Content uploaded by Robert H. Schmidt
Author content
All content in this area was uploaded by Robert H. Schmidt on May 02, 2018
Content may be subject to copyright.
666 DROWNING, EUTHANASIA,
AND CARBON-DIOXIDE
NARCOSIS
Drowning
is
not
euthanasia
John W
Ludders,
Robert
H. Schmidt,
F Joshua
Dein,
and Patrice
N. Klein
Historically, there has been considerable discus-
sion within the nuisance wildlife control and trap-
ping communities as to whether drowning is a
humane method for killing animals. The issue
received more attention in 1993, when the
American Veterinary Medical Association's Panel on
Euthanasia reaffirmed its position that drowning is
an unacceptable method (Andrews et al. 1993). For
this article, we make a distinction between euthana-
sia, a "good death" that occurs without pain or dis-
tress (Andrews et al. 1993), and death due to killing
by other methods. The central issue in this debate
is whether drowning animals are rendered uncon-
scious by great levels of carbon dioxide (CO2 car-
bon-dioxide-induced narcosis) early in the drown-
ing process and thus are insensitive to the distress
and pain associated with drowning.
Proponents of drowning cite an article by Gilbert
and Gofton (1982) in which the authors stated that
drowning animals die from carbon-dioxide-induced
narcosis. However, Gilbert and Gofton (1982) did
not report any information on levels of carbon
dioxide in blood, which is needed before a deter-
mination can be made about the acceptability of
drowning as a method of euthanasia. We wish to
introduce and clarify information concerning
effects of carbon dioxide that have been absent in
the debate on drowning.
In their laboratory investigations, Gilbert and
Gofton (1982) determined time to death by drown-
ing in mink (Mustela vison), muskrat (Ondatra
zibethica), and beaver (Castor canadensis).
Readings of the electrical activity of the brain (elec-
troencephalograph, EEG) and of the heart (electro-
cardiograph, ECG) were recorded from each animal
during drowning, and time of death was taken to be
the moment when electrical activity of the brain
ceased (EEG signal became flat). On average, the
EEG signal became flat in mink after 4 minutes, 37
seconds; in muskrats after 4 minutes, 3 seconds; and
in beaver after 9 minutes, 11 seconds. However,
neither arterial nor venous blood samples were col-
lected before, during, or after the animals drowned,
so the partial pressures of carbon dioxide (PCO2)
or oxygen (PO2) in blood from these animals were
not measured. The authors stated that "[d]eath by
CO2 induced narcosis (submersion asphyxia) was
evident in beaver, about 50% of muskrats, but 'wet'
drowning (defined below) occurred in mink"
(Gilbert and Gofton 1982:835). A review article
written by Timperman (1972) was referenced to
corroborate their conclusion. Timperman's (1972)
paper discussed the forensic diagnosis of drowning
through identification of diatoms in the lungs of
victims. The author mentioned that carbon-dioxide-
induced narcosis could be a possible cause of death
during drowning, but he also acknowledged that
death could be from anoxia. However, he did not
provide substantiating data, such as blood gas analy-
ses, to support either factor as the cause of death by
drowning.
Proponents of drowning make a distinction
between "wet" or "dry"
drowning, the former occur-
ring when water enters the lungs and the latter
when the lungs remain relatively dry. To some,
"dry" drowning implies that because the animal
does not inhale water, then death is from CO2-
induced narcosis, although this is most likely incor-
rect. According to reports of incidents involving
human drownings, 2 events may occur following
submersion: 1) during the ensuing panic and strug-
gle, water is swallowed and aspiration occurs in
Address for John W. Ludders:
College of Veterinary
Medicine, Cornell University, Ithaca, NY 14853, USA. Address for Robert H.
Schmidt: Department of Fisheries and Wildlife, Utah State University, Logan UT 84322-5210, USA. Address for F.
Joshua Dein:
USGS-BRD National Wildlife Health Center, Madison, WI 53711, USA. Address for Patrice N. Klein: Humane Society of the
United States, 700 Professional Drive, Gaithersburg,
MD 20879, USA.
