The ecology of tick-transmitted infections in the redwood chipmunk (Tamias ochrogenys)

Abstract

The redwood chipmunk contributes to the maintenance of tick-borne diseases in northern California. The range of redwood chipmunks overlaps that of western black-legged ticks and tick-borne disease, including granulocytic anaplasmosis and Lyme borreliosis. Chipmunks have high Anaplasma phagocytophilum PCR- and seroprevalence, are infested with a diversity of Ixodes spp. ticks, and are reservoir competent for Borrelia burgdorferi. We hypothesized that chipmunks could maintain tick-borne disease on the forest floor while also potentially bridging infection to arboreal sciurids as well. We used radio-telemetry to evaluate chipmunk movement and use of trees, characterized burrows, described prevalence of tick-borne disease, and identified ticks on these chipmunks. A total of 192 chipmunks from Hendy Woods, Mendocino County, California, USA, was evaluated between November 2005 and April 2009. The mean density was 2.26–5.8 chipmunks/ha. The longest detected life span was 3 years. Female weights ranged from 80–120 g and males from 80–180 g. The A. phagocytophilum and Borrelia spp. seroprevalence was 21.4% and 24.7%, respectively, and PCR prevalence for these pathogens was 10.6% and 0%, respectively. Ixodes spp. ticks included I. angustus, I. ochotonae, I. pacificus, and I. spinipalpis. The mean infestation level was 0.92 ticks/chipmunk. Based on telemetry of 11 chipmunks, the greatest distance traveled ranged from 0.14–0.63 km for females and 0.1–1.26 km for males. Areas occupied by chipmunks ranged from 0.005–0.24 km² for females and 0.006–0.73 km² for males. On 3 occasions, chipmunks were found in trees. Burrows were identified under a moss-covered redwood log, deep under a live redwood tree, under a Douglas fir log, in a clump of huckleberry, in a root collection from an overturned Douglas fir tree, and in a cluster of exposed huckleberry roots. The biology of the redwood chipmunk has multiple features that allow it to be an important reservoir host for tick-borne disease in northwestern California.

Introduction

The redwood or yellow-cheeked chipmunk (Tamias ochrogenys, Sciuridae, Merriam 1897) is a medium-sized, locally common rodent that was elevated from a subspecies of T. townsendii to full species status (Sutton and Nadler, 1974), supported on the basis of genetic, call, and anatomical distinction, especially in the baculum and os clitoris (Hoffman et al., 1993). These animals occupy coast redwood (Sequoia sempervirens) and mixed coniferous/Douglas fir (Pseudotsuga menziesii) forests in a restricted geographical range in California south of the Eel River in Humboldt County, south to central Sonoma County, and 40 km east, from 0–1280 m above sea level (Gannon et al., 1993). Closely related species, including T. sonomae (Sciuridae, Grinnell 1915) and T. senex (Sciuridae, Allen 1890), are found just south and north to northeast of T. ochrogenys. Within the forest, redwood chipmunks rely on dense undergrowth and downed woody debris for burrowing, feeding, and evasion of predators (Adams, 1967). We have trapped them frequently on forest floor and at 1 m height in trees in second- and old-growth redwood and Douglas fir forests as well as in campgrounds (Nieto and Foley, unpubl. data). They are active year-round [although less active from November to February (Adams, 1967)] and considered diurnal. Chipmunk populations are characterized by high densities and relatively low turnover (Brand, 1974; Carey, 1991). For some western chipmunk species, old-growth forests appear to support more individuals that move shorter distances, compared with second-growth forest (Rosenberg and Anthony, 1993).

