Euthanasia of meat rabbits with carbon dioxide: Behavioral and physiologic responses to gas chamber gradual- and fast-fill rates

Abstract

The use of CO₂ inhalation with different gas chamber fill rates has not been evaluated for euthanasia in commercial meat rabbits. Our objectives were to evaluate the behavioral and physiologic responses of rabbits (pre-weaned to adult) when exposed to gradual- and fast-fill rates of CO₂, and to determine the time to onset of insensibility and death. Cull rabbits (n = 81) were randomly assigned to either a gradual-fill chamber displacement rate of 28% volume change/min (n = 42) or a fast-fill chamber displacement rate of 58% volume change/min (n = 39). The fast-fill rate resulted in a more rapid loss of sensibility at a lower CO₂ chamber concentration and in a faster death than for gradual-fill. There were minimal differences in behavioral responses between fill rates with no clear signs of distress. These findings indicate that CO₂ at the studied displacement rates is suitable for commercial meat rabbit euthanasia.

Introduction

Protocols identifying optimal euthanasia methods should be in place on-farm for all livestock species, so that animals are humanely killed in a prompt, safe, and effective manner (1). Few methods have been validated for on-farm euthanasia of commercial meat rabbits (2). When surveyed, producers have indicated an interest in using gassing methods if operator safety can be assured (3). Carbon dioxide (CO₂) gas inhalation is a common euthanasia method for poultry and mammalian species because it can be applied to multiple animals at once and the technique minimizes animal handling. However, CO₂ inhalation does not cause immediate insensibility and distress may occur before loss of sensibility, depending on gas flow and chamber fill rates (4).

Carbon dioxide inhalation has been evaluated for wild rabbit warren fumigation (5) and in meat rabbits for stunning prior to slaughter (6,7), but it has not been investigated extensively as a method of euthanasia. Currently, the European Food Safety Authority (8) and the American Veterinary Medical Association (AVMA) (4) do not recommend use of CO₂ for euthanasia of conscious rabbits. It has been hypothesized that because rabbits are a burrowing species, they are regularly exposed underground to higher CO₂ levels (5). As a result, there is speculation that rabbits have a high CO₂ tolerance, leading to prolonged induction periods with increased potential for distress (5). This hypothesis has not been tested experimentally.

The objectives of this research were to study the behavioral and physiologic responses of rabbits (pre-weaned to adult) when exposed to gradual- and fast-fill rates of CO₂, and to determine the time to onset of insensibility and death. Welfare was evaluated based on behavioral and physiologic responses of rabbits prior to loss of sensibility. Time to loss of sensibility, death, and irreversibility were also evaluated. Based on anecdotal reports from others using 100% chamber volume change/min CO₂ fill rates to kill rabbits, we hypothesized that both fill rates would cause potential distress behaviors and vocalizations in rabbits, but these signs would be less common when gradual-fill methods were used.

Materials and methods

Study animals

Eighty-one New Zealand white and California mix rabbits (11 d to 25 mo old) were used. Rabbits were obtained from an Ontario commercial meat rabbit farm and identified for cull by the producer based on clinical signs, such as unthriftiness, injury, illness (e.g., enteritis), and end of reproductive use. Rabbits with clinically obvious respiratory disease (e.g., active sneezing, respiratory distress or labored breathing, mucopurulent nasal discharge) were excluded; however, pasteurellosis was enzootic on this farm. Three different ages of rabbits [pre-weaned kits (11 d to 5 wk), n = 28; growers (6 to 12 wk), n = 28; and adult does (> 12 wk), n = 25] were randomly assigned to 2 different treatments using a random number generator (random. org): gradual-fill (n = 42) and fast-fill (n = 39) CO₂ chamber exposure (Table 1). The initial study plan included evaluation of kits < 11 d of age; however, 3 pilot trials with kits of this age suggested that kits 10 d old or younger required prolonged exposure periods to CO₂ gas (> 45 min), indicating that this technique is unsuitable for this age group. Further trials were not conducted with pre-weaned kits < 11 d old. Pre-weaned kits (11 d to 5 wk) and growing rabbits were of mixed gender; no adult bucks required euthanasia during the study. The procedures and protocol for this research were reviewed and approved by the University of Guelph Animal Care Committee (AUPe3366). The University of Guelph is in compliance with the Animals for Research Act of Ontario (OMAFRA, 1990) and holds a Good Animal Practice certificate issued by the Canadian Council on Animal Care.

