Journal of Wildlife Diseases, 41(1), 2005, pp. 48–57
? Wildlife Disease Association 2005
AVIAN CHOLERA IN WATERFOWL: THE ROLE OF LESSER SNOW
AND ROSS’S GEESE AS DISEASE CARRIERS IN THE PLAYA
Michael D. Samuel,1,2,4Daniel J. Shadduck,1Diana R. Goldberg,1and William P. Johnson3
1US Geological Survey, National Wildlife Health Center, 6006 Schroeder Road, Madison, Wisconsin 53711, USA
2Wisconsin Cooperative Wildlife Research Unit, 204 Russell Laboratories, 1630 Linden Drive,
Madison, Wisconsin 53706, USA
3Texas Parks and Wildlife Department, PO Box 659, Canyon, Texas 79015, USA
4Corresponding author (email: email@example.com)
(USA) during the winters of 2000–01 and 2001–02 to determine whether carriers of Pasteurella
multocida, the bacterium that causes avian cholera, were present in wild populations. With the
use of methods developed in laboratory challenge trials (Samuel et al., 2003a) and a serotype-
specific polymerase chain reaction method for identification of P. multocida serotype 1, we found
that a small proportion of 322 wild birds (?5%) were carriers of pathogenic P. multocida. On the
basis of serology, an additional group of these birds (?10%) were survivors of recent avian cholera
infection. Our results confirm the hypothesis that wild waterfowl are carriers of avian cholera and
add support for the hypothesis that wild birds are a reservoir for this disease. In concert with
other research, this work indicates that enzootic infection with avian cholera occurs in lesser snow
goose (Chen caerulescens caerulescens) populations throughout their annual cycle. Although fewer
Ross’s geese (Chen rossii) were sampled, we also found these birds were carriers of P. multocida.
Even in the absence of disease outbreaks, serologic evidence indicates that chronic disease trans-
mission and recent infection are apparently occurring year-round in these highly gregarious birds
and that a small portion of these populations are potential carriers with active infection.
Avian cholera, carriers, Chen caerulescens caerulescens, Chen rossii, lesser snow
geese, Pasteurella multocida, Ross’s geese, wild waterfowl.
We collected samples from apparently healthy geese in the Playa Lakes Region
Avian cholera, caused by the bacterium
Pasteurella multocida, was initially report-
ed in North America in domestic poultry
during the late 1800s (Botzler, 1991) and
in wild waterfowl in Texas and California
(USA) in 1944 (Quortrup et al., 1946; Ro-
sen and Bischoff, 1949). Since that time,
the disease has been reported in wild birds
from all flyways and has become the most
important infectious disease affecting
North American waterfowl (Friend, 1999).
Avian cholera outbreaks occur almost an-
nually as acute outbreaks at waterfowl con-
centration areas in the Central Valley of
California, the Rainwater Basin of Nebras-
ka, areas of Texas and Minnesota, and
western Canada (Wobeser et al., 1982;
Botzler, 1991) and in snow goose (Chen
caerulescens caerulescens) breeding colo-
nies (Samuel et al., 1999a). In addition,
chronic transmission and infection might
occur year-round in some snow goose pop-
ulations (Samuel et al., 1999b). During
avian cholera outbreaks, it is not uncom-
mon for thousands of waterfowl to die
(Brand, 1984; Botzler, 1991; Samuel et al.,
1999a), with mortality from the most se-
vere epizootics exceeding 20,000 birds
(Brand, 1984; Friend, 1999; Samuel et al.,
Two competing hypotheses have been
proposed to explain the recurrent pattern
of avian cholera outbreaks: 1) P. multocida
persists in specific wetlands year-round in
water, soil, or other reservoirs and 2) wa-
terfowl carriers of P. multocida initiate dis-
ease outbreaks as migratory birds congre-
gate in staging and wintering areas. The
absence of P. multocida in these wetland
ecosystems provides evidence that wet-
lands are not a likely reservoir for cholera
outbreaks (Samuel et al., 2004). In con-
trast, researchers have argued that carrier
birds (healthy birds with virulent P. mul-
tocida) might serve as reservoirs (long-
term source) for the bacteria and initiate
SAMUEL ET AL.—AVIAN CHOLERA CARRIERS IN THE PLAYA LAKES REGION49
sas, New Mexico, Oklahoma, and Texas (USA).
Playa Lakes region of Colorado, Kan-
epizootics (Botzler, 1991; Samuel et al.,
2003a); however, convincing evidence that
waterfowl are an important reservoir has
also been weak. Serologic samples ob-
tained from snow geese at breeding colo-
nies on Wrangel Island (Russia) and at
Banks Island (Canada) when avian cholera
outbreaks did not occur indicated that
birds had been recently (within 3–4 mo)
infected with P. multocida and survived in-
fection (Samuel et al., 1999b). These sum-
mer outbreaks in colonial nesting species
might help maintain the disease cycle
through carrier birds (Botzler, 1991; Wob-
eser, 1992; Samuel et al., 1999b). Although
serologic techniques might be useful in as-
sessing recent infection from P. multocida
(Samuel et al., 1999b), they only suggest
that wild birds that survive infection could
retain bacteria and become potential car-
riers of P. multocida, which could be trans-
mitted later to susceptible birds. Pasteu-
rella multocida obtained from oral swab
samples showed that carrier snow geese
occurred (Samuel et al., 1997), but prev-
alence of detected carriers was very low
(Samuel et al., 1999b).