The range of redwood chipmunks at both coarse and fine scales considerably overlaps regions of high risk for western black-legged tick (I. pacificus, Ixodidae, Cooley and Kohls 1943) infestation and human and animal tick-borne disease (Madigan, 1993; Foley et al., 2001; Eisen et al., 2003; VBDS, 2006). Two important bacteria transmitted by this tick cause granulocytic anaplasmosis (GA) and Lyme borreliosis (LB). Anaplasma phagocytophilum, which causes GA, is an emerging pathogen across the Holarctic in humans, domestic animals, and wildlife (Madigan, 1993; Greig et al., 1996; Foley, 2000; Foley et al., 2001, 2004). I. pacificus is a 3-host tick, feeding on large mammals only as adults and utilizing small mammals, reptiles, or birds in immature stages (Casher et al., 2002). Because there is no transovarial transmission of A. phagocytophilum (Munderloh and Kurtti, 1995), infection must be acquired by the tick during larval or nymphal stages before it can be transferred to humans or domestic animals by adult ticks. Reported reservoirs are white-footed mice (Peromyscus leucopus, Muridae, Rafinesque 1818) in the eastern US (Telford et al., 1996), woodrats (Neotoma fuscipes, Muridae, Baird 1857), squirrels (Sciurus spp.), and chipmunks in the western US (Nicholson et al., 1999; Foley et al., 2002; Nieto and Foley, 2008, 2009), and bank voles (Myodes glareolus, Cricetidae, Schreber 1780) and wood mice (Apodemus agrarius, Muridae, Pallas 1771) in the Old World (Blanco and Oteo, 2002; Bown et al., 2003; Cao et al., 2006).

LB is caused by the spirochete Borrelia burgdorferi sensu lato. LB can be a chronic, debilitating human disease and is the most common vector-borne disease in the US (CDC, 2007). It is often a mild disease with nonspecific clinical manifestations in infected people and dogs, but it can be severe, with arthritis, neurological or cardiac dysfunction in humans, and fatal glomerulonephritis in dogs (Shadick et al., 1994; Dambach et al., 1997). Reservoirs in North America include white-footed mice (P. leucopus), woodrats, and squirrels (Brown and Lane, 1992; Lane et al., 2005; Ostfeld et al., 2006). Some work suggests that birds may be reservoir competent as well (Ginsberg et al., 2005).

There is strong evidence that chipmunks are key players in the ecology of GA and LB. Redwood chipmunks are found in habitats where A. phagocytophilum is common, have high PCR- and seroprevalence, and are infested with a diversity of Ixodes spp. ticks (Nieto and Foley, 2009). In a large survey of northern and central coastal California, seroprevalence of up to 28% and PCR prevalence of up to 34% were detected in chipmunks (Foley et al., 2008). Five different species of ticks were found on redwood chipmunks, representing the highest tick species richness found on any rodent species in the study. Experimental infection of 7 wild-caught redwood chipmunks induced chronic waxing and waning rickettsemia (Nieto and Foley, 2009). Of 3 animals for which xenodiagnosis was attempted, all successfully infected pools of I. pacificus larvae. These findings are consistent with results from east of the Rocky Mountains, where a PCR-positive eastern chipmunk (T. striatus, Sciuridae, Linnaeus 1758) was reported from Minnesota (Walls et al., 1997). In addition to maintaining active infection, chipmunks in the eastern US support high tick numbers, with a strong correlation between chipmunk numbers and the density of nymphal deer ticks one year later (Ostfeld et al., 2006).

Several members of the genus Tamias are important in the maintenance of B. burgdorferi in nature. Introduced Siberian chipmunks (T. sibiricus, Sciuridae, Laxmann 1769) in France had a PCR prevalence of B. burgdorferi sensu lato of 33%, including 9 individuals infected with B. afzelii, one with B. burgdorferi sensu stricto, and one coinfected (Vourc’h et al., 2007). Infection has also been reported from this species in China (Wan et al., 1999). However, among 8 tested Siberian chipmunks from far eastern Russia, all were B. burgdorferi PCR-negative (Sato et al., 1996). In Illinois, eastern chipmunks (T. striatus) appear to be an important reservoir for B. burgdorferi on the basis of high prevalence by culture (26%) and high immature I. scapularis (Say 1821) tick load (Slajchert et al., 1997). Positive culture also has been obtained from eastern chipmunks in southern Connecticut and Wisconsin (Anderson et al., 1985; Godsey et al., 1987), xenodiagnosis was positive in 75% of eastern chipmunks from Massachusetts (Mather et al., 1989), and 5% of eastern chipmunks in Wisconsin were antibody-positive (Godsey et al., 1987). This species was documented reservoir competent based on positive culture in blood and ear tissue and transmission to ticks after inoculation in a laboratory setting (McLean et al., 1993). Interestingly, chipmunks are an important reservoir for relapsing fever (RF) Borrelia spp. as well (Burgdorfer and Mavros, 1970). Most RF Borrelia spp. are transmitted by argasid ticks and tend to be more common in higher altitude and drier climates than B. burgdorferi sensu lato (Beck, 1942), but human cases have been reported in central coast range (Beck, 1942; Wynns, 1942). Thus, surveillance of chipmunks from coast range sites should reasonably include assays for GA, LB, and RF as well.

