EPIDEMIC PASTEURELLOSIS IN A BIGHORN SHEEP POPULATION

COINCIDING WITH THE APPEARANCE OF A DOMESTIC SHEEP

Janet L. George,1 Daniel J. Martin,2 Paul M. Lukacs,2 and Michael W. Miller2,3

1 Colorado Division of Wildlife, 6060 Broadway Street, Denver, Colorado 80216, USA2 Colorado Division of Wildlife, Wildlife Research Center, 317 West Prospect Road, Fort Collins, Colorado 80526-2097,USA3 Corresponding author (email: mike.miller@state.co.us)

ABSTRACT: A pneumonia epidemic reduced bighorn sheep (Ovis canadensis) survival andrecruitment during 1997–2000 in a population comprised of three interconnected wintering herds(Kenosha Mountains, Sugarloaf Mountain, Twin Eagles) that inhabited the Kenosha and TarryallMountain ranges in central Colorado, USA. The onset of this epidemic coincided temporally andspatially with the appearance of a single domestic sheep (Ovis aires) on the Sugarloaf Mountainherd’s winter range in December 1997. Although only bighorns in the Sugarloaf Mountain herdwere affected in 1997–98, cases also occurred during 1998–99 in the other two wintering herds,likely after the epidemic spread via established seasonal movements of male bighorns. In all, welocated 86 bighorn carcasses during 1997–2000. Three species of Pasteurella were isolated invarious combinations from affected lung tissues from 20 bighorn carcasses where tissues wereavailable and suitable for diagnostic evaluation; with one exception, b-hemolytic mannheimia(Pasteurella) haemolytica (primarily reported as biogroup 1G or 1aG) was isolated from lung tissuesof cases evaluated during winter 1997–98. The epidemic dramatically lowered adult bighornmonthly survival in all three herds; a model that included an acute epidemic effect, differingbetween sexes and with vaccination status, that diminished linearly over the next 12 mo bestrepresented field data. In addition to the direct mortality associated with epidemics in these threeherds, lamb recruitment in years following the pneumonia epidemic also was depressed ascompared to years prior to the epidemic. Based on observations presented here, pasteurellosisepidemics in free-ranging bighorn sheep can arise through incursion of domestic sheep onto nativeranges, and thus minimizing contact between domestic and bighorn sheep appears to be a logicalprinciple for bighorn sheep conservation.

Key words: Bighorn sheep, domestic sheep, epidemic, Mannheimia spp., Ovis canadensis,Pasteurella spp., pasteurellosis, survival, vaccine.

INTRODUCTION

Epidemics caused by Pasteurella spp.and other pathogens have plagued bighornsheep (Ovis canadensis) populations forat least a century, and have played asignificant role in declines of bighornpopulations throughout western NorthAmerica (Warren, 1910; Grinnell, 1928;Shillinger, 1937; Buechner, 1960; Post,1962; Miller, 2001). Disease epidemics inbighorn sheep were reported to followEuropean settlement and establishment ofdomestic livestock grazing (Warren, 1910;Grinnell, 1928; Shillinger, 1937). Thistrend may reflect historic introduction ofnovel pathogens (including some Pasteu-rella spp. strains) into naive bighornpopulations beginning in the late 1800s(Warren, 1910; Grinnell, 1928; Shillinger,1937; Marsh, 1938; Honess and Frost,

1942; Buechner, 1960; Goodson, 1982;Miller, 2001). Although pasteurellosis wasfirst diagnosed in free-ranging bighorns in1935 (Potts, 1937), earlier unconfirmedepidemics (e.g., Warren, 1910; Rush,1927; Grinnell, 1928; Honess and Frost,1942; Spencer, 1943; Buechner, 1960)seem likely (Post, 1962).

Native North American wild sheepspecies appear exquisitely susceptible topasteurellosis (Onderka and Wishart,1988; Onderka et al., 1988; Foreyt, 1989;Foreyt et al., 1994, 1996). Based oncontemporary taxonomic classifications,three species of Pasteurella, Pasteurellahaemolytica (recently reclassified andrenamed Mannheimia haemolytica; Angenet al., 1999), Pasteurella trehalosi (former-ly P. haemolytica, biotype T; Sneath andStevens, 1990), and Pasteurella multocida,have been isolated from both healthy and

Journal of Wildlife Diseases, 44(2), 2008, pp. 388–403# Wildlife Disease Association 2008

388

ill bighorns (Jaworski et al., 1998; Miller,2001; Kelley et al., 2007). All three speciescan cause pneumonic and septicemicpasteurellosis in bighorn sheep, but thecommensal nature, ubiquitous distribu-tion, and heterogeneity of Pasteurellaspp. among bighorn sheep and othermammalian hosts obscure several featuresof their transmission and epidemiology.Whether highly pathogenic Pasteurellaspp. strains are ‘‘normal’’ flora in NorthAmerican wild sheep populations on anevolutionary timescale seems questionable(Miller, 2001; Jenkins et al., 2007).

Because Pasteurella spp. can function asendemic, opportunistic pathogens, predis-posing factors like trauma, stress, orintercurrent disease may contribute toisolated or epidemic pasteurellosis in big-horn sheep (Feuerstein et al., 1980;Spraker et al., 1984; Festa-Bianchet,1988; Ryder et al., 1992). In some cases,however, pasteurellosis can arise in big-horn sheep in the apparent absence ofsignificant predisposing factors. In thesesituations, it appears that individual carri-ers (most likely either a domestic ruminantor a bighorn sheep) introduce a patho-genic and perhaps novel Pasteurella spp.strain into a susceptible host population(Foreyt and Jessup, 1982; Miller et al.,1991; Miller, 2001). Although stressorscould still play a role in precipitating theshedding of Pasteurella spp. or clinicaldisease in carriers under such circum-stances, they are not necessary to sustainsuch epidemics. Some M. haemolytica andP. trehalosi strains carried as normalcommensal flora by healthy domesticsheep are highly pathogenic in bighornsheep and Dall sheep (Ovis dalli; Onderkaand Wishart, 1988; Onderka et al., 1988;Foreyt, 1989; Foreyt et al., 1996). Itfollows that introduction of such strainscould lead to catastrophic epidemics insusceptible bighorn populations, and thatsome of these strains may become endem-ic and continue cycling in affected popu-lations (Miller et al., 1991; Hobbs andMiller, 1992; Miller et al., 1995; Miller,

2001). Here, we describe a pasteurellosisepidemic and its effects on populationperformance in a Colorado bighorn sheeppopulation; the onset of this epidemiccoincided in both time and space with theappearance of a single domestic sheep(Ovis aires) on occupied bighorn winterrange.

METHODS

Population structure and demography

Radiotelemetry studies of free-rangingbighorn sheep residing in the Tarryall andKenosha Mountains located in Park Coun-ty, Colorado, USA (39uN, 105uW) (Fig. 1)were conducted from 1991–2000 in con-junction with a series of field studies(George et al., 1996; George, 1997; Milleret al., 2000). From 1991 to 1998, wecaptured bighorns and marked them withradiocollars, radio ear tags, and/or uniqueplastic ear tags annually to maintainadequate sample sizes for these studies;capture and associated field methods weredescribed in detail elsewhere (George etal., 1996; George, 1997; Miller et al.,2000). All radiomarks were equipped withmortality signals. Radiocollared adult fe-males ($3 yr old) were present in thestudy area during the entire time period;radiocollared adult males ($3 yr old) werepresent from 1995 to 2000, and radio-collared subadults (,3 yr old) of bothsexes were present from 1996 to 2000.

