Prevalence and Potential Impact Of

PREVALENCE AND POTENTIAL IMPACT OF TOXOPLASMA GONDII ON THE ENDANGERED AMARGOSA VOLE (MICROTUS CALIFORNICUS SCIRPENSIS), CALIFORNIA, USA Authors: Amanda Poulsen, Heather Fritz, Deana L. Clifford, Patricia Conrad, Austin Roy, et. al. Source: Journal of Wildlife Diseases, 53(1) : 62-72 Published By: Wildlife Disease Association URL: https://doi.org/10.7589/2015-12-349

BioOne Complete (complete.BioOne.org) is a full-text database of 200 subscribed and open-access titles in the biological, ecological, and environmental sciences published by nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Complete website, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/terms-of-use.

Usage of BioOne Complete content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research.

PREVALENCE AND POTENTIAL IMPACT OF TOXOPLASMA GONDII ON THE ENDANGERED AMARGOSA VOLE (MICROTUS CALIFORNICUS SCIRPENSIS), CALIFORNIA, USA

Amanda Poulsen,1,5 Heather Fritz,2,4 Deana L. Clifford,1,3 Patricia Conrad,4 Austin Roy,1,2 Elle Glueckert,1 and Janet Foley1 1 Department of Veterinary Medicine and Epidemiology, University of California, One Shields Avenue, Davis, California 95616, USA 2 Department of Veterinary Microbiology and Pathology, Washington State University, PO Box 647040, Pullman, Washington 99164, USA 3 Wildlife Investigations Lab, California Department of Fish and Wildlife, 1701 Nimbus Road Suite D, Rancho Cordova, California 95670, USA 4 Department of Pathology, Microbiology, and Immunology, University of California, One Shields Avenue, Davis, California 95616, USA 5 Corresponding author (email: [email protected])

ABSTRACT: We investigated the prevalence of Toxoplasma gondii in 2011–15 to assess its potential threat on the endangered Amargosa vole (Microtus californicus scirpensis) in California, US. Surveillance was simultaneously performed on populations of syntopic rodent species. We detected antibodies to T. gondii in sera from 10.5% of 135 wild-caught Amargosa voles; 8% of 95 blood samples were PCR-positive for the T. gondii B1 gene, and 5.0% of 140 sympatric rodent brain samples were PCR-positive. Exposure to T. gondii did not change the probability that an animal would be recaptured in the field study. Behavioral response to domestic cat (Felis catus) and bobcat (Lynx rufus) urine was evaluated in five nonendangered Owens Valley voles (Microtus californicus vallicola) as surrogates for Amargosa voles and seven uninfected controls. Voles showed mild attraction to mouse urine and had neutral reactions to domestic cat urine whether or not infected. Time spent near bobcat urine was approximately twice as high in infected than in uninfected voles (although not statistically significant). The presence of T. gondii in wild Amargosa vole and sympatric rodent populations may hinder the endangered Amargosa vole population’s ability to recover in the wild. Key words: Amargosa vole, behavior, Microtus californicus, Toxoplasma gondii.

INTRODUCTION voles may maintain T. gondii in nature (e.g., Hejlicek and Literak 1998). The Amargosa vole (Microtus californicus Toxoplasma gondii is a zoonotic protozoan, scirpensis) is a federally and state-endangered infecting 30–40% of humans worldwide (Ten- California vole subspecies under the US ter et al. 2000). Most human infections are Endangered Species Act. It inhabits Mojave asymptomatic but can be fatal to immuno- Desert marshes only near Tecopa, Inyo compromised individuals (Nissapatorn 2009), County, California (Cudworth and Koprowski and T. gondii causes birth defects and 2010). Little research has been done on this miscarriages (Jones et al. 2003). Toxoplasma animal (US Fish and Wildlife Service gondii only sexually reproduces in felines, [USFWS] 1997), but its populations have a including cats (Felis catus) and bobcats (Lynx rufus) (Kikuchi et al. 2004). Oocysts passed in narrow niche breadth, long isolation from felid feces may remain viable in feces, water, other California voles, and low numbers or soil for months to years (Yilmaz and (Klinger et al. 2013). Habitat degradation, Hopkins 1972; Baldursson and Karanis 2011; predation, low genetic diversity, and disease Lelu et al. 2012), serving as a source of are important threats (USFWS 1997). In infection to intermediate hosts, which include 2013, Toxoplasma gondii DNA was found in most warm-blooded animals (Tenter et al. 13% of Amargosa voles and antibodies in 8% 2000). Infection leads to formation of latent (Ott-Conn et al. 2014), suggesting Amargosa tissue cysts in muscle and brain (Dubey 2010)

