Breithaupt A., Rautenschlein S. and Gruber A. D. (2014). Anatomical and histologic pathology induced by cervical dislocation following blunt head trauma for on-farm euthanasia of poultry. The Journal of Applied Poultry Research, 23(3): 546–556. doi: 10.3382/japr.2014-00977. Baggott J. (1982). Gas Transport and pH Regulation, in Textbook of Biochemistry with Clinical Correlations (Devlin, T. M., Ed.) pp 10981123, John Wiley and Sons, New York Baker B.I, Torrey S., Widowski T.M.,Turner P.V., Knezacek T.D., Nicholds J., Crowe T.G Schwean-Lardner K. (2019). Evaluation of carbon dioxide induction methods for the euthanasia of day-old cull broiler chicks. Poultry Science 98:2043–2053. http://dx.doi.org/10.3382/ps/pey581

Banzett R. B, Pedersen S. H, Schwartzstein R. M., Lansing R.W. 92008). The affective dimension of laboratory dyspnea: Air hunger is more unpleasant than work/effort. American Journal of Respiratory and Critical Care Medicine 177, 1384–90

Basbaum A. I., Bautista D M, Scherrer G, and Julius D. (2009). Cellular and Molecular Mechanisms of Pain. Cell, 139:267-284. DOI 10.1016/j.cell.2009.09.028 Beausoleil N.J., and Mellor D.J. (2015). Introducing breathlessness as a significant animal welfare issue. New Zealand Veterinary Journal 63:44-51. Benson, E.R.1., Alphin, R.L., Rankin, M.K., Caputo M.P., Kinney C.A. and Johnson A.L. (2012). Evaluation of EEG based determination of unconsciousness vs. loss of posture in broilers. Research in Veterinary Science. 93(2):960-964. doi: 10.1016/j.rvsc.2011.12.008 Blackman N., Cheetham K. and Blackmore D. (1986). Differences in blood supply to the cerebral cortex between sheep and calves during slaughter. Research in Veterinary Science, 40:252–254. Blackmore D. K., Newhook J. C. and Grandin T. (1983). Time of onset of insensibility in four- to six-week-old calves during slaughter. Meat Science. 9(2):145–149. doi: 10.1016/0309- 1740(83)90024-4.

189

BMJ (1976). Diagnosis of brain death: statement issued by the honorary secretary of the Conference of Medical Royal Colleges and their Faculties in the United Kingdom on 11 October 1976. BMJ 1976; 2: 1187-1188.29. Bock Industries Inc. Captive Bolt Devices for Humane Livestock Euthanasia (2016). Available online: http://www.bock-industries.com/ Bove S. E., Flatters S. J. L., Inglis J. J. and Mantyh P.W. (2009). New advances in musculoskeletal pain. Brain Research Reviews. Elsevier B.V., 60(1): pp. 187–201. doi: 10.1016/j.brainresrev.2008.12.012. Breazile J. E. and Hartwig H.G. (1989). Central nervous system. In: King AS, McLelland J (eds) Form and function in birds. Academic, London, pp 485–566 Brown R. E., Basheer R., McKenna J. T., Strecker R. E. and McCarley R. W. (2012). ‘Control of Sleep and Wakefulness’, Physiological Reviews, 92(3), pp. 1087–1187. doi: 10.1152/physrev.00032.2011. Bryant B., Silver W.L. (2000). Chemesthesis: the common chemical sense. In: Finger TE, Silver WL, Restrepo D, editors. The neurobiology of taste and smell. New York: Wiley-Liss. p 73–100. Butler A. B. and Hodos W. (2005a). Comparative Vertebrate Neuroanatomy- Evolution and Adaptation. John Wiley & Sons INC. publications Butler, A. B., Manger, P. R., Lindahl, B. I. B. and Arhem, P. (2005b) ‘Evolution of the neural basis of consciousness: A bird-mammal comparison’, BioEssays, 27(9), pp. 923–936. doi: 10.1002/bies.20280. Cabanac, M., (1979). Sensory pleasure. Quarterly Review of Biology 54,1–129. Carstens E. and Moberg G. P. (2000). ‘Recognizing pain and distress in laboratory animals’, ILAR Journal, 41(2), pp. 62–71. doi: 10.1093/ilar.41.2.62. Casey-Trott T. M., Millman S. T., Turner P. V., Nykamp S. G. and Widowski T. M. (2013). Effectiveness of a nonpenetrating captive bolt for euthanasia of piglets less than 3 d of age. J Anim Sc., 91:5477-5484. doi: 10.2527/jas.2013-6320 Casey-Trott T. M., Millman S. T., Turner P. V., Nykamp S. G., Lawlis P. C. and Widowski T. M. (2014). Effectiveness of a nonpenetrating captive bolt for euthanasia of 3 kg to 9 kg pigs. J Anim Sc., 92:5166-5174. doi: 10.2527/jas.2014-7980 Chapman C. R. (2005). Pain perception and assessment. Minerva Anestesiologica, 71(7-8):413- 417

Chen, X., J. Gallar, M.A. Pozo, M. Baeza, and C. Belmonte (1995). CO2 stimulation of the cornea: a comparison between human sensation and nerve activity in polymodal nociceptive afferents of the cat. European Journal of Neuroscience 7: 1154-1163.

190

Claassen, J., Carhuapoma, J.R., Kreiter, K.T., Du, E.Y., Connolly, E.S., Mayer, S.A. (2002). Global cerebral edema after subarchnoid hemorrhage: frequency, predictors, and impact on outcome. Stroke 33, 1225-1232

Clark L, Hagelin Julie, and Werner S J. (2014). "The Chemical Senses in Birds". USDA National Wildlife Research Center -Staff Publications. 1618. https://digitalcommons.unl.edu/icwdm_usdanwrc/1618

Clarke K. W. and Hall, L. W. (1990). A survey of anaesthesia in small animal practice: AVA/BSAVA report’, Veterinary Anaesthesia and Analgesia, 17(1): 4–10. doi: 10.1111/j.1467-2995.1990.tb00380.x.