Radiocollared bighorn sheep were bothobserved and monitored remotely formortality signals year-round. For eachvisual observation, technicians recordeddate, time, group size, age and sex ofanimals in the group, identification ofmarked animals, habitat type, and loca-tion. Visual observations were made froma distance with the use of binoculars andspotting scopes to minimize disturbance.The frequency of visual observationsvaried and was dependent on design ofspecific studies. From May to October1991–97, females and males were ob-served approximately once every 2 wk,

GEORGE ET AL.—DOMESTIC SHEEP–ASSOCIATED PASTEURELLOSIS IN BIGHORN SHEEP 389

although this was not possible in winter.From 1998 to 2000, visual observationswere made less frequently, varying fromonce every 2 wk to once every severalmonths. Mortality signals were monitoredmore frequently in all seasons and years.All mortality signals were investigatedpromptly and, when possible, carcasseswere necropsied either in the field or atthe Colorado State University DiagnosticLaboratory (CSUDL; Fort Collins, Color-ado, USA).

Based on data gathered during thesestudies and on prior knowledge aboutbighorn sheep in our study area (Bear andJones, 1973; Bailey, 1990), the Tarryall-Kenosha Mountains bighorn populationwas comprised of three distinct femalegroups (referred to here as the KenoshaMountains, Sugarloaf Mountain, and TwinEagles herds; Fig. 1). The Sugarloaf andTwin Eagles herds occupied largely sepa-rate ranges in the Tarryall Mountains with

elevations ranging from 2,400 to 3,800 m.The Sugarloaf Mountain herd ranged overabout 165 km2, but bighorns congregatedat lower elevations along Tarryall Creeknear Sugarloaf Mountain during winter.The Twin Eagles herd ranged over about139 km2 south and east of the SugarloafMountain herd, also congregating at lowerelevations along Tarryall Creek duringwinter about 15 km downstream fromthe Sugarloaf Mountain herd. Althoughthe Sugarloaf Mountain and Twin Eaglesherds’ ranges overlapped slightly duringseasonal movements, interchange betweenherds appeared minimal, particularlyamong females. The Kenosha Mountainsherd ranged over about 65 km2 in theKenosha and Platte River Mountainsbetween 2,800 and 3,400 m in elevation.This herd primarily used alpine habitatthroughout the year, but occasionallycongregated in subalpine habitats in latewinter or spring in Black Canyon and

FIGURE 1. Spatial and temporal distribution of 41 carcasses attributed to pneumonia among radiocollaredbighorn sheep in the Tarryall and Kenosha Mountains, Colorado, during the winters of (A) 1997–1998, (B)1998–1999, and (C) 1999–2000, and their relationship to the location where a single domestic sheep wasassociated with bighorns on range in December 1997. Primary winter ranges (KM 5 Kenosha Mountains, SM5 Sugarloaf Mountain, TE 5 Twin Eagles) and two areas of intermediate winter use were delineated withminimum convex polygons by using all radiocollared bighorn sheep locations between December and March,1992–2000 (females) and 1994–2000 (males). Locations of four radiocollared carcasses were not reported.

390 JOURNAL OF WILDLIFE DISEASES, VOL. 44, NO. 2, APRIL 2008

Long Gulch. The Kenosha Mountainsherd’s range was separated from the othertwo herds by at least 10 km during allseasons. Although male movements large-ly overlapped with those of their corre-sponding female herds during winter(Fig. 1), there also were some seasonalmale movements among herds, primarilybetween the Kenosha Mountains and theSugarloaf Mountain herds. A portion ofthe males associated with the SugarloafMountain females in winter migratedand associated with resident KenoshaMountains males during summer. Othermales that associated with the KenoshaMountains females during the fall breed-ing season moved to the north end ofthe Tarryall Mountains and associatedwith males from Sugarloaf Mountainon a distinct late winter/spring range(Fig. 1).

Bighorn sheep numbered about 150(90% confidence interval [CI]5136–164)in the Tarryall Mountains and about 100(90% CI597–115) in the Kenosha Moun-tains in March 1995 based on mark-resightinventory methods (George et al., 1996).Annual adult female survival rates from1991–95 in the Tarryall Mountainherds were 0.82–0.98 (standard error[SE]50.02–0.09) based on an annualsample size of 30 radiocollared females(15/herd; Miller et al., 2000). Of the 10mortalities among radiocollared Tarryallbighorns from 1991–1995, six resulted fromprobable mountain lion predation and theother four were from unknown causes(Miller et al., 2000). Of the 17 radiocollaredbighorns (10 females, seven males) in theKenosha Mountains, there were threemortalities during 1995–1997. One of thesewas killed by lighting, another by a hunter,and one from an unknown cause that mayhave been related to recapture. No consis-tent health problems were detected amongmarked animals and no cases of pneumoniawere diagnosed from 1991 to 1996 withinthe study area (George, 1997; Miller et al.,2000; Colorado Division of Wildlife[CDOW], unpubl. data).

Management intervention

Soon after the epidemic began inDecember 1997 as described below, weconsidered possible management actionsthat might be undertaken to minimize itsspread and adverse effects. Based onknowledge of bighorn movement patternsin our study area gained from radiotelem-etry data and ongoing field studies, weanticipated that late winter and summermovements of male bighorns wouldspread the responsible pathogen(s) fromthe Sugarloaf Mountain herd to theneighboring Kenosha Mountains andTwin Eagles herds by the winter of1998–99. In an effort to lessen the effectsof this epidemic by boosting herd immu-nity to pasteurellosis, 20 bighorns at TwinEagles, 19 in the Kenosha Mountains, andseven at Sugarloaf Mountain were vacci-nated with a multivalent P. haemolyticasubcomponent vaccine (Miller et al., 1997;Kraabel et al., 1998; McNeil et al., 2000)in February and March 1998 in anticipa-tion of the epidemic spreading in themonths to follow. Vaccine doses equiva-lent to 2 ml were delivered by handinjection, projectile syringe, or biobulletto bighorns uniquely identified by ear tagsor radiocollars.

Laboratory assessment

Carcasses that were not too heavilyscavenged or decomposed to yield tissuesfor diagnostic evaluation were eithernecropsied in the field or submitted tothe CSUDL for postmortem examination.In 20 cases where adequate tissues wereavailable and suitable for diagnostic eval-uation, lung and other representativetissues were collected and subsamplespreserved either refrigerated for bacteri-ology and virus isolation or in 10% neutralbuffered formalin for histopathology. Forsome cases, samples for bacteriology alsowere submitted to the University of IdahoCaine Veterinary Teaching Center(CVTC; Caldwell, Idaho); there, Pasteu-rella spp. were isolated and characterized

GEORGE ET AL.—DOMESTIC SHEEP–ASSOCIATED PASTEURELLOSIS IN BIGHORN SHEEP 391

via a modified biogrouping classificationsystem with the use of established meth-ods (Kilian and Frederiksen, 1981; Bis-gaard and Mutters, 1986; Ward et al.,1986, 1997; Jaworski et al., 1998). Wereport all isolates as Pasteurella spp. as perthe original laboratory reports, althoughsome of these strains are now classified asMannheimia spp. Virus and Mycoplasmaspp. isolations were attempted at CSUDLand CVTC with the use of establishedmethods (Spraker et al., 1984). Forhistopathology, fixed tissues were embed-ded in paraffin, cut into 5–6-mm sections,mounted on glass slides, and stained withhematoxylin and eosin.

Survival data analysis

To assess the effects of this pasteurel-losis epidemic on the Tarryall-Kenoshabighorn population, we analyzed radiote-

lemetry data with the use of a raggedtelemetry approach in Program MARK(White and Burnham, 1999). The raggedtelemetry analysis allowed for the fact thatsome bighorns may not have been founddead during the month that they diedbecause field conditions interfered withcarcass recovery; this was accommodatedby allowing for all possible combinationsof potential months of death in thestatistical likelihood, an approach mathe-matically analogous to the nest survivalmodel of Dinsmore et al. (2002). Weconstructed explanatory models of ob-served mortality patterns that includedan effect of the epidemic event in theyears that it occurred, variation amongsexes and ages (categorical 1, 2, 3,$4 years old), with an effect of vaccina-tion on survival (Table 1). Variationamong the three herds was included in

TABLE 1. Models and hypotheses considered for the bighorn sheep radiotelemetry survival analysis; models20 and 21 were developed post hoc to consider additional hypotheses about bighorn sheep survival.