We hypothesize that chipmunks have important roles maintaining sylvatic GA and LB on the forest floor while also potentially bridging infection via ticks to arboreal sciurids. To some extent, chipmunks co-occupy habitats of both dusky-footed woodrats and western gray squirrels (Sciurus griseus, Sciuridae, Ord 1818). The importance of chipmunks in the ecology of these tick-borne diseases depends not only on the prevalence of infection and tick infestation, but also on their patterns of movement and nest occupation (where they encounter ticks) and to what degree their space utilization overlaps that of other reservoirs. In this study, we used radio-telemetry to evaluate how much individual chipmunks move and their use of trees, characterize burrow sites, describe the prevalence of tick-borne disease, and identify ticks on these chipmunks that could be relevant in the ecology of tick-borne disease in north-coastal California.

Materials and methods

Trapping and sample collection

Extralarge (4 × 4.5 × 15 inch) Sherman (HB Sherman, Tallahassee, FL) live traps were set overnight in spring and fall in Hendy Woods State Park, central Mendocino County, California (39.1284 N,−123.7121 W) from November 2005 to April 2009. Predominant vegetation included redwood, true oaks (Quercus spp.), tan oak (Lithocarpus densiflorus), Douglas fir, bay laurel (Umbellularia californica), blackberry (Rubus sp.), grasses, poison oak (Toxicodendron diversilobum), and evergreen huckleberry (Vaccinium ovatum). Traps were set at locations of observed active rodent usage or dens and baited with peanut butter and oats. Rodents were anesthetized with 20 mg/kg ketamine and 3 mg/kg xylazine delivered SC, examined visually for ectoparasites and body condition, and given a permanent individually numbered metal ear tag.

Body condition was scored as 1 if the animal was emaciated, 2 if thin, 3 if in excellent condition (well-muscled over the spine, moderate fat over ribs), 4 if overweight, and 5 if obese. Blood was obtained by retroorbital abrasion and anticoagulated with EDTA. All blood was kept cool or frozen at −20 to −80°C until plasma could be separated by centrifugation. Ectoparasites were preserved in 70% ethanol. Ixodes spp. ticks were identified to species using keys (Furman and Loomis, 1984; Webb et al., 1990). Larvae were viewed under a compound microscope in a depression slide as well as with a dissecting microscope before identification was confirmed. All chipmunks were screened for current infection and exposure to A. phagocytophilum using PCR and with serology.

Serology

Plasma anti-A. phagocytophilum IgG was assayed by an indirect immunofluorescent antibody (IFA) assay (Dumler et al., 1995), using A. phagocytophilum-infected HL-60 cells as substrate, fluorescein isothiocyanate-labeled goat anti-rat heavy and light chain IgG secondary antibody (Kirkegaard & Perry, Gaithersburg, MD), and eriochrome black as a counter stain. Samples were tested starting at dilutions of 1:25, and positive and negative control sera were included on each run. Samples were considered positive if strong fluorescence was detected at dilutions of at least 80, consistent with previously published cut-off values (Dumler et al., 1995). IFA was performed for exposure to Borrelia spp. using commercially available slides (VMRD, Pullman, WA), also using the anti-rat secondary antibody, eriochrome black as a counter stain, 1:25 initial dilution, and a 1:80 cut-off. This assay will detect antibodies to both LB and RF Borrelia spp.