that are transmitted through tissue consump- Trapping and sample collection tion. Transplacental transmission is also a Small mammals were live-trapped intermittent- major route in small mammals (e.g., Marshall ly over approximately 5-d intervals in December et al. 2004). 2013, every month in 2014 except February and Low-dose tachyzoite infection can be sub- December, and January 2015 using large Sher- clinical in rats, mice, and voles, whereas man traps (HB Sherman Traps, Tallahassee, Florida, USA) in marsh groups Shoshone, One, infection with oocysts can cause muscle NC Trailer, Grimshaw, and Trans (Fig. 1). Traps lesions and locomotory or breathing difficul- were baited with peanut butter, oats, and alfalfa, ties (Dubey and Frenkel 1973; Sedlak et al. with an apple slice during summer. Traps were set 2001). Toxoplasma gondii causes reduced in covered areas and checked every 4–6 h. Voles aversion to felids in some rodents (Berdoy et were ear-tagged with numbered bands (1005-1 Monel, National Band and Tag Co., Newport, al. 2000; Vyas et al. 2007), a form of parasite- Kentucky, USA). A maximum of 0.05 mL of blood increased trophic transmission (PITT) was taken using retro-orbital abrasion (Mills et al. (Seppal¨ a¨ et al. 2008) that can increase 1995), and animals were released at the trap site. transmission to the definitive feline host. If Western harvest mice (Reithrodontomys mega- PITT occurs in voles, then T. gondii may lotis), house mice (Mus musculus), and cactus mice (Peromyscus eremicus) were euthanatized contribute to Amargosa vole population de- with a ketamine/xylazine overdose (.100 mg cline. ketamine and .10 mg xylazine) and cervical We assessed the effect of T. gondii on dislocation and kept frozen before postmortem Amargosa voles, specifically to 1) determine examination at the laboratory for collection of the prevalence of T. gondii in voles and blood from the chest cavity and brain tissue. sympatric rodents, 2) evaluate whether infect- Bobcats were assessed with remote cameras and scat surveys. We set 3–5 motion-activated ed voles from a related subspecies exhibit HyperFire PC700 cameras (Reconyx, Holmen, attraction to felids, and 3) determine whether Wisconsin, USA) around perimeters facing pred- naturally infected Amargosa voles have re- ator trails in each marsh (depending on marsh duced survival. size) for 13 mo and checked them monthly. Scat surveys were done along 200–700-m transects at the periphery of marshes monthly for 12 mo. MATERIALS AND METHODS Photographs from cameras were examined for bobcats by a trained researcher. Feces was Study area attributed to bobcats if it was 1–2.537.5–23 cm, We captured rodents in marshes near Tecopa smooth and tubular, and segmented with a twisted Hot Springs (3585202000 N, 11681305700 W) and pattern and had a blunt or slightly pointed end Shoshone (3585802300 N, 11681601600 W), Califor- (Elbroch 2003). If bobcats were observed on nia (Fig. 1). The climate is arid with average camera, bobcat scat was confirmed, or a bobcat summer temperatures of 31.2 C, winter of 10.7 sighting was reported by a resident, that marsh C, and 12.3 cm precipitation annually (NOAA was classed ‘‘bobcat present.’’ 2010). Vegetation is primarily bulrush (Schoeno- Blood was tested for T. gondii antibodies by plectus americanus), salt grass (Distichlis spica- indirect immunofluorescent assays (IFA) (Fritz et ta), rush (Juncus spp.), alkali heath (Frankenia al. 2012). Serum was diluted 1:25 in phosphate- salina), and arrow grass (Triglochin concinna). buffered saline (PBS), placed on slides, and The surrounding desert is harsh alkali playa, with incubated at 37 C for 24 h. After slides were the marshes minimally disturbed except for a hot washed in PBS, fluorescein isothiocyanate–la- spring used by people and dogs for recreation. beled anti-rat immunoglobulin G (Kirkegaard & The marshes are near town and residents’ cats Perry, Gaithersburg, Maryland, USA) diluted 1:25 may occasionally enter. Work with voles was in PBS was applied, and slides were incubated at done under University of California, Davis 37 C for 45 min. Slides were again washed with Institutional Animal Care and Use Committee PBS, stained with eriochrome black, and scanned authorization, with a State of California Scientific by two independent readers for fluorescence of Collection permit, Bureau of Land Management crescent-shaped T. gondii. DNA from brain and letter of authorization, and USFWS 10a.1.a blood was extracted using a Qiagen Blood and permit for take of endangered species. Author- Tissue kit (Qiagen, Valencia, California, USA) and ities also approved work with Owens Valley voles, used for PCR targeting the T. gondii B1 gene except that a USFWS permit was not required. (Rejmanek et al. 2010).

FIGURE 1. Sites for collection of Amargosa vole and sympatric rodent samples for Toxoplasma gondii testing. Field data were collected in 2011–15 from marsh sites in Shoshone and Tecopa (city centers marked with stars), which are adjacent towns in Inyo County, California, USA. Amargosa voles (Microtus californicus scirpensis), cactus mice (Peromyscus eremicus), house mice (Mus musculus), and harvest mice (Reithrodontomys megalotis) were collected from marshes within these areas. Marsh sites are shown in black; marshes used speciﬁcally for this study are circled by marsh group. Buildings within these areas are represented by hatched polygons. TNC¼NC Trailer.

