Sophie Atkinson, Anne Larsen, Pol Llonch, Antonio Velarde and Bo Algers

Transcription

1 Department of Animal Environment and Health Group stunning of pigs during commercial slaughter in a Butina pasternoster system using 80% nitrogen and 20% carbon dioxide compared to 90% carbon dioxide Sophie Atkinson, Anne Larsen, Pol Llonch, Antonio Velarde and Bo Algers

2 CONTENTS Page 1. INTRODUCTION BACKGROUND METHOD AND MATERIALS Technical aspects of gas filling 90%CO Technical aspects of gas filling 80%N 2 20% CO Exposure time 80%N 2 20% CO Filming 90%CO 2 and 80%N 2 20% CO Stun quality assessments 90%CO 2 and 80%N 2 20% CO Analysis of filmed behaviour during stunning RESULTS Gas analysis 90%CO Analysis of filmed behaviour during 90%CO 2 stunning Stun quality 90%CO Gas analysis 80%N 2 20% CO Analysis of filmed behaviour during 80%N 2 20% CO 2 stunning Stun quality 80%N 2 20% CO Comparisons between 90%CO 2 and 80%N 2 20% CO DISCUSSION CONCLUSIONS REFERENCES 41. ACKOWLEDGMENTS 41 APPENDIX 1: Temperatures recordings averaged for every 5 minutes 42. APPENDIX 2: O 2 concentrations averaged for every 5 minutes 43. APPENDIX 3: CO 2 concentrations averaged for every 5 minutes 44. APPENDIX 4: N 2 concentrations averaged for every 5 minutes 45. APPENDIX 5: O 2 concentrations averaged for every 5 minutes at 2.2m during 80%N 2 20% CO 2 & frequency of inadequate stunning 46. 2

3 1. INTRODUCTION All larger abattoirs in Sweden use Butina (Denmark), group-wise carbon dioxide (CO 2 ) stunning for pigs. Approximately 2.5 million pigs are slaughtered annually (Official Statistics of Sweden 2013) in these systems. CO 2 stunning is now favoured over other methods such as electricity and captive bolt due to benefits for animal welfare (Terlouw et al 2008) and meat quality (Velarde et al 2000). Pigs can be handled and stunned in small groups rather than individually, which reduces separation anxiety, pig refusal and consequential electric prod use from the handler, which is often a problem during other stun methods. However, the high concentration of carbon dioxide (>90%) used for stunning pigs, causes a degree of animal suffering prior to the loss of consciousness. There is also little documentation on how long it takes for pigs to lose consciousness during commercial slaughter, but laboratory tests on Butina Diplift or single box (cage) systems, confirm it is not immediate. CO 2 gas at high concentration is acidic when inhaled causing severe irritation of the eyes, nasal mucosa, lungs, and an overall painful experience due to the presence of chemoreceptors in the throat (Raj and Gregory 1995). The lack of oxygen (O 2 ) also causes a severe sense of breathlessness which may cause severe distress. For these reasons EFSA (2004) concluded that CO 2 stunning is not optimal from an animal welfare perspective. It is also in conflict with EU and Swedish animal slaughter and welfare regulations which state abattoirs must take measures to avoid pain and minimise distress during the killing and slaughter process; and animals should not show signs of avoidable pain or fear (Chapter II, Article 3 of the EC 2009) and the stun should be immediate (section 30 of the Swedish Animal welfare Act L22). The presence of nitrogen (N 2 ) in the atmosphere is around 79% and could be a more suitable gas for stunning pigs. The composition of air contains approximately 78.09% N 2, 20.95% O 2, 0.93% Argon and 0.39% CO 2 (Wikipedia 2015). Llonch et al (2013) found that when CO 2 is lower than 20% and contained in an anoxic atmosphere (<2%O 2 ) with a high concentration of N 2 (80%), the aversion in pigs as measured by behaviour tests is significantly reduced. However in N 2 /CO 2 gas mixes, a longer exposure time (at least 5 minutes), was required to induce brain inactivity and maintain it until sticking, compared to 90%CO 2 group stunning (Llonch et al 2012). Also the relative density of N 2 is slightly lower than air and its stability, defined as the capability of the gas to be sustained within the pit without being displaced by oxygen, is much lower than high levels of CO 2. At the Catalan Agricultural Research Institute (IRTA), Dalmau et al (2010) assessed the stability and uniformity of gas mixtures with nitrogen (N 2 ) and CO 2 in a individual dip-lift stunning system under laboratory conditions. The gas mixtures of 85% N 2 and 15% CO 2; 80% N 2 and 20% CO 2; and 70% N 2 and 30% CO 2 by volume in atmospheric air, could be sustained within the pit. From the gas mixtures assessed, 80% N 2 and 20% CO 2 caused a more rapid induction and the minimum aversion to pigs as well as providing reasonable gas stability. According to Llonch et al (2013), it is critical that residual O 2 at the first stop within the Butina remains below 2%. Based on these studies, this project aimed to assess level and duration of discomfort and quality of stunning in pigs exposed to 80% N 2 and 20% CO 2 (80N20C) compared to pigs exposed to 90% CO 2 (90C) in a Butina -6 box paternoster system during commercial slaughter. 3