Table 1.

Number of rabbits per age group randomly assigned to different treatments.

Methoda	Age categoryb	Number	Body weight (kg) (mean ± SE)
Gradual-fill (n = 42)	Pre-weaned kits	14	0.27 ± 0.01
	Growers	15	1.6 ± 0.03
	Adults	13	3.1 ± 0.06
Fast-fill (n = 39)	Pre-weaned kits	14	0.28 ± 0.01
	Growers	13	1.6 ± 0.03
	Adults	12	3.1 ± 0.03
Total		81

^aGradual-fill = 10% ± 0.3 chamber CO₂ increase in the first minute. Fast-fill = 39% ± 0.7 chamber CO₂ increase in the first minute.

^bPre-weaned kits = 11 d to 5 wk, Growers = 6 to 12 wk, Adults = > 12 wk.

Pre-weaned kits and growers were mixed gender, and adults were female only.

SE — standard error.

Experimental apparatus

To facilitate behavioral observations, a 14.3 L acrylic, anesthetic induction chamber (VetEquip, Livermore, California, USA) (internal dimensions: 38.5 cm long × 19.4 cm wide × 19.1 cm high) was modified to allow for reflex testing after loss of posture, as described by Moody et al (9). Briefly, a 7.5-cm diameter hole was created in the front chamber wall 4 cm from the bottom. A hand input was created using a clear poly boot cover (Fisher Scientific, Ottawa, Ontario) with a wire ring added for structural integrity, and the hole was sealed with duct tape.

Carbon dioxide gas was supplied from a 22.7-kg cylinder (Praxair Canada, Mississauga, Ontario) equipped with a single stage CO₂ regulator (Harris, Gainesville, Georgia, USA) and connected to a M3000 table-top veterinary anesthesia machine (Supera Anesthesia Innovations, Clackamas, Oregon, USA). The anesthesia machine was modified to solely utilize the flow meter and was connected to the induction chamber inflow port. The inflow and outflow ports were located on the same wall of the chamber 15 cm from the floor bottom and 3 cm from the side walls. The outflow waste gas tube was 6 m long and emptied downwind outside of the building (Figure 1).

Figure 1.

Equipment setup for carbon dioxide gassing euthanasia study. A — Overview of setup. B — Gas monitor (Draeger Safety Canada, Mississauga, Ontario) placement for measuring chamber concentrations of carbon dioxide and oxygen.

Chamber gas concentrations were monitored and manually recorded every 10 s using an X-am 7000 multi-gas monitor (Draeger Safety Canada, Mississauga, Ontario) configured to measure CO₂ concentration in the range of 0% to 100% and O₂ concentration of 0% to 25%. The gas monitor sat outside the induction chamber with the pump connected to a 0.7 cm diameter 14 cm long tube entering the chamber through a 1 cm diameter hole drilled 9.3 cm from the floor bottom and centered on the same chamber wall as the inflow and outflow ports. Design and testing of equipment were conducted under the guidance of an agricultural mechanical engineer (J VdeV). The unit was calibrated after each trial and sensitivity was within 0.1%.

Gas concentration analysis — empty chamber testing

Before starting in vivo trials, 24 CO₂-only trials were completed to evaluate safety risks for personnel and to determine the equipment settings needed to achieve the desired chamber concentration fill rates. Chamber CO₂ and O₂ concentrations were recorded every 10 s from the gas monitor. A second identical monitor was placed immediately outside the chamber to assess the ambient air, to check for chamber leaks, and to ensure personnel safety.

Gas concentration analysis — rabbit CO₂ exposure trials

Chamber temperature during CO₂ filling was recorded for the first 17 trials and discontinued afterwards due to minimal fluctuation. Temperature was monitored using an indoor/outdoor thermometer (RadioShack, Fort Worth, Texas, USA) with a remote sensor probe inserted through a 1 cm diameter hole (9.3 cm from the floor bottom) on the opposite chamber wall to the inflow and outflow ports.

Animal trials were conducted on a large (n = 600 doe) farm in southwestern Ontario over 7 research trial days from July 14 to August 9, 2016 in a drive-in shed open to the outdoors, allowing for good air flow. Order of treatment was assigned randomly the day before the study was initiated (random.org). Rabbits were euthanized individually in the chamber while the other rabbits waited in holding crates.