Experimental infection the expectation that some might not survive or become infected. Power analysis indicated that six Because Amargosa voles are endangered, we infected and six uninfected voles would be used Owens Valley voles (Microtus californicus necessary to detect a difference between 25% vallicola), the most phylogenetically proximal and 50% of time spent within a bobcat odor–laced subspecies, as surrogates in the infection exper- quadrant (SD¼14; 95% conﬁdence; 80% power). iment (Neuwald 2010). Of 16 Owens Valley voles, All Owens Valley voles initially were T. gondii six were wild-caught near Bishop, California, four antibody– and PCR–negative. Four female and were born into captivity to a wild-caught dam four male voles were administered 100 T. gondii (F1), and six were bred in captivity (F2). The site sporulated oocysts in 200 lL of Clinicare liquid has montane (Microtus montanus) and Owens diet (Abbot Laboratories, Columbus, Ohio, USA) Valley voles, so we conﬁrmed species identity via oral gavage (Rejmanek et al. 2010; University using DNA sequencing (Perrine et al. 2007). The of Minnesota Research Animal Resources 2014). study was initialized with 16 animals because of Dosage was based on a previous study (Sedlak et

diagonal from home quadrants to avoid bias toward or against feline (i.e., because voles might tend to move only one quadrant from ‘‘home’’). The direction of the arena and the positioning of feline versus mouse quadrants were randomized per trial. Each vole was subject to three bobcat and three domestic cat trials in random order. Trials occurred from 0900 to 1600 hours. To start a trial, a vole was placed into the arena center facing a blank quadrant, and activity was captured on video for 24 min. The number of minutes in each quadrant, the number of times voles moved among quadrants, and the final quadrant were recorded. Between trials, tiles were soaked in 10% bleach for 15 min, rinsed, and dried. Straw was replaced before each trial, FIGURE 2. Still image showing the metal arena and the arena was wiped with 10% bleach and designed for Owens Valley vole (Microtus californicus 70% ethanol and air-dried. After trials were vallicola) behavioral trials. The arena is divided into completed, voles were euthanatized using a four quadrants (A–D). The orientation of the metal ketamine/xylazine overdose and cervical disloca- arena within the room and the orientation of scented tion, and DNA was extracted postmortem from tiles within the quadrants were randomized for each brain, heart, and spleen for B1 gene PCR. These trial. Scents included feline (bobcat [Lynx rufus]or tissues were chosen because data from other domestic cat [Felis catus]), home (preconditioned in rodents indicate that they often harbor T. gondii each vole’s home cage for previous 2 wk), blank (no (Rejmanek et al. 2010). tile), or mouse (Mus musculus). A remote camera to assess behavior remained in the same position throughout all trials. Data analysis and mapping Data were analyzed using ArcMap (Esri, Red- al. 2001) in which 1–100 oocysts produced lands, California, USA) and R (R Development nonfatal infection of the common vole (Microtus Core Team 2014). For hypothesis testing, P0.05 arvalis). Three female and four male voles were was interpreted as statistically significant. The uninoculated controls. Voles were housed indi- analysis combined data from 2013 to 2015 with vidually and given rodent chow (Teklad Global, previously published data from March 2011– Madison, Wisconsin, USA) and water daily. Voles November 2012 (Ott-Conn et al. 2014), which were monitored daily for lethargy, labored used the same marshes and all protocols. Preva- breathing, loss of appetite, and behavioral chang- lence of T. gondii exposure and infection and 95% es. confidence intervals were calculated using the R prop.test function. Marshes were grouped into Behavioral trials five groups (Fig. 1). We used univariate logistic regression to assess whether T. gondii infection of Trials were conducted after 1 mo to allow the voles was associated with species, sex, age, season, voles to develop patent infection (Rejmanek et al. marsh group, distance to surface water and 2010) in a straw-lined 76.2376.2-cm metal arena human structures, and presence of bobcat sign. divided by wooden panels into home, mouse, and Because no size criteria has been established for feline quadrants, with a ceramic tile in each aging this species, we dichotomized into imma- quadrant center (Fig. 2), extensively modified ture and mature (testicles descended or palpable, from a prior design for rats (Vyas et al. 2007). A vagina perforate). A vole was classified as fourth blank control had no ceramic tile. A home ‘‘infected’’ if it was PCR-positive, IFA-positive, tile was conditioned in each vole’s cage 2 wk or both based on literature that indicates that before trials. The mouse quadrant tile had one small mammals become chronically infected drop of house mouse urine as a baseline (Derouin and Garin 1991). A sympatric rodent ‘‘response to a foreign urine substance,’’ modeled was classified as ‘‘infected’’ if its brain tissue was after the inclusion of rabbit urine by Vyas et al. PCR-positive. Dry seasons had average rainfall (2007). Mouse urine was used because house ,0.9 cm (April–October), whereas wet seasons mice are common in the vole marshes. The feline had average rainfall .0.9 cm (November–March) quadrant tile had either one drop of bobcat urine (NOAA 2010). Bobcat sign was present or absent (Kishels Bobcat Urine, East Aurora, New York, (Elbroch 2003). Predictor variables for distance to USA) or urine from a 1-yr-old, healthy, male, surface water and human structures were calcu- neutered domestic cat. Blank quadrants were set lated using Euclidean distances (m) in ArcMap.

