The ecology of Lyme borreliosis is complex in northwestern California, with several potential reservoir hosts, tick vectors, and genospecies of Borrelia burgdorferi sensu lato. The primary objective of this study was to determine the fine-scale spatial distribution of different genospecies in four rodent species, the California ground squirrel (Otospermophilus beecheyi), northern flying squirrel (Glaucomys sabrinus), dusky-footed woodrat (Neotoma fuscipes), and Allen’s chipmunk (Neotamias senex). Rodents were live-trapped between June 2004 and May 2005 at the Hoopa Valley Tribal Reservation (HVTR) in Humboldt County, California. Ear-punch biopsies obtained from each rodent were tested by polymerase chain reaction (PCR) and sequencing analysis. The programs ArcGIS and SaTScan were used to examine the spatial distribution of genospecies. Multinomial log-linear models were used to model habitat and host-specific characteristics and their effect on the presence of each borrelial genospecies. The Akaike information criterion (AICc) was used to compare models and determine model fit. Borrelia burgdorferi sensu stricto was primarily associated with chipmunks and B. bissettiae largely with woodrats. The top model included the variables “host species”, “month”, and “elevation” (weight = 0.84). Spatial clustering of B. bissettiae was detected in the northwestern section of the HVTR, whereas B. burgdorferi sensu stricto was clustered in the southeastern section. We conclude that the spatial distribution of these borreliae are driven at least in part by host species, time-of-year, and elevation.
Citation: Hacker GM, Brown RN, Fedorova N, Girard YA, Higley M, Clueit B, et al. (2018) Spatial clustering of Borrelia burgdorferi sensu lato within populations of Allen's chipmunks and dusky-footed woodrats in northwestern California. PLoS ONE 13(4): e0195586. https://doi.org/10.1371/journal.pone.0195586
Editor: Adriana Calderaro, Universita degli Studi di Parma, ITALY
Received: August 1, 2017; Accepted: March 26, 2018; Published: April 10, 2018
Copyright: © 2018 Hacker et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This research was supported by a National Institutes of Health grant AI22501 to Robert Lane. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. There was no additional external funding received for this study.
Competing interests: The authors have declared that no competing interests exist.
Lyme borreliosis is currently recognized as the most commonly reported vector-borne disease of humans in the United States . It is caused by a group of more than 20 related genospecies of bacteria referred to as Borrelia burgdorferi sensu lato (s. l.) that are transmitted to vertebrate hosts via the bite of Ixodes spp. ticks [2–4]. In North America, there is considerable diversity among genospecies within the B. burgdorferi s. l. complex, but only a small subset is known to cause Lyme borreliosis in humans [5–11]. Recent studies have characterized previously ungrouped genospecies belonging to the B. burgdorferi s. l. group, such as B. bissettiae [12–13], B. californiensis , B. carolinensis , B. americana , B. kurtenbachii [4, 17], and B. lanei . The only genospecies known to cause Lyme borreliosis in humans in North America was B. burgdorferi sensu stricto (s. s.) until recently [10, 19]. That changed in 2011, when B. bissettiae DNA was detected in human sera and subsequently in cardiac tissue indicating its importance as a potential disease-causing agent in California and the southeastern United States [20–21]. More recently, B. americana and B. andersonii were incriminated as human pathogens in the southeastern United States , and B. mayonii was found to cause clinical disease in the upper Midwestern United States . According to Clark et al. , B. americana-like strains have also been recovered from patients in the northeastern, northwestern, southwestern, and southeastern United States. The latter findings await confirmation, however.
In California, the ecology of Lyme borreliosis spirochetes is more complex, and therefore less understood, than its ecology in the northeastern United States [3, 22–24] because of multiple potential reservoir hosts, tick vectors, and genospecies of B. burgdorferi s. l. [7, 13, 19, 25–31]. In the far-western United States, B. burgdorferi s. s. is transmitted to humans by the western black-legged tick (Ixodes pacificus) [32–33]. Other enzootic cycles involving ticks and spirochetes other than I. pacificus and B. burgdorferi s. s. have been described [24–26]. For example, Ixodes spinipalpis was initially thought to contribute to the enzootic maintenance of B. burgdorferi s. s., but that non-human-biting tick is now known to be more important in the enzootic maintenance of B. bissettiae [24–25, 27, 34–35].