Close, B., Banister K., Baumans V., Bernoth E. M., Bromage N., Bunyan J., Erhardt W., Flecknell P., Gregory N., Hackbarth H., Morton D., Warwick C. (1997). Recommendations for euthanasia of experimental animals: part 2. Lab. Anim. 31:1–32 Coenen, A., A. Smith, L. Zhonghua, and G. Vanluijtelaar (2000). Gas mixtures for anesthesia and euthanasia in broiler chickens. World’s Poultry Science Journal 56: 225-234. Coenen A. M. L., Lankhaar, J., Lowe, J. C. and McKeegan, D. E. F. (2009). Remote monitoring of electroencephalogram, electrocardiogram, and behavior during controlled atmosphere stunning in broilers: implications for welfare.’, Poultry science, 88(1), pp. 10–19. doi: 10.3382/ps.2008-00120. Collins J. G., Kendig J. J. and Mason P. (1995). Anesthetic actions within the spinal cord: contributions to the state of general anesthesia. Trends in Neurosciences, 18(12): 549– 553. doi: 10.1016/0166-2236(95)98377-B. Cook C. J., Devine C. E., Gilbert K.V., Smith D. D. and Maasland S. A. (1995). The effect of electrical head-only stun duration on electroencephalographic-measured seizure and brain amino acid neurotransmitter release. Meat Sci. 40:137–147. Cors J.C., Gruber A.D., Gunther R., Meyer-Kuhling B., Esser K.H., Rautenschlein S. (2015). Electroencephalographic evaluation of the effectiveness of blunt trauma to induce loss of consciousness for on-farm killing of chickens and turkeys. Poult. Sci. 94:147–155. doi: 10.3382/ps/peu038 Cruccu G. and Deuschl G. (2000).The clinical use of brainstem reflexes and hand-muscle reflexes. Jounal of animal Sc. 91 (11):5477-87.doi: 10.2527/jas.2013-6320 Cullen L.K. (1996). Medetomidine sedation in dogs and cats: a review of its pharmacology, antagonism and dose. British Veterinary Journal 152:519–535 Damasio A. and Meyer K. (2009). Consciousness: An Overview of the Phenomenon and of Its Possible Neural Basis. S. Laureys & G. Tononi (Eds.) The Neurology of Consciousness: 3-14. Online access : https://www.sciencedirect.com/sdfe/pdf/download/eid/3-s2.0- B9780123741684000010/first-page-pdf

191

Danneman, P. J., Stein S, Walshaw S.O. (1997). Humane and practical implications of using carbon dioxide mixed with oxygen for anesthesia or euthanasia of rats. Lab. Anim. Sci. 47:376–385 Dawkins M. S. (1980). Animal Suffering. Chapman and Hall: London, UK Dawkins M. S. (1990). From an animal's point of view: Motivation, fitness, and animal welfare. Behavioral and Brain Sciences 13(1):1-9 Dawson M. D., Lombardi M. E., Benson E. R., Alphin R. L. and Malone G. W. (2007). Using accelerometers to determine the cessation of activity of broilers. Journal of Applied Poultry Research, 16(4), pp. 583–591. doi: 10.3382/japr.2007-00023. Derr R. F. (1991). Pain perception in decapitated rat brain. Life Sciences, 49(19): 1399–1402. doi: 10.1016/0024-3205(91)90391-N. Desmond J. E., Rosenfield M. E. and Moore J. W. (1983). An HRP study of the brainstem afferents to the accessory abducens region and dorsolateral pons in rabbit: Implications for the conditioned nictitating membrane response. Brain Research Bulletin, 10(6): 747– 763. doi: 10.1016/0361-9230(83)90208-3. DeTroyer A., Estenne M., (1984). Coordination between rib cage muscles and diaphragm during quiet tidal breathing in humans. J Appl Physiol, 57:899–906.

Drew L.B. and Drew W.E. (2004). The contrecoup-coup phenomenon. Neurocritical Care 1: 385-390. doi.org/10.1385/NCC:1:3:385 Dumont R.J., Okonkwo D.O., Verma S., Hurlbert R.J., Boulos P.T., Ellegala D.B., Dumont A.S. (2001). Acute Spinal Cord Injury, Part I: Pathophysiologic Mechanisms. Clinical Neuropharmacology 24: 254–264 Duncan I. J. H. (1996). Animal welfare defined in terms of feelings. Acta Agriculturae Scandinavica Section A, Animal Science, Supplement 27: 29-35. Duncker H. R. (1974). Structure of the avian respiratory tract. Respiration Physiology, 22(1–2): 1–19. doi: 10.1016/0034-5687(74)90044-9. Eccleston C. (1995). Chronic pain and distraction: an experimental investigation in to the role of sustained and shifting attention in the processing of chronic persistent pain. Behaviour research therapy 33:391-405. Edelman D. B., Baars B. J. and Seth A. K. (2005). Identifying hallmarks of consciousness in non-mammalian species. Consciousness and Cognition, 14(1): 169–187. doi: 10.1016/j.concog.2004.09.001. Erasmus M. A., Turner P. V. and Widowski T. M. (2010a). Measures of insensibility used to determine effective stunning and killing of poultry. Journal of Applied Poultry Research, 19(3): 288–298. doi: 10.3382/japr.2009-00103.

192

Erasmus M. A., Lawlis P., Duncan I. J. H. and Widowski T. M. (2010b). Using time to insensibility and estimated time of death to evaluate a nonpenetrating captive bolt, cervical dislocation, and blunt trauma for on-farm killing of turkeys. Poultry Science, 89(7): 1345–1354. doi: 10.3382/ps.2009-00445. Erasmus M.A., Turner P.V., Nykamp S.G., Widowski T.M. (2010c). Brain and skull lesions resulting from use of percussive bolt, cervical dislocation by stretching, cervical dislocation by crushing and blunt trauma. Veterinary Records 167: 850–858. European Council. European Council Regulation (EC) 1099/2009 of 24 September 2009 on the Protection of Animals at the Time of Killing. Available online: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/325000/re gulation_1099_2009_en.pdf Evans K.C., Banzett R.B., Adams L., McKay L., Frackowiak R.S., Corfield D.R., (2002). BOLD fMRI identified limbic, paralimbic, and cerebellar activation during air hunger. Journal of Neurophysiology 88, 1500–11, 2002 Farmer C.G. (2006). On the origin of avian air sacs. Respiratory Physiology & Neurobiology 154: 89–106. Feldman J. L., Mitchell G. S. and Nattie E. E. (2003). Breathing: Rhythmicity, Plasticity, Chemosensitivity. Annual Review of Neuroscience, 26(1): 239–266. doi: 10.1146/annurev.neuro.26.041002.131103. Feng Y., Simpson T.L. (2003). Nociceptive sensation and sensitivity evoked from human cornea and conjunctiva stimulated by CO2. Invest. OPhth. Vis. Sci. 44., 529-532.