4 2. BACKGROUND Figure 1 details a diagram of a Butina paternoster system used at the slaughter plant where the experiment took place. Six cages (or boxes) which can take normally 4 (but up to 5) pigs in each, moved below ground level through a motorised gondola system within a 6.8m deep, 3.2m wide and 2.8m long rectangular cement pit (volume 63m 3 ). Once pigs were closed into the system the cage moved 2.2m down where it stopped for a pause for some seconds before descending 5.6m to the bottom of the pit exposing pigs to a higher CO 2 gas gradient for a longer period until the cage ascended back to 2.2m. The cage paused again while pigs were loaded on the other side of machine and then ascended to the pit top where stunned pigs were tipped out on the other side of the machine onto a conveyer belt where hoisting took place. Each pig was then hoisted up onto an assembly line for transfer to sticking. Figure 2 shows a photo of a cage in the exit position where stunned pigs are delivered on to a conveyer belt. Figure 3 shows the pit where the cages rotate during operation. Previous to the current experiment, two pilot studies were conducted to see if it was possible to fill the paternoster system with an external gas source consisting of 80% N 2 and 20% CO 2. In contrast to the IRTA studies where the stun pit was 8m 3, the size of the stun pit was nearly 8 times larger (63m 3 ) and proved challenging to fill with the desired gas concentrations within a reasonable time frame. Initially, the trials tested taking gas into the pit via separate bottles of nitrogen and carbon dioxide pumped through a pipe system and regulator, which was then mixed through a Dansensor MapCheck machine, before being pumped further into the stun pit. Both trials revealed that a larger capacity regulator and pumping system was required to successfully fill the pit volume. Better systems to measure the gas were also needed. With the help of technicians from Butina, it was realised that the gas could be pumped directly through the Butina machine increasing pump capacity from 450 litres gas per minute to 1500 litres. It was also decided to change the participating gas company as the initial company chosen, lacked technological knowledge to full fill our requirements. The second gas company was better equipped with wider more robust hoses and larger regulators for delivering the gas at high pressure into the Butina. The gas company also recommended using 80%N 2 and 20% CO 2 already premixed and contained in bottles. This was rather than taking N 2 and CO 2 from separate bottles and mixing them on site through a mixer machine, which was unsuccessfully trialled in one of the pilot tests. A large heater was also recommended to help warm the gas to improve its distribution through the pit. During one of the pilot tests, eventual problems occurred with setting the Butina to run at longer exposure times. During this test it was established that pigs should be in the Butina for at least 6 minutes (from time when the box starts to move, until time when door opens and pigs exit the Butina) to ensure they remained deeply stunned until sticking was complete. The slaughter plant however, wanted to reduce this time to keep the slaughter rate feasible. Several groups of pigs were tested and the slaughter plant decided that they could work with a target exposure time of 5 minutes and 30 seconds. 4

5 Stunned pigs exit Live pigs enter Top of stun pit Gas measure point 1 at 2.2 m or stop 1 Volume = 63m3 (6.8 x 3.2 x 2.8m) Gas measure point 2 at 5.6m or stop 2 Bottom of stun pit Gas injection valve Figure 1: Butina paternoster construction used in experiment 5

6 Figure 2: Cage for holding pigs during stunning in the delivery position of the Butina Figure 3: Photo looking down into the pit where the cages descend and ascend during operation 6

7 3. METHOD AND MATERIALS Studies were conducted during routine slaughter on 1000 pigs stunned with conventionally used 90% CO 2 (90C) on one day, and on 825 pigs with 80% N 2 and 20% CO 2 (80N20C) on another day. During the trial there were two expert technicians attending from the gas company YARA Prax Air, one from ProReg control who supplied the gas monitoring devices, and a technician from Butina (Denmark). Three researchers from IRTA and two from SLU, assisted with stun quality and meat quality data recordings. 3.1 Technical aspects of gas filling 90%CO 2 Permanently installed sensors positioned in the pit at a height corresponding to the top of the stun box when in the lowest position (i.e. approximately 1m from pit bottom) and at the first stop, gave digitally displayed readings for CO 2 concentrations on the machine. To maintain a constant CO 2 gradient during processing (at least 80% in the first stop and 90% at the bottom), built in sensors caused periodic automatic filling of CO 2 gas located half a meter from the base of the Butina. The exposure times varied and had to be individually calculated per filmed batch from recording the time each box was seen moving through the gas gradient. 3.2 Technical aspects of gas filling 80%N 2 20% CO 2 The requirements for the gas company were to achieve <2 % oxygen (O 2 ) at the first stop (2.2 m down) and <1 % O 2 at bottom stop (5.6 m down) in the Butina pit. The gas had to be filled within one hour, and was to be maintained throughout full processing. Gas was supplied in clusters of 12 bottles containing a mixture of 20 % CO 2 and 80% N 2 (Mapcon ND20), supplied and delivered by Yara Praxair AS. The gas bottles contained m 3 gas registered for food quality usage. Pressure in the cylinders was 200 bars. To transfer the gas into the Butina from the bottles during the experiment, flexible metal hoses with an inner diameter of 25 mm were used with regulators (U13 F20 200/20 bar) operated at 3.5 bars. Three batches of bottles with the Mapcon ND20 gas were connected in parallel using U13 regulators and 25 mm hoses to a T-piece (Figure 4). From the T-piece, a 12 mm hose placed through a ventilation hole in the wall (Figure 5), carried the gas from the bottles into the control cabinet of the Butina Backloader machine to increase the flow rate. In this cabinet the hose was further attached to an extra heater supplied by Butina to help warm the gas (Figure 6). From the heater, three hoses connected into the base of the pit delivering gas into the system via two -12 mm hoses and one -14 mm hose. Two Dansensor machines (MAP Check 3), were used for measuring CO 2, N 2 and O 2 levels inside the pit. These were rented and supplied by ProReg control, Stockholm, Sweden. The Map Check devices (Figure 6) were connected with tubes to sample and measure at 2.2m and 5.6m from the top of the pit which measured 6 gas samples every minute. Temperature loggers were placed at 2.2 and 5.6m in the pit sampling every minute. 7

8 N/CO 2 premixed gas bottles and bundles connected in parallel with U13 regulator at 0-20 bar T piece regulator Figure 4: Photo showing NCO 2 gas bundles located outside connected in parallel with a large regulator for delivery of gas into Butina machine Ventilation hole where pipe from gas cylinders delivered gas through wall from outside into Butina machine room Main pipe delivering gas from bottles outside into Butina machine located inside slaughter plant Figure 5: Photos showing ventilation inlet where pipes were passed through to deliver gas from outside to inside 8