We used the Google Earth polygon feature prevalence of 10% (CI 0.02–0.10). Among all (Google, Mountain View, California, USA) to rodents, the Grimshaw marsh group had the create map layers of buildings. Distance to surface highest prevalence (32%), 4.6 times higher water was measured using map layers for surface water from the Topologically Integrated Geo- than marsh One (odds ratio [OR] 4.62, graphic Encoding and Referencing 2013 database P¼0.009, CI 1.429–14.91). Toxoplasma gondii (US Census Bureau 2015). Significant (P,0.1) infection in voles was not associated with univariate variables were used in a multivariable bobcat sign or proximity to surface water logistic regression (Hosmer et al. 2013). The most (Table 2). Exposure risk was reduced in voles parsimonious model was chosen backward step- wise to minimize the Akaike information criterion. living further from buildings (OR 0.73, A Hosmer–Lemeshow test was used to assess fit of P¼0.05, CI 0.53–0.99). The highest prevalence the model (Hosmer et al. 2013). of T. gondii in voles occurred in 2011 (15.6%). We used a Student’s t-test to compare the Overall, more voles (58.6%) were captured number of times infected voles were captured compared with uninfected voles. Although more than sympatric rodents (24.5%) during the elaborate statistics such as mark–recapture or wet season, and the wet season was signifi- survival analysis could help inform whether vole cantly associated with infection in rodents survivorship differed among these voles, the t-test (OR 2.8, P¼0.01, CI 1.21–6.47). Adult voles better detects simple differences between infec- (20.6%) were eight times more likely to be tion status without assumptions as to the cause. For analysis of behavioral trials, exposure to cat exposed than subadults (2.9%) (OR 8.58, and bobcat urine was analyzed separately, sum- P¼0.04, CI 1.06–69.69) (Table 2), but age ming individual 24-min trials to yield a total of 72 was not a risk factor for sympatric rodents. Sex min. By chance, a vole would be expected to was not significantly associated with vole or spend 25% of the 72 min in each quadrant. A sympatric rodent infection status. The most paired t-test was used to determine whether the seven uninfected voles exhibited differences in parsimonious multivariable logistic regression time spent near feline versus mouse urine. We model retained age and distance to buildings then calculated the time spent near feline urine as risk factors but removed season (R2¼11%, minus time near mouse urine (urine difference); a Hosmer–Lemeshow P¼0.7). negative urine difference suggests an aversion to Individual voles were recaptured 0–12 felid urine or preference for mouse urine, and a positive urine difference suggests an attraction to times during the study (Fig. 3). Average felid urine. number of recaptures did not differ between We used analysis of variance to test the infected and uninfected voles (P¼0.71). In relationship between infection status and urine fact, of 20 infected voles, three were recap- difference and analyze the number of times infected versus uninfected voles moved among tured more than five times, with a maximum quadrants. of 12 captures.

Behavioral trials RESULTS All Owens Valley voles were negative for T. Prevalence and risk factors gondii antibodies at the start of trials. Of the Voles were found in all marsh groups except eight experimentally infected voles, five were Shoshone; sympatric species were found in all confirmed by serology and PCR of brain to be marsh groups. Eight of 127 voles (8.4%, infected, and one died before the trials. confidence interval [CI] 0.04–0.16) were Because postmortem examination on that vole PCR-positive for T. gondii in blood, and 13 showed no evidence of toxoplasmosis and the (10.5%, CI 0.06–0.18) were antibody positive cause of death was not determined, this vole (Table 1). Of the 141 sympatric rodents (81 was omitted from analyses. By the end of the house mice, 36 western harvest mice, and 22 trials, all but one vole were PCR negative on cactus mice) sampled, five house mice from spleen and heart, including three brain PCR- the Trans marsh group and two western positive voles and two antibody-positive but harvest mice from the Grimshaw group were brain PCR-negative voles. Therefore, results PCR-positive in brain tissue, for a PCR were analyzed using an infected group of five

Downloaded From: https://bioone.org/journals/Journal-of-Wildlife-Diseases on 16 Sep 2019 Terms of Use: https://bioone.org/terms-of-use Access provided by Universidade de Sao Paulo (USP) POULSEN ET AL.—TOXOPLASMA GONDII IN THE ENDANGERED AMARGOSA VOLE 67 , voles by combined IFA and PCR of brain sting. results (one female, four males) and a control group of seven uninfected voles (three fe- interval Toxoplasma males, four males). Mus musculus 95% Conﬁdence , Among uninfected voles, the average across trials of urine difference (comparing time spent in quadrants with house cat or mouse urine) was positive (6.64), suggesting unin- (%) fected voles are not repelled by domestic cat urine. In contrast, the average when using Total prevalence bobcat urine was negative (9.89), indicating a tendency (although not statistically signiﬁcant) Reithrodontomys megalotis for uninfected voles to avoid bobcat urine. However, when infected voles were compared 27 (268) 10 0.07–0.14 with uninfected voles, almost twice as much time was spent in quadrants with bobcat urine (26% compared with 15%), although results