Feng, Y., and T. L. Simpson (2003). Nociceptive sensation and sensitivity evoked from human cornea and conjunctiva stimulated by CO2. Investigative Ophthalmology & Visual Science 44: 529-532.

Finnie J. W., Blumbergs P. C., Manavis J., Summersides G. and Davies R. (2000). Evaluation of brain damage resulting from penetrating and non‐penetrating captive bolt stunning using lambs. Australian Veterinary Journal 78: 775–778.

Finnie J.W., Manavis J., Summersides G.E., Blumbergs P.C. (2003). Brain damage in pigs produced by impact with a non-penetrating captive bolt pistol. Aust. Vet. J. 81:153–155. doi: 10.1111/j.1751-0813.2003.tb11078.x Forslid A., Ingvar M., Rosen I., Ingvar D.H. (1986). Carbon dioxide narcosis: influence of short- term high concentration carbon dioxide inhalation on EEG and cortical evoked responses in the rat. Acta Physiolgica 127: 281-287. Fraser D. and Duncan I. J. (1998). Pleasures, Pains and Animal Welfare: Toward a Natural History of Affect. Animal Welfare, 7: 383–396.

193

Fraser D., Weary D. M., Pajor E. A., and Milligan B. N. (1997). A scientific conception of animal welfare that reflects ethical concerns. Animal Welfare, 6:187-205. Freeman, B. M. and Vince M. A. (1974). Development of the Avian Embryo. Chapman and Hall, London. Freeman K. T., Koewler N. J., Jimenez-Andrade J. M., Buus R. J., Herrera M. B., Martin C. D., Ghilardi J. R., Kuskowski M. A., Mantyh P. W. (2008). A fracture pain model in the rat: adaptation of a closed femur fracture model to study skeletal pain. Anesthesiology, 108(3): 473–483. doi: 10.1097/ALN.0b013e3181649351. Gentle M.J. (2001). Attentional shifts alter pain perception in the chicken. Animal Welfare, 10:187-194 Gentle M.J. (2011). Pain issues in poultry. Applied Animal Behaviour Science 135:252– 258 doi:10.1016/j.applanim.2011.10.023 Gerritzen M. A., Lambooij E, Hillebrand S.J., Lankhaar, J.A. and Pieterse C. (2000). Behavioral responses of broilers to different gaseous atmospheres. Poultry Science 79(6): 928-933. http://www.ncbi.nlm.nih.gov/pubmed/10875779. Gerritzen M. A, Lambooij B., Reimert H., Stegeman A., Spruijt B. (2004). On-farm euthanasia of broiler chickens: effects of different gas mixtures on behavior and brain activity. Poultry Science, 83(8): 1294–1301. doi: 10.1093/ps/83.8.1294. Gerritzen, M., B. Lambooij, H. Reimert, A. Stegeman, and B. Spruijt (2007). A note on behaviour of poultry exposed to increasing carbon dioxide concentrations. Applied Animal Behaviour Science 108(1–2): 179–185. doi: 10.1016/j.applanim.2006.11.014. Gerstenbrand F. and Stepan C. A (2001). Mild traumatic brain injury. Brain injury, 15(2): 95–97. doi: 10.3171/2010.11.FOCUS.Intro. Getty, R. (1975). Sisson and Grossman’s the Anatomy of the Domestic Animals.W.B. Saunders Company, Philadelphia Gibson T. J., Johnson C. B., Stafford K. J., Mitchinson S. L. and Mellor D. J. (2007). Validation of the acute electroencephalographic responses of calves to noxious stimulus with scoop dehorning. New Zealand veterinary journal, 55(4): 152–7. doi: 10.1080/00480169.2007.36760. Gibson T. J., Johnson C. B., Murrell J. C., Hulls C. M., Mitchinson S. L., Stafford K. J., Johnstone A. C. and Mellor D. J. (2009).Electroencephalographic responses of halothane - anaesthetised calves to slaughter by ventral-neck incision without prior stunning. New Zealand veterinary journal, 57(2): 77–83. doi: 10.1080/00480169.2009.36882. Gibson T. J., Ridler A. L., Lamb C. R., Williams A., Giles S., and Gregory, N. G. (2012). Preliminary evaluation of the effectiveness of captive-bolt guns as a killing method without exsanguination for horned and unhorned sheep. Animal Welfare, 21, 35–42

194

Gibson T. J., Dadios N., Gregory N. G. (2015a). Effect of neck cut position on time to collapse in halal slaughtered cattle without stunning. Meat Science. Elsevier B.V., 110: 310–314. doi: 10.1016/j.meatsci.2015.03.026. Gibson T. J., Mason C. W., Spence J. Y., Barker H., Gregory N. G. (2015b). Factors Affecting Penetrating Captive Bolt Gun Performance. Journal of Applied Animal Welfare Science, 18 (3):222-238 Gibson T.J., Rebelo C.B., Gowers T.A., Chancellor N.M. (2018). Electroencephalographic assessment of concussive non-penetrative captive bolt stunning of turkeys. British Poultry Science 59 (1):13-20. DOI: 10.1080/00071668.2017.1401215 Gleed R. D., and Ludders J.W. (eds). (2001). Recent Advances in Veterinary Anesthesia and Analgesia: Companion Animals. Ithaca, NY, International Veterinary Information Service Goldberg M.E. (2014). Pain Management for Veterinary Technicians and Nurses. John Wiley and Sons Grandin T. (2002). Return-to-sensibility problems after penetrating captive bolt stunning of cattle in commercial beef slaughter plants. Journal of the American Veterinary Medical Association, 221:1258–1261 Grandin T. (2013). Making Slaughterhouses More Humane for Cattle, Pigs, and Sheep. Annual Review of Animal Biosciences, 1(1): 491–512. doi: 10.1146/annurev-animal-031412- 103713. Grandin, T., S.E. Curtis, T.M. Widowski, and J.C. Thurmon (1986). Electro-immobilization versus mechanical restraint in an avoid-avoid choice test for ewes. J Animal Science 62(6):1469-1480. Gregory N. G., Fielding H. R., von Wenzlawowicz M. and von Holleben K. (2010).Time to collapse following slaughter without stunning in cattle. Meat Science. Elsevier Ltd, 85(1): 66–69. doi: 10.1016/j.meatsci.2009.12.005. Gregory N. G., Lee C. J. and Widdicombe J. P. (2007). Depth of concussion in cattle shot by penetrating captive bolt. Meat Science, 77:499–503 Gregory N.G. and Wotton S.B. (1990a). Comparison of neck dislocation and percussion of the head on visual evoked responses in the chicken's brain. Veterinary Records, 126:570–572 Gregory N.G. and Wotton S.B. (1990b). Effect of stunning on spontaneous physical activity and evoked activity in the brain. British Poultry Science 31:215-220 Guais A., Brand G., Jacquot L., Karrer M., Dukan S., Grévillot G., Molina T. J., Bonte J., Regnier M., Schwartz L. (2011). Toxicity of carbon dioxide: A review, Chemical Research in Toxicology, 24(12): 2061–2070. doi: 10.1021/tx200220r.