9 Gas heating machine Two tubes connected to the Map check were inserted through a small bolt hole and lowered 2.2m and 5.6 m down in the pit Figure 6: Photos showing Butina machine cabinet with attached heating device where gas flowed through, and the Map Check 3 loggers used to sample and measure gas concentrations inside the Butina A technician from ProReg control AB operated the MapCheck devices during the 80N20C experiment. The Map check logged data 6 times per minute, recording to an Excel file. After the experiment the data was downloaded and analysed using Excel functions and graphs. Due to the large amount of data, the gas concentration recordings for N 2, CO 2 and O 2 were plotted on a graph showing the average levels for every 5 minutes. Battery operated temperature loggers of the type Ibutton data logger -ThermoChron, IDC, Barcelona, Spain, were placed 2.2m and 5.6m down in the pit. These loggers recorded temperature every 5 minutes and stored the information on a memory device within the logger which could be down loaded into a computer. 3.3 Exposure time 80%N 2 20% CO 2 The intended exposure time in the trial was 6 minutes (min) defined as time in which the boxes were moving (from box start to box stop just prior to tip out). This time frame was selected because during the limited pilot study, 2 of 5 groups of pigs were successfully stunned until sticking when this exposure time was used. 3.4 Filming 90%CO 2 and 80%N 2 20% CO 2 The environment inside the paternoster Butina system contained a deep cement walled pit which housed 6 metal cages connected in a gondola system which moved down, around and up through the gas gradient 9

10 inside the pit. It was therefore not possible to have cables or a wireless internet connection for recording film data. As the pit was also dark and wet, with moisture build up from condensation and pig urine, a particularly robust camera system was required for filming. The camera system also had to be powered by batteries powerful enough to run the cameras for up to 12 hours with capacity to film in the dark at high definition. Therefore, 2 vandal and water proof surveillance cameras with infra-red capacity were used (Zavio, D5210 with 2 Megapixel 1080p Full HD resolution, auto-iris lens, and IR LEDs). Two cages were fitted with each of the cameras, which were powered via four -12 volt water proof motorcycle batteries attached in parallel, rewired from the original 230 Volt AC system. The batteries were housed in custom made metal boxes, bolted onto the top of the roof of 2 of the cages. The camera domes were bolted in the upper corner inside the cage (Figure 7) aimed to film towards the opposite corner. The cameras recorded to a memory card and after each experiment (90C and 80N20C stunning) the film data was downloaded onto the computer for analysis. The cameras could not record sound. Therefore a separate digital recorder enclosed in a plastic bag for water protection was fixed inside the cage next to the cameras. The sound and the film cameras were set to record at the same time. The sound files were then added to the film files using the film editing program Adobe Premier. Figure 7: Camera set up with battery pack bolted on top of cage connected to camera which was attached under roof of cage 3.5 Stun quality assessments 90%CO 2 and 80%N 2 20% CO 2 Pigs were assessed for stun quality from delivery out of the Butina until sticking. During 90C stunning one person made the stun quality assessments and clocked the stun to stick intervals. Group sizes in the stun-boxes were recorded by counting the number of pigs in each group as they came out of the stun-box. The stun-to-stick interval was timed for every pig in the group using a stopwatch. The time when the end of the stun occurred for all pigs in a group began when the Butina box stopped just before the gate opened to release the pigs. All pigs were chest stuck (severing all major blood vessels in the thorax) with one of six blood collecting knives (Rotary Stick, Butina Anitec ). Sticking was considered to be the point at which the knife was pushed into the chest and signalled the end of the stun-to-stick time. These times were recorded sequentially for each pig in the group. During the 80N20C trial, more thorough checks were made due to the higher risk for pigs to recover. This included continuous checks for corneal 10

11 reflexes, and close examinations for signs of breathing and blinking up until sticking by 2 people. Another person recorded stun to stick intervals. Pigs were continually observed for physical symptoms that could indicate consciousness or a risk that recovery was imminent. The stun level was given a score from 0 to 4 based on the behavioural reactions observed after pigs exited the stunner. If pigs were in a state of whole body relaxation, and there was no evidence of rhythmic breathing, righting reflex, vocalisations, convulsions, blinking, pain or eye responses to stimulation, pigs were considered deeply and adequately stunned and scored 0. Pigs that showed symptoms outside this criterion were closely examined and the eyes tested by carefully touching the corneal and eyelid area with a fingertip continually until sticking was complete. Pigs showing regular gasping were always tested for reflexes and closely monitored until sticking. If there were any doubts as to the state of unconsciousness, pigs were immediately re-stunned with a backup captive bolt weapon to the head. Table 1 shows a stun quality assessment protocol used in this study for defining signs and scoring them based on an estimated risk level for inferior animal welfare rated from 0 (deep stunning no welfare risk) to 4 (poor stunning and high welfare risk). Pigs with a stun level of 2 or 3 were not considered to be conscious, but the stun depth indicated a recovery risk that justified re-stunning. Level 4 indicated consciousness was probable. Level 1 indicated closer monitoring was necessary to ascertain if inadequate stunning might occur. 11

12 Table 1: Stun quality assessment protocol defining signs after stunning and scoring them based on an estimated risk level for inferior animal welfare rated from 0 (deep stunning no welfare risk) to 4 (poor stunning and high welfare risk) Stun quality level Description of signs 0 If a pig showed no reflexes or signs mentioned below, it was considered as being in a state of deep unconsciousness and posed no risk for poor animal welfare 1 If a pig showed kicks or other movements, or infrequent gasps but had no eye or pain reflexes when checked, it was considered adequately stunned but justified continual monitoring 2 If a pig displayed frequent gasps (opening and closing of the mouth with or without stretching of neck), kicks or body convulsions, but was found to have no eye or pain reflexes, it was re-stunned as a precaution to avoid recovery. The stun depth was considered as unacceptable due to the risk that the animal could recover 3 If a pig showed corneal or cilia blink reflex at sticking, with or without kicking or convulsions, it was immediately re-stunned and the recovery risk thought to be imminent and the stun was considered inadequate 4 If a pig showed spontaneous blinking, righting reflex, vocalisation or pain reflex, it was considered as indicating some form of consciousness and a high risk for poor welfare and the stun considered inadequate 3.6 Analysis of filmed behaviour during stunning A list of behaviours with the help of previously described definitions described in Llonch et al (2013) was developed and prepared in an ethogram (Table 2). Using Excel these behaviours were timed when each first appeared and disappeared in the pig group during the film sequence. The period of discomfort was estimated as time when the first reaction (retreat, sway, neck stretch, gasp or struggling) occurred, until when the last pig fell and the most violent of the muscle excitations were complete. These times were compared between the 2 gases. As pigs were often held inside the stunner without the cage moving, exposure times were calculated by watching the time code displayed in the films and timing the duration in which the cage could be seen moving through the gas gradient. 12