Nucleic acid extraction and real-time TaqMan polymerase chain reaction (PCR)

Rodents and all ticks were assessed for active infection by PCR. DNA was extracted from 200 μl of whole blood using a kit (DNeasy Tissue kit, Qiagen, Valencia, CA) according to manufacturer’s instructions. DNA was extracted from ticks by surface cleaning with 70% ethanol, followed by grinding in a mortar and pestle, and then boiling in TE buffer for 10 min. Tick extracts were diluted 1:100 in water for PCR. Real-time quantitative PCR was performed for A. phagocytophilum as previously described (Drazenovich et al., 2006) and for B. burgdorferi sensu stricto (Taqman Services, UC Davis). For both tests, each 12 μl reaction contained 5 μl DNA, 1X TaqMan Universal Master Mix (Applied Biosystems), 2 nmol each primer, and 400 pmol of probe. The thermocycling conditions consisted of 50°C for 2 min, 95°C for 10 min, and 50 cycles at 95°C for 15 s, followed by 60°C for 1 min. Samples were considered positive if they had a cycle threshold (CT) value <50 and characteristic amplification plots.

Telemetry methods

From October 2007 to July 2008, 11 chipmunks were fitted with radio-collars for observation of movement and substrate utilization. Under anesthesia, each chipmunk was fitted with a 4-g (with collar) VHF transmitter on a plastic ziptie (Wildlife Track, Caldwell, ID) which was cinched snugly around the neck. A model R1000 receiver (Communications Specialists, Orange, CA) was used to obtain locations during day and night acquisitions for at least 5 days per animal. All acquisitions were recorded by global positioning system (GPS), and a note was made of the time of acquisition, substrate type, vegetation present, and whether any burrows or other indications of animal activity were detected.

Data summary and statistical analysis

Data were maintained in Excel (Microsoft, Redmond, WA) and analyzed with the statistical package ‘R’ (R-Development Core Team, http://www.r-project.org). For all tests, the cut-off for statistical significance was P=0.05. Significant skew from random in the male:female ratio was evaluated using a chi-square test. A Wilcoxon test was used to determine whether males or females had significantly different body condition scores and a t-test evaluated whether weights were different in males and females. Population size was estimated from recaptured animals for each year separately using a Lincoln-Peterson estimator (Braun, 2005) and density calculated for the area that encompassed all points in which chipmunks were detected. Descriptive statistics were done to summarize sero- and PCR prevalence for each pathogen. Odds ratios (OR) and 95% confidence intervals (C.I.) for each disease risk factor (year, season, sex, and body condition score) were estimated using univariate logistic regression: A multivariate model was not justified because significant univariate risk factors were not detected. Whether or not coexposed individuals were statistically overrepresented compared to the probability of getting this level (or higher) of coexposed animals by chance was calculated using a Fisher exact test. Telemetry points were visualized with the software Topo! (National Geographic, Evergreen, CO). From plots, we calculated the greatest distance traveled as well as the area that encompassed all points for that animal. Distance traveled and areas were compared between male and female chipmunks by t-test.

Results

A total of 192 chipmunks was evaluated from Hendy Woods from November 2005 to April 2009. Based on recapture rates, estimated population sizes in the area sampled ranged from 99.6–256.2 with a mean and standard deviation across all years of 167.7 and 68.0, respectively (Table 1). These translate to density estimates from 2.26–5.8 chipmunks/ha (mean = 3.8/ha). Twelve chipmunks were recaptured one year after initial capture, and 3 were recaptured 2 years later. The male:female ratio was 1.35:1 which was not significantly different from 1:1 (P=0.37). The mean and variance in body condition scores were 2.2 and 0.42; scores were not significantly different between male and female chipmunks (P=0.87). Female weights ranged from 80 to 120 g and males from 80 to 180 g. Although males were a little heavier (mean weight 102.4 g) than females (101 g), this difference was not statistically significant (P=0.74).

Table 1.

Annual capture numbers and estimated population sizes for redwood chipmunks captured at Hendy Woods State Park, Mendocino County, California, from 2005 to 2009.

	Estimated population size	Estimated density (per hectare)
2005	Baseline	Baseline
2006	99.6	2.26
2007	256.2	5.8
2008	182	4.1
2009	133	3.02

Of 168 chipmunks that were evaluated by serology for A. phagocytophilum exposure, 36 were seropositive (21.4%; 95% C.I. = 15.6–28.6%). Of 141 chipmunk DNA samples available for PCR, 15 were PCR-positive for A. phagocytophilum (10.6%; 95% C.I. = 6.2–17.1%); 46.7% of PCR-positive chipmunks were also seropositive, 3 from May and 4 from August. Twenty chipmunks were captured at least one additional time ≥6 months after the initial capture date. All were initially seronegative, and 11 seroconverted to A. phagocytophilum. Of these, 5 seroconverted between fall and the next spring, 4 between spring and the next fall, and 2 between spring and the next spring. Based on logistic regression, no significant risk factors were found for A. phagocytophilum seropositivity. A. phagocytophilum PCR-positive chipmunks were more likely to be male (OR=3.8, 95% C.I. = 1.1–13.2, P=0.03). The risk of being infected with A. phagocytophilum was lowest in animals with an excellent (2+) body condition score (OR=0.07, 95 C.I. = 0.008–0.68, P=0.03).