Behavioral and physiologic analyses

An ethogram (Table 2) was developed based on predicted and confirmed behaviors. Behavioral and physiologic responses were constantly observed during the period of exposure by one person and recorded at 10 s intervals by the second person; neither blinded to treatment. Baseline behavior was observed for at least 2 min for each rabbit to establish response to chamber conditions when filled with air. Time to onset of a behavior as well as the time and gas concentration data were recorded. The hand input was used to check for reflexes from loss of posture until the last reflex disappeared. Start and stop times for vocalizations were recorded as this behavior was the only one to persist in some animals after loss of righting reflexes. Behavioral observations were made directly and video recording using a GZ-E200 full HD Everio Camcorder (JVC, Yokohama, Japan) was used for verification, if needed. Chamber CO₂ and O₂ concentrations were recorded every 10 s from the gas monitor. All rabbits, regardless of treatment group, were exposed to CO₂ for up to 420 s, a minimum of 1 min after last breath (4). Heartbeat was palpated upon removal from the chamber to confirm death. Rabbits were observed for return to breathing or sensibility during the period of time in between trials (5 to 19 min) after removal from the chamber and a non-penetrating captive bolt device (Zephyr-E, Bock Industries, Philipsburg, Pennsylvania, USA) was then used as a secondary euthanasia method. The induction chamber was emptied and cleaned between uses.

Table 2.

Ethogram for evaluating rabbit behavioral and physiologic responses during carbon dioxide gas exposure.

Behavior	Definition
Loss of balance	Uncoordinated movement or falling over.
Loss of posture	Rabbit is no longer able to hold itself up; falls to the side, head sags, belly on the ground.
Loss of righting reflex	Turning the animal on its side to see if it can right itself.
Corneal reflex	Blink response to touching the surface of the eye.
Palpebral reflex	Blink response to touching the area around the eye.
Ear pinch reflex	Pinching the rabbit’s ear looking for voluntary movement, i.e., pulling head away.
Toe pinch reflex	Pinching the skin in between the rabbit’s toes looking for voluntary movement, i.e., pulling foot away.
Clonic convulsions	Leg paddling.
Tonic convulsions	Rigid extension of the limbs.
Increased respiration	Increased speed of nostril flaring.
Decreased respiration	Long pauses in between breaths.
Last breath	Nostril flaring and open mouth breathing stop occurring.
Open mouth breathing	Mouth opening to breathe.
Gasping	Deep full body breath through a wide open mouth.
Vocalizations	High pitch scream.
Grunts	Undefined sounds emitted by the animal.
Head shake	Rapid movement of the head.
Head lift	Attempts to get above chamber concentration; stretching neck and lifting head up, nose pointed up, smaller rabbits stretch upwards against chamber wall.
Head bob	Moves head up and down and side to side.
Turn around	Rabbit switches direction facing.
Chewing	Moving mouth and tongue in pattern, not in attempt to manipulate or eat material.
Defecation	Visible stool is released from the body.
Urination	Visible urine is released from the body.

Statistical analysis

Statistical analyses were conducted using SPSS (SPSS Statistics for Windows, Version 23.0. 2014; IBM, Armonk, New York) and P < 0.05 was accepted for significance. All values are reported as mean ± standard error (SE). Dependent variables were checked for normality using the Shapiro-Wilk test. To check for differences in time and concentration between fill rate treatments, a non-parametric independent sample Mann-Whitney U-test was used for variables not normally distributed and an independent samples t-test was used for variables that were normally distributed. The Kruskal-Wallis 1-way analysis of variance (ANOVA) was used to check for differences between age categories within treatment. Data was split by treatment to allow for comparisons within treatment groups with age as the independent variable. To determine if behaviors such as urination, open mouth breathing, and convulsions occurred before a loss of sensibility, time of behavior occurrence was subtracted from time to loss of righting reflex. A Chi-square test of independence was used to test for associations between occurrence of a behavior and treatment. Column properties were compared using a Bonferroni correction for multiple analyses. Cramer’s V-test was run to determine the strength of the association.