9 were not statistically signiﬁcant (Fig. 4, 10 (113) P¼0.96). Time spent in the domestic cat ) and sympatric rodents ( quadrant was comparable for infected voles (23%) and uninfected voles (23%) (P¼0.47). 0

0 (7) Infected voles moved an average of 49.0 times, and uninfected voles moved 65.7 times in trials with domestic cat urine (P¼0.257). Infected voles moved 35.8 times and unin- 9 fected voles 62.7 times when exposed to 11 (121) bobcat urine (P¼0.03). Microtus californicus scirpensis 0

0 (8) DISCUSSION

Toxoplasma gondii infection is prevalent in

32 the remaining wild population of Amargosa 6 (19) voles, representing a concern for their conservation. The Amargosa vole lives obligatorily in isolated marshes near moisture with minimal water ﬂow. These conditions could Test Grimshaw TNC One Shoshone Trans Total tested

IFACompositeIFA 4 (8) 3 (8) 0 (3) 0 (3) NA 11 (104) 6 (101)promote NA NA NA NA accumulation 5 (12) 4 (12) NA 20 (127) 13 (124) of T. NA gondii 16 11oocysts NA 0.10–0.24 in 0.06–0.18 NA NA using both PCR and IFA. The total marsh group prevalence only counts each individual rodent once, regardless of number of recaptures and subsequent te water and mud, where oocysts can remain a viable for up to 21.5 mo (Lelu et al. 2012). T. gondii Voles could experience pathology from T. gondii infection, abortion (Cowen and Wolf

) captured from 2011 to 2015 within marsh groups (Grimshaw, NC Trailer [TNC], One, Shoshone, Trans, as depicted in Fig. 1) tested for 1950), and altered behaviors that increase their risk of predation. We documented T. gondii in marshes and infection in Amargosa voles, western harvest mice, and house mice, by PCR or indirect immunoﬂuorescent assay (IFA). NA indicates that the table category does not apply to the given data. 1. Number (and percentages of all tested) of Amargosa voles ( and data suggest a possibility of attraction of infected Owens Valley voles to bobcat urine, ABLE Many of the voles were tested for T a Peromyscus eremicus gondii VoleSympatric rodentsTotal marsh group prevalence (%) PCR PCR 2 (11) 1 (8) 0 (5) 0 (3) 0 (17) 6 (76) 0 (7) NA 5 (101) 1 (8) 7 (141) 8 (95) 5 8 0.02–0.10 0.04–0.16 95% Confidence intervalsalthough not statistically 0.14–0.57 NA 0.05–0.16 signiﬁcant. NA 0.05–0.16 These 0.07–0.14

TABLE 2. Results of univariate logistic regression tests to assess potential risk factors for Toxoplasma gondii infection in Amargosa voles (Microtus californicus scirpensis) and sympatric rodents (Reithrodontomys megalotis, Mus musculus, Peromyscus eremicus) in Tecopa, California, USA, 2011–15. NA indicates that the table category does not apply to the given data.

Rodent Reference Odds 95% Conﬁdence Risk factor species group ratio interval P

Sex (male, female) Vole Male 1.59 0.58–4.40 0.36 Sympatric Female 5.48 0.61–48.86 0.12 All rodents Male 1.08 0.48–2.43 0.86 Age (adult, subadult) Vole Subadult 8.58 1.06–69.69 0.04* Sympatric Subadult 6.93106 NA 0.99 All rodents Subadult 6.89 0.88–54.16 0.61 Marsh with/without bobcat (Lynx rufus) Vole NA NA NA NA Sympatric Marshes without bobcat 1.23 0.13–11.37 0.85 All rodents Marshes with bobcat 1.88 0.23–15.36 0.55 Season (wet, dry)a Vole Dry 2.01 0.70–5.75 0.19 Sympatric Dry 2.49 0.51–12.12 0.25 All rodents Dry 2.8 1.21–6.47 0.01* Distance to water Vole NA 0.94 0.56–1.61 0.82 Distance to buildings Vole NA 0.73 0.53–0.99 0.05*

a Monthly normal values for rainfall were provided by National Oceanic and Atmospheric Administration (2010). Dry seasons consisted of months with average rainfall ,0.9 cm (April–October); wet seasons were months with average rainfall .0.9 cm (November–March). * Statistically signiﬁcant at P0.05.