Model Hypothesis Model structure

1 Four-month acute epidemic effect epi(acute)2 Four-month acute epidemic effect with differences by vaccination status epi(acute)+vaccine3 Four-month acute epidemic effect with differences by sex epi(acute)+sex4 Four-month acute epidemic effect with differences by age epi(acute)+age5 Four-month acute epidemic effect with differences by sex and age epi(acute)+sex+age6 Acute effect with 12-mo linearly decreasing epidemic effect epi(trend)7 Acute effect with 12-mo linearly decreasing epidemic effect with

differences by vaccination statusepi(trend)+vaccine

8 Acute effect with 12-mo linearly decreasing epidemic effect withdifferences by sex

epi(trend)+sex

9 Acute effect with 12-mo linearly decreasing epidemic effect withdifferences by age

epi(trend)+age

10 Acute effect with 12-mo linearly decreasing epidemic effect withdifferences by sex and age

epi(trend)+sex+age

11 Chronic epidemic effect epi(chronic)12 Chronic epidemic effect with differences by vaccination status epi(chronic)+vaccine13 Chronic epidemic effect with differences by sex epi(chronic)+sex14 Chronic epidemic effect with differences by age epi(chronic)+age15 Chronic epidemic effect with differences by sex and age epi(chronic)+sex+age16 Constant survival (null) Constant17 Survival differences by sex Sex18 Survival differences by age Age19 Survival differences by sex and age sex+age20 Four-mo acute epidemic effect with differences by sex and by

winter seasonepi(acute)+sex+winter

21 Acute effect with 12-mo linearly decreasing epidemic effect withdifferences by sex and by vaccination status

epi(trend)+sex+vaccine

392 JOURNAL OF WILDLIFE DISEASES, VOL. 44, NO. 2, APRIL 2008

all models. The epidemic effect wasmodeled to reflect three hypotheses aboutthe form of the effect: 1) as an acute effectlasting for 4 mo (December–March) dur-ing the winter of first occurrence, withsurvival then returning to a ‘‘normal’’ level;2) as an acute effect with a linearlyincreasing trend in survival that returnedto normal survival after 12 mo; and 3) as achronic effect on survival that continueduntil the end of the study period in 2000.We also included a null hypothesis of noeffect of the epidemic on survival. TheSugarloaf Mountain herd suffered threeepidemic events that were modeled sepa-rately to account for the potential differ-ences in the repeated effect of theoutbreak on the population. After arrivingat the ‘‘best’’ model, we added an effect ofwinter on survival to test whether allwinters or only epidemic winters wereaffecting survival, and we also added avaccination effect to the top-rankedmodel. We used a logit link function tomodel survival as a function of theeffects listed above, basing model selec-tion on the minimum Akaike’s informationcriterion (AIC) corrected for small samplesize (AICc; Burnham and Anderson,2002). Thirty six, 52, and 55 bighornsheep were radiocollared in the Kenosha,Sugarloaf and Twin Eagles herds, respec-tively.

RESULTS

Bighorn sheep population structureand demography

An unusual pattern of mortality forbighorn sheep residing in the study areabegan on 2 December 1997, when aradiocollared yearling female was founddead on Sugarloaf Mountain. The carcasswas necropsied at CSUDL the followingday and diagnosed with ‘‘severe hemor-rhagic necrotizing suppurative broncho-pneumonia’’; the microscopic lesions were‘‘suggestive of bronchopneumonia due toPasteurella hemolytica [sic]’’ (CSUDLNo. 978-17470). From 8–19 December,

eight more bighorn sheep carcasses (fourradiocollared, one ear tagged, and threeunmarked) were found dead within about1 km of the first carcass. Two of theseeight mortalities, a 3-yr-old male and amale lamb found on 18 December, werenecropsied and also diagnosed with ‘‘acutefibrinous bronchopneumonia’’; the othersix carcasses were heavily scavenged andunsuitable for necropsy.

On 18 December, while radiotrackingbighorn sheep, a field technician observeda male domestic sheep on SugarloafMountain. The domestic sheep was moredifficult to see on the snow-coveredterrain than the darker bighorn sheep.When first observed, the domestic sheepappeared to be following the technician.However, when the technician tried toapproach the domestic sheep it fled andjoined a nearby group of bighorn sheep.According to his notes, ‘‘Several attemptswere made by the bighorns to keep thedomestic male away but it was persistentand eventually allowed to graze withthem.’’ (T. Verry, unpubl. field notes,CDOW and United States Forest Service).We made unsuccessful attempts to cap-ture the domestic sheep and to locate itsowner later that day and again on themorning of 19 December. We subse-quently shot the domestic sheep on 19December while it was still associated witha group of bighorn sheep (Fig. 1A). Thecarcass of the domestic sheep was trans-ported to CSUDL for necropsy. This wasthe first (and only) time during our 10-yrstudy that a domestic sheep was foundwith bighorn sheep on range in the studyarea.

After removal of the domestic sheep,pneumonia-related bighorn mortalitiescontinued at Sugarloaf Mountain for3 wk, with seven additional bighorn car-casses (one radiocollared, two ear tagged,and four unmarked) found between 23December 1997 and 13 January 1998(Fig. 1A). Of these seven, the relativelyintact carcass of a 3-yr-old female foundon 23 December was transported to

GEORGE ET AL.—DOMESTIC SHEEP–ASSOCIATED PASTEURELLOSIS IN BIGHORN SHEEP 393

CSUDL for necropsy and also was diag-nosed with acute fibrinous bronchopneu-monia. No additional pneumonia-relatedmortalities were documented between 13January 1998 and 24 November 1998 atSugarloaf Mountain or elsewhere in thestudy area.

Unusually high numbers of bighornmortalities resumed in the SugarloafMountain herd during the winter of1998–99, and also began as predicted inthe Twin Eagles and Kenosha herds(Fig. 1B). We found 13 bighorn carcasses(five radiocollared, two ear tagged, and sixunmarked) on the Sugarloaf Mountainwinter range between 16 December 1998and May 1999. At Twin Eagles, we found27 bighorn carcasses (17 radiocollared,four ear tagged, and six unmarked) andone sick unmarked female was shot,between 24 November 1998 and May1999; seven of these were necropsied anddiagnosed with multifocal to confluentnecrosuppurative pneumonia. In the Ke-nosha Mountains, ten of 23 radiocollaredbighorns (43%) died between January andMay 1999, and the carcasses of two eartagged and four unmarked bighorns alsowere located the same winter. Because ofdifficult access and greater snow accumula-tions in the Kenosha Mountains, no carcass-es were necropsied, and four of thecarcasses were not recovered until springor summer 1999. The unusually highmortality rate continued on the SugarloafMountain winter range during the winter of1999–2000, with 13 carcasses (seven radio-collared, one ear tagged, and five unmarked)located from 22 to 30 December 1999, butsubsided on the Twin Eagles and Kenoshawinter ranges (Fig. 1C).

In all, we found 86 bighorn carcasses(including 45 radiocollared individuals)during the course of the epidemic. Un-doubtedly there were additional mortali-ties of nonradioed animals that were notdetected, especially in the less accessibleKenosha Mountains. The source of thedomestic sheep was never determined—although we did subsequently identify a

small private collection of domestic sheepabout 14 km from Sugarloaf Mountainthat could have been the source, we couldneither confirm nor eliminate the possi-bility that the animal came from else-where.