Previous investigators have suspected
that lesser snow and Ross’s geese (Chen
rossii) (light geese) could be the primary
source of P. multocida because outbreaks
have been associated with their movement
and distribution patterns (Wobeser et al.,
1979, 1982; Brand, 1984; Samuel et al.,
2001), they suffer outbreaks and chronic
mortality every year (Mensik and Samuel
1995), outbreaks that might perpetuate
the disease cycle occur in snow goose
breeding areas (Samuel et al., 1999a), and
the magnitude of mortality in other species
has been associated with snow goose mor-
tality. This evidence suggests that snow
and Ross’s geese, and possibly other spe-
cies, could serve as a reservoir for P. mul-
tocida. However, to confirm that carrier
birds occur and are important in the trans-
mission of avian cholera, additional re-
search was needed that focused on isolat-
ing live P. multocida from the tissues of
birds and determining how and when the
organism is transmitted to susceptible
birds (Samuel and Mensik, 2000). The ob-
jectives of this study were to determine
whether lesser snow and Ross’s geese are
carriers of P. multocida, the frequency of
carriers in healthy bird populations, and
what tissues should be sampled to suc-
cessfully recover the bacterium.
MATERIALS AND METHODS
Our research targets primarily lesser snow
geese of the western Central Flyway which
breed in the western Canadian Arctic (Kerbes
et al., 1999) and are often involved in avian
cholera outbreaks in wintering and migration
areas. These geese winter in the Playa Lakes
Region of the United States (Fig. 1), an area
encompassing 364,000 km2in northwestern
Texas and adjacent portions of eastern New
Mexico, southeastern Colorado, southwestern
Kansas, and western Oklahoma (Playa Lakes
Joint Venture, 1994). This area has a semiarid
climate and is relatively flat and open, and pla-
ya lakes are the dominant wetland habitat fea-
ture. The region is affected heavily by agricul-
ture, including both cultivation and grazing.
Detailed descriptions of the area are found in
reports by the US Fish and Wildlife Service
(1988), Bolen et al. (1989a, b), and Haukos and
Smith (1992). The Playa Lakes Region is im-
portant to migrating and wintering waterfowl in
the Central Flyway (Buller, 1964; Bolen et al.,
1989a), and light geese are increasing in this
area (Ray and Miller, 1997). Since the 1970s,
estimates of light geese in this region have in-
creased from less than 100 to more than
150,000 in 1999 (Texas Parks and Wildlife De-
partment, unpubl. data). These increases have
been especially apparent during the last 10 yr,
when harvest estimates in the Playa Lakes Re-
50 JOURNAL OF WILDLIFE DISEASES, VOL. 41, NO. 1, JANUARY 2005
gion have averaged over 12,000 light geese per
Field collection and testing
Light geese from the Playa Lakes Region
were collected during the winters of 2000–01
and 2001–02. Geese were collected primarily
by shooting live, apparently healthy birds. A
blood sample for antibody analysis was collect-
ed from the body cavity or heart of each carcass
within 8 hr and centrifuged (10–15 min at
1,500 ? G). Sera were removed and frozen at
?20 C before shipment on dry ice to the Na-
tional Wildlife Health Center (NWHC; Madi-
son, Wisconsin, USA), where they were stored
frozen until processing. On the basis of a lab-
oratory study of mallards (Anas platyrhynchos;
Samuel et al., 2003a), two swab samples were
taken for P. multocida culture from the oral,
nasal, and cloacal cavities; leg joint; and eye
surface from each goose collected. Swabs were
swirled in cryovials containing 1.25 ml of either
brain heart infusion (BHI) broth (Difco Labo-
ratories, Detroit, Michigan, USA) or 10% di-
methylsulfoxide (DMSO), and swab tips were
broken off into the vial. The cryovials were
stored in liquid nitrogen (?210 C) within 2 hr
of sampling and sent in a liquid nitrogen dry
shipper to the NWHC for subsequent labora-
Frozen samples were thawed at room tem-
perature (16–20 C), and the contents of each
vial were transferred into a culture tube con-
taining 5 ml of BHI broth. Each tube was in-
cubated for 2–2.5 hr at 37 C with shaking (100
revolutions per minute) for pre-enrichment.