195

Haas D.A. and Harper D.G. (1992). Ketamine: A review of its pharmacologic properties and use in ambulatory anesthesia. Anesthesia Progress 39:61-68

Hall J. B. and McShane P.J. (2013). Ventilatory Failure. in The Merck Manual. Sharp and Dohme Corp, Whitehouse Station, NJ.

Hameroff S. (2001). Anesthesia: The “Other Side” of Consciousness’. Consciousness and Cognition, 10(2): 217–229. doi: 10.1006/ccog.2001.0512. Harrop J., Sharan A., Vaccaro A.R., Przybylski G.J. (2001). The cause of neurologic deterioration after acute cervical spinal cord injury. Spine 26:340–346. Hari R., Portin K., Kettenmann B., Jousmaki V., Kobal G., (1997). Right-hemisphere preponderance of responses to painful CO stimulation of the human nasal mucosa. Pain 72, 145–151. Haselsberger K., Pucher R., Auer L. M. (1988). Prognosis after acute subdural or epidural haemorrhage. Acta Neurochirurgica, 90(3–4): 111–116. doi: 10.1007/BF01560563. Heard D. (2000). Perioperative supportive care and monitoring. Veterinary Clinics of North America: Exotic Animal Practice 3(3): 587-615. https://doi.org/10.1016/S1094- 9194(17)30064-6 Hernandez E. (2018). Electroencephalographic and behavioral evaluation of physical methods for on-farm euthanasia of poultry. Doctoral thesis, University of Guelph, Canada Hiestand W.A. and Randall W.C. (1941). Species differenciation in the respiration of birds following carbon dioxide administration and the location of inhibitory receptors in the upper respiratory tract. J Cell Comp Physiol17:333. Hogg D. A. (1982). Fusions occurring in the postcranial skeleton of the domestic fowl. J Anat. ,135(3):501-512 Hogg D.A. (1984). The development of pneumatisation in the postcranial skeleton of the domestic fowl. J. Anatomy. 139(1): 105-113 HSA (2004). Practical Slaughter of Poultry: A Guide for the Small Producer. Humane Slaughter Association. Hughes B.O (1983). Head shaking in fowls: the effect of environmental stimuli. Applied Animal Ethology 11:45–53. Husband S. and Shimizu T. (1999). Evolution of the avian visual system. http://luna.cas.usf.edu/~husband/evolve/default.htm. International Association for the study of pain (IASP). Pain terms: A current list with definitions and notes on usage. Pain Suppl. 3, S217. https://www.iasp-pain.org/ Jacobs L., Bourassa D.V., Harris C.E., Buhr R.J. (2019). Euthanasia: Manual versus mechanical cervical dislocation for broilers. Animals 9:47. doi:10.3390/ani9020047

196

Jaksch W. (1981). Euthanasia of day-old male chicks in the poultry industry. Int. J. Stud. Anim. Prob. 2:203–213. Jarvis E.D., Gu¨ ntu¨ rku¨ n, O., Bruce, L., Csillag, A., Karten, H., Kuenzel, W., Medina, L., Paxinos, G., Perkel, D.J., Shimizu, T., Striedter, G., Wild, J.M., Ball, G.F., Dugas- Ford, J., Durand, S.E., Hough, G.E., Husband, S., Kubikova, L., Lee, D.W., Mello, C.V., Powers, A., Siang, C., Smulders, T.V., Wada, K., White, S.A., Yamamoto, K., Yu, J., Reiner, A., Butler, A.B. (2005). Avian Brain Nomenclature Consortium. Avian brains and a new understanding of vertebrate brain evolution. Nature Reviews Neuroscience 6, 151–159 Jaynes J. (1985). Sensory pain and conscious pain. Behavioural and brain sciences 8:61-63 Jiang C., Rojas A., Wang R. and Wang X. (2005). CO2 central chemosensitivity: Why are there so many sensing molecules?. Respiratory Physiology and Neurobiology, 145(2–3): 115– 126. doi: 10.1016/j.resp.2004.07.005. KC L., A. K. Johnson, T. A. Shepherd, J. P. Stinn, H. Xin, K. J. Stalder, L. A. Karriker, M. A. Sutherland, D. C. Lay, Jr., and S. T. Millman. (2016). Aversion to carbon dioxide gas in pigs using approach–avoidance and conditioned place avoidance paradigms. Journal of Animal Science, Volume 94, Issue suppl 2: 9-10. doi.org/10.2527/msasas2016-020 KC L. (2018). Investigation of Aversion Associated with Gas Inhalants in Pigs Using Cognitive Paradigms. Graduate Theses and Dissertations, Iowa State University Digital Repository, 16726. Accessed online: https://lib.dr.iastate.edu/etd/16726. KC L., A. K. Johnson, T. A. Shepherd, J. P. Stinn, H. Xin, K. J. Stalder, L. A. Karriker, M. A. Sutherland, D. C. Lay, Jr., and S. T. Millman. (2016). Aversion to carbon dioxide gas in pigs using approach–avoidance and conditioned place avoidance paradigms. Journal of Animal Science, Volume 94, Issue suppl 2: 9-10. doi.org/10.2527/msasas2016-020 Keegan R.D., Greene S.A., Bagley R.S., Moore M.P., Weil A.B., Short C.E. (1995). Effects of medetomidine administration on intracranial pressure and cardiovascular variables of isoflurane-anesthetized dogs. American Journal of Veterinary Research 56:193–198. Kinney H., Korein J., Panigrahy A., Dikkes P., Goode R. (1994). Neuropathological findings in the brain of Karen Ann Quinlan. The New England Journal of Medicine, 330: 1469-1475. Kirkden R. D. and Pajor E. A. (2006). Using preference, motivation and aversion tests to ask scientific questions about animals’ feelings, Applied Animal Behaviour Science, 100(1– 2): 29–47. doi: 10.1016/j.applanim.2006.04.009. Kirkden R.D., Niel L., Lee G., Makowska I. J., Pfaffinger M.J., Weary D.M. (2008). The validity of using an approach-avoidance test to measure the strength of aversion to carbon dioxide in rats. Applied Animal Behaviour Science 114: 216–234 Koechner M.F.G. Co., Inc. Tipton, MO (2016) Available online at: http://www.turkeycoops.com/euthanizingdevice.html.