13 Table 2: List of defined behaviours which were timed and registered in Excel after viewing recorded films of pigs during stunning in the Butina with 90C or 80N20C Registered and timed parameter when first observed in the group Retreat/escape Neck stretch Gasping Sway Struggle Sits First fall First reaction Gagging Last fall First violent muscle excitations begin Last violent muscle excitations Probable discomfort period Definition First pig in group anxiously backs up, turns around or moves forward in cage First pig in group lifts its head outwards, stretching the neck, either when standing, sitting or laying First pig in group opens and closes mouth while standing upright First pig in group to loose balance First pig in group fights against falling and jumps or jerks body around the box while remaining upright First pig in group sits with rump on floor First pig falls and is laying on floor with whole body Any of the above behaviours that could first be identified as the pig reacting to the gas First pig in group opens and closes mouth periodically while laying on ground Last pig falls and is laying on floor with whole body When first pig starts struggling and showing violent muscular contractions and convulsions of the whole body or part of it including violent head movements When last pig completes struggling and violent muscular contractions and spasms and convulsions of the whole body or part of it including violent head movements Period of time in which first reaction is noted (any of the above behaviours except gagging) until when last pig falls and severe muscle excitations have ceased 13

14 4. RESULTS In total 1000 pigs stunned in 90C were studied for stun quality and of those 393 pigs in 115 groups filmed for behavioural analysis. In total 825 pigs stunned in 80N20C were studied for stun quality and of those 153 pigs in 46 groups filmed for behavioural analysis. Thirty groups of pigs that were filmed during 80N20C stunning could not be analysed due to dirt covering the camera. The average live-weight of pigs studied in both gas systems was 92.5kg. 4.1 Gas analysis 90%CO 2 To maintain a constant CO 2 gradient during processing (minimum 80% in the first stop and 90% at the bottom), a built in sensor caused periodic automatic filling of CO 2 gas. The gas inlet was located half a meter from the base of the Butina. During the trial, the gas concentrations remained between 91 and 93% at the pit base, and 81% at the first stop; verified by computer logged data readings from the Butina on the day of the experiment. Recordings collected from 57 observed films from the surveillance cameras installed in the Butina cages, showed that the total duration of the box movement around the Butina ranged from 3mins 13 s to 4 min 59 s (average 3min 57 s). Recordings were collected from 5 specific films to track how long the pigs stayed at each stop (Table 3). The time the box stayed at the bottom stop varied from 1 min 11 s to 1 min 38 s. The temperature in the stun pit was always 17 C during this trial as seen on a display on the Butina machine cabinet. Table 3: Five groups of pigs timed during the cycle within the Butina during 90C stunning (minutes: seconds) Time where cage is in Butina Sample 1(m:s) Sample 2(m:s) Sample 3(m:s) Sample 4(m:s) Sample 5(m:s) Mean (m:s) Time for cage to move to stop 1 00:24 00:24 00:24 00:24 01:17 00:35 Time cage stays at stop 1 00:28 00:28 00:25 00:23 00:30 00:27 Time for cage to move to stop 2 00:25 00:24 00:24 00:25 00:25 00:25 Time cage stays at stop 2 01:11 01:37 01:12 01:19 01:38 01:23 Time for cage to move back up to stop 1 00:24 00:25 00:23 00:23 00:26 00:24 Time cage stays at stop 1 00:19 00:25 01:17 00:25 00:24 00:34 Time for cage to stop at top 00:13 00:12 00:12 00:12 00:14 00:13 Total time cage was moving 03:24 03:55 04:17 03:31 04:54 04:00 14

15 4.2 Analysis of filmed behaviour during 90%CO 2 stunning A total of 393 pigs and 115 separate batches were film and sound recorded. All films successfully showed behaviour of the pigs during stunning, though some films were much clearer based on the direction and position pigs took inside the cage. All pigs showed resistance to being pushed into the stun box. It took approximately 10 s for each box to be loaded with the mechanical stun gate from start of gate push until door was closed. Two pigs fell as a result of being forced by the mechanical gate, and 1 pig was hit by the gate as it closed. From analysis of digitally recorded sound the noise level inside the Butina was loud (well over 100 decibels). Mostly the sound came from when the CO 2 gas was sprayed intermittently from the pit bottom. This made a high pitched noise which muffled all other sounds (including pig vocalisations). The loud hissing was picked up by the sound recorders even when the box was at the highest point (the gas was sprayed in from the bottom of the pit 6m down). The machine also made a loud sound when the boxes were moving. Pigs could be seen in the films with ears pointed forward and alert upon entry into the cage, often refusing to enter the box, indicating they could be affected by the noise level. The sound files often picked up the sound of pigs urinating once the door closed inside the Butina which was a probable stress response to the challenging environment inside the Butina machine. During the gassing, pigs could be heard making snoring or a low pitched vocalisation sound once in the prone position. Table 4 shows when certain behaviours and events occurred in a film of 1 pig during CO 2 stunning. This film can be observed in film file 1 (CO2-1 pig only). In film file 2 (CO2 ear posture), a film is sped up to show ear posture changes after the pig is prone. Film file 3 (CO2 4 groups) shows film footage of the reactions of several pigs during 90C stunning, and film file 4 (CO2 slow motion) shows footage of pigs during stunning with the film sequence slowed to 40% speed. This enables the viewer to see more clearly how the pigs are reacting before falling. In these films pigs can be seen lifting their head and stretching out the neck, lifting the snout and opening up the mouth, jumping upwards, and struggling with aggressive body movements; presumably in attempt to breathe. The eyes are often wide open with the scleral apparent (possible indications of fear). These behaviours are suggestive of the animals survival instincts being triggered to the maximum capacity, which probably induced the highest level of fear and distress possible in the animals attempt to survive. 15