Following incubation, 2 ml of the broth was
transferred to a tube containing 5 ml of P. mul-
tocida selective broth (PMSB; Moore et al.,
1994) and incubated for 12–16 hr at 37 C with
5–10% CO2. Following this second incubation,
a portion of the liquid culture was streaked
onto a blood agar plate (BAP; Becton Dickin-
son, Sparks, Maryland, USA). The BAP were
incubated for 20–24 hr at 37 C with 5–10%
CO2and examined for colonies resembling P.
multocida. Suspect P. multocida colonies were
examined by Gram stain, serotyped by the aga-
rose gel precipitin test (Heddleston et al.,
1972), and identified with the analytical profile
index (API) 20E identification system (bio-
Merieux, St. Louis, Missouri, USA) as indicated
by Samuel et al. (2003b).
Each P. multocida isolate was tested for vir-
ulence in four adult mallard ducks (Whistling
Wings, Hanover, Illinois, USA) by subcutane-
ously injecting 0.2 ml of the challenge inocula
(6.2 ? 105to 1.7 ? 106cells) in the dorsal
caudal region of the neck (Samuel et al.,
2003b), a method previously used to standard-
ize our challenge trials with similar results to
naturally infected ducks (Samuel et al., 2003a).
Birds challenged with serotype 1 isolates were
housed in individual stainless steel cages (76 ?
61 ? 41 cm; Lab Products, Inc., Aberdeen,
Maryland) to reduce the risk of disease trans-
mission among birds. Mallards challenged with
non–serotype 1 isolates were maintained to-
gether, but in a separate 22-m2isolation room
on 1.3 ? 2.5 cm gauge diamond Tenderfoot?
(Tandem Products, Inc., Minneapolis, Minne-
sota, USA) because mortality was not expected
from these isolates. Ducks were observed for
illness, morbidity, and mortality for 7 days, and
challenged survivors were euthanized by cer-
vical dislocation. The livers of all birds that died
or were euthanized were cultured for P. mul-
tocida, and isolates were identified by API 20E
and serotyped as described in the previous par-
We performed a polymerase chain reaction
(PCR) analysis on 1 ml of PMSB cultures that
were frozen at the same time initial culturing
was started. We used the PCR procedure de-
scribed by Rocke et al. (2002) with the follow-
ing modifications. Briefly, PMSB enrichment
culturing time was increased to 12–16 hr, and
normal flora from the oral cavity of the mallard
ducks, consisting of five or six bacterial species,
was used as a negative control. We spiked the
negative control with a mixture of 25 P. mul-
tocida cultures to provide our positive test con-
trol and used a second positive control to en-
sure that DNA extraction procedures worked
correctly. Electrophoresis of the PCR products
was performed with the Horizon 11-14 gel
electrophoresis apparatus (Life Technologies,
Inc., Rockville, Maryland, USA).
Sera from wild geese were tested with the
same enzyme-linked immunosorbent assay
(ELISA) procedures described for mallard
ducks (Samuel et al., 2003a), with three chang-
es to improve test consistency. The antigen and
conjugate preparations were done gravimetri-
cally, rather than by pipetting. Room temper-
ature was standardized to 18.5 C ? 1.5 C,
which required a reduction of the substrate in-
cubation time to 35 min. These procedural
changes were tested with the use of our snow
goose reference sera (Samuel et al., 1999b) to
ensure consistency in detecting positive anti-
body levels with our previous ELISA proce-
Fifty-five Ross’s and 266 snow geese
were collected in the Playa Lakes Region
and tested for P. multocida during January,
SAMUEL ET AL.—AVIAN CHOLERA CARRIERS IN THE PLAYA LAKES REGION51
Ross’s and snow geese collected for Pasteurella multocida isolation from the Playa Lakes Region (USA) in January–March 2001 and 2002.
March 2001 and March 2002
February–March 2001February 2001 and February–March 2002
March 2001 and February 2002
February–March 2001 and January–March 2002
aPasteurella multocida serotype 1 was isolated from several swabs from one bird.
bPasteurella multocida serotype 1 was isolated from one swab from one bird.
cPasteurella pneumotropica/haemalytica isolated from one swab from one bird.
dPasteurella multocida (three cultures, three distinct serotypes: 8/13, 3/4, and 3) was isolated from one swab from one bird.
ePasteurella multocida serotype 1 was isolated from one swab from each of two birds.
fPasteurella multocida serotype 3/12 was isolated from two swabs from one bird.
gAn additional negative sample was collected from an adult snow goose in Texas; however, gender was listed as undetermined for that bird, so it was not included in this table.