data suggest toxoplasmosis could represent a PCR-positive because tachyzoites may not stressor to vole population health. remain long in circulation (Derouin and Garin Rodents, including Amargosa voles, are 1991). We cannot directly compare preva- intermediate hosts of T. gondii, maintaining lence between rodent species because of the the pathogen in tissue cysts and representing a different tests, but given that 6–7% of house source of infection for offspring and preda- mice and harvest mice were PCR-positive in tors. Toxoplasma gondii infection kinetics brain, we expect that the true infection seen in laboratory mice (Mus musculus) and prevalence in Amargosa voles might be very rats (Rattus norvegicus) vary depending on high if brain could be tested. The infection in parasite stage and strain, dosage, and host house mice is of public health concern, species (Zenner et al. 1998). In laboratory because they can be abundant and perido- mice, T. gondii tachyzoites typically dissemi- mestic and thus a source of infection for nate within 4 h of ingestion and spread via humans (Marshall et al. 2004). blood and lymphatics to muscle and neural Toxoplasma gondii infection risk differed tissue (Dubey 2010). Cysts can be detected in significantly among marshes. Before recent tissues as early as 6 d postinfection, but roadwork activity that resulted in inadvertent generally by 15–28 d postinfection (Dubey et marsh drainage, marsh One with an 11% al. 1997). We used a composite of PCR and prevalence in voles housed the majority of serology for voles to define a ‘‘case,’’ because Amargosa voles remaining in the wild. Possi- we could not access vole muscle or neural ble differences could include abundance of tissue for testing, acknowledging also that the felids or rodents, particular substrates to PCR is highly specific but there could be support oocyst persistence, or random historic undetected cross-reactivity in the serology. contingencies. Considering that Trans and Our estimate of prevalence in voles was 10%. Shoshone (a site under construction for future For sympatric rodents, we had access to brain, translocation) were the only marshes with a tissue that is more likely than blood to be confirmed bobcat sign, it is interesting that

FIGURE 3. Number of times Amargosa voles (Microtus californicus scirpensis) from California, USA, were recaptured during field sampling. Toxo- plasma gondii–infected and uninfected wild-caught Amargosa voles were compared using the percentage of recaptures from 2011 to 2015. A recapture number FIGURE 4. Results of a behavioral trial of Owens of 0 means that the animal was only caught once. Valley voles (Microtus californicus vallicola) in Cal- Values .0 mean the animal was caught that many ifornia, USA, experimentally infected with Toxoplasma additional times after the initial capture. Number of gondii oocysts and exposed to felid scent (domestic cat recaptures between infected and uninfected Amargosa [Felis catus] or bobcat [Lynx rufus]). The study arena voles was not significantly different (t-test; P¼0.71). was divided into four quadrants (Fig. 2). Each vole experienced six 24-min trials: three with domestic cat urine in the feline urine quadrant and three with Shoshone had no evidence of T. gondii. bobcat urine in the feline urine quadrant. The amount However, although positive confirmation of of time spent in each quadrant was summed by trial bobcats was reliable, marshes where bobcats type (bobcat or cat) for a total of 72 min per vole per were not detected may constitute false nega- trial. The average percentage of time calculated for each infected and uninfected vole and separated by tives, so we may have underestimated the trial type is shown. probability that bobcats served as a risk factor for T. gondii exposure. Voles that lived farther limiting vole recovery through attraction to from buildings showed lower exposure to T. felid predators. Data suggest no reaction to gondii, consistent with pet cats, which some- domestic cat or bobcat urine and attraction to times venture into marshes (A. Roy, California house mouse urine, regardless of infection Fish and Wildlife Service pers. comm.), as a status, but moderate increases in time spent in source. Future studies should assess for T. a quadrant with bobcat urine and reduction in gondii in bobcats and domestic cats directly, times moved when infected. The lack of and wildlife managers might consider preda- statistical significance may be due to sample tor exclusion infrastructure to limit felid size, true lack of effect, or heterogeneity in the access and T. gondii contamination in vulner- study animals. Lack of an effect may be able regions. affected by oocyst dose or use of a proxy On univariate analysis, age was a significant species. Susceptibility to T. gondii varies factor for infection in voles, suggesting across species and infection intensity (Dubey cumulative risk over time, and the wet season 2010). For example, rats tolerate higher oocyst had higher risk. These marshes expand with doses (Dubey 1996), whereas brown hares rain, providing more habitat and facilitating (Lepus europaeus) have high mortality at very parasite persistence. low oocyst doses (Sedla´k et al. 2000). There The behavioral trials on experimentally was likely heterogeneity in exposure history infected voles were designed to inform our and age of wild-caught voles used. Because knowledge on the role T. gondii may play in most of the subjects were wild-caught, we