Key
words: animals, carbon dioxide, drowning, euthanasia, killing
Wildlife
Society
Bulletin
1999, 27(3):666-670 Peer edited
Drowning,
euthanasia,
and carbon-dioxide narcosis
* Ludders et al. 667
85% of the victims, which leads to "wet" drowning,
i.e., the lungs fill with water (Newman and Stewart
1995) and hypoxia and cardiac arrest occur rapidly,
the latter probably because the vagal nerve, in
response to water contacting the mucous mem-
branes of the larynx or trachea, causes a reflex
slowing and arrest of the heart (Suzuki 1996); or 2)
during drowning, the act of swallowing water may
lead to laryngospasm (an involuntary closure of the
glottis or entrance to the airway), thus sealing the
airway and preventing water from being aspirated
into the lungs (Yagil et al. 1983, Suzuki 1996).
Approximately 15% of human drowning victims
experience "dry" drowning, in which the lungs
remain relatively free of water (Newman and
Stewart 1995). Hypoxia and cardiac arrest develop,
but often this process is protracted compared to
the victims experiencing "wet" drowning. In fact,
current research strongly suggests that death
occurs more rapidly when water is inhaled because
it initiates a reflex vagal inhibition of the heart
(Suzuki 1996). Thus, a longer period of conscious-
ness may be associated with "dry"
drowning than
with "wet" drowning. The accumulated evidence
(as discussed below) indicates that the cause of
death during drowning is hypoxia and anoxia, not
CO2-induced narcosis.
Stedman's Medical Dictionary (1995:1176)
defines narcosis as a "[g]eneral and nonspecific
reversible depression of neuronal excitability, pro-
duced by a number of physical and chemical
agents, usually resulting in stupor rather than in
anesthesia." Hypercarbia, or an excess of carbon
dioxide (CO2) in blood, can cause narcosis. In ani-
mals, CO2 is a normal byproduct of oxygen (02)
metabolism, and it is eliminated from the body
through the lungs and the process of pulmonary
ventilation (Guyton 1991). The relationship of CO2
production to 02 utilization is-expressed as the res-
piratory exchange ratio, generally accepted to be
around 0.8; it indicates that in general, less CO2 is
produced for a given amount of metabolized 02
(Guyton 1991).
Several studies, involving numerous animal
species in which blood gases were measured, indi-
cate that carbon-dioxide narcosis does not occur
until the partial pressure of carbon dioxide in arte-
rial blood (PaCO2) exceeds 95 millimeters of mer-
cury (mm Hg) and true anesthesia occurs only
when PaCO2 exceeds 200 mm Hg. For example,
laboratory rats exposed to 100% CO2 at various
chamber fill rates started to show evidence of CO2
narcosis (they became uncoordinated) after PaCO2
exceeded 123 mm Hg (Hewett et al. 1993). The
same rats became immobile only after PaCO2
exceeded 212 mm Hg, and they finally lost the
pedal reflex to painful stimulation (toe pinch) after
PaCO2 exceeded 332 mm Hg (Hewett et al. 1993).
A study of the narcotic properties of carbon diox-
ide in dogs sheds more light on the issue of CO2-
induced narcosis (Eisele et al. 1967). In this study,
the narcotic and anesthetic properties of CO2 were
determined in 2 ways: 1) by determining the MAC
(the minimum alveolar concentration of an inhalant
anesthetic that prevents purposeful movement by
an animal exposed to a painful stimulus) for the
inhalant anesthetic halothane (2-bromo-2-chloro-
1,1,1-trifluoroethane), and then, in a step-wise man-
ner, replacing the halothane with CO2 while main-
taining a constant plane of anesthesia; and 2) by
administering only CO2 to dogs and recording the
PaCO2 when each dog was anesthetized and unre-
sponsive to a painful stimulus. The results indicat-
ed that increasing levels of PaCO2 above 95 mm Hg
were increasingly narcotic. At a PaCO2 of 95 mm
Hg the narcotic effect of CO2 was minimal as it
reduced the MAC of halothane by only 0.08%. In
this study, anesthesia was produced at an average
PaCO2 of 222 mm Hg.