February, and March 2001 and 2002 (Ta-
ble 1). We recovered 11 P. multocida se-
rotype 1 isolates from five geese. Six se-
rotype 1 isolates were recovered from a
single adult male Ross’s goose from Texas
in 2001; the bacterium was isolated from
nasal, oral, cloacal, and both leg joint sam-
ples. Three serotype 1 isolates were recov-
ered from two adult male snow geese from
Colorado in 2002: from an eye swab from
one goose and the cloacal and leg joint
swabs from the second bird. Serotype 1
isolates were recovered from two adult fe-
male snow geese from Texas in 2001, from
a nasal swab, and from an oral swab. Other
P. multocida serotypes were also isolated
from an adult snow goose collected in
Kansas in 2001 and Texas in 2002 (Table
1); however, these serotypes are not com-
monly pathogenic in waterfowl (Friend,
1999). Estimated prevalence of carriers in
light geese on the basis of culture samples
was approximately 2% in Texas and 4% in
Colorado, compared with 0% in the re-
maining states. Population prevalence was
2% in Ross’s geese compared with 2% in
snow geese. Prevalence of carriers was 2%
in adult geese compared with 0% in im-
mature geese and 2% in males compared
with 1% in females. The PCR and cultur-
ing techniques produced similar results; all
culture-positive samples were PCR posi-
tive. The PCR analysis detected one ad-
ditional positive bird from oral swabs tak-
en from an immature female snow goose
from Texas in 2002.
We recovered at least one P. multocida
serotype 1 isolate from each of the tissues
we sampled in wild geese (Table 2). We
also recovered the bacterium from swabs
preserved in both BHI and DMSO. Se-
rotype 1 isolates were recovered from both
preservation methods (BHI and DMSO)
for two tissues from the positive Ross’s
goose. All other serotype 1 isolates were
recovered from either the DMSO or BHI
swabs, but not both. The number of pos-
itive samples was insufficient to determine
which preservation method or tissue swab
was preferred for detection of carriers.
52 JOURNAL OF WILDLIFE DISEASES, VOL. 41, NO. 1, JANUARY 2005
Playa Lakes Region (USA) in January–March 2001 and 2002. Swab samples were preserved in brain heart
infusion broth (BHI) and dimethylsulfoxide (DMSO) and processed with Pasteurella multocida selective broth
after liquid nitrogen storage. Pathogenicity trials were conducted by challenging four mallard ducks with each
Pasteurella multocida isolate recovered.
Pasteurella multocida serotype 1 isolates recovered from Ross’s and lesser snow geese from the
P. multocida serotype 1 isolates
No. of birds sampled
aNumber of serotype 1 isolates killing one or more of four mallards/number of isolates tested.
bPasteurella multocida serotype 1 isolates were recovered from BHI and DMSO swabs for a Ross’s goose.
cThree non–serotype 1 isolates were also recovered from a BHI swab in one bird. These isolates were nonpathogenic in
dNon–serotype 1 P. multocida isolates were recovered from BHI and DMSO swabs from one bird. One of these isolates was
nonpathogenic in mallard ducks; the other was not tested.
eNumber of birds sampled, with ?1 recovered P. multocida serotype 1 isolates and with ?1 pathogenic isolate.
Ten of the 11 P. multocida serotype 1
isolates recovered from wild birds were
tested for pathogenicity in mallards; we ex-
cluded one of two leg joint isolates recov-
ered from a Ross’s goose. Mortality oc-
curred in six of the 10 challenge groups
(Table 2). We tested five of the six isolates
recovered from the Ross’s goose from Tex-
as. The oral swab isolate in DMSO killed
three ducks, and the BHI isolate killed
one; the leg joint isolate in BHI killed two
ducks, and the isolate in DMSO was not
tested; the isolate from the nares (DMSO)
killed one duck. No mortality occurred in
the challenge group inoculated with the
cloacal isolate. The isolate from a snow
goose oral swab (DMSO) collected in Tex-
as caused mortality in three ducks, but the
isolate from the nares (BHI) of the other
Texas snow goose was nonpathogenic. Two
isolates from one snow goose collected in
Colorado were tested; the cloacal (DMSO)
isolate killed one duck, and no deaths oc-
curred from the leg joint (DMSO) isolate.
No mortality occurred in the challenge
group inoculated with the eye swab (BHI)
isolate from Colorado. Pasteurella multo-
cida serotype 1 was recovered from all the
ducks that died during the challenge trial,
and three ducks surviving challenge with
serotype 1 isolates (an indication of infec-
tion), including one from a group in which
no mortality occurred. No mortality or ill-
ness was observed in any of the ducks in-
oculated with the non–serotype 1 isolates
(one serotype 3, one serotype 3/4, one se-
rotype 3/12, and one serotype 8/13), and
no P. multocida was recovered from these
groups on necropsy.
(r?0.84, n?96, P?0.0001) between our
previous antibody titer levels measured for
our reference sera from Bank’s Island
snow geese (Samuel et al., 1999b) and the
antibody titers from the ELISA test used
in this study. A threshold ELISA value of
17.5 to indicate seroconversion in this
study provided good agreement (kappa ?
0.75) between the two ELISA procedures.
This threshold produced a slightly lower
seropositive prevalence (41% positive)
from the revised ELISA procedure than
we found for our original ELISA (49%) in
the 96 snow goose reference samples.
