675J. Parasitol., 94(3), 2008, pp.

675
J. Parasitol., 94(3), 2008, pp. 675–683
! American Society of Parasitologists 2008
RISK FACTORS FOR TOXOPLASMA GONDII INFECTION IN WILD RODENTS FROM
CENTRAL COASTAL CALIFORNIA AND A REVIEW OF T. GONDII PREVALENCE
IN RODENTS
Haydee A. Dabritz*, Melissa A. Miller†, Ian A. Gardner‡, Andrea E. Packham, E. Robert Atwill§, and Patricia A. Conrad"
School of Veterinary Medicine, University of California, Vet Med III-A, Room 4206, One Shields Avenue, Davis, California 95616.
e-mail: [e-mail]
ABSTRACT: Sera from 523 wild rodents were tested for Toxoplasma gondii antibodies using either an indirect fluorescent antibody
test (IFAT) (rats and mice, with titer !80 considered positive) or a latex agglutination test (LAT) (voles, squirrels, and pocket
mice, with titer !32 considered positive). Seventeen percent (88/523) of the rodents, including 26% (85/328) of the Peromyscus
sp. and 8% (3/37) of Spermophilus beecheyi, were seropositive. Fourteen percent (23/161) of rodents captured in trap sites next
to Morro Bay (California) and 15% (16/109) of rodents from sites adjacent to riparian habitats had antibodies to T. gondii,
compared to 19% (49/253) of rodents captured in habitats not associated with water; this difference was not statistically significant
(P ! 0.32). Significantly fewer rodents were captured "200 m from residential housing compared to locations further away (11%
vs. 30%, respectively). Factors associated with an increased risk for T. gondii seropositivity in rodents were capture location
!200 m from residential housing and adult age.
The protozoan parasite T. gondii infects most species of
warm-blooded vertebrates, including rodents. Like humans, ro-
dents serve as intermediate hosts for T. gondii. Parasites form
a latent cyst stage in the tissues of intermediate hosts, notably
in the striated muscle and brain. The definitive hosts for T.
gondii are domestic and feral cats (Felis catus) and other mem-
bers of the Felidae (Dubey et al., 1970; Aramini et al., 1999;
Kenny et al., 2002). Cats also serve as intermediate hosts. Toxo-
plasma gondii undergoes sexual reproduction in the feline in-
testine, resulting in the production of millions of environmen-
tally resistant oocysts that are shed in cat feces (Dubey et al.,
1970). Rodents become infected after ingesting soil, vegetation,
or water contaminated with T. gondii oocysts. Toxoplasma gon-
dii can be transmitted horizontally between warm-blooded an-
imals, including humans, when they consume the raw or un-
dercooked flesh of intermediate or definitive hosts (Tenter et
al., 2000). Vertical transmission also occurs and is of particular
concern for humans, because infection in utero can cause abor-
tion or congenital defects in the fetus (Jones et al., 2003). Both
modes of transmission appear to be important to ensure the
survival of T. gondii in nature.
Small rodents play an important role in the life cycle of T.
gondii, because they are believed to represent the main source
of infection for domestic and feral cats. Studies of cat predation
on wildlife suggest that rodents comprise about two-thirds of
the prey consumed, although this may vary according to season,
rodent abundance, and the availability of other prey (McMurry
and Sperry, 1941; Hubbs, 1951; Eberhard, 1954; Molsher et al.,
1999). On islands where cats were not present, T. gondii in
intermediate hosts was rare or absent (Wallace, 1969; Garcelon
Received 1 June 2007; revised 19 September 2007; accepted 24 Oc-
tober 2007.
* Current affiliation: California Department of Public Health, Infant
Botulism Treatment and Prevention Program, 850 Marina Bay Pkwy,
E-361, Richmond, California 94804.
† California Department of Fish and Game, Marine Wildlife Veterinary
Care and Research Center, 1451 Shaffer Road, Santa Cruz, California
95060.
‡ Department of Medicine and Epidemiology, University of California,
Davis, California 95616.