Drowning animals, of course, are not breathing
100% C02, let alone air;
in fact, they are not breath-
ing at all. Because the drowning animal cannot
breathe, it uses all of the 02 available in its blood,
and CO2 accumulates because of oxygen metabo-
lism. As previously noted, the respiratory exchange
ratio indicates that the rate of 02 utilization is
greater than the rate of CO2 production (Guyton
1991), and this fact is demonstrated by numerous
animal studies. In dogs that were drowned with
either cold salt water (CSW) or cold fresh water
(CFW), PaCO2 increased significantly, but after 10
minutes of immersion it never exceeded 64.8?4.9
mm Hg in either group (Conn et al. 1995).
However, PaO2 significantly decreased in both
groups; after 4 minutes of immersion, PaO2 was
16.4?1.5 mm Hg in the CFW group and 18.8?21.6
mm Hg in the CSW group, and after 10 minutes of
immersion it was 9.6?3.8 and 8.8+1.9 in the CFW
and CSW groups, respectively. Similar results were
found in another study involving anesthetized, intu-
bated dogs that inhaled a fixed quantity (20 ml/kg)
of fresh water (Rai et al. 1980). Prior to inhaling
water, the PaO2 and PaCO2 were 100 mm Hg and
35 mm Hg, respectively. Five minutes after inhaling
668 Wildlife
Society Bulletin 1999, 27(3):666-670
water, the PaO2 and PaCO2 were 35 mm Hg and 52
mm Hg, respectively. During 40 minutes of obser-
vation, PaCO2 never exceeded 60+0.5 mm Hg
(mean + SEM) and the PaO2 did not exceed 47+5.5
mm Hg. The results from these 2 studies show that
PaCO2 levels were well below those necessary to
induce CO2 narcosis and that the dogs were hypox-
emic (inadequate oxygen in blood).
In a study that measured cerebral blood flow and
arterial blood gases in ducks (Anas platyrhynchos)
held under water for more than 4 minutes, the
average PaO2 was 52 mm Hg (minimum recorded
was 37 mm Hg) at 4.61 minutes, while the average
PaCO2 was 51 mm Hg (Stephenson et al. 1994).
These numbers indicate that the ducks were
hypoxemic and hypercarbic and that PaCO2 was
not at levels known to produce narcosis. However,
PaO2 had decreased to hypoxemic levels, and had
the ducks not been killed by decapitation, the PaO2
would have continued to decrease to levels incom-
patible with life, i.e., the ducks would have died
from anoxic asphyxiation.
A study in which blood gases were measured in
beaver during submersion sheds more light on the
drowning issue, especially as it relates to furbear-
ers. After venous and arterial catheterization to
sample blood, European beaver (Castor fiber) were
forcefully submerged in water for up to 10 minutes
(Clausen and Ersland 1970). From the authors' fig-
ures, the following conclusions can be drawn.
Throughout the period of submersion, PaCO2
increased but never exceeded 100 mm Hg; it took
7.5 minutes of submersion before PaCO2 exceeded
95 mm Hg. The PaO2 rapidly decreased during the
first 7 minutes of submersion, but both PaO2 and
arterial hemoglobin saturation with oxygen were at
hypoxemic levels (PaO2<50 mm Hg and satura-
tion<50%) within 5 minutes from the start of sub-
mersion. Thus the beavers were hypoxemic 2-3
minutes before PaCO2 reached 95 mm Hg.
The method by which great CO2 concentrations
kill animals is anesthesia-induced respiratory arrest
and the ensuing tissue hypoxia-anoxia (Mullenax
and Dougherty 1963,Andrews et al. 1993). In fact,
the time to death is prolonged when oxygen is used
with CO2. When a gas mixture consisting of
approximately 70% CO2, 24% N2 and 6% 02 was
used to kill mink, for example, the 5 test animals
survived for at least 15 minutes in the gas mixture
(Hansen et al. 1991). One animal died 6 minutes
after being removed from the gas mixture, but the
4 other animals fully recovered.