In addition to the direct mortalityassociated with epidemics in these threeherds, lamb recruitment in years followingthe pneumonia epidemic decreased. Inthe Kenosha Mountains, winter lamb:eweratios dropped from an average of 46lambs:100 females (range 39–56lambs:100 females) during 1992–1996 tozero lambs:100 females from 1999–2001,then improved to a 5-yr average (2002–2006) of 32 lambs:100 females (range 20–50). Similar poor lamb recruitment wasqualitatively noted in the Tarryall Moun-tains following epidemics in the TwinEagles and Sugarloaf Mountain herds:recruitment had been consistently high($0.53; Miller et al., 2000) prior to theepidemic, but by 2006–2007 winter ratioswere only about 23 lambs:100 females.

Although mark-resight survey methodswere not applied after the epidemic toestimate bighorn population size, a stan-dardized annual winter composition sur-vey showed a 50% decline in the KenoshaMountain herd after the epidemic: duringthe 5 yr of surveys in 1992–1996, prior tothe epidemic, an average of 90 bighorn(range 79–96) were observed each year; incontrast, during the 5 yr following theepidemic (1999–2003) an average of 42(range 39–49) bighorns were observedannually. A survey conducted in winter2006–2007 using the same methods asprevious surveys located 45 bighorns,indicating that 7 yr postepidemic this herdhad not recovered to numbers observed inyears prior to the epidemic. Populationrecovery in the Tarryall Mountains herdswas comparable to recovery in the Ke-nosha Mountains: The numbers of big-horn classified on standardized surveyswere 66 in 2003, 76 in 2004, 84 in 2005,and 64 in 2006, indicating that after 9 yrbighorn abundance was only about half of

394 JOURNAL OF WILDLIFE DISEASES, VOL. 44, NO. 2, APRIL 2008

that estimated in the Tarryall Mountainsbefore the epidemic.

Laboratory findings

Among bighorn carcasses examinedeither in the field or at CSUDL, grosslesions consistently included varying de-grees of consolidation and adhesion ofdependent lung lobes, as well as variousdefects consistent with scavenging inmany of the cases examined. Histologiclesions of severe, hemorrhagic, necrotiz-ing, suppurative bronchopneumonia oracute fibrinous bronchopneumonia weredescribed in all cases from the SugarloafMountain herd examined in the winter of1997–98; in some cases, accumulations ofoat-shaped macrophages were observed.Similarly, cases from the Twin Eagles herdalso showed acute to subacute necrotizing,suppurative bronchopneumonia. Eithergross or histologic evidence of lungworms(Protostrongylus spp.) were noted in somecases. The domestic sheep appearedgrossly normal, and histopathology of lungtissue revealed only congestion and atel-ectasis, with some small areas of neutro-phils or lymphoid cell infiltration.

Three species of Pasteurella were iso-lated in various combinations from affect-ed lung tissues from pneumonic bighorns.With one exception, b-hemolytic P. (m.)haemolytica (primarily reported asbiogroup 1G or 1aG) was isolated fromlung tissues of cases evaluated duringwinter 1997–1998; in the exceptional case,only ‘‘very hemolytic’’ Escherichia coli wasisolated from lung (and tonsil) tissue. b-hemolytic P. (m.) haemolytica, hemolyticand nonhemolytic P. trehalosi, P. multo-cida multocida b, nonhemolytic or b-hemolytic E. coli, and Pseudomonas fluo-rescens were isolated in various combina-tions from pneumonic bighorns sampledopportunistically throughout the epidem-ic; in some cases, culture results appearedto be influenced by postmortem carcasscondition or delays in shipping andlaboratory processing. The domestic sheepyielded various combinations of b-hemo-

lytic P. (m.) haemolytica (reported asbiogroup 3A), nonhemolytic P. trehalosi,Streptococcus bovis, and P. fluorescensfrom nasal, sinus, and pharyngeal swabsand tonsil tissue, but no bacteria wereisolated from lung tissue. No viruses wereisolated from any of the bighorn ordomestic sheep samples, but an unchar-acterized Mycoplasma spp. was isolatedfrom two of the seven carcasses from theTwin Eagles herd.

Epidemic effects on survival

The model that included an acute effectwith a 12-mo linearly decreasing effect ofthe outbreak on survival and differingsurvival by sex was selected as the bestmodel by AICc (Table 2). The AICc

weights suggested the acute effect with a12-mo linearly decreasing effect was 2.23

more likely than an acute effect alone(Table 2). Parameter estimates suggesteda strong negative effect of the epidemicon survival (slope520.18, SE50.02,Fig. 2). Survival patterns after epidemicswere similar across herds, and the subse-quent outbreaks on the Sugarloaf Mountainwinter range were similar to the first. Malesshowed lower survival than females (20.50,SE50.23, Fig. 2). Models that included along-term chronic effect on survival or noepidemic effect on survival received nosupport (Table 2). There was no supportfor winter survival differing from nonwintermonths in nonepidemic years (0.464,SE50.37, Table 3). There was support fora small, positive influence of vaccination onbighorn survival (0.007, SE50.003, 4.39AICc units smaller than the model withouta vaccine effect; Table 3).

DISCUSSION

Epidemic pasteurellosis and the result-ing depression in the Tarryall-KenoshaMountains bighorn population’s survival,recruitment, and size followed the appear-ance of a single domestic sheep on nativebighorn winter range and occurred in theabsence of other known or suspected

GEORGE ET AL.—DOMESTIC SHEEP–ASSOCIATED PASTEURELLOSIS IN BIGHORN SHEEP 395

inciting factors, illustrating the potentialconsequences of contact between thesespecies under natural conditions. Ourobservations provide one more replicateof epidemiologic and mortality patterns

reported following association of these twospecies in Colorado and elsewhere forover a century (Warren, 1910; Grinnell,1928; Shillinger, 1937; Buechner, 1960;Bear and Jones, 1973; Foreyt and Jessup,

FIGURE 2. Monthly survival estimates for male (gray diamonds) and female (black triangles) bighornsheep in the Kenosha Mountains herd from October 1998 to December 1999, showing an example of thechange in survival prior to, during, and following the disease outbreak. Bars are 95% confidence intervals onpoint estimates of survival.

TABLE 2. A priori models, model selection results (AICc), model weights (w), numbers of estimatedparameters (K), and deviance for the bighorn sheep radiotelemetry analysis.

Model AICc DAICc wi K Deviance

epi(trend)+sex 785.22 0.00 0.29 6 773.20epi(acute)+sex 786.80 1.58 0.13 6 774.79epi(acute)+sex+age 786.99 1.77 0.12 9 768.96epi(trend)+vaccine 787.06 1.84 0.12 6 775.05epi(trend) 787.67 2.45 0.09 5 777.66epi(trend)+sex+age 788.33 3.11 0.06 11 766.28epi(trend)+age 789.44 4.23 0.04 10 769.40epi(acute) 790.07 4.85 0.03 5 780.06epi(acute)+vaccine 791.37 6.15 0.01 6 779.35epi(acute)+age 792.18 6.96 0.01 10 772.13epi(chronic)+sex 815.28 30.06 0.00 5 805.27epi(chronic)+sex+age 819.39 34.17 0.00 10 799.34epi(chronic)+vaccine 819.39 34.17 0.00 5 809.38epi(chronic) 820.28 35.06 0.00 4 812.28epi(chronic)+age 821.86 36.64 0.00 9 803.82Sex 829.99 44.77 0.00 4 821.98sex+age 837.00 51.78 0.00 9 818.96Constant 838.03 52.81 0.00 3 832.02Age 840.15 54.93 0.00 8 824.12

396 JOURNAL OF WILDLIFE DISEASES, VOL. 44, NO. 2, APRIL 2008

1982; Goodson 1982; Coggins, 1988;Brown, 1989, deVos, 1989; Beecham etal., 2007). Previous studies have consis-tently linked contact with domestic sheepto occurrences of pneumonia in captivebighorns (Foreyt and Jessup, 1982; On-derka and Wishart, 1988; Foreyt, 1989;Callan et al., 1991; Foreyt et al. 1994), butthis phenomenon has not been producedexperimentally in natural bighorn popula-tions because intentional manipulationsthat likely would result in the loss of alarge number of free-ranging bighornsheep are neither politically, ethically,nor logistically feasible. Although thesituation we described here was not a trueexperiment, the spatial and temporalrelationship between epidemic onset anddiscovery of a domestic sheep on theaffected winter range was clear (Fig. 1).Ecologic patterns that occur repeatedly—like pneumonia in bighorn sheep followingassociation with domestic sheep—aremeaningful even though they may notarise from experimentation, and suchpatterns should not be ignored in conser-vation and resource management decisionmaking.