16 Table 4: An example of the time when certain behaviours and events occurred during the stunning of 1 pig in 90C (minutes: seconds) Time (minutes: seconds) Observed behaviour or event Figure (photo) number 00:00 Pig enters stun box 00:18 Box starts to move 00:21-22 Pig turns head one side and then other 8 00:28-33 Gas spray can be heard switching on and pig suddenly moves forward 00:29 Pig licks lips or chews 00:30 Pig lowers head 00:32 Pig lurches forward 00:35 Pig backs into pen corner with mouth open and front legs 9 splayed 00:38-43 Pig stretches neck and tosses head upwards 10 00:45 Pig throws body to right side of pen 00:46 Pig tosses head upwards 00:47 Box arrives to stop 1 00:47 Front legs of pig collapse 11 00:48 Back legs collapse 00:43 to 00:52 Head and snout remain stretched backwards with mouth gaping open 00:51 Ears are pointed backwards 00:52 to 00:59 Pig remains seized in same position with front legs stretched apart 00:57 Head rolls backwards as pig lays on side 00:58 Head flexes back with snout pointing up 00:58-01:05 Head rolls down on cage floor and ears move downwards 14 01:09 Pig in full prone position 01:20 First gag 15 01:23 Second gag 01:21 Third gag 01:24, 01:26, Fourth, fifth, sixth, seventh gag 16 01:27, 1:29; 01:39 01:59 Mouth remains open and body twitches intermittently 02:08 Slight movements in back leg 02:09 Eighth gag 01:45 to 02:01 Ear moves from backward position to forward position Film file 2 02:09 Ninth gag 02:12 02:15 Intermittent back leg and body movements 02:20 Tenth Gag 04:38 Tip out

17 Figure 8: CO 2 stunning head moves side to side Figure 9: CO 2 stunning Pig backs into corner with open mouth 17

18 Figure 10: CO 2 stunning Neck stretching and gasping for Figure 11: CO 2 stunning Pig falls down with front legs first but keeps head and snout upwards 18

19 Figure 12: CO 2 stunning Pig rolls head backwards with snout Figure 13: CO 2 stunning Pig in seized position for several seconds with stiff front legs 19

20 Figure 14: CO 2 stunning Note backward ear posture Figure 15: CO 2 stunning Pig gags (note forward ear posture) 20

21 During group stunning pigs showed similar behaviours but it was more difficult to see individuals because there was so much movement amongst the group (Film file 3 (Co2 4 groups). Once pigs entered the pen they showed behaviours indicating fear and alertness i.e. forward and backward ear posture (figure 17) and wide open eyes (Figure18 and 20). Many pigs jumped up on one another (figure 19), blocking the camera view. Pigs often lifted head and snout upwards which appeared like they were trying to get air (figure 20 to 22). 21

22 Forward alert ear posture Backward (worried) ear posture Figure 17: Alert ear posture once closed in the Butina which changed to backward ear posture Figure 18: Wide open eyes may be a possible signs of fear 22

23 Figure 19: Pigs often jumped on one another during early stages of the cage decent Figure 20: Neck stretching, lifting head upwards and gasping before arriving to the first stop, note scleral area in eyes possibly indicating fear and distress 23

24 Figure 21: Neck stretching, lifting head upwards and wide opening of mouth before arriving to the first stop Figure 22: Neck stretching, lifting head upwards and wide opening of mouth 24

25 Most pigs showed that fearful and stress related behaviours (wide open eyes showing scleral areas, urination, perched ears, moving around and sniffing the cage floor) from the time they entered the Butina. However, behaviours indicating they were reacting specific to the gas defined as first reaction, included suddenly backing up in the pen, lurching forward, climbing on other pigs, stretching the neck and pointing the snout upwards, tossing the head from side to side, violent paddling movements with the legs, jumping around the pen, and opening the mouth wide open. We assessed the total time of the discomfort behaviour to be the time from when the box started to move until after pigs fell and muscle exertions began. This time ranged between 2 min 32s to 3 min 39s and averaged 2 min 07s. According to Rodriguez et al. 2008, muscle excitation occurs before significant changes in brain function appear which could indicate that pigs were conscious longer than during this period. 4.3 Stun quality 90%CO 2 In total 500 pigs were assessed for stun quality and 300 pigs for stun to stick interval. All pigs showed signs of deep stunning (100% pigs had a stun quality score of 0). Group sizes varied but more than half of the batches contained 4 pigs per group with 3 pigs the next most common batch size. For the 4 th (last pig) in the group, the stun to stick intervals varied between 1 min 12 s to 1 min 29 s (average 1 min 23 s) (Table 5). Table 5: Stun to stick intervals for each pig in the group during 90C stunning Time Pig 1 Pig 2 Pig 3 Pig 4 (minutes: seconds) Minimum 0:21 0:51 01:03 01:12 Maximum 0:59 01:11 01:30 01:29 Average 0:48 01:00 01:12 01:23 25