§ Department of Population Health and Reproduction, University of
California, Davis, California 95616.
" To whom correspondence should be addressed.
et al., 1992; Dubey, Rollor et al., 1997). Conversely, on U.S.
swine farms where cats were abundant, the prevalence of T.
gondii infection in mice was significantly associated with a
higher mean density of cats on the premises (Smith et al., 1992;
Dubey, Weigel et al., 1995).
There is a paucity of data on T. gondii infection in small
mammals native to California. The most abundant rodent spe-
cies in coastal California ecosystems south of San Francisco
Bay are Peromyscus spp., including P. maniculatus (deer mice),
Microtus spp. (voles), and Perognathus spp. (pocket mice) (Ja-
meson and Peeters, 1988). It was not known which species may
be important in maintaining T. gondii infection in near shore-
dwelling wild carnivores and domestic cats, so a variety of
small mammals was included in the sampling protocol. In ad-
dition to constituting an important prey item for domestic cats,
rodents were also suitable for study because they are easily
captured, and tissues harvested at necropsy can later be tested
for the presence of parasites by immunohistochemistry or poly-
merase chain reaction (PCR).
The Morro Bay area of central coastal California was se-
lected for study because of the high proportion of southern sea
otters (Enhydra lutris nereis) with evidence of T. gondii infec-
tion (Miller et al., 2002; Conrad et al., 2005) off the adjacent
coastline. In an updated analysis of the risk for T. gondii ex-
posure in 562 live or dead otters (Conrad et al., 2005), otters
sampled between San Simeon and Morro Bay were 5 times
more likely to be infected with T. gondii than were otters living
along the more remote and rocky Big Sur coastline. An epi-
demiologic investigation of T. gondii infection prevalence in
owned and feral cats was being concurrently conducted in the
Morro Bay area (Dabritz et al., 2007). The objectives of this
study were to determine demographic and habitat-related risk
factors for T. gondii seropositivity in wild rodents and to obtain
tissue samples for future T. gondii genotyping. We hypothesized
that there would be a greater risk for T. gondii seropositivity in
wild rodents captured near residential housing, because of the
association with owned domestic cats. A previous study had
shown that T. gondii seropositivity in southern sea otters was
associated with high freshwater outflow (Miller et al., 2002),
so we also hypothesized that rodents captured near bodies of
water (Morro Bay) or water channels would be more likely to
be serologically positive.
676 THE JOURNAL OF PARASITOLOGY, VOL. 94, NO. 3, JUNE 2008
TABLE I. Toxoplasma gondii seroprevalence by species for rodents
trapped in the Morro Bay area of California.
Species No. tested
No.
seropositive*
(%)
Peromyscus maniculatus 74 25 (34)
Peromyscus boylii 29 9 (31)
Peromyscus californicus 214 50 (23)
Peromyscus truei 11 1 (9)
Spermophilus beecheyi 37 3 (8)
Mus musculus 15 0 (0)
Microtus californicus 25 0 (0)
Neotoma fuscipes 50 0 (0)
Perognathus californicus 14 0 (0)
Reithrodontomys megalotis 44 0 (0)
Rattus rattus 10 0 (0)
Total 523 88 (17)
* For Peromyscus spp., Rattus spp., N. fuscipes, R. megalotis, and M. musculus,
a positive test was defined as an IFAT titer !80; for voles, squirrels, and P.
californicus, a positive test was defined as an LAT titer !32.
MATERIALS AND METHODS
Rodent trapping
Trapping was conducted at 18 different locations in the Morro Bay
area between July 2004 and February 2006. Study sites were selected
to be either "200 m from residential housing or feral cat colonies,
where cat densities were expected to be high (n ! 13), or situated on
public lands !200 m from private residences (n ! 5). Sites were clas-
sified as being adjacent to Morro Bay (n ! 3), adjacent to riparian
habitats (n ! 6), or neither (n ! 9). Trapping occurred twice in each
wet (December–May) and dry (June–November) season. Each site was
trapped for a minimum of 2 nights. In the first trapping session of 2004
(July–August), only mice were sampled; other species were released.