screened using serology and PCR; however, of infection-related morbidity. Although these serology may sometimes be negative with results should be explored with larger sample congenital transmission, whereas blood is sizes and more clinically significant infections, often negative later in infection (Rejmanek they do suggest that risk aversion behavior in et al. 2010). Therefore, we perhaps did not infected endangered Amargosa voles may vary detect all infected individuals. The voles used from uninfected voles. in our trials included those reared in the Amargosa voles are profoundly endangered colony and known to be 5 mo old and wild- and occupy a very remote, patchy, and caught animals presumed to be at least 10 mo specialized habitat deep in the Mojave Desert. old. Overall most voles appeared weakly They face considerable challenges associated infected, with PCR of heart and spleen with disease, predation, competition, and negative for T. gondii, except one individual, habitat loss. As marshes are restored and possibly because tissues were not homoge- managed for Amargosa vole conservation, this neously infected, and cysts may have been study provides valuable information about T. unevenly dispersed in tissues. Other voles gondii prevalence and possible behavioral were antibody- or PCR-positive in brain, but changes due to parasitism. Combining the not both. The antibody-positive voles should effect of T. gondii with the pressures of have had time for infection to develop in predation, PITT, habitat destruction, and brain, because they were tested after almost 2 competition with sympatric and invasive mo, much longer than the 6–28 d required for species could inhibit species recovery. Limit- tissue cysts to form in mice (Dubey et al. ing invasive house mice, bobcat–vole contact, 1997). and domestic cat access to wild marshes could Although Owens Valley voles are the closest reduce the prevalence of T. gondii in the known genetic relative to Amargosa voles, marshes and improve Amargosa vole recovery. they are still a distinct subspecies with different habitat requirements. In general, ACKNOWLEDGMENTS Owens Valley voles occupy a wider habitat breadth than Amargosa voles to include drier Funding for this study was provided by the US Fish and Wildlife Service Traditional Section 6 shrub-dominated zones in addition to marsh Grant program, California Department of Fish (Nelson et al. 2006). Our goal was to use and Wildlife, and the Bureau of Land Manage- laboratory exposure to reflect influences on ment. We thank the California Department of Amargosa voles in the wild, but in reality, wild Fish and Wildlife, Wildlife Investigations Lab, for behaviors such as predator avoidance are also field support, Kendra McAlister for assistance with behavior trials, Risa Pesapane for animal influenced by cover, sounds, and interactions care, Andrea Packham for experimental infection with other organisms, in addition to scent support, Patrick Foley for statistics consultation, stimuli. Trial findings indicated that urine in and Alex Blaney and Alan Mendoza for help with general was a source of attraction or curiosity postmortem examinations. to voles. Felid (particularly bobcat) urine influenced infected voles disproportionately LITERATURE CITED to uninfected voles, and infected voles moved Baldursson S, Karanis P. 2011. Waterborne transmission fewer times than uninfected voles. The reason of protozoan parasites: Review of worldwide out- for the response difference between domestic breaks—An update 2004–2010. Water Res 45:6603– 6614. cat and bobcat trials is not clear. Likely Berdoy M, Webster JP, Macdonald DW. 2000. Fatal contributors include low sample size, low attraction in rats infected with Toxoplasma gondii. level of infection, use of a strain that was not Proc Biol Sci 267:1591–1594. infectious for voles, and responses to bobcats Cowen D, Wolf A. 1950. Experimental congenital being more likely found in nature than toxoplasmosis II, transmission of toxoplasmosis to the placenta and fetus following vaginal infection in domestic cats. The number of times voles the pregnant mouse. J Exp Med 92:403–416. moved can be interpreted as predator avoid- Cudworth NL, Koprowski JL. 2010. Microtus californicus ance behavior or a change in mobility because (Rodentia: Cricetidae). Mamm Species 42:230–243.