The preceding evidence demonstrates that in
drowning animals,
hypercarbia
lags behind hypoxia
and anoxia and that drowning animals die from
hypoxia and anoxia. All of this suggests that drown-
ing animals experience hypoxemia-induced dis-
comfort and distress before CO2
narcosis occurs, if
narcosis occurs at all. This raises the question: do
animals experience distress during drowning? For
the following reasons,
we believe that the answer is
yes. The classic stress response consists of changes
in heart rate and increases in blood pressures and
circulating blood levels of epinephrine and norepi-
nephrine and other stress-related hormones
(Moberg 1985). In rats breathing 100%
CO2 (CO2
anoxia), plasma norepinephrine increased signifi-
cantly and was released from the sympathetic nerv-
ous system and not the adrenal
medulla (Borovsky
et al. 1998). The authors concluded that the
response was mainly from hypoxia, not from CO2
in and of itself (Borovsky
et al. 1998).
In a model of asphyxia in which rats were stran-
gled (anoxic asphyxia), mean serum norepineph-
rine and epinephrine concentrations were signifi-
cantly greater in the strangled group compared to
the non-strangled group (norepinephrine=5.4+2.6
ng/mL vs. 2.8+0.1 ng/mL, P<0.001 and epineph-
rine=6.0+3.4 ng/mL vs. 3.8+3.0 ng/mL, P<0.05;
Hirvonen et al. 1997). The author concluded that
the data supported the idea that catecholamine
concentrations increased in blood upon suffocation
and could be used as indicators of hypoxia
(Hirvonen et al. 1997).
In dogs that were drowned with either cold salt
water (CSW) or cold fresh water (CFW),
epineph-
rine and norepinephrine concentrations (pg/mL)
increased significantly after immersion and contin-
ued to rise throughout the experimental period
(Conn et al.
1995). Prior
to immersion,
epinephrine
was 206?25 in the CFW
group and 133?67 in the
CSW
group. After 10 minutes of immersion, it had
risen to 174,650+1,750 in the CFW group and
153,250+4,585 in the CSF
group. Prior to immer-
sion, norepinephrine was 224?46 in the CFW
group and 374+182 in the CSW
group, and by 10
minutes it had reached 63,025?4,946 in the CFW
group and 50,400+1,796 in the CSF group. The
authors noted that though the greater values
reported in their study could be partly
attributed
to
sudden cold stress that has been described after
cold-water immersion, a more important etiological
factor is likely to be anoxic-ischemic stress pro-
ducing a catecholamine surge (Conn et al. 1995).
Drowning,
euthanasia,
and
carbon-dioxide
narcosis
* Ludders
et al. 669
Thus, the accumulated data indicate that hypoxia-
anoxia readily elicit the stress response in a variety
of animal species.
To summarize, data from several studies and a
variety of animal species indicate that CO2 can
produce narcosis, but only at partial pressures in
arterial blood exceeding 95 mm Hg. Furthermore,
data from rats and dogs suggest that a level of CO2-
induced narcosis sufficient to render an animal
insensible to the discomfort, anxiety, and stress
associated with hypoxemia is probably above 123
mm Hg; true C02-induced anesthesia, and thus
insensibility, does not occur until PaCO2 exceeds
200 mm Hg.
We recognize that drowning has been a tradi-
tional wildlife management technique, especially
for trapping aquatic mammals such as beaver,
muskrat, nutria (Myocastor coypus), mink, and river
otters (Lontra canadensis). In some states, trap-
pers have been encouraged to drown non-aquatic
mammals captured in cage traps, including rac-
coons (Procyon lotor), striped skunks (Mephitis
mephitis), and opossums (Didelphis virginiana).
Drowning is a method of killing animals that is con-
venient for humans. However, the concept of
euthanasia is independent of traditions and con-
venience, and drowning can not be considered
euthanasia. As we noted at the beginning of this
article, euthanasia is a "good death" that occurs
without pain or distress. Time is an important ele-
ment in euthanasia, and any technique that requires
minutes rather than seconds to produce death can
not be considered euthanasia. We encourage
wildlife administrators, researchers, animal care and
use committees, managers, and trappers to consid-
er these findings as they develop wildlife euthana-
sia technique guidelines and Best Management
Practices for Trapping (Proulx and Barrett 1989,
Friend et al. 1994, Hamilton et al. 1998).