The mortality and subsequent de-pressed recruitment patterns observedduring and after the 1997–2000 Tarryall-Kenosha Mountains pneumonia epidemicsare similar to patterns reported by others(Coggins, 1988; Festa-Bianchet, 1988;Coggins and Matthews, 1992; Ryder etal., 1994; Cassirer et al., 1996; Enk et al.,2001). Mortality during the Tarryall-Ke-nosha Mountains epidemic resulted in thedirect loss of at least 72 individuals,representing about 28% of the estimated

population. Bighorn losses incurred dur-ing recent pneumonia epidemics reportedelsewhere ranged from about 25% atWhiskey Mountain, Wyoming (Ryder etal., 1994), 35–40% at Sheep River, Alberta(Festa-Bianchet, 1988), 50–75% in Hell’sCanyon, Idaho (Cassirer et al., 1996),about 67% of an Oregon herd (Coggins,1988), to about 80% of a Montanapopulation (Enk et al., 2001). In theTarryall-Kenosha Mountains bighorn pop-ulation and elsewhere, pneumonia epi-demics tended to occur in fall and winterwith index cases detected in November orearly December (Coggins, 1988; Festa-Bianchet, 1988; Ryder et al., 1992; Cas-sirer et al., 1996; Enk et al. 2001; thisstudy). Although epidemics depressedadult survival for ,1 yr, lamb recruitmentwas very low for 2–3 yr following epidem-ics in our herds and in others (Onderkaand Wishart, 1984; Festa-Bianchet, 1988;Coggins and Matthews, 1992; Ryder et al.,1994). Consequently, as of winter 2006–2007 the Tarryall-Kenosha Mountains big-horn population’s estimated size remainedabout half of that estimated before theepidemic, similar to trends observed inother bighorn populations (Coggins andMatthews, 1992; Enk et al., 2001).

The seasonal pattern of pneumonia-associated mortalities in the Twin Eaglesand Kenosha Mountains herds, as well asthe recurrence of pneumonia-associatedmortality in the Sugarloaf Mountain herdin the two winters following the initialepidemic, suggest that environmental and/or social conditions coinciding with orsubsequent to the introduction of novelpathogens can contribute to the develop-

TABLE 3. Post hoc models considering additional hypotheses about bighorn sheep survival, model selectionresults (AICc), model weights (w), numbers of estimated parameters (K), and deviance for the bighorn sheepradiotelemetry analysis. The epi(trend)+sex model was the top-ranked model from the a priori analysis(Table 2).

Model AICc DAICc wi K Deviance

epi(trend)+sex+vaccine 780.82 0.00 0.87 7 766.802epi(trend)+sex 785.22 4.39 0.10 6 773.203epi(acute)+sex+winter 787.06 6.24 0.04 5 773.042

GEORGE ET AL.—DOMESTIC SHEEP–ASSOCIATED PASTEURELLOSIS IN BIGHORN SHEEP 397

ment of clinical pneumonia in bighornsheep. In the Kenosha Mountains herd,exposure most likely first occurred duringthe summer following the initial 1997–98epidemic in the Sugarloaf Mountain herdas males made seasonal movements andjoined with males from other herds onsummer range (Fig. 1); however, pneu-monia-associated mortalities were notdetected until December–January. Cog-gins (1988) reported a similar delaybetween suspected exposure and subse-quent mortalities in the Lostine bighornherd: There, exposure likely occurred inearly October when a female domesticsheep was seen associated with threebighorn sheep (two males and one female)on summer range, but the epidemic wasnot detected until late November afterbighorns returned to winter range and thebreeding season was in progress. Afterobserving that chronic, sporadic pneumo-nia-associated mortality in adult bighornsoccurred in fall and early winter (Octo-ber–January) in the Hells Canyon popula-tion, Cassirer and Sinclair (2007) hypoth-esized that ‘‘seasonal increase in diseasecould be due to variation in immunocom-petence caused by energy availability orstressors or to seasonal behavior patternsthat might facilitate pathogen transmis-sion.’’ Seasonal declines in immunitywithin individuals in late fall and winterand increased contact between individualsbecause of breeding behavior may explainthe relatively common timing of mostpneumonia epidemics in free-ranging big-horn populations, regardless of whetherthese epidemics arise from endemic orintroduced pathogens.

Our experiences suggest that contactbetween bighorn sheep and domesticsheep may go undetected on nativeranges. The Tarryall and Kenosha Moun-tains, like most bighorn ranges, includeremote areas of broken terrain that havelittle human presence in winter. At thetime of the initial epidemic at SugarloafMountain, a full-time technician was livingwithin 4 km of where bighorn mortalities

occurred, and radiocollars were monitoredon a regular basis. With the exception ofone small band, this entire herd hadremained within ,1.6 km of SugarloafMountain since early November 1997, andof the seven marked bighorns that died inDecember, all were seen associated withone another in a group on either 5 or 13November. Even with such intensivemonitoring, given the domestic sheep’srecalcitrance and the difficulty of observ-ing it against the snow pack, we believethis animal may have been present some-where on the Sugarloaf Mountain winterrange for several weeks prior to beingdetected. If the same contact and mortal-ity event had occurred outside of ourstudy, we doubt that the epidemic wouldhave been detected until some time afterthe domestic sheep had disappearedbecause of predation or another cause. Itfollows that when pneumonia epidemicsoccur in free-ranging native North Amer-ican wild sheep it may not be possible toconclude that contact with domestic sheepdid not occur, but only that contact wasnot detected.

Our attempted management interven-tions in the Tarryall-Kenosha Mountainsbighorn population, including removingthe domestic sheep and vaccinating aproportion of the bighorns in two herdsprior to exposure and onset of epidemics,did not prevent significant mortality orpoor recruitment in following years. Al-though the subcomponent vaccine weused had previously been shown effectiveunder experimental conditions (Kraabel etal., 1998), a single dose of vaccineprovided perhaps 8 mo or more prior toexposure appeared to confer only limitedprotection under field conditions. Re-search should continue toward developingeffective vaccines or therapeutic tools, butuntil effective intervention or treatmentprotocols are available management ef-forts are best focused on disease preven-tion. Research and management effortsalso should focus on long-term studies inselected free-ranging bighorn populations

398 JOURNAL OF WILDLIFE DISEASES, VOL. 44, NO. 2, APRIL 2008

(e.g., Jorgenson et al., 1997; Cassirer andSinclair, 2007). Long-term studies couldprovide valuable information on popula-tion dynamics in both the absence andpresence of epidemics. In retrospect,either annual or biennial population esti-mates based on mark-resight techniques,as well as more careful estimation ofrecruitment in all three affected herds,would have afforded us more certainty inmeasuring the overall effects of thisepidemic on the Tarryall-Kenosha Moun-tains bighorn population.