Several species of rodents are important reservoirs of B. burgdorferi s. l. in California including the California kangaroo rat (Dipodomys californicus), the western gray squirrel (Sciurus griseus), and the dusky-footed woodrat. The dusky-footed woodrat is the primary reservoir host in an enzootic cycle involving B. bissettiae and I. spinipalpis [25, 27–28, 31, 34–35]. Similarly, there is a high prevalence (30% to 80%) of infection of B. burgdorferi s. s. in western gray squirrels in oak-woodland habitats in northwestern California [30, 36–37], which suggests that other sciurid species occupying similar niches likewise may support cycles of B. burgdorferi s. s. in ecologically diverse California. One such group, the chipmunks (Neotamias spp.) are likely candidates–they are active diurnally and occupy a range of habitats from coastal rain forests to arid sagebrush steppe, many of which overlap the biotopes of B burgdorferi s. l. and I. pacificus [38–39]. Both redwood (Neotamias ochrogenys) and Siskiyou (Neotamias siskiyou) chipmunks reportedly host B. burgdorferi s. l. in California , hinting at their possible importance in the maintenance of such spirochetes. For rodents that inhabit forested environs, any activity that fragments the landscape and alters the composition of a forest (e.g., logging, fire, or disease) also may affect the distribution and density of rodents, their associated ticks, and pathogens [38–39, 41–51].
Forestry practices in California have created a mosaic-like landscape that differs dramatically from that prior to European settlement . Such a patchwork of habitats doubtless has influenced the contemporary ecology of B. burgdorferi s. l [48–49]. Landscapes managed for timber generally consist of forest units classified by the number of years since the most recent harvest. Structural stage is defined as the stage of a forest unit that is related to the size and orientation (vertical and horizontal) of trees and tree limbs . Forest management practices often result in complex forest structure depending on the intended goals of the management, ranging from clear-cut units and units with simple structure to older forest units with complex structure. The response of many rodent communities to forest management practices depends on a variety of factors including plant species composition and forest structure [43–46], the amount of downed woody debris , and stand age since last harvest. These factors also may impact the presence or absence of pathogens in rodent communities. In New York State, for example, a greater prevalence of B. burgdorferi s. l. in Ixodes scapularis nymphs was associated with smaller forest units, which may have a direct impact on the prevalence of B. burgdorferi s. l. in rodents .
The focal nature of many vector-borne diseases is typically due to the interactive effects of various biotic and abiotic processes resulting in temporal and spatial clustering in hosts and vectors. In California, the density of questing I. pacificus nymphs irrespective of their infection status with B. burgdorferi s. l., was found to vary both temporally and spatially [49–50]. This indicates that risk of human exposure to this pathogen is driven in part by location and time-of-year. By extending these findings to rodents, we ask whether B. burgdorferi s. l. similarly may be clustered within rodent populations. Hence, to further evaluate the role of rodents in the ecology of B. burgdorferi s. l. in forested areas, we evaluated environmental factors that may affect the distribution of Lyme borreliosis spirochetes within known or potential rodent reservoirs in northwestern California. To that end, we sought to compare the prevalence of B. burgdorferi s. l. among rodent species at the Hoopa Valley Tribal Reservation (HVTR) in northwestern California to determine whether rodents infected with B. burgdorferi s. l. are geographically clustered, and to ascertain if host specific and/or forest structure attributes influence the prevalence and spatial distribution of B. burgdorferi s. l.
Materials and methods
The HVTR (Latitude = 41.049141, Longitude = -123.677064) encompasses approximately 360 km2 in northern Humboldt County. The climate is characterized by cool, wet winters and warm, dry summers. The reservation is bisected by the Trinity River that runs south to north and flows into the Klamath River to the north. Elevation ranges from 100 m above sea level in the Trinity River valley to more than 1,500 m above sea level along its eastern border. The principal habitat type is montane hardwood-conifer forest with dominant tree species including Douglas fir (Pseudotsuga menziesii), tanoak (Notholithocarpus densiflorus), incense cedar (Calocedrus decurrens), Pacific madrone (Arbutus menziesii), California black oak (Quercus kelloggii), and Oregon white oak (Q. garryana). Evergreen huckleberry (Vaccinium ovatum) and salal (Gaultheria shallon) are the dominant understory shrubs . Recent forestry practices guided by the tribe’s Forest Management Plan include the retention of snags (dead trees) and large trees, and “thin and release” practices in which a subset of trees are thinned from a forest unit to decrease competition and accelerate tree growth. The result of these forestry practices is a mixture of old growth units (more than 80 years post-harvest) and second growth units (up to 80 years post-harvest) having increased structure compared to typical clear-cuts (Table 1). Access to the HVTR was granted by Hoopa Tribal Forestry and the Hoopa Tribal Council.