Results

Gas concentration analysis — empty chamber testing

From the 24 dry CO₂ runs, it was determined that a cylinder regulator output pressure of 30 psi was needed to overcome tubing resistance and maximize air flow. The gradual-fill displacement rate used was 28% chamber volume change/min and was most accurately achieved with the flow meter set to a rate of 4 L/min. The gas monitor recorded a mean chamber CO₂ concentration of 10 ± 0.3% (range: 6.8% to 15.5%) at 60 s during gradualfill. The fast-fill rate was established based on the flow meter’s maximum capacity calculated at 8.3 L/min with a calculated displacement rate of 58% chamber volume change/min. The gas monitor recorded a mean chamber CO₂ concentration of 39 ± 0.7% (range: 32% to 54%) at 60 s during fast-fill. During the fast-fill trials, maximum CO₂ chamber concentrations occurred within 3 min. The flow rate was reduced to 4 L/min at 3 min, allowing the CO₂ concentration to be held constant.

Analysis of the ambient air outside the chamber determined that there were minimal safety risks for personnel working near the chamber as CO₂ levels remained below the gas monitor’s safety alarm level of 3%.

Gas concentration analysis — rabbit CO₂ exposure trials

Chamber temperature monitoring during CO₂ filling of the first 17 trials indicated minimal temperature change from ambient. The greatest temperature fluctuation inside the induction chamber was 3.9°C during a gradual-fill trial (Figure 2). The ambient temperature on the trial days was between 19°C and 30°C with high humidity and a natural breeze.

Figure 2.

Change in chamber temperature across the first 17 gradual- and fast-fill rabbit gassing trials. Temperatures were recorded every 10 s. The average ambient outside temperature across these trials was 26°C.

Chamber gas concentrations were consistent across all trials of the same treatment, forming tight gradual-fill curves and fast-fill curves (Figure 3). The maximum CO₂ concentration achieved for gradual-fill was 58.4% at 360 s. The fast-fill rate achieved higher CO₂ concentrations with a maximum of 73.6% reached at 230 s. Oxygen concentrations were reduced as CO₂ levels increased. The lowest O₂ concentration achieved during a gradual-fill trial was 5.2% at 420 s (end of exposure) and 1.1% in 180 s during a fast-fill trial.

Figure 3.

Chamber gas concentration curves during rabbit gassing trials. Chamber concentrations of carbon dioxide and oxygen were recorded every 10 s. A — Carbon dioxide curves for the gradual-fill trials (n = 42). B — Carbon dioxide curves for the fast-fill trials (n = 39).

Physiologic analyses

Average CO₂ and O₂ concentrations at each time point were calculated for treatments (Figures 4 and 5). Rabbits progressed from a loss of balance to loss of posture to loss of righting reflex (Table 2). Loss of righting reflex occurred within seconds (9 ± 1 s) of loss of posture and was recorded for all but 1 larger rabbit for which lack of space in the chamber prohibited reflex testing. Loss of righting reflex (inability to right itself after being turned on its side) was used as the primary variable to judge loss of sensibility. Loss of righting reflex occurred significantly earlier (P < 0.001) during fast-fill (40 ± 1 s) than gradual-fill (99 ± 3 s). Time to loss of righting reflex significantly differed (P < 0.001) between pre-weaned kits and growers during gradual-fill. Pre-weaned kits lost their righting reflex earlier (84 ± 3 s) compared to growers (113 ± 5 s) during gradual-fill. There was a significant difference (P = 0.02) in the mean overall CO₂ chamber concentration to which rabbits were exposed before loss of righting reflex, 20.7 ± 0.9% for gradual-fill and 16.5 ± 1.5% for fast-fill. The maximum CO₂ concentration to which any rabbit was exposed before loss of righting reflex was 36% for gradual-fill and 39% for fast-fill.

Figure 4.

Mean concentration of carbon dioxide (CO₂) and oxygen (O₂) in the chamber during gradual-fill experiments (n = 42). Standard error across time points ranged from 0.02% to 0.8%. Concentration was recorded every 10 s. Median time and CO₂ concentration at onset of euthanasia stages: A — Increased respiration 30 s, 2.4%. B — Loss of righting reflex 100 s, 20.5%. C — Loss of corneal reflex 135 s, 29%. D — Last breath 315 s, 56%. F — Gas off and rabbit removed from chamber 420 s.

Figure 5.

Mean concentration of carbon dioxide (CO₂) and oxygen (O₂) in the chamber during fast-fill experiments (n = 39). Standard error across time points ranged from 0.09% to 1.8%. Concentration was recorded every 10 s. Median time and CO₂ concentration at onset of euthanasia stages: A — Increased respiration 10 s, 1.4%. B — Loss of righting reflex 40 s, 14.3%. C — Loss of corneal reflex 50 s, 27%. D — Last breath 120 s, 66%. E — Gas flow rate reduced to 4 L/min to maintain CO₂ levels at 70%, 180 s. F — Gas off and rabbit removed from chamber 420 s.