Recent exposure to P. multocida on the
basis of positive ELISA antibody titer was
indicated in three of the 52 Ross’s geese
and eight of the 266 snow geese sampled
SAMUEL ET AL.—AVIAN CHOLERA CARRIERS IN THE PLAYA LAKES REGION53
Serum samples collected from Ross’s and lesser snow geeseafor Pasteurella multocida antibody from the Playa Lakes Region (USA) in January–March
2001 and 2002.
March 2001 and March 2002
February–March 2001 and February–March 2002February 2001 and February–March 2002
March 2001 and February 2002
February–March 2001 and January–March 2002
aF ? female; M ? male.
bA positive antibody response to P. multocida was found for one bird in this group.
cA positive antibody response to P. multocida was found for two birds in this group.
dAn additional negative sample was collected from an adult snow goose in Texas; however, gender was listed as undetermined for that bird, so it was not included in this table.
(Table 3). Positive Ross’s geese included
one adult female from New Mexico and
one adult female and one immature male
from Texas. Positive snow geese included
one adult male from Kansas; one adult
male from New Mexico; and two adult
males, two adult females, one immature
male, and one immature female from Tex-
as. As expected, the estimated seropreva-
lence of geese to P. multocida was low: 0%
in Colorado, 3% in Kansas, 4% in New
Mexico, 0% in Oklahoma, and 5% in Tex-
as. Seroprevalence was 6% in Ross’s geese
compared with 3% in snow geese; 3% in
adults compared with 4% in juveniles, and
4% in males compared with 3% in fe-
males. The number of samples collected
by species, age, sex, or state was insuffi-
cient to statistically assess differences in
prevalence. Pasteurella multocida was not
isolated from any of the geese with posi-
tive ELISA antibody titers.
Samuel et al. (2003a) found that swab
samples (oral, nasal, cloacal, eye, and leg
joint) preserved in liquid nitrogen with ei-
ther BHI or DMSO and cultured in
PMSB provided the most successful meth-
od for recovering P. multocida. On the ba-
sis of these methods, wild geese in the Pla-
ya Lakes Region were collected and sam-
pled for P. multocida during the winters of
2000–01 and 2001–02. Swab cultures from
these birds confirmed that wild, apparently
healthy lesser snow and Ross’s geese can
be carriers of P. multocida and therefore
can transmit the bacterium that causes avi-
an cholera to other birds and spread the
disease to other locations. A persistent is-
sue in the epizootiology of avian cholera
has been the identification of a reservoir
for the disease agent. Our study provides
the strongest available evidence that birds
serve as a reservoir for avian cholera.
These findings suggest that further re-
search might be warranted to determine
how long birds can carry P. multocida and
how the bacterium is transmitted to sus-
54JOURNAL OF WILDLIFE DISEASES, VOL. 41, NO. 1, JANUARY 2005
We found no association between P.
multocida culture-positive and serologic-
positive geese. In a previous laboratory
challenge study, we only found a weak as-
sociation between the detection of carriers
and seroconversion (Samuel et al., 2003a).
In this and previous studies, we found that
detectable antibodies peaked approximate-
ly 2 wk postinfection and declined to be-
low background levels within 3–4 mo;
thus, seroconversion represents recent in-
fection with P. multocida. In challenged
ducks, we found that 90% seroconverted
within 2 wk postinfection (Samuel et al.,
2003a). We suspect there are several pos-
sible reasons that none of our carrier birds
had seroconverted: 1) these birds were
commensal carriers of P. multocida but
had not become infected, 2) carriers had
been infected for at least 3–4 mo and an-
tibody titers had declined below the de-
tectable threshold, and 3) carriers were re-
cently infected with P. multocida and had
not seroconverted. We suspect the latter
explanation is less probable because most
birds would have detectable antibody
within 1–2 wk postinfection. Further re-
search is needed to understand the rela-
tionship between infection, seroconver-
sion, and carrier status.
As in previous studies we conducted
with P. multocida isolates (Samuel et al.,
2003b), none of the non–serotype 1 iso-
lates were virulent in ducks. Six of the 10
serotype 1 isolates we recovered from wild
birds were pathogenic to mallards (at least
one bird of four died). The remaining four
isolates caused some illness, but no mor-
tality. Four of the five isolates tested from
the Ross’s goose, all except the cloacal
swab, were pathogenic. Two of five isolates
tested from the snow geese were patho-
genic. The cloacal isolate from a snow
goose collected in Colorado was pathogen-
ic, but not the leg joint isolate from this
bird. Isolates from oral swabs were all
pathogenic, but pathogenicity varied for
cloacal, nasal, and leg joint swabs (one of
two pathogenic for each). The single iso-
late from the eye swab was nonpathogenic.