The dominant strain of P. (m.) haemo-lytica isolated from bighorn carcasses inDecember 1997 had not been isolatedpreviously from either the Tarryall-Ke-nosha Mountains bighorn population orfrom other Colorado bighorn populations(Miller et al., 1995; Green et al., 1999),suggesting it may have been a novelpathogen introduced at the time theepidemic began. (Alternatively, some oth-er unidentified pathogen may have beenintroduced.) Mannheimia haemolyticabiogroup 1 is the most common biovariantisolated from domestic sheep, and iscomparatively rare in wild sheep (Jaworskiet al., 1998). Although these findings wereconsistent with other field evidence impli-cating the domestic sheep that appearedon the Sugarloaf Mountain herd winterrange as the inciting cause of this epidem-ic, only P. (m.) haemolytica biogroup 3, aremarkably rare biovariant among domes-tic sheep (Jaworski et al., 1998), wasisolated from the carcass of that animal.Thus, despite gathering overwhelmingcircumstantial epidemiologic evidence re-garding the most plausible and parsimo-nious explanation for the origins of thisepidemic, laboratory findings failed to‘‘prove’’ this apparent explanation.

Our laboratory findings raise questionsabout interpreting data from fine-scalephenotypic strain classifications like bio-grouping in field investigations of pasteu-rellosis epidemics in bighorn sheep. Al-though such approaches have been shownas useful adjuncts to studies of pasteurel-

losis in captive bighorn sheep (e.g.,Kraabel et al., 1998), their ability to trackan introduced Pasteurella spp. strainthrough the course of an epidemic undernatural conditions has not been demon-strated experimentally. We identified sev-eral potential sources of bias in laboratoryresults arising from our field samples.With respect to recovery of isolates forcharacterization, we noticed that Pasteu-rella spp. strains originally isolated atCVTC and CSUDL were quite differentdespite temporal and spatial proximity ofcases, suggesting that differences in tech-niques between laboratories may influ-ence primary culture results. In addition,delays in shipping of some specimens alsoappeared to hamper recovery of somePasteurella spp. strains. Differences indisease progression and carcass preserva-tion may have further influenced recoveryof Pasteurella spp. Beyond the limitationsof recovery methods, the use of pheno-typic strain classification data in fieldinvestigations is based on assumptions thatrelative strain abundance is static withinhosts and that trait expressions withinstrains are not altered with interspeciesor serial intraspecies passage; the formerdoes not appear to be true, at least incaptive bighorn sheep (Miller et al., 1997),and the latter has not been studied undernatural conditions. Interpretations of phe-notypic strain classification data are fur-ther confounded by observations (Kelleyet al., 2007) that apparently identicalbiovariants can have divergent phyloge-netic relationships and yet different bio-variants can be quite closely related: Forexample, in comparing three biovariant 1isolates (I26, I45, and I49), Kelley et al.(2007, Fig. 1) genetically identified one asa different species (P. trehalosi) andshowed that two isolates (I30, I26) classi-fied as biovariants 3 and 1, respectively,appeared more closely related than thethree aforementioned biovariant 1 isolates.Moreover, evidence of horizontal transferof the lktA gene—the gene that encodesleukotoxin production in Pasteurellace-

GEORGE ET AL.—DOMESTIC SHEEP–ASSOCIATED PASTEURELLOSIS IN BIGHORN SHEEP 399

ae—among Pasteurella spp. and Mannhei-mia spp. isolates from bighorn and do-mestic sheep (Kelley et al., 2007) suggeststhat endemic and introduced strains canexchange genetic material (including cod-ing for virulence factors), thereby alteringboth phenotypic traits and pathogenicityand further confounding epidemiology.Based on the findings presented hereand elsewhere, it follows that the failureto match Pasteurella spp. and Mannheimiaspp. strains between domestic sheep andaffected bighorns precisely does not nec-essarily preclude the involvement of do-mestic sheep in precipitating pasteurello-sis epidemics in bighorns.

Epidemic pasteurellosis can cause wide-spread mortality that may hamper effec-tive population or species management, aswe observed in the Tarryall-KenoshaMountains bighorn population and as hasbeen observed elsewhere (Jorgenson et al.,1997; McCarty and Miller, 1998; Singer etal., 2000). In some cases the effects ofepidemics are limited and/or ephemeral,and present only temporary obstacles toresource management. Overall, however,bighorn sheep abundance appears to belimited by recurrent pasteurellosis epi-demics (Hobbs and Miller, 1992; Jorgen-son et al., 1997; McCarty and Miller, 1998;Singer et al., 2000). The immediate impactof significant mortality exacted across allage classes during these epidemics iscompounded by pneumonia and septice-mia in neonatal lambs that may suppressrecruitment for years afterward, therebyimpairing population recovery and stabil-ity. In addition to intensive populationmanagement designed to keep some big-horn herds below perceived density-de-pendent epidemic thresholds, livestockgrazing policies on some public lands inthe western United States (Bureau ofLand Management, 1992; Schommer andWoolever, 2001; Beecham et al., 2007;Western Association of Fish and WildlifeAgencies, 2007) have been modified toprevent contact between bighorn anddomestic sheep in an attempt to reduce

the frequency and severity of theseepidemics via introduction of novel path-ogenic Pasteurella spp. and Mannheimiaspp. strains and other respiratory patho-gens. Such segregation is not likely toprevent all pasteurellosis epidemics inbighorns, but should help reduce boththe frequency and, in many cases, theseverity of epidemics in free-rangingpopulations.

ACKNOWLEDGMENTS

Our study was supported by Federal Aid inWildlife Restoration Project W-153-R, Color-ado bighorn sheep auction and raffle funds,the Rocky Mountain Bighorn Society, theFoundation for North American Wild Sheep,the United States Forest Service and theColorado Division of Wildlife. We thank T.Verry, J. Vayhinger, S. Roush, V. Jurgens, A.Torres, N. Howard, S. Tracey, P. Krumm, S.Berry, T. Sandmeier, J. Duran, B. Davies, M.Elkins, R. Green, M. Lamb, R. Myers, R.Zaccagnini, S. Ogilvie, G. Schoonveld, K.Green, R. Mason, S. Werner, W. Adrian, K.Larsen, L. Carpenter, and others for fieldassistance. We also thank T. Spraker, D.Gould, A. Ward, G. Weiser, E. Williams, andothers for diagnostic assistance, D. Brownnefor geographic information system (GIS)assistance, and M. Schuette for revising andimproving Figure 1. B. Davies, D. Walsh, F.Cassirer, and two anonymous reviewers pro-vided valuable comments on earlier drafts ofthis manuscript.

LITERATURE CITED

ANGEN, O., R. MUTTERS, D. A. CAUGANT, J. E. OLSEN,AND M. BISGAARD. 1999. Taxonomic relationshipsof the [Pasteurella] haemolytica complex asevaluated by DNA-DNA hybridizations and16S rRNA sequencing with proposal of Mann-heimia haemolytica gen. nov., comb. nov.,Mannheimia granulomatis comb. nov., Mannhei-mia glucosida sp. nov., Mannheimia ruminalissp. nov. and Mannheimia varigena sp. nov.International Journal of Systematic Bacteriology49: 67–86.

BAILEY, J. A. 1990. Management of Rocky Mountainbighorn sheep herds in Colorado. ColoradoDivision of Wildlife Special Report Number 66.24 pp.

BEAR, G. D., AND C. W. JONES. 1973. History anddistribution of bighorn sheep in Colorado.Colorado Division of Wildlife Report. 232 pp.

BEECHAM, J. J., JR., C. P. COLLINS, AND T. D.REYNOLDS. 2007. Rocky Mountain bighorn sheep

400 JOURNAL OF WILDLIFE DISEASES, VOL. 44, NO. 2, APRIL 2008

(Ovis canadensis): A technical conservationassessment, United States Department of Agri-culture, Forest Service, Rocky Mountain Region,http://www.fs.fed.us/r2/projects/scp/assessments/rockymountainbighornsheep.pdf. Accessed Oc-tober 2007.