197

Kohler I., Meier R., Busato A., Neiger-Aeschbacher G., Schatzmann U. (1999). Is carbon dioxide (CO2) a useful short acting anaesthetic for small laboratory animals? Laboratory Animals Ltd. Laboratory Animals 33: 155-161 Kristensen H. H., Burgess L. R., Demmers T. G. H., Wathes C. M. (2000). The preferences of laying hens for different concentrations of atmospheric ammonia. Applied Animal Behaviour Science 68:307-318 Label L. (1997). Injuries and disorders of the head and brain, Mosby, New York.

Lambooij E., Gerritzen M. A., Engel B., Hillebrand S. J. W., Lankhaar J., Pieterse C. (1999). Behavioural responses during exposure of broiler chickens to different gas mixtures. Applied Animal Behaviour Science 62:255–265

Lambooij E., Reimert H., Van De Vis J. W., Gerritzen M. A. (2008). Head-to-Cloaca Electrical Stunning of Broilers. Poultry Science, 87: 2160–2165. doi: 10.3382/ps.2007-00488.

Lancet (1976). Conference of the Royal Colleges and Faculties of the United Kingdom: Diagnosis of brain death. Lancet 1976; 2: 1069-1070.

Lansing R.W., Gracely R.H., Banzett R.B. (2009). The multiple dimensions of dyspnea: Review and hypotheses. Respiratory Physiology and Neurobiology 167: 53–60, 2009 Leach M. C., Bowell V. A, Allan T. F., Morton D. B. (2002). Aversion to gaseous euthanasia agents in rats and mice. Comparative medicine, 52(3): 249–257. Lee J., Taira T., Pihlaja P., Ransom B. R., Kaila K. (1996). Effects of CO2 on excitatory transmission apparently caused by changes in intracellular pH in the rat hippocampal slice. Brain Research, 706(2): 210–216. doi: 10.1016/0006-8993(95)01214-1. Lierz M. and Korbel R. (2012). Anesthesia and Analgesia in Birds. Journal of Exotic Pet Medicine. Elsevier Inc., 21(1): 44–58. doi: 10.1053/j.jepm.2011.11.008. Llonch P., Rodríguez P., Jospin M., Dalmau A., Manteca X., Velarde A. (2012). Assessment of unconsciousness in pigs during exposure to nitrogen and carbon dioxide mixtures. Animal, pp. 1–7. doi: 10.1017/S1751731112001966. Magnussen H., Wilmer H., Scheid P. (1979). Gas exchange in air sacs: contribution to respiratory gas exchange in duck lungs. Respiratory Physiology. 27: 129–146. Maina J.N. (2003). Developmental dynamics of the bronchial (airway) and air sac systems of the avian respiratory system from day 3 to day 26 of life:a scanning electron microscopic study of the domestic fowl, Gallus gallus variant domesticus. Anatomy and Embryology 207 (2):119–134. DOI 10.1007/s00429-003-0333-6

Martin J. E., McKeegan D. E. F., Sparrey J., Sandilands V. (2016). Comparison of novel mechanical cervical dislocation and a modified captive bolt for on-farm killing of poultry on behavioural reflex responses and anatomical pathology. Animal Welfare, 25(2): 227– 241. doi: 10.7120/09627286.25.2.227.

198

Martin J.E., McKeegan D.E.F., Sparrey J., Sandilands V. (2017). Evaluation of potential killing performance of novel percussive and cervical dislocation tools in chicken cadavers. Br. Poult. Sci. 58:216–223. Martin J.E., Sandilands V., Sparrey J., Baker L., McKeegan D.E. (2018a). On farm evaluation of a novel mechanical cervical dislocation device for poultry. Animals 8: 10. doi:10.3390/ani8010010 Martin J.E., Sandercock D.E., Sandilands V., Sparrey J., Baker L., Sparks N.H.C., McKeegan D.E.F. (2018b). Welfare risks of repeated application of on-farm killing methods for poultry. Animals 8(3): 39; doi:10.3390/ani8030039 Martin J.E., Sandilands V., Sparrey J., Baker L., Dixon L.M., McKeegan D.E.F. (2019) Welfare assessment of novel on-farm killing methods for poultry. PLoS ONE 14(2): e0212872. https://doi.org/10.1371/journal.pone.0212872

Martoft L., Stødkilde-Jørgensen H., Forslid A., Pedersen H. D. and Jørgensen P. F. (2003). CO2 induced acute respiratory acidosis and brain tissue intracellular pH: a 31P NMR study in swine. Laboratory animals, 37: 241–248. doi: 10.1258/002367703766453092.