Samples from 73 chipmunks were available for evaluation of Borrelia spp. exposure, of which 18 were positive, i.e. 24.7% (95% C.I. = 15.6–36.4%). Coexposure with A. phagocytophilum was detected in 11.6% (95% C.I. = 5.1–21.6), which is not statistically significant, i.e. the probability of being coexposed was not higher than that expected by chance given the probabilities of exposure to the 2 agents separately. No chipmunks were B. burgdorferi PCR-positive. Significant risk factors for B. burgdorferi seropositivity were not detected.

Ixodes spp. ticks on chipmunks included 4 adult, 16 nymphal, and 4 larval I. angustus (Ixodidae, Neumann 1899), 2 adult I. ochotonae (Ixodidae, Gregson 1941), 9 nymphal and 117 larval I. pacificus, and 2 nymphal I. spinipalpis (Ixodidae, Hadwen and Nuttall 1916). The mean infestation level was 0.92 ticks/chipmunk or 3.1 ticks on chipmunks that were infested. Of these ticks, only one, an I. pacificus larva, was PCR-positive for A. phagocytophilum. No ticks were PCR-positive for B. burgdorferi.

Of the 11 chipmunks fitted with radio-collars, 6 were female and 5 were male. One was A. phagocytophilum PCR-positive, 2 were seropositive, and none was positive for B. burgdorferi antibodies or DNA. Infection status for these individuals did not change over the period of observation. The extent of movement of some individual chipmunks was considerable. Often animals were found in washes and creek ravines, generally upslope from where they were captured. The greatest distance traveled ranged from 0.14 to 0.63 km (mean of 0.35) for females and 0.1–1.26 km (mean 0.54) for males, with an overall mean of 0.43 km. The difference in distance traveled between males and females was not significant (P=0.19). Areas occupied by a given chipmunk ranged from 0.005 to 0.24 km² (mean 0.08) for females and from 0.006 to 0.73 km² (mean 0.19) for males. The overall mean area was 0.13 km², and differences between males and females also were not significant (P=0.22).

On 3 occasions, chipmunks were found in trees. One animal was localized high in a redwood tree and then the next day in a Douglas fir tree. A second chipmunk was located in the morning in a Douglas fir even though the night before it was detected nearby in a burrow under a clump of evergreen huckleberry. Apparent burrows could be identified for 6 chipmunks based on repeated occupation of the same site. One was under a moss-covered large redwood log, one deep under a live redwood tree, one under a large Douglas fir log, one in a clump of huckleberry, one in a root collection from an overturned Douglas fir tree, and one in a cluster of exposed huckleberry roots. When investigated, sites where chipmunks tended to burrow consisted of numerous small interconnected tunnels dissecting under the structure (e.g. log). No nest cups were identified in any of the excavated burrows sites.

Discussion

Chipmunks are important reservoirs for tick-borne disease in the western US and likely other areas as well, but their ability to support an enzootic disease within a community and increase the risk of disease in other species (such as humans, horses, livestock, and pets) depends on ecological interactions with other vertebrates and tick vectors. In the present study, we document infection with A. phagocytophilum and Borrelia spp. in redwood chipmunks from a known enzootic site in Mendocino County, California. We show extensive infestation with known and likely vector ticks. Chipmunks have demographic features including high population density, a relatively long lifespan, and the ability to become infected for long periods of time that contribute to their reservoir competence. Additionally, the animals’ movement behaviors promote sharing of ticks with other vertebrate reservoirs and could facilitate spread of infection. Together with our previous studies documenting that redwood chipmunks can be chronically infected with A. phagocytophilum and transmit this infection via ticks to laboratory mice (Nieto and Foley, 2009), the present findings clarify the role of this species in tick-borne disease maintenance in north coastal California forest communities.