26 4.4 Gas analysis 80%N 2 20% CO 2 It was 9 C outdoors, cloudy and with some rain periods during the day of the trial which lasted from 7am until 3.30pm. It took 50 minutes to fill the pit with the correct 80N20C. The temperature in the stun pit varied between 14.6 and 18.5 C during the trial at the first stop and 13 and 16 C at the bottom (Appendix 1). The indoor temperature of the slaughter plant ranged between 17 and 18 C (averaging 18 C). At the start of slaughter at 7am the temperature inside the stun pit at the first stop was between 14 and 15 C, and at the bottom 13 and 14 C (Butina recommended 17 C). To try to warm up the gas and temperature inside the pit, the gas technicians suggested spraying hot tap water over the T connection/regulator transferring the gas from the bottles outside (Figure 23). By 9.30am the temperature within the Butina increased at the first stop to 17 C and the bottom stop to 14.3 C. Appendix 2 to 4 shows the average gas concentrations collected every 5 minutes at 2.2m and 5.6m down in the pit for O 2, N 2 and CO 2 during the trial. Oxygen levels fluctuated above the critical 2% level on 3 occasions between approximately 6.39 and 8.30am and 1.39 and 2.15pm (Appendix 1). CO 2 levels also fluctuated during these times (Appendix 3) but the N 2 levels remained relatively stable throughout the trial (Appendix 4). Figure 23: Warm tap water being sprayed over the main regulator to help warm up the transferred 80N20C gas coming from the bottles Recordings collected from 51 films showed that pigs were inside the Butina from 5min 10s to 7min 24 s (average time 5min 20 s). It had been requested by the research team that the exposure time be set at least to 6 minutes (360 s) during the trial. However the slaughter plant wanted to minimise the time pigs were in the Butina in order to keep up the line speed. Exposure time in some batches was therefore less than optimal. Table 6 shows the time durations from film analysis of 5 batches of pigs during 80N20C stunning within the Butina. The time at the second stop ranged between 2 min 22 s and 2 min 42 s (a variation difference of 20 s) compared to a variation of 27 s for 90C stunning. A total of 9 bottles of Mapcon ND20 were used during the trial. Table 7 shows the gas consumption during the 80N20C trial. 26

27 Table 6: Five groups of pigs timed during the cycle within the Butina during 80N20C stunning (minutes: seconds) N Gas Sample 1 Sample 2 Sample 3 Sample 4 Sample 5 Average Time for cage to move to stop 1 00:23 00:25 00:23 00:25 00:22 00:24 Time cage stays at stop 1 00:52 00:51 00:54 00:51 00:55 00:53 Time for cage to move to stop 2 00:24 00:24 00:24 00:23 00:23 00:24 Time cage stays at stop 2 02:33 02:22 02:42 02:28 02:28 02:28 Time for cage to move back to stop 1 00:19 00:21 00:26 00:20 00:18 00:21 Time cage stays at stop 1 02:28 00:52 00:36 00:51 00:50 01:03 Time for cage to stop at top 00:13 00:12 00:12 00:13 00:13 00:13 Total time cage was moving 06:33 05:18 05:20 05:20 05:18 05:34 Table 7: Detailed breakdown of gas consumption during the 80N20C stunning Description No. of bottles m 3 gas m 3 CO 2 kg CO 2 m 3 N 2 kg N 2 Total ,3 516,3 1117,2 1306,7 Filling pit ,5 48,9 105,8 123,8 Stunning ,8 467,4 1011, Stunning/pig 1,542 0,308 0,570 1,233 1, Analysis of filmed behaviour during 80%N 2 20% CO 2 stunning A total of 164 pigs and 47 separate batches were film and sound recorded successfully. Twenty-nine films failed because dirt covered one of the cameras half way through the trial, and the pigs could not be seen. The sound inside the Butina was not as often so loud as during 90C stunning. Pigs could be seen in the films often chewing and opening their mouth immediately upon entry into the Butina, suggesting they could already feel the lack of oxygen inside the cage before descending. As in 90C stunning pigs showed individual variations in behavioural reactions to the gas. Film file 5 (N-lower aversion) shows examples of several groups of pigs where before they fell, they appeared to show less escape behaviours than pigs filmed in film file 6 (N-higher aversion). Table 8 describes the time certain behaviours appeared during 80N20C stunning of 1 pig filmed during the end of the trial when gas and temperature conditions were optimal (film file 7: N-1 pig). 27

28 Table 8: An example of the time when certain behaviours and events occurred during the stunning of 1 pig in with corresponding figures showing photos of certain behaviours Time (minutes: seconds) Observed behaviour or event Figure 00:00 Pig enters stun box 24 00:14-35 Sniffing around floor 00:40 Ear posture moves from forward to sideward then and 25 pig backs to corner of pen turning head side to side 00:48 Box starts to move 00:54 Pig starts to sway on back legs 00:58 Pig starts to fall 00:59 Pig tries to keep balance and scrambles with front legs 26 01:00 Front legs splay and pig falls to chest 00:38-43 Fallen pig paddles with all 4 legs while on side and head is tossed around with open mouth 01:11 Leg padding stops and pig is still :23 Pig is still 01:22 Pig gags and flexes body 01:27 Second gag 01:30, 35, 44,53,59 Third to seventh gag 02:10, 20, 31,42,53 Eighth to twelfth gag 03:01,10,18, Ninth to nineteenth gag 26,36,45,54, 04:05, 16,28,39,51 Twentieth to twenty-fourth gag 05:02 Mild body muscle twitches 05:04 Twenty-fifth gag and mild paddling with legs 05:16 and 05:29 Weak gags 05:30 Tip out 27 28

29 Figure 24: Pig just entered cage during 80N20C Figure 25: Pig shows ear posture moving back and backs to back of cage during 80NC stunning before box arrives to stop 1 29

30 Figure 26: Pig sways and tries to keep balance during 80N20C stunning Figure 27: Pig commences violent muscle excitations, leg paddling and gagging during 80N20C stunning 30