Small mammal sampling
Rodents were trapped during 24 sampling occasions from June 2004 to May 2005. Sampling occasions lasted from 1–5 days, depending on weather, and all sites were sampled only once. Traps (Tomahawk brand, Tomahawk Live Trap Company, Tomahawk, WI; and Sherman brand, H. B. Sherman Traps, Tallahassee, FL) were placed at locations that maximized the probability of capturing the following rodent species: the dusky-footed woodrat, Allen’s chipmunk, and the California ground squirrel. Typically, one Sherman and one Tomahawk trap was set per sampling site near rodent burrows. Traps were checked at dawn for overnight sampling sessions and were checked every four hours during daytime sampling. Variation in the size of sampling sites prevented the use of equal numbers of traps in all locations. All animal handling methods were approved by the Humboldt State University Institutional Animal Care and Use Committee. Captured animals were anesthetized with a subcutaneous injection of ketamine hydrochloride (40 mg/kg). Two 2-mm ear punch biopsies (EPBs) were collected from each captured animal (one per pinna) and stored in 95% ethanol until they could be tested for presence of borreliae. Species, sex, age, morphometric measurements, and location of capture (GPS coordinates) were collected from each rodent. Monel metal-fingerling tags (National Band and Tag Company, Newport, KY) were used to uniquely mark both ears of all animals captured. Animals were released at their sites of capture following recovery from anesthesia. Recaptured animals were noted and immediately released at their point of capture. Data on tick abundance on hosts and infection status were not gathered during this study.
Molecular analyses were conducted in the Lane Laboratory at the University of California, Berkeley. Genomic DNA was extracted from ear tissues using the DNeasy Blood and Tissue Kit (Qiagen, Valencia, CA) according to the manufacturer’s protocol with overnight digestion in ATL buffer with proteinase K at 56°C. Amplification of the 5S-23S rRNA intergenic spacer region of Borrelia burgdorferi s. l. was performed as described previously . Multiple negative (deionized water) and positive controls (B. burgdorferi isolate CA4) were included with each PCR run. Products from positive samples were purified using the QIAquick PCR purification kit (Qiagen, Valencia, CA). Sequencing was performed at the University of California, Berkeley DNA Sequencing Facility using previously described nested PCR primers . Sequences were assembled and manually edited using Sequencher 4.5 (Gene Codes Corp., Ann Arbor, MI). A 158 base-pair PCR product was used for phylogenetic analysis by the neighbor-joining method implemented in PAUP* 4.0 (uncorrected p distances)  and sequences were determined to belong to Borrelia genospecies based on similarity to sequences available in GenBank (National Center for Biotechnology Information, 2008). Amplicons that could not be classified to genospecies due to mixed sequences were eliminated from further phylogenetic analysis and categorized as unclassified B. burgdorferi s. l.
The programs ArcGIS (version 10.0, Environmental Systems Research Institute, Inc., Redlands, CA) and SaTScan (version 9.0, www.satscan.org, accessed 3 January 2012) were used to assess the spatial relationships between hosts and associated borreliae . SaTScan uses a spatial scan statistic to identify the most likely clusters of rodents infected with B. burgdorferi s. s. or B. bissettiae. An elliptical scanning window was used to include clusters with shapes other than circles to allow for differences in geography (i.e., ellipses that run along riparian corridors or mountain ridges). The Bernoulli model was used to assess clustering of each borrelial genospecies within rodent hosts. To evaluate both small and large clusters, the maximum number of capture locations (i.e., population) within a cluster was set from 35% to 50% of the total samples and an alpha of 0.05 was set as the criterion for significance in all tests. By setting the maximum population size to 35% of the total, we restricted the size of clusters so they do not cross the Trinity River. ArcGIS was used to map statistically significant clusters and individual locations of borrelial infections in rodents across the HVTR.