Mason C., Spence J., Bilbe L., Hughes T., Kirkwood J. (2009). Methods for dispatching backyard poultry. Veterinary Record. 164: 220 Mason J.R. and Clark L. (2000). The chemical senses in birds, in G.C. Whittow (Ed.), Sturkies Avian Physiology, Academic Press, London McKeegan D.E.F. (2004). Mechano-chemical nociceptors in the avian trigeminal mucosa. Brain Research Reviews, 46: 146-154 McKeegan D.E.F., Smith F. S., Demmers T.G.M, Wathes C.M. and Jones R. B. (2005). Behavioural correlates of olfactory and trigeminal gaseous stimulation, Gallus domesticus. Physiology and Behavior, 84: 761-776 McKeegan D.E.F., Mclntyre J., Demmers T.J.M, Wathes C.M., Jones R.B. (2006). Behavioural responses of broiler chickens during acute exposure to gaseous stimulation. Applied Animal Behaviour Science, 99:271-286. Mclelland J., (1990). A Colour Atlas of Avian Anatomy, Edinburgh, Scotland McLeod W.M., Trotter, D.M., Lumb, J.W. (1964). Avian Anatomy. Burgess Publishing Company, United States of America Mellor D.J. (2016). Updating Animal Welfare Thinking: Moving beyond the “Five Freedoms” towards “A Life Worth Living”. Animals (Basel) 6 (3): 21. doi: 10.3390/ani6030021 Mellema M.S (2008). The neurophysiology of dyspnea. Journal of Veterinary Emergency and Critical Care 18: 561–71

199

Melzack R. and Wall P. (1965). Pain Mechanisms: A New Theory. Science: 971–979. doi: 10.1126/science.150.3699.971. Menon D. K., Schwab K., David W. Wright D.W., Maas A.I. (2010). Position Statement: Definition of Traumatic Brain Injury. Arch Phys Med Rehabil Vol 91, doi:10.1016/j.apmr.2010.05.017. Accessed online https://www.archives-pmr.org/article/S0003-9993(10)00650-7/pdf

Merck M. Veterinary Forensics: Animal Cruelty Investigations. 2nd ed. New York, NY: Wiley; 2012. Moberg G.P. (2000). Biological responses to stress: implications of animal welfare. In: Moberg, G.P., Mench J.A. (eds), Biology of Animal Stress. CABI Publishing. New York, pp 1-21. Mohandas C. (1971). Brain Death: a clinical and pathological study. J Neurosurg. :35-211 Mollaret P., Bertrand I., Mollaret H. (1959). Comadepasse et necroses nerveuses centrales massives. Revue Neurologique, 101, 116-139. Nasr M. A. F., Nicol C. J., Murrell J. C. (2012). Do laying hens with keel bone fractures experience pain? PLoS ONE, 7(8). doi: 10.1371/journal.pone.0042420. National Farm Animal Care Council Code of Practice for the Care and Handling of Pullets and Laying Hens (2016). Available online: http://www.nfacc.ca/pdfs/pullets_and_laying_hens_code_of_practice.pdf National Farm Animal Care Council Code of Practice for the Care and Handling of Hatching Eggs, Breeders, Chickens, and Turkeys. (2017).http://www.nfacc.ca/codes-of- practice/chickens-turkeys-and-breeders Nevarez J. G. (2005). Monitoring during avian and exotic pet anesthesia. Seminars in Avian and Exotic Pet Medicine, 14(4): 277–283. doi: 10.1053/j.saep.2005.09.011. Nickel R. et al. In: Anatomy of Domestic Birds. Verlag Paul Parey. Berlin; pp 118-121. (1977) Orosz S. E. (1996). Principles of avian clinical neuroanatomy. Seminars in Avian and Exotic Pet Medicine, 5(3): 127–139. doi: 10.1016/S1055-937X(96)80002-8. Parshall M.B., Schwartzstein R.M., Adams L., Banzett R.B., Manning H.L.,Bourbeau J., Calverley P.M., Gift A.G., Harver A.., Lareau S. et al. An official American thoracic society statement: Update on the mechanisms, assessment, and management of dyspnea. American Journal of Respiratory and Critical Care Medicine 185, 435–52, 2012 Patterson J.L., Mullinax P.F., Bain T., Kreuger J.J. Richardson D.W. (1962) Carbon dioxide- induced dyspnea in a patient with respiratory muscle paralysis. American Journal of Medicine, 32:811-816.

200

Paul-Murphy J., Fialkowski J. (2001). Injectable anesthesia and analgesia of birds. In: Gleed RD, Ludders JW, editors. Recent Advances in Veterinary Anesthesia and Analgesia: Companion Animals. Ithaca, NY: International Veterinary Information Service (2001). Pavlov I.P. (1927). Conditioned Reflexes: An Investigation of the Physiological Activity of the Cerebral Cortex. London: Oxford Univ. Press. Pearson R., (1972), The Avian Brain, Academic Press, London. Prinz S., Van Oijen G., Ehinger F., Bessei W., Coenen A. (2010). Effects of water bath stunning on the electroencephalograms and physical reflexes of broilers using a pulsed direct current. Poultry Science, 89(6): 1275–1284. doi: 10.3382/ps.2009-00136. Prys-Roberts C. (1987). Anaesthesia- A practical or impractical construct. Br J Anaesth. 59:1341– 45. Raj A. B. M. and N.G. Gregory. (1991). Preferential feeding behaviour of hens in different gaseous atmospheres. British Poultry Science, 32(1): 57-65. doi.org/10.1080/00071669108417327 Raj A. B. M., Wotton S. B. and Gregory N. G. (1992). Changes in the somatosensory evoked potentials and spontaneous electroencephalogram of hens during stunning with a carbon dioxide and argon mixture. British Veterinary Journal, 148(2):147–156. doi: 10.1016/0007-1935(92)90106-B. Raj A.B.M., and Gregory N.G. (1993). Time to loss of somatosensory evoked potentials and onset of changes in the spontaneous electroencephalogram of turkeys during gas stunning. Veterinary Record. 133: 318–320.