At present, we are uncertain why some of
the serotype 1 isolates were not pathogen-
ic in ducks. However, previous studies
have concluded that pathogenicity of P.
multocida strains can be highly variable
(Christensen and Bisgaard, 2000) and that
pathogenicity increases over a short time
with repeated transmission (Matsumoto
and Strain, 1993). We did not attempt to
further characterize these isolates or to
test isolates recovered from our challenged
Although P. multocida isolates have
been recovered previously from wild wa-
terfowl (see Botzler, 1991; Wobeser, 1992),
only Samuel et al. (1997) determined the
serotypes of the isolates and demonstrated
that the isolates were virulent. Samuel et
al. (1997) reported recovering only a single
P. multocida serotype 1 isolate from oral
swabs of more than 3,400 adult snow geese
(Samuel et al., 1999b). On the basis of re-
sults presented in this study, we suspect
the sampling and preservation methods
used by Samuel et al. (1997) were insuf-
ficient to ensure adequate isolation of P.
multocida in carrier birds. The methods
we developed appear suitable for detect-
ing avian cholera carriers among wild wa-
terfowl; however, we also suspect the sam-
pling and preservation methods developed
in this study will fail to detect all birds that
might be carriers. For most of the tissues
we sampled for P. multocida isolates, we
were only successful in recovering the bac-
terium in one of the two swabs collected
(seven of nine tissues). Fortunately, in
most of the positive geese (four of five),
we were successful at recovering P. mul-
tocida in tissues (oral, nasal, cloacal) from
which collection would be easy in live or
dead birds. To increase the probability of
detecting P. multocida in carrier water-
fowl, we recommend testing at least two
samples from a variety of tissues, with
preservation at low temperatures in either
BHI or DMSO.
Polymerase chain reaction provided
only minimal enhancement of our ability
to detect P. multocida organisms in our
SAMUEL ET AL.—AVIAN CHOLERA CARRIERS IN THE PLAYA LAKES REGION55
field samples. As a result, we only recom-
mend using PCR as a potential supple-
ment to culture when the primary goal is
to determine prevalence or number of car-
riers. The PCR alone would not be suit-
able if viable isolates were desired for vir-
ulence testing as recommended by Samuel
et al. (2003b), if isolates were desired for
further characterization, or if pre-enrich-
ment with PMSB was not possible.
Avian cholera is of particular concern to
wildlife managers because most species of
waterfowl, raptors, and other birds of wet-
land ecosystems are susceptible (Botzler,
1991; Friend, 1999). Although the factors
that trigger an outbreak are poorly under-
stood, it is commonly believed that weath-
er, stress, and high densities of susceptible
birds are important contributors (Botzler,
1991; Windingstad et al., 1998). Increased
densities of waterbirds, especially gregari-
ous light goose species, probably increase
the risk of disease transmission and out-
break events (Wobeser, 1992). Once an
outbreak starts, wetland contamination
from diseased birds is the primary source
of infection to susceptible birds of all spe-
cies, although other routes of transmission
such as bird-to-bird contact are likely
(Wobeser, 1992). Our research demon-
strates that some species of waterfowl, es-
pecially light geese, are carriers of P. mul-
tocida and might be more disposed to avi-
an cholera outbreaks that concurrently or
subsequently affect other susceptible spe-
cies. In addition, the increased abundance
of light geese and the large-scale mixing of
these populations could enhance the ex-
change and spread of avian cholera and
other disease agents (Wobeser, 1992). Loss
of habitat, increased abundance of light
geese and other waterfowl, and increased
densities of waterbirds are all factors that
likely contribute to increasing the risk of
avian cholera outbreaks, increasing the risk
of infecting other waterbirds in the same
wetlands, and increasing the continental
distribution of this infectious disease.
Current management strategies to con-
trol avian cholera losses are reactive, con-
sisting primarily of collecting and dispos-
ing of carcasses when outbreaks occur
(Wobeser, 1992). Development of proac-
tive or alternative disease management ap-
proaches to avian cholera has likely been
hindered by uncertainty about the reser-
voir for the disease. The apparent impor-
tance of snow geese as carriers of avian
cholera, coupled with the dramatic in-
crease in abundance of midcontinent snow
goose populations (Ankney, 1996; Abra-
ham and Jefferies, 1997) and their propen-
sity to occur in large aggregations through-
out the year, amplifies the potential role of
this species in avian cholera outbreaks. Al-
though the proportion of P. multocida car-
riers in the midcontinent light goose pop-
ulation is relatively low (3–5%), these pop-
ulations likely exceed 2.5 million birds
(Kelley et al., 2001), containing an esti-
mated 75,000–125,000 P. multocida carri-
ers. Results from our research have impli-
cations for other areas in which snow
geese occur in large numbers (e.g., Cali-
fornia, coastal regions of Texas and Loui-
siana) and for areas in which avian cholera
epizootics are problematic in waterfowl
(e.g., Nebraska’s Rainwater Basin). Strat-
egies for prevention and control of avian
cholera should consider that carrier birds
are a likely source of disease outbreaks and
disease spread. Management actions that
decrease potential disease transmission by
separating light geese (and other carrier
species) from other species, reducing
stress factors that might precipitate epi-
zootic events, and reducing high concen-
trations of waterfowl on a limited number
of wetlands might minimize the effect of
avian cholera on waterfowl populations.