The palpebral reflex disappeared after loss of the righting reflex in all rabbits. Toe pinch and ear pinch were lost next. The corneal reflex was the last to disappear and was lost at significantly different times based on fill rate; 142 ± 4 s for gradual-fill and 50 ± 2 s for fast-fill (P < 0.001). Carbon dioxide concentration did not vary significantly between fill rates (P = 0.11) when corneal reflex was absent; 30.5 ± 0.8% for gradual-fill and 27 ± 2% for fast-fill. When analyzed within gradual-fill treatment groups, time to loss of corneal reflex differed significantly between pre-weaned kits and growers (P = 0.01), and adults and growers (P = 0.001). The corneal reflex was absent at 129 ± 7 s in adults, 164 ± 8 s for growers, and 132 ± 3 s for pre-weaned kits.

Respiration changed during prolonged CO₂ exposure from an increased rate to open mouth breathing to decreased respiration to last breath (Figure 6). Increased respiration occurred at significantly different times (P < 0.001), and concentrations (P = 0.002) by fill rate, specifically 35 ± 3 s and 4.2 ± 0.7% CO₂ for gradual-fill, and 15 ± 1 s and 1.8 ± 0.3% CO₂ for fast-fill. Breath holding was not observed. When open mouth breathing occurred, in 65% of trials it started after the rabbit had lost its righting reflex. Distribution of open mouth breathing across age groups and fill rates is shown in Table 3. There was no association between open mouth breathing before loss of righting reflex and fill rate, χ² (1, n = 81) = 0.48, P = 0.49. Duration of open mouth breathing before loss of righting reflex for the 28 rabbits that exhibited this behavior was 25 ± 0.002 s with the longest duration being 1 min 10 s.

Figure 6.

Timeline of progression of physiologic responses (median time to occurrence) during gradual-fill (n = 42) and fast-fill (n = 39) carbon dioxide exposures.

Table 3.

Frequency of behavioral response, grouped by age and fill-rate.

Behavior	Fill ratea	Total	Number	Number of pre-weaned kitsb	Number of growersb	Number of adultsb
Open mouth breathing	Gradual-fill	42	42	14	15	13
	Fast-fill	39	38	14	13	11
	Total	81	80	28	28	24
Open mouth breathing before loss of righting reflex	Gradual-fill	42	16	2	9	5
	Fast-fill	39	12	5	5	2
	Total	81	28	7	14	7
Chewing	Gradual-fill	42	19	3	9	7
	Fast-fill	39	8	1	3	4
	Total	81	27	4	12	11
Gasping	Gradual-fill	42	16	8	7	1
	Fast-fill	39	17	4	6	7
	Total	81	33	12	13	8
Head shake	Gradual-fill	42	5	0	2	3
	Fast-fill	39	11	6	2	3
	Total	81	16	6	4	6
Head bob	Gradual-fill	42	14	4	4	6
	Fast-fill	39	10	2	6	2
	Total	81	24	6	10	8
Urination	Gradual-fill	42	18	8	5	5
	Fast-fill	39	10	5	5	0
	Total	81	28	13	10	5
Coughing	Gradual-fill	42	1	1	0	0
	Fast-fill	39	2	0	1	1
	Total	81	3	1	1	1
Defecation	Gradual-fill	42	5	1	3	1
	Fast-fill	39	2	0	0	2
	Total	81	7	1	3	3
Head lifting	Gradual-fill	42	33	11	14	8
	Fast-fill	39	23	9	8	6
	Total	81	56	20	22	14

^aGradual-fill = 10% ± 0.3 chamber CO₂ increase in the first minute; Fast-fill = 39% ± 0.7 chamber CO₂ increase in the first minute.

^bPre-weaned kits = 11 d to 5 wk; Growers = 6 to 12 wk; Adults = > 12 wk; Pre-weaned kits and growers were mixed gender, and adults were female only.

Some rabbits (n = 10) were observed squinting before loss of righting reflex. After death, some rabbits had red coloration of the nictitating membrane and occasional minor bleeding around the eyes and nostrils. During exposure, there was no discharge from the eyes, nose, or mouth, and no facial grooming.