Little ecological research has been reported previously for the redwood chipmunk, although it is likely that life span and other demographic characteristics are similar between this and closely related species. The longest period chipmunks were documented to live in our data was 2 years, although this is certainly a minimum estimate. Eastern chipmunks have been reported to live 3–4 years (McLean et al., 1993). Most chipmunks at Hendy Woods were in good body condition, regardless of gender, consistent with sufficient food resources. Males tended to be larger than females but the sex ratio was 1:1. Sex ratios in other studies of chipmunks in Douglas fir forests have shown a skew towards males, although the magnitude of the skew was greater in studies in second-growth compared with old-growth (Gashwiler, 1959; Rosenberg and Anthony, 1993). Although population size estimates in the present study were crude, they establish approximate densities of 2–6 chipmunks/ha. In Oregon, T. townsendii (Sciuridae, Bachman 1839) density ranged from 0.4 to 10.3/ha, from 0.4 to 1.1 in second-growth Douglas fir forest, and 3.6–7.7 in old-growth (Rosenberg and Anthony, 1993), or 0.56–1.52/ha in another study in Douglas fir habitat (Gashwiler, 1959).

In the present study, chipmunk movement was considerable, especially along drainages. Males tended to move farther than females, consistent with previous research (Rosenberg and Anthony, 1993). The greatest single distance moved by any chipmunk in the study was 1.26 km, compared with 0.09–0.52 km in a previous study (Gashwiler, 1959) and 0.05–0.12 km in second-growth compared with 0.05–0.1 km in old-growth in another study (Rosenberg and Anthony, 1993). Statistics in the present study cannot be translated directly into home ranges due primarily to low sample size for telemetry locations. However, it is useful to compare space used in this study with reported home ranges from other studies. The space used by animals in the present study, from 0.005 to 0.73 ha, was somewhat less than the prior reports from 0.04 ha to 1.1 ha for eastern chipmunks (also typically showing larger home ranges for males compared with females) (Lacki et al., 1984). It is interesting that our approach may have allowed us to detect much of the area used by chipmunks even with relatively few locations.

Little is known about the structure of redwood chipmunk dens and nests. Dens were constructed in diverse locations such as under logs, under trees, and in root clusters. When investigated, burrow tunnels were extensive and interconnecting deep into the substrate, possibly allowing for occupation by multiple animals. The apparent use of the tree canopy was minimal. Thus, the habitat and substrate encountered by chipmunks varied from the occasional tree canopy to burrows in duff and under trees and shrubs. These are locations where chipmunks may encounter questing ticks and become infected or infect tick vectors that could then transmit the pathogen to another susceptible host.

Exposure to ticks and tick-borne infection in the present study was extensive. Approximately 25% of chipmunks were A. phagocytophilum-seropositive, of which about half were PCR-positive. It is possible that some PCR-positive chipmunks had recently been infected, but not had time to develop detectable antibody titers [which would be expected by about 14 days based on our earlier studies (Nieto and Foley, 2009)], although this could not be discerned by examining the seasonality of the PCR- and seropositive individuals. The exact sensitivity and specificity of the serology test for chipmunks is not known, although we had used this test before with experimentally infected chipmunks and found 100% sensitivity and specificity (Nieto and Foley, 2009). However, because there was not a specific anti-chipmunk secondary antibody, it is possible that some weakly seropositive chipmunks were overlooked. Exposure to Borrelia spp. was also high, suggesting that this host species could be an important player in the ecology of these pathogens in this area. Unfortunately, PCR tests for B. burgdorferi were not positive, and it was thus not possible to determine specifically what the serological test was reacting to. Possibilities include other related genospecies (e.g. B. bissettii, B. californicus) or members of the relapsing fever complex of spirochetes such as B. hermsii, all of which cross-react using IFA serology. B. bissettii is in the B. burgdorferi sensu lato complex and is present in Mendocino County but is rarely pathogenic in people (Eisen et al., 2009). Relapsing fever is rarely reported from the coast range, with the exception of the Monterey area south of the San Francisco Bay Area and in cases in far southwestern Oregon, not far north of the present study site (Beck, 1942; Wynns, 1942; Dworkin et al., 1998). Some of the apparent absence could be due to a lack of surveillance. In areas where relapsing fever occurs in California, chipmunks which are important reservoirs include T. merriami (Sciuridae, Allen 1889), T. speciosus (Sciuridae, Allen 1890), and T. amoenus (Sciuridae, Allen 1890) (Burgdorfer and Mavros, 1970), but not the redwood chipmunk in the present study. More work on Borrelia spp. infection in chipmunks in the coast range mountains of California would be valuable. It would also be interesting to examine risk of infection in chipmunks as a function of their specific movement patterns, but this was not possible with the small sample size in this study.