Because avian cholera affects many water-
bird species, further research is needed to
determine whether other species, such as
white-fronted geese (Anser albifrons) and
northern pintails (Anas acuta), frequently
associated with avian cholera outbreaks
can also serve as a reservoir for this disease.
We appreciate the dedicated assistance of
many individuals who contributed to this proj-
56JOURNAL OF WILDLIFE DISEASES, VOL. 41, NO. 1, JANUARY 2005
ect. M. O’Meilia, J. Neal, J. Gammonley, T.
Sanders, H. Hands, B. Price, A. Price, M. Sex-
son, K. Sexson, C. Swank, S. Baugh, T. Unruh,
M. Torres, D. Fisher, T. Mitchusson, S. Denny,
D. Cook, J. Zotter, J. Gentz, M. Gentz, and the
Garden City Zoo assisted with collecting sam-
ples from wild geese. Funding for this study
was provided by the US Fish and Wildlife Ser-
vice, the Playa Lakes Joint Venture, Texas Parks
and Wildlife, Ducks Unlimited, and the US
Geological Survey. We thank J. Blanchong, D.
Blehert, R. Botzler, and an anonymous referee
for manuscript review.
ABRAHAM, K. F., AND R. L. JEFFERIES. 1997. High
goose populations: Causes, impacts, and impli-
cations. In Arctic ecosystems in peril: Report of
the arctic goose habitat working group, B. D. J.
Batt (ed.). Arctic goose joint venture special pub-
lication. US Fish and Wildlife Service, Washing-
ton, D.C., and Canadian Wildlife Service, Otta-
wa, Ontario, Canada, pp. 7–72.
ANKNEY, C. D. 1996. An embarrassment of riches:
Too many geese. Journal of Wildlife Manage-
ment 60: 217–223.
BOLEN, E. G., L. M. SMITH, AND H. L. SCHRAMM,
JR. 1989a. Playa lakes: Prairie wetlands of the
Southern High Plains. BioScience 39: 615–623.
, G. A. BALDASSARRE, AND F. S. GUTHERY.
1989b. Playa lakes. In Habitat management for
migrating and wintering waterfowl in North
America, L. M. Smith, R. L. Pederson, and R.
M. Kaminski (eds.). Texas Tech University Press,
Lubbock, Texas, pp. 341–365.
BOTZLER, R. G. 1991. Epizootiology of avian cholera
in waterfowl. Journal of Wildlife Diseases 22:
BRAND, C. J. 1984. Avian cholera in the Central and
Mississippi Flyways during 1979–80. Journal of
Wildlife Management 48: 399–406.
BULLER, R. J. 1964. Central flyway. In Waterfowl
tomorrow, J. P. Linduska (ed.). US Fish and
Wildlife Service, Washington, D.C., pp. 209–232.
CHRISTENSEN, J. P., AND M. BISGAARD. 2000. Fowl
cholera. Revue Scientifique et Technique de Le
Office International Des Epizooties 19: 626–637.
FRIEND, M. 1999. Avian cholera. In Field manual of
wildlife diseases, general field procedures and
diseases of migratory birds, M. Friend and J. C.
Franson (eds.). US Geological Survey, Informa-
tion and Technology Report 1999–001, pp. 74–
HAUKOS, D. A., AND L. M. SMITH. 1992. Ecology of
playa lakes. In Waterfowl management hand-
book. US Department of the Interior, Fish and
Wildlife Service, Fish and Wildlife Leaflet
HEDDELSTON, K. L., J. E. GALLAGER, AND P. A RE-
BERS. 1972. Fowl cholera: Gel diffusion precip-
itin test for serotyping Pasteurella multocida
from avian species. Avian Diseases 16: 925–936.
KELLEY, J. R., D. C. DUNCAN, AND D. R. YPARRA-
GUIRRE. 2001. Distribution and abundance. In
The status of Ross’s geese, T. J. Moser (ed.). Arc-
tic goose joint venture special publication. US
Fish and Wildlife Service, Washington, D.C., and
Canadian Wildlife Service, Ottawa, Canada, pp.
KERBES, R. H., K. M. MEERES, AND J. E. HINES,
eds. 1999. Distribution, survival and numbers of
lesser snow geese of the western Canadian Arctic
and Wrangel Island, Russia. Canadian Wildlife
Service, Occasional Paper 98, Ottawa, Ontario.
MATSUMOTO, M., AND J. G. STRAIN. 1993. Patho-
genicity of Pasteurella multocida: Its variable na-
ture demonstrated by in vivo passages. Avian
Diseases 37: 781–785.
MENSIK, G., AND M. D. SAMUEL. 1995. Studying
snow geese and avian cholera. California Water-
fowl (October/November): 16.