Subsequently, sampled species were expanded to include rats and voles.
Trapping of S. beecheyi (California ground squirrels) began in 2005 and
was conducted at 4 sites. At 2 of these sites, ground squirrels were the
predominant species trapped. Ten to 60 traps were set per site, depend-
ing on the space available. For nocturnal rodents, kangaroo rat–sized
folding live traps (HB Sherman, Tallahassee, Florida) were set in a
transect layout 5–15 m apart within 2 hr of sunset and baited with
hamster/rat feed (Radco, San Jose, California). If temperatures were
expected to fall below 12 C, traps were bedded and over-baited. Traps
were collected within 3 hr of sunrise. At 4 sites, Tomahawk 48 # 15
# 15 cm live traps (Tomahawk Live Trap, Tomahawk, Wisconsin) were
set during daylight hours and monitored for 2–3 hr for the presence of
ground squirrels. Rodents were identified to species using a key (Ja-
meson and Peeters, 1988). The location of each captured animal was
recorded on a handheld Garmin GPS locator (Olathe, Kansas). All an-
imals were humanely handled in accordance with an animal use pro-
tocol approved by the Institutional Animal Care and Use Committee at
the University of California (UC) Davis, which is accredited by the
Association for the Assessment and Accreditation of Laboratory Animal
Care International. In 2004 and 2005, captured rodents were killed using
CO2 asphyxiation and immediately bled by intracardiac puncture. In
2006 only, animals trapped on nights 1 and 2 were bled via the retro-
orbital sinus, ear-tagged, and released; rodents captured on the third
night were killed and bled via intracardiac puncture. After collection,
whole blood was allowed to clot and was centrifuged, and serum was
aliquoted into separate vials. Serum was either frozen and transported
to UC Davis on dry ice for testing or refrigerated until transport and
shipped in coolers at 4 C. Serum was

Hasil (Bahasa Indonesia) 1:

675J. Parasitol., 94(3), 2008, ms. 675-683! American Society Parasitologists 2008FAKTOR RISIKO UNTUK INFEKSI GONDII TOXOPLASMA DALAM LIAR TIKUS DARICALIFORNIA PESISIR TENGAH DAN REVIEW T. GONDII PREVALENSIDI TIKUSHaydee A. Dabritz *, Melissa A. Miller†, Ian A. Gardner‡, Andrea E. Packham, E. Robert Atwill§ dan Patricia A. Conrad"Sekolah kedokteran hewan, Universitas California, dokter hewan Med III-A, Ruang 4206 satu perisai Avenue, Davis, California 95616.e-mail: [e-mail]: Sera dari 523 liar tikus diuji Toxoplasma gondii antibodi menggunakan baik antibodi neon langsunguji (IFAT) (tikus dan tikus, dengan titer! 80 dianggap positif) atau sebuah tes aglutinasi lateks (LAT) (tikus, Tupai dan sakutikus, dengan titer! 32 dianggap positif). Tujuh belas persen (88/523) tikus, termasuk 26% (85/328) dari PeromyscusSP. dan 8% (3/37) Spermophilus beecheyi, yang seropositif. Empat belas persen (23/161) dari tikus yang ditangkap di menjebak situs berikutnyaMorro Bay (California) dan 15% (16/109) dari tikus dari situs-situs yang berdekatan dengan habitat riparian memiliki antibodi terhadap T. gondii,dibandingkan dengan 19% (49/253) dari tikus yang ditangkap di habitat yang tidak berhubungan dengan air; perbedaan ini adalah tidak signifikan secara statistik(P! 0,32). Secara signifikan lebih sedikit tikus ditangkap "200 m dari perumahan dibandingkan dengan lokasi lebih jauh (11%vs 30%, masing-masing). Faktor-faktor yang terkait dengan peningkatan risiko untuk T. gondii seropositivity di tikus yang menangkap lokasi!200 m from residential housing and adult age.The protozoan parasite T. gondii infects most species ofwarm-blooded vertebrates, including rodents. Like humans, ro-dents serve as intermediate hosts for T. gondii. Parasites forma latent cyst stage in the tissues of intermediate hosts, notablyin the striated muscle and brain. The definitive hosts for T.gondii are domestic and feral cats (Felis catus) and other mem-bers of the Felidae (Dubey et al., 1970; Aramini