31 During 80N20C group stunning pigs showed similar behaviours to 90C. The pig movements were so fast that it was often difficult to assess individual behaviours. Film file 8 (Nstun2 groups_slow motion) demonstrates 2 groups of pigs filmed during stunning towards the end of the trial (when gas conditions were optimal), with the film sequences slowed down to 40% speed. It shows the pigs struggled before falling but the most violent struggle behaviour occurred after the pigs fell as opposed to before falling which generally occurred in pigs during 90C stunning. As in 90C stunning, once pigs entered the pen, they showed behaviours indicating fear and alertness (forward and backward ear posture and retreating to back of pen) and behaviours indicative of respiratory distress (i.e. signs of trying to breath) - Figure Figure 28: Pig showing first sign of respiratory distress during 80N20C group 31

32 Figure 29: Group showing respiratory distress before falling over during 80N20C group stunning Figure 30: Pigs stretch neck, head and snout upwards; as if in attempt to find air mimicking a drowning response before latency during 80NC group stunning 32

33 Figure 31 and 32: Pigs with neck stretched and wide open mouth during 80N20C group stunning 33

34 4.6 Stun quality 80%N 2 20% CO 2 In total 825 pigs were assessed for stun quality and stun to stick interval. Group size varied from 1 to 5 pigs. The number of batches with respective group size was as follows; 140 batches with 4 pigs (57%), 76 batches with 3 pigs (31%), 4 batches with 2 pigs (1.5%), 24 batches with 1 pig (10%), and 1 batch with 5 pigs (0.5%). For the 4 th (last pig) in the group, the stun to stick intervals varied between 1 min 2 s to 2 min 21 s (average stick time was 1min 21 s) (Table 9). In total 762 (92.4%) pigs were adequately stunned and 63 (7.6%) were inadequately stunned. Table 10 shows the symptoms observed indicating inadequate stunning during 80N20C stunning. There were 40 pigs showing corneal reflex at sticking (24 with only corneal reflex and 16 with corneal reflex and regular gasping). In total 39 pigs showed regular gasping (20 with regular gasping only). Table 9: Stun to stick intervals for each pig in the group during 80N20C stunning Time Pig 1 Pig 2 Pig 3 Pig 4 (minutes : seconds) Minimum 0:16 0:32 0:45 01:02 Maximum 1:40 01:47 02:01 02:21 Average 0: :11 01:21 Table 10: Symptoms of inadequate stunning observed during 80N20C stunning and allocated stun quality level Description of observed symptoms Number (percent) pigs Stun quality level Regular gasping only 20 2 Headshaking and regular gasping 1 2 Corneal reflex only 24 3 Corneal reflex and regular gasping 16 3 Regular gasping and convulsion 2 3 Total Inadequate stun 63 (7.6%) 34

35 Between 7 and 8.30am, and 1.30 to 2.15pm there were problems with O 2 levels rising above 2% within the pit (Appendix 5). Of pigs all pigs found inadequately stunned, 62% occurred during these times. Between 7 and 8.30am 30 pigs were inadequately stunned (22 pigs showing a stun quality level of 2 and 19 with level 3). Between the and 2.15pm the gas remained relatively stable as did the pit temperature. Comparing stun quality during the suboptimal and optimal gas conditions it appeared that when the oxygen increased above 2%, the frequency of inadequate stunning also increased (Table 11) and (Appendix 5). Table 11: Time periods when the O 2 levels were above or below 2% at the first stop (2.2 meters down in the pit) corresponding to total number of pigs inadequately stunned Time Oxygen levels Total pigs 7 to 08.30am & 1.30 to 2.15pm Total Inadequately stunned stunned >2% % inadequately stunned 8.30am to 1.29pm and 2.16 to 3.30pm <2% Comparisons between 90%CO 2 and 80%N 2 20% CO 2 Table 12 shows a comparison of the time in which a batch of pigs could be seen moving through the stun pit from film examination between pigs stunned in the 2 gas types. In 90C stunning 57 batches could be viewed and 51 batches during 80N20C. Table 12: Exposure times presented in minutes and seconds when box moved through pit observed from 54 analysed films during 90C and 51 analysed films during 80N20C stunning 90C 80N20C Minimum 3min 13s 5min 10s Maximum 4min 59s 7min 24s Average 3min 57s 5min 20s The average time to the first reaction pigs made to the gas was similar in both 90C and 80N20C stunning; however the maximum time was longer in 80N20C stunning (12 compared to 21 s respectively. There was not much difference in the time it took before pigs fell between the 2 gas mixes but it took longer for the last pig to fall during 80N20C (Table 13). Pig behaviour before falling generally appeared less vigorous in 80N20C than 90C but after falling muscle excitations and convulsions were often more vigorous (film file 9: Both NvCO2_1 pig and group). Pigs also showed longer periods of vigorous muscle excitations during 80N20C stunning compared to 90C. Film file 9 (both NvCo2 1 pig and group), shows 1 pig and several batches of pigs for both 90C and 80N20C presented simultaneously. Gagging was more frequent in 80N20C stunning and lasted for a longer duration (Table 13). 35

36 Table 13: Comparison of different behavioural phases during the stunning of pigs with 90C and 80N20C gas from when cage with pigs began moving in the Butina (minutes: seconds) 90C Minimum time (minutes:seconds) Maximum time (minutes:seconds) Average time (minutes:seconds) 80N20C Minimum time (minutes:seconds) Maximum time (minutes:seconds) Average time (minutes:seconds) First reaction to gas First pig fall Last pig fall First muscle excitation Last muscle excitation First gag Last gag 00:02 00:11 00:17 00:20 01:30 00:18 01:18 00:12 01:14 00:52 01:50 03:39 01:11 03:37 00:06 00:18 00:27 00:55 02:42 00:30 02:21 00:05 00:15 00:17 00:27 02:46 00:22 03:37 00:21 00:35 01:07 02:50 05: :12 00:07 00:22 00:37 01:04 04:31 00:37 04:51 36