Raj A.B.M., and Gregory N.G. (1994). An evaluation of humane gas stunning methods for turkeys. Veterinary Records 135:222-223. Raj A. B. M. and Gregory N. G. (1995). Welfare implications of the gas stunning of pigs 1. Determination of aversion to the initial inhalation of carbon dioxide or argon. Anim. Welfare 4:273280. Raj B. M. (1996). Aversive reactions of turkeys to argon, carbon dioxide and a mixture of carbon dioxide and argon. Veterinary Record 138(24): 592-593. doi: 10.1136/vr.138.24.592 Raj B. M. and O’Callaghan M. (2001). Evaluation of a pneumatically operated captive bolt for stunning/killing broiler chickens. British poultry science, 42(3): 295–9. doi: 10.1080/00071660120055232. Ramsaya D. S., Watson C. H., Leroux B. G., Prall C. W., Karl J., Kaiyala K. J. (2003). Conditioned place aversion and self-administration of nitrous oxide in rats. Pharmacology Biochemistry and Behavior, Volume 74, (3): 623- 633.https://doi.org/10.1016/S0091-3057(02)01048-1

201

Rattenborg N. C., Martinez-Gonzalez D. and Lesku J. A. (2009). Avian sleep homeostasis: Convergent evolution of complex brains, cognition and sleep functions in mammals and birds, Neuroscience and Biobehavioral Reviews, 33(3), pp. 253–270. doi: 10.1016/j.neubiorev.2008.08.010. Revzin A. M. (1965). Characteristics of the spontaneous electrical activity in the neostriatum of the pigeon. Fed. Proc. 24: 338. Ries C. R. and Puil E. (1999). Mechanism of anesthesia revealed by shunting actions of isoflurane on thalamocortical neurons. Journal of neurophysiology, 81(4):pp. 1795–1801. Rosenthal M. (1993). Mild traumatic brain injury syndrome. Annals of Emergency Medicine, 22, (6):1048-1051. DOI: https://doi.org/10.1016/S0196-0644 (05)82749-0 Sandercock D. A., Auckburally A., Flaherty D., Sandilands V. and McKeegan D. E. F. (2014) Avian reflex and electroencephalogram responses in different states of consciousness. Physiology and Behavior. doi: 10.1016/j.physbeh.2014.05.030. Sandilands V., Abm R., Baker L., Nhc S. (2011). Aversion of chickens to various lethal gas mixtures. Animal Welfare, 20(2): 253–262. Schepens B. (2004). Independent and Convergent Signals From the Pontomedullary Reticular Formation Contribute to the Control of Posture and Movement During Reaching in the Cat. Journal of Neurophysiology, 92(4): 2217–2238. doi: 10.1152/jn.01189.2003. Serpa L., Widowski. T. M., Schwean-Lardner K., Torrey S. Under pressure: determining optimum air pressure for humane euthanasia of mature broiler breeders with an air powered non-penetrating captive bolt device. Proceedings of the X th European symposium on poultry welfare. Ploufragan, France. (2017).77 Seth A. K., B. J. Baars, and D. B. Edelman. 2005. Criteria for consciousness in humans and other animals. Conscious. Cogn. 14:119–139 Sharp T. M., McLeod S. R. , Leggett K. E. A., Gibson T. J. (2014). Evaluation of a spring- powered captive bolt gun for killing kangaroo pouch young. Wildlife Research 41(7) 623-632 https://doi.org/10.1071/WR14094 Shaw N.A. (2002). The neurophysiology of concussion. Prog. Neurobiology 67: 281-344 Siegal A., and Sapru H. N. (2006). Essential neuroscience. Philadelphia, PA: Lippincott, Williams & Wilkins Silver W.L., Maruniak, J.A., (1981). Trigeminal chemoreception in the nasal and oral cavities. Chem. Senses 6, 295–305. Sinclair M.D. (2003). A review of the physiological effects of α2-agonists related to the clinical use of medetomidine in small animal practice. Canadian Veterinary Journal 44(11): 885– 897.

202

Sommerhoff G. and MacDorman K. (1994). An account of consciousness in physical and functional terms: A target for research in the neurosciences. Integrative Physiological and Behavioral Science, 29(2): 151–181. doi: 10.1007/BF02691012. Sparrey J, Sandercock D.A., Sparks N.H.C., Sandilands V. (2014). Current and novel methods for killing poultry individually on-farm. Worlds Poult Sci J. 70:737–58. doi: 10.1017/S0043933914000816 Starzl T. E., C. W. Taylor, and H. W. Magoun. 1951. Ascending conduction in reticular activating system, with special reference to the diencephalon. J. Neurophysiol. 14:461– 477. Stinson J. M., and Mattsson J. L. (1970). Tolerance of rhesus monkeys to graded increase in environmental CO2- Serial changes in heart rate and cardia rhythm. Aerosp. Med. 42:78 – 80. St-John W. M., Paton J. F. R. and Leiter J. C. (2004). Uncoupling of rhythmic hypoglossal from phrenic activity in the rat. Experimental Physiology, 89(6): pp. 727–37. doi: 10.1113/expphysiol.2004.028829. St-John W. M. (2009). Noeud vital for breathing in the brainstem: gasping--yes, eupnoea-- doubtful. Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 364(1529): 2625–33. doi: 10.1098/rstb.2009.0080. Sun M., Sun C., Yang Y. (1996). Effect of low-concentration CO2 on stereoacuity and energy expenditure Aviation, Space, and Environmental Medicine, 67(1):34-39 Taneichi H., Suda K., Kajino T., Kaneda K. (2005). Traumatically Induced Vertebral Artery Occlusion Associated with Cervical Spine Injuries: Prospective Study Using Magnetic resonance angiography. Spine (Phila Pa 1976), 30(17):1955-62. Terlouw C., Bourguet C. Deiss V. (2016a). Consciousness, unconsciousness and death in the context of slaughter. Part I. Neurobiological mechanisms underlying stunning and killing. Meat Science. Elsevier Ltd, 118: 133–146. doi: http://dx.doi.org/10.1016/j.meatsci.2016.03.010 Terlouw C., Bourguet C. Deiss V. (2016b). Consciousness, unconsciousness and death in the context of slaughter. Part II. Evaluation methods. Meat Science. Elsevier Ltd, 118: 147– 156. doi: 10.1016/j.meatsci.2016.03.010. Thomas P. D. (2000). The differential diagnosis of fixed dilated pupils: a case report and review [online]. Critical Care and Resuscitation: journal of the Australasian Academy of Critical Care Medicine, 2(1): 34–37. ISSN: 1441-2772.