et al., 1999;Kenny et al., 2002). Cats also serve as intermediate hosts. Toxo-plasma gondii undergoes sexual reproduction in the feline in-testine, resulting in the production of millions of environmen-tally resistant oocysts that are shed in cat feces (Dubey et al.,1970). Rodents become infected after ingesting soil, vegetation,or water contaminated with T. gondii oocysts. Toxoplasma gon-dii can be transmitted horizontally between warm-blooded an-imals, including humans, when they consume the raw or un-dercooked flesh of intermediate or definitive hosts (Tenter etal., 2000). Vertical transmission also occurs and is of particularconcern for humans, because infection in utero can cause abor-tion or congenital defects in the fetus (Jones et al., 2003). Bothmodes of transmission appear to be important to ensure thesurvival of T. gondii in nature.Small rodents play an important role in the life cycle of T.gondii, because they are believed to represent the main sourceof infection for domestic and feral cats. Studies of cat predationon wildlife suggest that rodents comprise about two-thirds ofthe prey consumed, although this may vary according to season,rodent abundance, and the availability of other prey (McMurryand Sperry, 1941; Hubbs, 1951; Eberhard, 1954; Molsher et al.,1999). On islands where cats were not present, T. gondii inintermediate hosts was rare or absent (Wallace, 1969; GarcelonReceived 1 June 2007; revised 19 September 2007; accepted 24 Oc-tober 2007.* Current affiliation: California Department of Public Health, InfantBotulism Treatment and Prevention Program, 850 Marina Bay Pkwy,E-361, Richmond, California 94804.† California Department of Fish and Game, Marine Wildlife VeterinaryCare and Research Center, 1451 Shaffer Road, Santa Cruz, California95060.‡ Department of Medicine and Epidemiology, University of California,Davis, California 95616.§ Department of Population Health and Reproduction, University ofCalifornia, Davis, California 95616." To whom correspondence should be addressed.et al., 1992; Dubey, Rollor et al., 1997). Conversely, on U.S.swine farms where cats were abundant, the prevalence of T.gondii infection in mice was significantly associated with ahigher mean density of cats on the premises (Smith et al., 1992;Dubey, Weigel et al., 1995).There is a paucity of data on T. gondii infection in smallmammals native to California. The most abundant rodent spe-cies in coastal California ecosystems south of San FranciscoBay are Peromyscus spp., including P. maniculatus (deer mice),Microtus spp. (voles), and Perognathus spp. (pocket mice) (Ja-meson and Peeters, 1988). It was not known which species maybe important in maintaining T. gondii infection in near shore-dwelling wild carnivores and domestic cats, so a variety ofsmall mammals was included in the sampling protocol. In ad-dition to constituting an important prey item for domestic cats,rodents were also suitable for study because they are easilycaptured, and tissues harvested at necropsy can later be testedfor the presence of parasites by immunohistochemistry or poly-merase chain reaction (PCR).The Morro Bay area of central coastal California was se-lected for study because of the high proportion of southern seaotters (Enhydra lutris nereis) with evidence of T. gondii infec-tion (Miller et al., 2002; Conrad et al., 2005) off the adjacentcoastline. In an updated analysis of the risk for T. gondii ex-posure in 562 live or dead otters (Conrad et al., 2005), otterssampled between San Simeon and Morro Bay were 5 timesmore likely to be infected with T. gondii than were otters livingalong the more remote and rocky Big Sur coastline. An epi-demiologic investigation of T. gondii infection prevalence inowned and feral cats was being concurrently conducted in theMorro Bay area (Dabritz et al., 2007). The objectives of thisstudy were to determine demographic and habitat-related riskfactors