BISGAARD, M., AND R. MUTTERS. 1986. Reinvestigationof selected bovine and ovine strains previouslyclassified as Pasteurella haemolytica and descrip-tion of some new taxa within the Pasteurellahaemolytica-complex. Acta Pathologica Micro-biologica et Immunologica Scandinavica—Sec-tion B: Microbiology 94: 185–193.

BROWN, D. 1989. Early history. In The desert bighornsheep in Arizona, R. M. Lee (ed.). Arizona Gameand Fish Department, Phoenix, Arizona, pp. 1–11.

BUECHNER, H. K. 1960. The bighorn sheep in theUnited States, its past, present, and future.Wildlife Monographs 4: 1–174.

BUREAU OF LAND MANAGEMENT. 1992. Guidelines formanaging domestic sheep in bighorn sheephabitats. Information Bulletin 92-212, Bureauof Land Management, Washington, D.C., 2 pp.

BURNHAM, K. P., AND D. R. ANDERSON. 2002. Modelselection and multimodel inference: A practicalinformation-theoretic approach. 2nd Edition.Springer-Verlag, New York, New York, 488 pp.

CALLAN, R. J., T. D. BUNCH, G. W. WORKMAN, AND R.E. MOCK. 1991. Development of pneumonia indesert bighorn sheep after exposure to a flock ofexotic domestic sheep. Journal of the AmericanVeterinary Medical Association 198: 1052–1056.

CASSIRER, E. F., AND A. R. E. SINCLAIR. 2007.Dynamics of pneumonia in bighorn sheepmetapopulations. Journal of Wildlife Manage-ment 71: 1080–1088.

———, L. E. OLDENBURG, V. L. COGGINS, P. FOWLER,K. RUDOLPH, D. L. HUNTER, AND W. J. FOREYT.1996. Overview and preliminary analysis of abighorn sheep dieoff, Hells Canyon 1995–96.Proceedings of the Biennial Symposium of theNorthern Wild Sheep and Goat Council 10: 78–86.

COGGINS, V. L. 1988. The Lostine Rocky Mountainbighorn sheep die-off and domestic sheep.Proceedings of the Biennial Symposium of theNorthern Wild Sheep and Goat Council 6: 57–64.

———, AND P. E. MATTHEWS. 1992. Lamb survivaland herd status of the Lostine bighorn herdfollowing a Pasteurella die-off. Proceedings ofthe Biennial Symposium of the Northern WildSheep and Goat Council 8: 147–154.

DEVOS, J. C., JR. 1989. The role of disease in Arizona’sbighorn sheep. In The desert bighorn sheep inArizona, R. M. Lee (ed.). Arizona Game andFish Department, Phoenix, Arizona, pp. 30–62.

DINSMORE, S. J., G. C. WHITE, AND F. L. KNOPF. 2002.

Advanced techniques for modeling avian nestsurvival. Ecology 83: 3476–3488.

ENK, T. A., H. D. PICTON, AND J. S. WILLIAMS. 2001.Factors limiting a bighorn sheep population inMontana following a die-off. Northwest Science75: 280–291.

FESTA-BIANCHET, M. 1988. A pneumonia epizootic inbighorn sheep, with comments on preventivemanagement. Proceedings of the Biennial Sym-posium of the Northern Wild Sheep and GoatCouncil 6: 66–76.

FEUERSTEIN, V., R. L. SCHMIDT, C. P. HIBLER, AND W.H. RUTHERFORD. 1980. Bighorn sheep mortalityin the Taylor River-Almont Triangle Area, 1978–1979: A case study. Colorado Division of WildlifeSpecial Report Number 48. 19 pp.

FOREYT, W. J. 1989. Fatal Pasteurella haemolyticapneumonia in bighorn sheep after direct contactwith clinically normal domestic sheep. AmericanJournal of Veterinary Research 50: 341–344.

———, AND D. A. JESSUP. 1982. Fatal pneumonia ofbighorn sheep following association with domes-tic sheep. Journal of Wildlife Diseases 18: 163–168.

———, K. P. SNIPES, AND R. W. KASTEN. 1994. Fatalpneumonia following inoculation of healthybighorn sheep with Pasteurella haemolytica fromhealthy domestic sheep. Journal of WildlifeDiseases 30: 137–145.

———, ———, AND J. E. LAGERQUIST. 1996.Susceptibility of Dall sheep (Ovis dalli dalli) topneumonia caused by Pasteurella haemolytica.Journal of Wildlife Diseases 32: 586–593.

GEORGE, J. 1997. Seasonal ranges, movements, andpopulation estimates of Rocky Mountain bighornsheep in the Kenosha Mountains. Unpublishedreport. Colorado Division of Wildlife, Denver,Colorado, 36 pp.

———, M. W. MILLER, G. C. WHITE, AND J.VAYHINGER. 1996. Comparison of mark-resightpopulation size estimators for bighorn sheep inalpine and timbered habitats. Proceedings of theBiennial Symposium of the Northern WildSheep and Goat Council 10: 20–25.

GOODSON, N. J. 1982. Effects of domestic sheepgrazing on bighorn sheep populations: A review.Proceedings of the Biennial Symposium of theNorthern Wild Sheep and Goat Council 3: 287–313.

GREEN, A. L., N. M. DUTEAU, M. W. MILLER, J.TRIANTIS, AND M. D. SALMAN. 1999. Polymerasechain reaction techniques for differentiatingcytotoxic and noncytotoxic Pasteurella trehalosifrom Rocky Mountain bighorn sheep. AmericanJournal of Veterinary Research 60: 583–588.

GRINNELL, G. B. 1928. Mountain sheep. Journal ofMammalogy 9: 1–9.

HOBBS, N. T., AND M. W. MILLER. 1992. Interactionsbetween pathogens and hosts: Simulation ofpasteurellosis epizootics in bighorn sheep pop-

GEORGE ET AL.—DOMESTIC SHEEP–ASSOCIATED PASTEURELLOSIS IN BIGHORN SHEEP 401

ulations. In Wildlife 2001: Populations, D. R.McCullough and R. H. Barrett (eds.). ElsevierScience Publishers, London, England, pp. 997–1007.

HONESS, R. F., AND N. M. FROST. 1942. A Wyomingbighorn sheep study. Wyoming Game and FishDepartment Cheyenne Bulletin Number 1,Cheyenne, Wyoming, 126 pp.

JAWORSKI, M. D., D. L. HUNTER, AND A. C. S. WARD.1998. Biovariants of isolates of Pasteurella fromdomestic and wild ruminants. Journal of Veter-inary Diagnostic Investigation 10: 49–55.

JENKINS, E. J., A. M. VEITCH, S. J. KUTZ, T. K.BOLLINGER, J. M. CHIRINO-TREJO, B. T. ELKIN, K.H. WEST, E. P. HOBERG, AND L. POLLEY. 2007.Protostrongylid parasites and pneumonia incaptive and wild thinhorn sheep (Ovis dalli).Journal of Wildlife Diseases 43: 189–205.

JORGENSON, J. T., M. FESTA-BIANCHET, J. M. GAIL-

LARD, AND W. D. WISHART. 1997. Effects of age,sex, disease, and density on survival of bighornsheep. Ecology 78: 1019–1032.

KELLEY, S. T., E. F. CASSIRER, G. C. WEISER, AND S.SAFAEE. 2007. Phylogenetic diversity of Pasteu-rellaceae and horizontal gene transfer of leuko-toxin in wild and domestic sheep. Infection,Genetics and Evolution 7: 13–23.