203

Tidswell S., Blackmore D., Newhook J. (1987). Slaughter methods: electro- encephalographs (EEG) studies on spinal cord section, decapitation and gross trauma of the brain in lambs. New Zealand Veterinary Journal, 35: 46–49. Tseng W.C., Shih H.M., Su Y.C., Chen H.W., Hsiao K.Y., Chen I.C. (2011). The association between skull bone fractures and outcomes in patients with severe traumatic brain injury. J. Trauma 71:1611–1614. doi: 10.1097/TA.0b013e31823a8a60 Turner B. H. and Knapp M. E. (1995). Consciousness: A neurobiological approach’, Integrative Physiological and Behavioral Science, 30(2):151–156. doi: 10.1007/BF02691683. Turner P.V., Kloeze H., Dam A., Ward D., Leung N., Brown E.E.L., Whiteman A., Ciappetta M.E., Hunter D.B. (2012). Mass depopulation of laying hens in whole barns with liquid carbon dioxide: Evaluation of welfare impact. Poult. Sci. (2012) 91:1558–1568. doi: 10.3382/ps.2012-02139 Vander A. J., Sherman J. H., Luciano D. S. (2001). Human physiology the mechanisms of body function, 8th ed. Boston: McGraw-Hill. Veras L., Pedraza-Gutiérrez, S., Castellanos, J., Capellades, J., Casamitjana J., Rovira- Cañellas A. (2000). Vertebral Artery Occlusiion After Acute Cervical Spine Trauma. Spine 25: 1171-1177. Verhoeven M. T. W., Gerritzen M. A, Hellebrekers L. J. and Kemp, B. (2015) ‘Indicators used in livestock to assess unconsciousness after stunning: a review.’, Animal: an international journal of animal bioscience, 9(2), pp. 320–30. doi: 10.1017/S1751731114002596. Von-Leupoldt A., Sommer T., Kegat S., Baumann H.J., Klose H., Dahme B., Buchel C. (2008). The unpleasantness of perceived dyspnea is processed in the anterior insula and amygdala. American Journal of Respiratory and Critical Care Medicine 177,1026–32, 2008

Walsh J., Percial A., Turner P.V. (2017). Efficacy of blunt force trauma, mechanical cervical dislocation, and a non-penetrating captive bolt device for on-farm euthanasia of pre- weaned kits, growers, and adult commercial meat rabbits. Animals 7:100 Waran N., Williams V. M., Clarke N., Bridge I. S. (2010). Recognition of pain and use of analgesia in horses by veterinarians in New Zealand. New Zealand veterinary journal, 58: 274–280. doi: 10.1080/00480169.2010.69402. Webster, A. B., and D.L. Fletcher. (2001). Reactions of laying hens and broilers to different gases used for stunning poultry. Poultry Science 80(9): 1371-1377. doi: 10.1093/ps/80.9.1371 Webster A. B. and Fletcher D. L. (2004). Assessment of the aversion of hens to different gas atmospheres using an approach-avoidance test. Applied Animal Behaviour Science, 88(3–4): 275–287. doi: 10.1016/j.applanim.2004.04.002.

204

White B.C. and Krause G.S. (1993). Brain injury and repair mechanisms: the potential for pharmacologic therapy in closed-head trauma. Annals of Emergency Medicine 22: 970- 979. doi.org/10.1016/S0196-0644 (05)82734-4 Whiting T. L., Steele G. G., Wamnes S., Green C. (2011). Evaluation of methods of rapid mass killing of segregated early weaned piglets. Can Vet J. 52:753–758 Whittow G.C. (ed): Sturkie’s Avian Physiology (ed 5). (2000). San Diego, CA, Academic Press Widowski T., Woolcott C., Turner P., Schwean-Lardner K., Caston L., Torrey S. (2018). Evaluation of a mechanical cervical dislocation tool for growing and mature turkeys (Abstract). Proceedings of Poultry Science Association 107th Annual Meeting, pg 97 Widowski T. M., Elgie R. H., Lawlis P. (2008). Assessing the effectiveness of a non-penetrating captive bolt for euthanasia of newborn piglets. Differences, 107-111. Wilberger J. E., Harris M. and Diamond D. L. (1991). Acute subdural hematoma: morbidity, mortality, and operative timing. Journal of Neurosurgery, 74(2): 212–218. doi: 10.3171/jns.1991.74.2.0212. Winslow C., Rozovsky J. (2003). Effect of spinal cord injury on the respiratory system. Am J Phys Med Rehabil, 82:803–814. DOI: 10.1097/01.PHM.0000078184.08835.01 Withrock I.C. (2015). The Use of Carbon Dioxide (CO2) as an Alternative Euthanasia Method for Goat Kids. Graduate Theses and Dissertations, Iowa State University Digital Repository, 14718. Accessed online: https://lib.dr.iastate.edu/etd/14718 Woolcott C.R., Torrey S., Turner P.V., Serpa L., Schwean-Lardner K., Widowski T.M. Evaluation of two models of non-penetrating captive bolt devices for on-farm euthanasia of turkeys. Animals (2018a) 8:42-58. doi:10.3390/ani8030042 Woolcott C.R., Torrey S., Turner P.V., Chalmers H., Levison L., Schwean-Lardner K., Widowski T.M. (2018b). Assessing a method of mechanical cervical dislocation as a humane option for on-farm killing using anesthetized poults and young turkeys. Front. Vet. Sci. 5:275 doi.org/10.3389/fvets.2018.00275 Yang Y., Sun C., Sun M. (1987). The effect of moderately increased CO2 concentration on perception of coherent motion. Aviation, space and environment medicine, 68:187-191 Yew D. T. W., Yu M. C., Kwong W. H. (1996). Basic Neuroanatomy. World Scientific Publishing Co. Pte. Ltd. Singapore. Zar J.H. (1998). Biostatistical Analysis. 4th ed. Prentice Hall, Upper Saddle River, NJ Zeman A. (2005). What in the world is consciousness?. Progress in Brain Research, 150:1–10. doi: 10.1016/S0079-6123(05)50001-3. Zerari-Mailly F., Dauvergne C., Buisseret P., Buisseret-Delmas C. (2003). Localization of Trigeminal, Spinal, and Reticular Neurons Involved in the Rat Blink Reflex. Journal of Comparative Neurology, 467(2): 173–184. doi: 10.1002/cne.10917.

205

Zornow M.H., Fleischer J.E., Scheller M.S. (1990). Dexmedetomidine, and α2-adrenergic agonist, decreases cerebral blood flow in the isoflurane-anesthetized dog. Anesthesia and Analgesia 70:624–630 Zulkifli I., Goh Y., Norbaiyah B., Sazili A., Lotfi M. and Small A. H. (2014). Changes in blood parameters and electroencephalogram as affected by different stunning methods in cattle.

206