37 5. DISCUSSION Butina recommend that the minimum time from when pigs arrive to stop 1 to when they leave stop 1 on other side of the stun pit is not less than 2 min 30 s for 90C stunning (Kildegaard, Butina personal communications 2013). Using this interval for 5 samples during the 90C study, the time averaged 3min 10 s (1 min 20 s more than Butina s recommendation). The CO 2 concentrations were also 80% at the first stop and between 90 and 93% at the bottom as recommended for 90C stunning. This explains why all pigs were deeply stunned at the stun level rate of 0, with not even the slightest sign of shallow stunning in any of the pigs, despite stun to stick times as long as 1 min 30 s. Most of the pigs were probably dead which is supported by studies completed by Llonch et al (2013) who found that the majority of pigs were clinically dead after a 3 min exposure in a Butina dip-lift system. This is a positive factor to safe guard animal welfare, and as Atkinson et al (2012) found, 90%CO 2 group stun systems consistently provided adequate stunning despite long stun to stick times in 6 Swedish slaughter plants. In 80N20C stunning the time in which the Butina boxes were moving through the pit were mostly shorter than our minimum desired target of 6 minutes (360s). Llonch et al (2013) recommended that pigs need to be in 80%N 2 and 20% CO 2 for at least 5 minutes. However this was during stunning of 1 pig in a 1 box system without stops in a pit volume of 8m 3 opposed to our study which encompassed 6 cages each loaded with 4 pigs, rotating in a pit volume of 63m 3. In 80N20C, one group of pigs due to technical reasosn were left in the Butina for a longer period and the film footage revealed that the pigs continued gagging and slight body movements until 6 minutes and 15 seconds. This is probably a good indication that this is the minimum time necessary to minimise recovery until sticking is complete. In both stun methods the stun to stick time for the last pig in the group exceeded the Swedish recommendation of 60 s (average was 1min 21 s for 90C and 1min 23 s for 80N20C). As 38 pigs showed gasping after exiting the Butina in 80N20C, in combination with exposure times below 5 minutes 30 seconds and extended stun to stick times, the physical action of gasping would have facilitated O 2 back into the body. This was likely as pigs with corneal reflexes showed this just before sticking and not when exiting the Butina. In 80N20C, the spraying of warm water did appear to reduce the frequency of inadequate stunning. While Butina stated that in their experience 17 o C appears to optimise stunning during 90C stunning, little is known how ambient temperature affects pigs uptake of gas. It did appear in the trial that when the gas was below 17 o C at the first stop, O 2 levels increased above 2% and corresponded to an increase in inadequate stun frequency. Heating the gas may have caused the CO 2 to expand pushing out and reducing the O 2 levels. In 90C stunning several pigs regained posture after lying still for some seconds. It is unknown if this was some sort of recovery or if it was an involuntary unconscious nerve reaction. Rodriguez et al (2008) suggests that during 90C stunning, pigs may be conscious during the muscle excitation period. Llonch et al (2013) also found that gagging occurred on average 5 s after brain activity measures started to decrease. Taking this into account, we could therefore consider that in some pigs the duration of discomfort may have been much longer than just at the time in which the pig fell. The purpose of this study was to investigate animal welfare during stunning therefore we must consider that as soon as the pigs entered the Butina, the discomfort and stress imposition begins; not just when they are exposed to the gas. If we take this into consideration, potentially the duration of suffering that the pigs were exposed to during 90C gas was as long as 3 min 39 s. This time interval and the photos in figures 19 to 21 showing signs of discomfort and suffering during conscious animals are certainly an animal welfare concern. The films revealed extremely high hissing sounds from the gas injection into the stun pit, as well as noises caused by the machine pulling cages and cables over cogs during decent and ascent in the gondola system. Pigs were seen listening intently to the sounds, and as the cage descended, behaviours such as 37

38 urination, defecation were triggered. Probably the most frightening aspects upon being closed into the Butina were lack of light, loud sounds, noxious smells and box movement. In both gas systems, pigs were often held inside the Butina for over a minute before the machine was operated. One batch of pigs during 90C stunning was held for nearly 2 minutes. The times the boxes were rotating through the gas gradient and for how long they paused at the bottom of the pit, also varied considerably during both 90C and 80N20C stunning (average 27 s and 20 s respectively). These variations were unexpected. Holding pigs inside the Butina increased the period of discomfort imposed on them, and the reason why this occurred was unknown. If the time in which brain activity decreases (start of unconsciousness) occurs during the onset of gagging, the total time from box entry until first gas should be considered as the potential period of discomfort. This was as long as 3 min 39 s for 90C stunning. During 80N20C pigs exhibited more pronounced paddling with the legs after falling than pigs in 90C. Though pigs lifted the head and snout and jumped upwards in both stun systems, pigs in 90C were often in a state of seizure after falling. Pigs in 80N stunning gasped and gagged more frequently and for longer periods of time. Although the behavioural indications of discomfort during gas induction were generally similar for both gas types, some pigs reacted for longer or with stronger behavioural indications of suffering compared to other pigs within the same group. The reason for these individual differences in reactions cannot be clearly explained other than perhaps gas uptake is affected by individual differences in metabolism, respiration rate and coping behaviours to stress. Pigs that were the first to react were also often the last to fall in the group. It was difficult to analyse vocalisations in both gas types due to high sounds inside the Butina from gas induction and machine function. There was also a lack of synchronisation between the film footage and the time of the digitalised sound recordings as the camera filmed a certain amount of frames per second not matching sound recorded in real time, causing a slight lag between the two. Although it appeared that in some pigs, vocalisations began at the time as high struggle behaviour, it was not possible in many cases to interpret state of consciousness. Although it was generally easier in 80N20C to see that some pigs were still conscious during vocalisations. In 90C pigs frequently made a low snoring sound rather than a high pitched squeal as in 80N20C stunning. This is plausible as high CO 2 levels are highly acidic probably reducing the pigs capacity to squeal in contrast to 80N20C stunning. During 80N20C it was a challenge in this study to keep O2 levels under 2% at the first stop inside the pit. It was also a challenge to keep the stun pit at a stabilised temperature of at least 17 degrees Celsius. However, during the majority of the day, it was possible to maintain optimal gas concentrations and temperature within the pit, as well as an acceptable stun quality. 38