KILIAN, M., AND W. FREDERIKSEN. 1981. Identificationtables for the Haemophilus-Pasteurella-Actino-bacillus group. In Haemophilus, Pasteurella, andActinobacillus, M. Kilian, W. Frederiksen andE. L. Biberstein (eds.). Academic Press, SanFrancisco, California, pp. 281–287.

KRAABEL, B. J., M. W. MILLER, J. A. CONLON, AND H.J. MCNEIL. 1998. Evaluation of a multivalentPasteurella haemolytica vaccine in bighornsheep: Protection from experimental challenge.Journal of Wildlife Diseases 34: 325–333.

MARSH, H. 1938. Pneumonia in Rocky Mountainbighorn sheep. Journal of Mammalogy 19: 214–219.

MCCARTY, C. W., AND M. W. MILLER. 1998. Modelingthe population dynamics of bighorn sheep: Asynthesis of literature. Colorado Division ofWildlife Special Report Number 73.

MCNEIL, H. J., M. W. MILLER, J. A. CONLON, I. K.BARKER, AND P. E. SHEWEN. 2000. Effects ofdelivery method on serological responses ofbighorn sheep to a multivalent Pasteurellahaemolytica supernatant vaccine. Journal ofWildlife Diseases 36: 79–85.

MILLER, M. W. 2001. Pasteurellosis. In Infectiousdiseases of wild mammals. 3rd Edition, E. S.Williams and I. K. Barker (eds.). Iowa StateUniversity Press, Ames, Iowa, pp. 330–339.

———, N. T. HOBBS, AND E. S. WILLIAMS. 1991.Spontaneous pasteurellosis in captive RockyMountain bighorn sheep (Ovis canadensis cana-densis): Clinical, laboratory, and epizootiological

observations. Journal of Wildlife Diseases 27:534–542.

———, B. J. KRAABEL, J. A. CONLON, H. J. MCNEIL,AND J. M. BULGIN. 1995. Strategies for managinginfectious diseases in mountain sheep popula-tions. Wildlife Research Report, MammalsResearch, Federal Aid Projects, Job ProgressReport, Project W-153-R-8, WP2a, J4. ColoradoDivision of Wildlife, Fort Collins, Colorado,pp. 151–161.

———, J. A. CONLON, H. J. MCNEIL, J. M. BULGIN,AND A. C. S. WARD. 1997. Evaluation of amultivalent Pasteurella haemolytica vaccine inbighorn sheep: Safety and serologic responses.Journal of Wildlife Diseases 33: 738–748.

———, J. E. VAYHINGER, D. C. BOWDEN, S. ROUSH, T.VERRY, A. TORRES, AND V. JURGENS. 2000. Drugtreatment for lungworm in bighorn sheep:Reevaluation of a 20-year-old managementprescription. Journal of Wildlife Management64: 505–512.

ONDERKA, D. K., AND W. D. WISHART. 1984. A majorbighorn sheep die-off from pneumonia insouthern Alberta. Biennial Symposium of theNorthern Wild Sheep and Goat Council 4: 356–363.

———, AND ———. 1988. Experimental contacttransmission of Pasteurella haemolytica fromclinically normal domestic sheep causing pneu-monia in Rocky Mountain bighorn sheep.Journal of Wildlife Diseases 24: 663–667.

———, S. A. RAWLUK, AND W. D. WISHART. 1988.Susceptibility of Rocky Mountain bighorn sheepand domestic sheep to pneumonia induced bybighorn and domestic livestock strains of Pas-teurella haemolytica. Canadian Journal of Vet-erinary Research 52: 439–444.

POST, G. 1962. Pasteurellosis of Rocky Mountainbighorn (Ovis canadensis canadensis). WildlifeDisease 23: 1–14.

POTTS, M. K. 1937. Hemorrhagic septicemia in thebighorn of Rocky Mountain National Park.Journal of Mammalogy 18: 105–106.

RUSH, W. M. 1927. Notes on diseases in wild gameanimals. Journal of Mammalogy 8: 163–165.

RYDER, T. J., E. S. WILLIAMS, K. W. MILLS, K. H.BOWLES, AND E. T. THORNE. 1992. Effect ofpneumonia on population size and lamb recruit-ment in Whiskey Mountain bighorn sheep.Proceedings of the Biennial Symposium of theNorthern Wild Sheep and Goat Council 8: 136–146.

———, ———, AND S. L. ANDERSON. 1994. Residualeffects of pneumonia on the bighorn sheep ofWhiskey Mountain, Wyoming. Proceedings ofthe Biennial Symposium of the Northern WildSheep and Goat Council 9: 15–19.

SCHOMMER, T., AND M. WOOLEVER. 2001. A processfor finding management solutions to the incom-patibility between domestic and bighorn sheep.

402 JOURNAL OF WILDLIFE DISEASES, VOL. 44, NO. 2, APRIL 2008

United States Department of Agriculture ForestService, Wallowa-Whitman National Forest, 40 pp.

SHILLINGER, J. E. 1937. Disease relationship ofdomestic stock and wildlife. In Transactions ofthe second North American Wildlife Confer-ence. American Wildlife Institute, Washington,D. C., pp. 298–302.

SINGER, F. J., E. WILLIAMS, M. W. MILLER, AND L. C.ZEIGENFUSS. 2000. Population growth, fecundity,and survivorship in recovering populations ofbighorn sheep. Restoration Ecology 8 (1 Suppl):75–84.

SNEATH, P. H. A., AND M. STEVENS. 1990. Actinoba-cillus seminis sp. nov., nom. rev., Pasteurellabettii sp. nov., Pasteurella lymphangitidis sp.nov., Pasteurella mairi sp. nov., and Pasteurellatrehalosi sp. nov. International Journal ofSystematic Bacteriology 40: 148–153.

SPENCER, C. C. 1943. Notes on the life history ofRocky Mountain bighorn sheep in the TarryallMountains of Colorado. Journal of Mammalogy24: 1–11.

SPRAKER, T. R., C. P. HIBLER, G. G. SCHOONVELD, AND

W. S. ADNEY. 1984. Pathologic changes andmicroorganisms found in bighorn sheep during astress-related die-off. Journal of Wildlife Diseas-es 20: 319–327.

WARD, A. C. S., L. R. STEVENS, B. J. WINSLOW, R. P.GOGOLEWSKI, D. C. SCHAEFER, S. K. WASSON, AND

B. L. WILLIAMS. 1986. Isolation of Haemophilus

somnus: A comparative study of selective media.Proceedings of the American Association ofVeterinary Laboratory Diagnosticians 29: 479–486.

———, D. L. HUNTER, M. D. JAWORSKI, M. P.DOBEL, J. B. JEFFRESS, AND G. A. TANNER. 1997.Pasteurella spp. in sympatric bighorn anddomestic sheep. Journal of Wildlife Diseases33: 544–557.

WARREN, E. R. 1910. The mountain sheep. In E. R.Warren (ed.). The mammals of Colorado. Anaccount of the several species found within theboundaries of the state, together with a record oftheir habits and of their distribution. G. P.Putnam’s Sons, The Knickerbocker Press, NewYork and London, pp. 9–12.

WESTERN ASSOCIATION OF FISH AND WILDLIFE AGEN-

CIES. 2007. Western Association of Fish andWildlife Agencies (WAFWA), Wild Sheep Work-ing Group, Initial Subcommittee, recommenda-tions for domestic sheep and goat managementin wild sheep habitat. http://www.mwvcrc.org/bighorn/wafwawildsheepreport.pdf. Accessed Octo-ber 2007.

WHITE, G. C., AND K. P. BURNHAM. 1999. ProgramMARK: Survival estimation from populations ofmarked animals. Bird Study 46: S120–S139.

Received for publication 13 April 2007.

GEORGE ET AL.—DOMESTIC SHEEP–ASSOCIATED PASTEURELLOSIS IN BIGHORN SHEEP 403