Derouin F, Garin YJ. 1991. Toxoplasma gondii: Blood and http://www.ncdc.noaa.gov/cdo-web/datatools/ tissue kinetics during acute and chronic infections in normals. Accessed June 2015. mice. Exp Parasitol 73:460–468. Nelson FC, Morrison ML, Lopez RR, Smeins FE, Silvy Dubey J, Frenkel J. 1973. Experimental Toxoplasma NJ. 2006. Ecology of Owens Valley vole (Microtus infection in mice with strains producing oocysts. J californicus vallicola). West N Am Naturalist 66:529– Parasitol 59:505–512. 534. Dubey J, Speer C, Shen S, Kwok O, Blixt J. 1997. Oocyst- Neuwald JL. 2010. Population isolation exacerbates induced murine toxoplasmosis: Life cycle, pathoge- conservation genetic concerns in the endangered nicity, and stage conversion in mice fed Toxoplasma Amargosa vole, Microtus californicus scirpensis. Biol gondii oocysts. J Parasitol 83:870–882. Conserv 143:2028–2038. Dubey JP. 1996. Pathogenicity and infectivity of Toxo- Nissapatorn V. 2009. Toxoplasmosis in HIV/AIDS: A plasma gondii oocysts for rats. J Parasitol 82:951–956. living legacy. Southeast Asian J Trop Med 40:1158– Dubey JP. 2010. Toxoplasmosis of animals and humans. 1178. 2nd Ed. CRC Press, Taylor & Francis Group, Ott-Conn CN, Clifford D, Branston T, Klinger R, Foley J. Beltsville, Maryland, 314 pp. 2014. Pathogen infection and exposure, and ectopar- Elbroch M. 2003. Mammal tracks & sign: A guide to asites of the federally endangered Amargosa vole North American species. Stackpole Books, Mechan- (Microtus californicus scirpensis), California, USA. J icsburg, Pennsylvania, pp. 485–487. Wildl Dis 50:767–776. Fritz HM, Buchholz KR, Chen X, Durbin-Johnson B, Perrine JD, Pollinger JP, Sacks BN, Barrett RH, Wayne Rocke DM, Conrad PA, Boothroyd JC. 2012. Tran- RK. 2007. Genetic evidence for the persistence of the scriptomic analysis of Toxoplasma development critically endangered Sierra Nevada red fox in reveals many novel functions and structures speciﬁc California. Conserv Genet 8:1083–1095. to sporozoites and oocysts. PLoS One 7:e29998. R Development Core Team. 2015. R: A language and Hejlicek K, Literak I. 1998. Long-term study of Toxo- environment for statistical computing. R Foundation plasma gondii prevalence in small mammals (Insec- for Statistical Computing, Vienna, Austria. http:// tivora and Rodentia). Folia Zool 47:93–102. www.R-Project.org/. Accessed July 2016. Hosmer DW Jr, Lemeshow S, Sturdivant RX. 2013. Rejmanek D, Vanwormer E, Mazet JA, Packham AE, Applied logistic regression. 3rd Ed. John Wiley & Aguilar B, Conrad PA. 2010. Congenital transmission Sons, Hoboken, New Jersey, pp. 157–168. of Toxoplasma gondii in deer mice (Peromyscus Jones JL, Kruszon-Moran D, Wilson M. 2003. Toxoplasma maniculatus) after oral oocyst infection. J Parasitol gondii infection in the United States, 1999–2000. 96:516–520. Emerg Infect Dis 9:1371–1374. Sedla´k K, Litera´k I, Faldyna M, Toman M, Bena´k J. 2000. Kikuchi Y, Chomel BB, Kasten RW, Martenson JS, Swift Fatal toxoplasmosis in brown hares (Lepus euro- PK, O’Brien SJ. 2004. Seroprevalence of Toxoplasma paeus): Possible reasons of their high susceptibility to gondii in American free-ranging or captive pumas the infection. Vet Parasitol 93:13–28. (Felis concolor) and bobcats (Lynx rufus). Vet Para- sitol 120:1–9. Sedlak K, Literak I, Pavlasek I, Benak J. 2001. Suscep- Klinger R, Cleaver M, Anderson S, Maier P, Clark J. 2013. tibility of common voles to experimental toxoplasmo- Short-term population dynamics, demography, and sis. J Wildl Dis 37:640–642. habitat selection by the Amargosa vole. Report to the Seppal¨ a¨ O, Valtonen ET, Benesh DP. 2008. Host BLM. United States Geological Survey, Western manipulation by parasites in the world of dead-end Ecological Research Station, Bishop, California, 72 predators: Adaptation to enhance transmission? Proc pp. Biol Sci 275:1611–1615. Lelu M, Villena I, Darde ML, Aubert D, Geers R, Dupuis Tenter AM, Heckeroth AR, Weiss LM. 2000. Toxoplasma E, Marnef F, Poulle ML, Gotteland C, Dumetre A, et gondii: From animals to humans. Int J Parasitol 30: al. 2012. Quantitative estimation of the viability of 1217–1258. Toxoplasma gondii oocysts in soil. Appl Environ University of Minnesota Research Animal Resources. Microbiol 78:5127–5132. 2014. Oral gavage in rodents. 2nd Ed. 30 April 2014, Marshall P, Hughes J, Williams R, Smith J, Murphy R, 3 pp. https://www.ahc.umn.edu/rar/documents/ Hide G. 2004. Detection of high levels of congenital RARClassHandoutforVS8008OralGavageinRodents. transmission of Toxoplasma gondii in natural urban pdf. Accessed March 2016. populations of Mus domesticus. Parasitology 128:39– US Census Bureau. 2015. 2013 TIGER Geodatabases. 42. https://www.census.gov/geo/maps-data/data/ Mills JN, Childs JE, Ksiazek TG, Peters C, Velleca WM tiger-geodatabases.html. Accessed August 2016. 1995. Methods for trapping and sampling small USFWS (US Fish and Wildlife Service). 1997. Recovery mammals for virologic testing. US Dept Health and plan for the Amargosa vole (Microtus californicus Human Services, Public Health Service, CDC, scirpensis). US Department of Interior, Fish and Atlanta, Georgia, 61 pp. Wildlife Service, Region One, Portland, Oregon, 43 NOAA (National Oceanic and Atmospheric Administra- pp. https://www.fws.gov/carlsbad/SpeciesStatusList/ tion). 2010. 1981–2010 normals, Shoshone, CA. RP/19970915_RP_AMVO.pdf. Accessed July 2016.

Vyas A, Kim SK, Giacomini N, Boothroyd JC, Sapolsky Zenner L, Darcy F, Capron A, Cesbron-Delauw M-F. RM. 2007. Behavioral changes induced by Toxoplas- 1998. Toxoplasma gondii: Kinetics of the dissem- ma infection of rodents are highly speciﬁc to aversion ination in the host tissues during the acute phase of cat odors. Proc Natl Acad of Sci U S A 104:6442– of infection of mice and rats. Exp Parasitol 90:86– 6447. 94. Yilmaz SM, Hopkins SH. 1972. Effects of different conditions on duration of infectivity of Toxoplasma Submitted for publication 30 December 2015. gondii oocysts. J Parasitol 58:938–939. Accepted 30 May 2016.

Downloaded From: https://bioone.org/journals/Journal-of-Wildlife-Diseases on 16 Sep 2019 Terms of Use: https://bioone.org/terms-of-use Access provided by Universidade de Sao Paulo (USP)