Habitat and host specific analyses
An information-theoretic approach using a priori-selected multinomial log-linear models (program R, version 3.1.1) and predictor variables was used to evaluate the effects of landscape and host-specific variables on the presence or absence of B. burgdorferi s. s. and B. bissettiae in rodent hosts. Variables used as predictors of infection status included the elevation of capture site, month of capture, forest unit of capture (i.e., structural stage), dominant overstory vegetation, dominant understory vegetation, area of the forest unit surrounding the capture location (m2), host species, host sex, host age (juvenile, subadult, or adult), and host weight. Rodent species that yielded no PCR-positive results were subsequently removed from further analysis. The variable “forest unit” consisted of seven categories corresponding to differences in structural stage (Table 1).
Akaike's Information Criterion corrected for small sample sizes (AICc) was used to select the best fitting model with the greatest predictive power. From the top ranked model, coefficients of variation were analyzed along with standard errors to determine which variables positively or negatively influenced the prevalence of each genospecies. The predicted probabilities of infection class (i.e., not infected, infected with B. burgdorferi s. s., or B. bissettiae) were calculated from the exponentiated coefficients (i.e., relative risk ratios) of individual variables obtained from the output of the top model. The predicted probabilities produced by the top model were used to determine the association between combinations of predictor variables and their effect on the probability of infection of an individual rodent.
A total of 284 rodents were trapped and tested including 74 California ground squirrels, seven northern flying squirrels, 114 dusky-footed woodrats, and 84 Allen’s chipmunks (hereinafter referred to as woodrats and chipmunks) (Fig 1). All California ground squirrels and northern flying squirrels were PCR negative for B. burgdorferi s. l. and excluded from subsequent analyses (Table 1). Overall, the prevalence of woodrats and chipmunks PCR-positive for B. burgdorferi s.l. was similar (Table 2). However, woodrats had a higher prevalence of B. bissettiae, whereas chipmunks had a higher prevalence of B. burgdorferi s. s. (Table 2).
Each location represents one to many individuals because of overlapping capture locations.
A total of 55 chipmunks and 74 woodrats were trapped in the northwestern section of the HVTR. A greater proportion of woodrats (n = 23, 31.1%) were positive for B. bissettiae as compared to chipmunks (n = 7, 12.7%). A cluster of 22 woodrats and 7 chipmunks PCR-positive for B. bissettiae (total population, 104) was detected in the northwestern section (RR = 9.01, P<0.001) (Fig 2). Five (9.1%) chipmunks also were positive for B. burgdorferi s. s in the northwestern section. All six unclassified B. burgdorferi s. l. samples were collected from woodrats in the northwestern section.
A total of 29 chipmunks and 45 woodrats were trapped in the southeastern section of the HVTR. Nine of 29 (31.0%) chipmunks trapped in the southeastern section were positive for B. burgdorferi s. s. and none were positive for B. bissettiae. Additionally, four (8.9%) of 45 woodrats were positive for B. burgdorferi s. s. No woodrats trapped in the southeastern section of the HVTR were positive for B. bissettiae. A cluster of 9 chipmunks and 3 woodrats PCR-positive for B. burgdorferi s. s. (total population, 42) was detected in the southeastern section including (RR = 8.1, P = 0.004) (Fig 2).
Habitat and host specific analyses
The top ranked model accounted for 84% of the cumulative weight of a model set that consisted of 32 models and included the variables “species”, “month”, and “elevation” (Tables 3 and 4, S3 Table). The fourth ranked model had the largest model likelihood (i.e. greatest fit) versus the fully parameterized model and included the same variables as the top model, with the addition of the “forest unit” variable.
Models were ranked via a corrected Akaike’s Information Criterion (AICc) value. The K* column indicates the number of parameters for each model.
The abbreviations Bbss and Bbis stand for Borrelia burgdorferi sensu stricto and Borrelia bissettiae, respectively. The term “Chipmunk” refers to the Allen’s Chipmunk (N. senex).
Analyses of relative risk ratios and predicted probabilities from the top ranked model demonstrate that, compared to chipmunks, woodrats were more often associated with infection with B. bissettiae than not infected (RR = 2.6, Table 4). Compared to woodrats, chipmunks were more often associated with B. burgdorferi s. s. infection than not infected (RR = 12.5, Table 4).