Clonic (n = 24) and tonic (n = 11) convulsions never occurred before loss of righting reflex for any age group or treatment. No rabbits had a palpable heartbeat when removed from the chamber, and none regained sensibility.

Behavioral analyses

Chewing was the only behavior that was not independent of treatment, χ² (1, n = 81) = 5.56, P = 0.02 with a medium strength of association by treatment (Cramer’s V = 0.3). More rabbits displayed this behavior during gradual-fill exposures (Table 3). Chewing occurred at 71 ± 8 s and 14 ± 2% CO₂, respectively, for gradual-fill (n = 19), and 27 ± 4 s and 8 ± 4% CO₂, respectively, for fast-fill (n = 8). Other behaviors observed, but with no findings of associations between occurrence and CO₂ fill rate were gasping (n = 33), head shake (n = 16), head bobbing (n = 24), urination (n = 28), coughing (n = 3), vocalizations (n = 6), defecation (n = 7), and head lifting (n = 56) (Table 3). Urination was generally observed before loss of righting reflex (n = 18). Table 3 summarizes the mean time to onset of measured behaviors by age and fill-rate.

Vocalizations occurred inconsistently for both fill rates but was a behavior displayed by more rabbits in the fast-fill group. In total, 5 of 39 (13%) rabbits vocalized during fast-fill exposure before loss of righting reflex, at a mean time of 12 ± 2 s and a mean CO₂ concentration of 1.5 ± 0.3%. One of 42 (2%) rabbits vocalized during gradual-fill at 140 s and 30% CO₂. The rabbit that vocalized during gradual-fill began vocalizations after loss of righting reflex and stopped shortly before the corneal reflex was lost. Similarly, for 1 of the rabbits in the fast-fill group, vocalizations started before loss of balance and persisted for a short period of time after loss of righting reflex. Vocalization was a rare occurrence, lacking a pattern when it did occur as well as any correlation to loss of righting reflex. Response to CO₂ gas flow being turned on varied amongst individuals with some rabbits (n = 14) turning away upon first exposure while other rabbits pushed their faces directly into the gas inflow experiencing 100% CO₂ with no obvious behavioral response.

Discussion

Exposure of rabbits to CO₂ at a gradual-fill rate (28% volume change/min) resulted in insensibility, judged by loss of righting reflex, in 99 ± 3 s. Insensibility occurred significantly earlier for fast-fill (58% volume change/min) at 40 ± 1 s. Rabbits stopped breathing and were dead at 328 ± 11 s during gradual-fill and at 116 ± 4 s during fast-fill. Irreversible insensibility was achieved for all rabbits regardless of flow rate. There were minimal differences between fill rates for behavioral and physiologic responses.

Identifying exact time to loss of sensibility after exposure to CO₂ has important welfare implications since it marks the end of potential distress associated with gas exposure. We found that fast-fill displacement limited potential distress by reducing time to loss of righting reflex. Whether gradual-fill rates limit distress is unclear and may vary by species. Niel and Weary (10) determined that rats avoided exposure to CO₂ faster as chamber displacement rates increased. Hickman et al (11) observed distress in rats at displacement rates as low as 7% volume change/min, and they found that a 30% volume change/min minimized pain and distress. In piglets, higher displacement rates resulted in a faster onset of insensibility, a lower frequency of distress behaviors, and enhanced euthanasia efficacy (12). Bolvin et al (13) found that for mice, displacement rate variation did not cause a significant difference in observed pain or distress. These results support the lack of significant behavioral and physiologic differences between gradual- and fast-fill chamber rates.

Our experimental design permitted direct reflex testing and loss of righting reflex was primarily used to judge insensibility in combination with other behaviors and reflexes. Loss of posture was validated as an indicator of insensibility in EEG-instrumented rabbits by Dalmau et al (7). Others have challenged the lack of consistency of loss of posture as a measure of insensibility (14). A study in rats determined that loss of righting reflex was a more sensitive measure of insensibility than loss of posture. Dalmau et al (7) were unable to assess loss of righting reflex as their observations were conducted visually from the top of a gassing pit. In our study, loss of posture occurred immediately prior to loss of righting reflex and the onset of each was often indistinguishable. Loss of righting reflex was used as a conservative estimate of loss of sensibility whenever there was a difference in onset between loss of posture and loss of righting reflex.