MOORE, M. K., L. CICNJAK-CHUBBS, AND R. J.
GATES. 1994. A new selective enrichment pro-
cedure for isolating Pasteurella multocida from
avian and environmental samples. Avian Diseases
PLAYA LAKES JOINT VENTURE. 1994. Guidelines for
the implementation of the North American wa-
terfowl management plan in the Playa Lakes
Joint Venture. Special report, Playa Lakes Joint
Venture Coordinator, Albuquerque, New Mexi-
co, 25 pp.
QUORTRUP, E. R., F. B. QUEEN, AND L. J. MEROVKA.
1946. An outbreak of pasteurellosis in wild
ducks. Journal of the American Veterinary Med-
ical Association 108: 94–100.
RAY, J. D., AND H. W. MILLER. 1997. A concentra-
tion of small Canada geese in an urban setting at
Lubbock, Texas. Southwestern Naturalist 42: 68–
ROCKE, T. E., S. R. SMITH, A. MIYAMOTO, AND D. J.
SHADDUCK. 2002. A serotype-specific polymer-
ase chain reaction for identification of Pasteurella
multocida serotype 1. Avian Diseases 46: 370–
ROSEN, M. N., AND A. I. BISCHOFF. 1949. The
1948–1949 outbreak of fowl cholera in birds in
the San Francisco Bay area and surrounding
counties. California Fish and Game 35: 185–192.
SAMUEL, M. D., AND G. MENSIK. 2000. New in-
sights on a deadly disease. California Waterfowl
, D. R. GOLDBERG, D. J. SHADDUCK, J. I.
PRICE, AND E. G. COOCH. 1997. Pasteurella
multocida serotype 1 isolated from a lesser snow
goose: Evidence of a carrier state. Journal of
Wildlife Diseases 33: 332–335.
, J. Y. TAKEKAWA, G. SAMELIUS, AND D. R.
SAMUEL ET AL.—AVIAN CHOLERA CARRIERS IN THE PLAYA LAKES REGION 57
GOLDBERG. 1999a. Avian cholera mortality in
lesser snow geese nesting on Banks Island,
Northwest Territories. Wildlife Society Bulletin
, D. J. SHADDUCK, D. R. GOLDBERG, V. BAR-
ANYUK, L. SILEO, AND J. I. PRICE. 1999b. An-
tibodies against Pasteurella multocida in snow
geese in the western arctic. Journal of Wildlife
Diseases 35: 440–449.
, K. A. CONVERSE, AND K. J. MILLER. 2001.
Disease mortality events involving Ross’s geese.
In The status of Ross’s geese, T. J. Moser (ed.).
Arctic Goose Joint Venture special publication.
US Fish and Wildlife Service, Washington, D.C.,
and Canadian Wildlife Service, Ottawa, Canada,
, D. J. SHADDUCK, D. R. GOLDBERG, AND W.
P. JOHNSON. 2003a. Comparison of methods to
detect Pasteurella multocida in carrier waterfowl.
Journal of Wildlife Diseases 39: 125–135.
, D. J. SHADDUCK, D. R. GOLDBERG, M. A.
WILSON, D. O. JOLY, AND M. A. LEHR. 2003b.
Characterization of Pasteurella multocida isolates
from wetland ecosystems during 1996 to 1999.
Journal of Wildlife Diseases 39: 798–807.
, D. J. SHADDUCK, AND D. R. GOLDBERG.
2004. Are wetlands the reservoir for avian chol-
era? Journal of Wildlife Diseases 40: 377–382.
UNITED STATES FISH AND WILDLIFE SERVICE. 1988.
Playa Lakes Region waterfowl habitat concept
plan, category 24 of the North American water-
fowl management plan. US Department of the
Interior, Albuquerque, New Mexico. 37 pp.
WINDINGSTAD, R. M., M. D. SAMUEL, D. D. THORN-
BURG, AND L. C. GLASER. 1998. Avian cholera
mortality in Mississippi flyway Canada geese. In
Biology and management of Canada geese. D.
H. Rusch, M. D. Samuel, D. D. Humburg, and
B. D. Sullivan (eds), Proceedings of the Inter-
national Canada Goose Symposium. pp. 283–
WOBESER, G. 1992. Avian cholera and waterfowl bi-
ology. Journal of Wildlife Diseases 28: 674–682.
, D. B. HUNTER, B. WRIGHT, D. J. NEIMAN,
AND R. ISBISTER. 1979. Avian cholera in water-
fowl in Saskatchewan, spring 1977. Journal of
Wildlife Diseases 15: 19–24.
, F. A. LEIGHTON, A. D. OSBORNE, D. J. NIE-
MAN, AND J. R. SMITH. 1982. Avian cholera in
waterfowl in western Canada, 1978–1981. Ca-
nadian Field-Naturalist 96: 317–322.
Received for publication 2 April 2004.