Literature cited
ANDREWS,
E.J.,
B.T.
BENNETT,J.
D. CLARK,
K.A.
HOUPT,
P.J.
PASCOE,
G.
W ROBINSON,AND
J.
R.
BOYCE.
1993. Report of the AVMA
panel
on euthanasia. Journal of the American Veterinary
Medical
Association 202: 229-249.
BOROVSKY, V, M. HERMAN,
G. DUNPHY,A.
CAPLEA,
AND D. ELY.
1998.
C02 asphyxia increases plasma norepinephrine in rats via
sympathetic nerves. American Journal of Physiology 274:
R19-22.
CLAUSEN, G., AND A. ERSLAND.
1970. Blood 02 and acid-base
changes in the beaver during submersion. Respiration
Physiology 11: 104-112.
CONN,
A. W, K. MIYASAKA,
M. KATAYAMA,
M. FUJITA,
H. ORIMA,
G.
BARKER,
AND D. BOHN. 1995. A canine study of cold water
drowning in fresh versus salt water. Critical Care Medicine
23:2029-2037.
EISELE,J.
H.,
E.
I.
EGER,AND
M.
MUALLEM.
1967. Narcotic properties
of carbon dioxide in the dog. Anesthesiology 28:856-865.
FRIEND,
M.,
D.
E.TOWEILL,
R.
L.
BROWNELL,V.
E
NETTLES,
D. S.
DAVIS,AND
W J. FOREYT.
1994. Guidelines for proper care and use of
wildlife in field research.
Pages 96-105 in
T.A.
Bookhout, edi-
tor. Research and Management
Techniques for Wildlife and
Habitats.The
Wildlife Society, Bethesda,
Maryland.
GILBERT,
F E, AND N. GOFTON. 1982. Terminal dives in mink,
muskrat,
and beaver. Physiology and Behavior 28:835-840.
GUYTON,
A. C. 1991. Textbook of medical physiology. WB.
Saunders,
Philadelphia, Pennsylvania.
HAMILTON,
D. A., B. ROBERTS,
G. LINSCOMBE,
N. R. JOTHAM,
H.
NOSEWORTHY,AND
J.
L.
STONE.
1998.The European
Union's wild
fur regulation:
A battle of politics, cultures, animal rights,
international trade and North America's wildlife policy.
Transactions of the North American Wildlife and Natural
Resources Conference 63: 572-588.
HANSEN,
N. E.,A. CREUTZBERG,
AND
H. B. SIMONSEN.
1991. Euthan-
asia of mink (Mustela vison) by means of carbon dioxide
(C02), carbon monoxide (CO) and nitrogen (N2). British
Veterinary
Journal 147: 140-146.
HEWETT,T.A.,
M. S. KOVACS,J.
E.ARTWOHL,AND
B.T. BENNETT.
1993. A
comparison of euthanasia methods in rats, using carbon
dioxide in prefilled and fixed flow rate filled chambers.
Laboratory
Animal Science 43: 579-582.
HIRVONEN,J.,
M. L.
KORTELAINEN,AND
P.
HUTUNEN. 1997. Pulmonary
and serum surfactant phospholipids and serum cate-
cholamines in strangulation.
An experimental study on rats.
Forensic Science International
90:17-24.
MOBERG,
G.
P. 1985. Biological response to stress:
Key to assess-
ment of animal
well-being? Pages 27-49 in G.
P.
Moberg,
edi-
tor.Animal Stress.American
Physiological Society, Bethesda,
Maryland.
MULLENAX,
C. H.,AND
R.W. DOUGHERTY.
1963. Physiologic respons-
es of swine to high concentrations of inhaled carbon diox-
ide. American
Journal of Veterinary
Research 24: 329-332.
NEWMAN,A.
B.,AND R. D. STEWART.