for T. gondii seropositivity in wild rodents and to obtaintissue samples for future T. gondii genotyping. We hypothesizedthat there would be a greater risk for T. gondii seropositivity inwild rodents captured near residential housing, because of theassociation with owned domestic cats. A previous study hadshown that T. gondii seropositivity in southern sea otters wasassociated with high freshwater outflow (Miller et al., 2002),so we also hypothesized that rodents captured near bodies ofwater (Morro Bay) or water channels would be more likely tobe serologically positive.676 THE JOURNAL OF PARASITOLOGY, VOL. 94, NO. 3, JUNE 2008TABLE I. Toxoplasma gondii seroprevalence by species for rodentstrapped in the Morro Bay area of California.Species No. testedNo.seropositive*(%)Peromyscus maniculatus 74 25 (34)Peromyscus boylii 29 9 (31)Peromyscus californicus 214 50 (23)Peromyscus truei 11 1 (9)Spermophilus beecheyi 37 3 (8)Mus musculus 15 0 (0)Microtus californicus 25 0 (0)Neotoma fuscipes 50 0 (0)Perognathus californicus 14 0 (0)Reithrodontomys megalotis 44 0 (0)Rattus rattus 10 0 (0)Total 523 88 (17)* For Peromyscus spp., Rattus spp., N. fuscipes, R. megalotis, and M. musculus,a positive test was defined as an IFAT titer !80; for voles, squirrels, and P.californicus, a positive test was defined as an LAT titer !32.MATERIALS AND METHODSRodent trappingTrapping was conducted at 18 different locations in the Morro Bayarea between July 2004 and February 2006. Study sites were selectedto be either "200 m from residential housing or feral cat colonies,where cat densities were expected to be high (n ! 13), or situated onpublic lands !200 m from private residences (n ! 5). Sites were clas-sified as being adjacent to Morro Bay (n ! 3), adjacent to riparianhabitats (n ! 6), or neither (n ! 9). Trapping occurred twice in eachwet (December–May) and dry (June–November) season. Each site wastrapped for a minimum of 2 nights. In the first trapping session of 2004(July–August), only mice were sampled; other species were released.Subsequently, sampled species were expanded to include rats and voles.Trapping of S. beecheyi (California ground squirrels) began in 2005 andwas conducted at 4 sites. At 2 of these sites, ground squirrels were thepredominant species trapped. Ten to 60 traps were set per site, depend-ing on the space available. For nocturnal rodents, kangaroo rat–sizedfolding live traps (HB Sherman, Tallahassee, Florida) were set in atransect layout 5–15 m apart within 2 hr of sunset and baited withhamster/rat feed (Radco, San Jose, California). If temperatures wereexpected to fall below 12 C, traps were bedded and over-baited. Trapswere collected within 3 hr of sunrise. At 4 sites, Tomahawk 48 # 15# 15 cm live traps (Tomahawk Live Trap, Tomahawk, Wisconsin) wereset during daylight hours and monitored for 2–3 hr for the presence ofground squirrels. Rodents were identified to species using a key (Ja-meson and Peeters, 1988). The location of each captured animal wasrecorded on a handheld Garmin GPS locator (Olathe, Kansas). All an-imals were humanely handled in accordance with an animal use pro-tocol approved by the Institutional Animal Care and Use Committee atthe University of California (UC) Davis, which is accredited by theAssociation for the Assessment and Accreditation of Laboratory AnimalCare International. In 2004 and 2005, captured rodents were killed usingCO2 asphyxiation and immediately bled by intracardiac puncture. In2006 only, animals trapped on nights 1 and 2 were bled via the retro-orbital sinus, ear-tagged, and released; rodents captured on the thirdnight were killed and bled via intracardiac puncture. After collection,whole blood was allowed to clot and was centrifuged, and serum wasaliquoted into separate vials. Serum was either frozen and transportedto UC Davis on dry ice for testing or refrigerated until transport andshipped in coolers at 4 C. Serum was