Numerous ticks of multiple species were recovered from chipmunks, although it was rare to find any that were PCR-positive for either pathogen. This was expected for A. phagocytophilum given that the level of this rickettsia in the tick is quite low (Foley and Nieto, 2007) and thus even a highly sensitive TaqMan PCR occasionally fails to detect very low-level infections. More surprising was the low rate of Borrelia spp. infection in these ticks, since spirochetes are often abundant in Ixodes spp. vectors (Piesman et al., 2001). Soft ticks which vector relapsing fever borreliae were not evaluated in the present study. The most likely Ornithodoros spp. tick in Mendocino County is O. coriaceus (Argasidae, Koch 1844), which is not a known vector of relapsing fever to humans, in contrast with O. parkeri (Argasidae, Cooley 1936), which possibly vectors B. parkeri in coast range sites south of the San Francisco Bay Area.

Ixodes spp. tick loads on these chipmunks were on average about 1 tick per individual. In Massachusetts, a mean of 3.2 larvae/chipmunk (T. striatus) was found (Mather et al., 1989). However, a study in France found 2130 I. ricinus (Ixodidae, Linnaeus 1758) on 29 Siberian chipmunks, for a mean of 73 ticks/individual (Vourc’h et al., 2007). An average of 2.21 nymphal I. persulcatus (Ixodidae, Schulze 1930) ticks were recovered from 14 Siberian chipmunks in northeastern China, and chipmunks were much more likely to be infested than other Lyme reservoirs in the area (A. speciosus, A. agrarius, Myodes rufocanus) (Ai et al., 1991). In studies in the US, chipmunks at times support as many or more immature ticks as other reported reservoirs such as white-footed mice (Godsey et al., 1987; Mannelli et al., 1993; Slajchert et al., 1997; Schmidt et al., 1999). Of the ticks that were recovered in the present study, most were I. angustus or I. pacificus. I. pacificus is a known vector for both GA and LB in the western US and could support enzootic transmission of these infections among chipmunks, other reservoirs such as squirrels and woodrats, and larger animals including humans and domestic animals. The importance of I. angustus in this ecology is less clear. I. angustus is a nidicolous tick that feeds on a variety of rodents and occasionally humans (Furman and Loomis, 1984). This species may be naturally infected with B. burgdorferi (Banerjee et al., 1994) and is a vector for B. burgdorferi sensu stricto (Peavey et al., 2000). Because much of the ecology of both LB and GA is similar (e.g. identical primary vectors and reservoir hosts), it is possible that this tick may also transmit infection among rodents and large mammals. Unfortunately, this work has not yet been performed.

In general, the biology of the redwood chipmunk has multiple features that allow it to be an important reservoir host for tick-borne diseases in northwestern California. It can share ticks with tree squirrels along tree trunks and canopies as well as along duff on the ground. While no chipmunks were observed in woodrat houses, locations at which they were documented by telemetry were often close to woodrat sign, and it is likely that chipmunks and woodrats could share ticks. This role of redwood chipmunks in tick-borne disease ecology can likely be extended to other chipmunks as well, given that the literature reports infection with GA and LB in Siberian and eastern chipmunks as well as tick infestation in these species. It would be interesting to know whether chipmunks that inhabit drier sites, such as Merriam’s chipmunks which often are found in chaparral habitats, could support infection with A. phagocytophilum and B. burgdorferi if they were exposed to vector ticks. Such data could help distinguish key roles of tick vectors feeding on chipmunks relative to the roles of the chipmunks themselves. Further ecological work on diverse sciurid species in areas where tick-borne disease occurs in humans and other animals would enrich our understanding of the multiple ecological factors that contribute to enzootic infections.