1995. Submersion Incidents.
Pages 1209-1233 in P. S. Auerbach, editor. Wilderness
Medicine: Management of Wilderness and Environmental
Emergencies,
Mosby, St.
Louis,
Missouri.
PROULX,
G., AND
M.W. BARRETT.
1989. Animal welfare concerns
and wildlife trapping: Ethics, standards and commitments.
Transactions of the Western Section of the Wildlife Society
25: 1-6.
RAI,
U. C., P K. BHARDWAJ,AND
M. MOHAN.
1980. Effects of aspirat-
ed and swallowed water in mongrel dogs subjected to fresh-
water drowning. Indian Journal of Physiology and
Pharmacology 24:197-204.
STEDMAN'S MEDICAL
DICTIONARY.
1995. Twenty-sixth Edition.
Williams
and
Wilkins,
Baltimore,
Maryland.
STEPHENSON,
R., D. R. JONES,
AND R. M. BRYAN,
JR. 1994. Regional
cerebral blood flow during submergence asphyxia in Pekin
duck. American
Journal of Physiology 266: R1162-1168.
SUZUKI,
T 1996. Suffocation and related problems. Forensic
Science International
80:71-78.
TIMPERMAN,
J. 1972. The diagnosis of drowning. A review.
Forensic Science 1:397-409.
YAGIL, R., Z. ETZION,
AND A. OREN. 1983. The physiology of drown-
ing. Comparative Biochemistry and Physiology 74A:
189-193.
670 Wildlife Society Bulletin 1999, 27(3):666-670
John W. Ludders
(top photo) is - -
an associate professor
and chief
of section in anesthesiology in
the Department of Clinical i
Sciences at the College of
Veterinary Medicine, Cornell
University.
He received his B.S.
(zoology) and his D.V.M. from
Washington State University
and did his residency in veteri-
nary anesthesiology at the
University of California, Davis. :
His research interests are in
analgesia and anesthesia for
birds. j
.
Robert H. Schmidt (bottom
photo) is an associate profes-
sor in the Department of
Fisheries and Wildlife at Utah
State University. He received
his B.S. in natural resources
from Ohio State University;
an
M.S. in forestry,
fisheries, and
wildlife from the University of
Nebraska, Lincoln; and an
M.S. and Ph.D. in biological
ecology from the University
of
California, Davis. His inter-
ests cover the spectrum of
wildlife policy, ecology, and
management. Robert was president of the Western Section of
The Wildlife Society in 1989 and currently serves as president
of the National Animal Damage Control Association and as an
Executive Board member of the Wildlife Damage Management
Working Group of TWS.
F. Joshua Dein is animal welfare officer at the United States
Geological Service-Biological Research Division, National
Wildlife Health Center. Trained
as a biologist and a veterinari-
an, he is responsible for providing technical assistance to man-
agers and researchers in areas such as captive animal manage-
ment, capture and immobilization of wildlife, biological sample
collection, telemetry implantation, euthanasia, and disease
monitoring. He also has interests in electronic information
resources, moderating the Wildlife Health mailing list and the
Wildlife Health information Partnership. Patrice M. Klein is the
wildlife veterinarian for the Humane Society of the United
States (HSUS) and the veterinary director of the HSUS Wildlife
Rehabilitation Training Center, Cape Cod, Massachusetts. She
received her B.A. in biology from Hofstra
University in 1976, an
M.S. in pharmacology-toxicology from St. John's University,
New York in 1983, and her V.M.D. from the University of
Pennsylvania School of Veterinary Medicine in 1988. She has
extensive training and experience in pathology and is a diplo-
mate in the American College of Poultry
Veterinarians. From
1990 to 1995, Pat was the center veterinarian at the United
States Fish and Wildlife Service, Patuxent Wildlife Research
Center, Laurel, Maryland, where she was responsible for the
health management of endangered species of birds such as
whooping cranes and Mississippi sandhill cranes and evaluated
the effects of environmental pollutants on avian species. She is
currently working with Humane Society International
on inter-
national wildlife rehabilitation programs in Central and South
America.