Similar to other studies, the corneal reflex was the last reflex to disappear in this study. Loss of corneal reflex has been validated as a measure of insensibility in rabbits by several others (15–19). The corneal reflex is known to disappear at a deeper stage of anesthesia (20). Thus, loss of righting reflex and loss of corneal reflex can be used in combination to determine insensibility during euthanasia of rabbits.

This study determined that time to loss of righting reflex (LORR) varied across different ages of rabbits. Within gradualfill rates there was a significant difference in time to loss of righting reflex between pre-weaned and growing rabbits in that pre-weaned rabbits became insensible faster. In contrast, time to LORR during fast-fill exposures did not vary between ages. Previous studies exposing rabbits to CO₂ used only market weight (i.e., growing) rabbits (5–7). In other mammalian species, neonates have been observed to be more tolerant to higher CO₂ exposure levels than adults, generally taking longer to lose sensibility and die (21). Piglets are an exception to this pattern as Sadler et al (12) identified that pre-weaned piglets succumbed to CO₂ earlier than grower piglets. Because CO₂ is denser than air, it may be that younger animals are exposed to higher concentrations faster than taller adults. Niel and Weary (10) examined the concentrations in different areas of a similar sized chamber and found that during the first few seconds while filling the chamber, there was a difference in CO₂ concentration between the top and bottom. After 20 s, the gas mixed more thoroughly and no difference was observed.

One potential cause of distress induced by CO₂ is air hunger, indicated by a change in respiration rate (22,23). Increased respiration was observed in our study at low CO₂ concentrations (1.8 ± 0.3% for fast-fill and 4.2 ± 0.7% for gradual-fill). It is unknown whether labored breathing at low CO₂ concentrations is associated with distress or pulmonary pain (22–24). More visibly labored breathing occurred when open mouth breathing began and this was used as the first sign of potential distress. In most rabbits (65%), open mouth breathing occurred after loss of righting reflex and thus was unlikely to be associated with distress or pain. Duration of open mouth breathing before loss of sensibility lasted 25 ± 0.002 s.

Carbon dioxide reacts with water to form carbonic acid, mildly acidifying mucous membranes. This may create an acidic taste and cause irritation to the eyes, nose, and mouth (25). Behaviors predicted for rabbits experiencing irritated mucous membranes included blinking, cleaning of the face, and nasal and eye discharge (26,27), none of which were observed in our study. Sensibility was lost at a CO₂ concentration below the threshold for mucosal nociceptor activation for mammals (28). Head lifting was observed in 56 of 81 rabbits and in both CO₂ groups. Head lifting also occurred during baseline observations when rabbits were breathing ambient air, suggesting that it represents an exploratory behavior rather than an attempt to escape. Head shaking was displayed by 20% (n = 16) of rabbits. Head shaking occurred independent of fill rate, but chewing motions were seen in 33% of 81 rabbits and significantly more commonly during gradual-fill trials. These behaviors could indicate an aversive response; however, lack of an air-only control group limits interpretation. A comparison could be made to rabbits exposed to isoflurane and sevoflurane for anesthesia. Flecknell et al (29) observed that rabbits exposed to isoflurane and sevoflurane avoid gas by raising their noses to the lid, breath-holding, and violently struggling (29). Compared to these observations, CO₂ exposure did not cause behavioral signs of distress in rabbits. Similarly, vocalizations were a rare occurrence with no correlation to fill rate, and generally occurred at low CO₂ concentrations. There is some uncertainty in interpreting vocalizations as 1 rabbit began vocalizing after loss of righting reflex when it was nonresponsive to other reflexes.

Limitations of the current study include variability in relative space per rabbit and lack of a compressed air-only control group. Personnel safety was not found to be a concern. Small gas leaks were found near the hand input; however, due to good airflow in the working space, safe levels of CO₂ in the ambient air were not breached. There are minimal concerns for personnel safety when using CO₂ on-farm for euthanasia if conducted in a well-ventilated area.

This study evaluated the use of CO₂ gas inhalation for meat rabbit euthanasia. Loss of sensibility and last breath were prolonged during gradual-fill and occurred significantly earlier during fast-fill CO₂ exposure. Both methods were highly effective in that all animals were verified to be dead after removal from the chamber and none returned to sensibility. There were minimal differences in behaviors observed prior to insensibility between rabbits exposed to gradual- and fast-fill and both fill rates provided an acceptable method for inducing a humane death.