Elevation and month of capture contributed significantly to the top model (Tables 3 and 4). Woodrats were trapped at elevations ranging from 175 m to 964 m (average = 723 m). Chipmunks were trapped at elevations ranging from 447 m to 961 m (average = 704 m). Additionally, the average elevation of rodents trapped in the northwestern section of the HVTR was 658 m while the average elevation of rodents trapped in the southeastern section of the HVTR was 818 m. The average elevation at which B. burgdorferi s. s.-positive rodents were trapped throughout the HVTR was 827 m (range = 638–961 m), whereas B. bissettiae-positive rodents (all in the northwestern section of HVTR) were trapped at an average elevation of 643 m (range = 514–923 m). Borrelia burgdorferi s. s.-positive rodents in the southeastern section were trapped at an average elevation of 891 m. Our top model revealed that, as elevation increased the relative risk of observing a rodent infected with B. burgdorferi s. s. also increased (RR = 1.01, Table 4). This pattern was amplified when examining the predicted probabilities of infection with B. burgdorferi s. s., given a chipmunk host, and increasing elevation (Fig 3). The relative risk of a rodent infection with B. bissettiae decreased with increasing elevation (RR = .99) (Table 4). A negative relationship was observed between the predicted probabilities of infection with B. bissettiae for woodrats and increasing elevation (Fig 3).
The predicted probabilities of no infection (1None), infection with B. burgdorferi s. s. (2Bbss), or B. bissettiae (3Bbis) from (a) 350 m to 1000 m above sea level at the HVTR. Blue triangles represent probabilities associated with Allen’s chipmunks (Ns, Neotamias senex) and gray circles represent probabilities associated with dusky-footed woodrats (Nf, Neotoma fuscipes). LOESS curves (solid blue and gray lines) were used to display the relationship between probability of infection, elevation, and host species.
The relative risk of infection and associated predicted probabilities of infection with B. burgdorferi s. s. were greatest in the summer months (June–August) with another peak in April (Table 4, Fig 4). In contrast, the relative risk and associated predicted probabilities of infection with B. bissettiae were greatest in the spring months with another peak in July (Table 4, Fig 4). The greatest probability of infection with B. burgdorferi s. s. was associated with a chipmunk host, summer months, and higher elevations, whereas the greatest probability of infection with B. bissettiae was associated with a woodrat host, spring months, and lower elevations.
Here, we report for the first time a high prevalence of infection with B. burgdorferi s. s. and the highest reported prevalence of B. bissettiae in Allen’s chipmunks in the western United States. Our findings demonstrate that at broad and fine scales, Allen’s chipmunks not only host these pathogens, but are likely to be integral in maintaining and distributing them (Tables 2 and 4). The prevalence of B. bissettiae observed in dusky-footed woodrats mirrors that reported previously for this rodent from chaparral and oak woodland habitats in northern California, and reconfirms the important role of woodrats in the enzootic maintenance of this spirochete [25,27–29, 31, 34]. Two of the six named genospecies, i.e., B. burgdorferi s. s., B. bissettiae, B. americana, B. californiensis, B. carolinensis, and B. lanei (formerly B. genomospecies 2) known to circulate in California were detected in this study. This compares favorably with a previous report that found seven genospecies of Borrelia burgdorferi s. l. including one novel borrelial spirochete in Alameda County, California . The six unclassified B. burgdorferi s. l. samples from woodrats had amplicons with nucleotide mixtures at one or more positions. This suggests that these woodrats were possibly infected with a mixture (i.e., co-infection) of B. burgdorferi s. s. and B. bissettiae, or mixtures of other borrelial genospecies known or heretofore unknown to circulate in California. However, further characterization of these genospecies was not possible for these six samples. The use of cloning techniques may have aided in the identification of genospecies in these samples, but were not used during this study.
This study revealed significant spatial segregation of B. burgdorferi s. l. within two rodent populations at the HVTR, which likely is due to the abundance and distribution of these rodents and their associated ticks across the landscape. In the northwestern cluster, the presence of B. bissettiae in both woodrats and chipmunks may demonstrate an ecological overlap between these species (Figs 1 and 2). The increased prevalence of B. bissettiae in woodrats versus chipmunks reconfirms that woodrats serve as the primary reservoir host for B. bissettiae [7, 25, 27–29, 31, 34, 55]. We conclude that biotic and abiotic factors affecting the distribution of woodrats likely determine the distribution and clustering of B. bissettiae.
A high prevalence of B. burgdorferi s. s. infection (31.0%) within the chipmunk population in the southeastern cluster contrasts with the relatively low prevalence (9.1%) in chipmunks throughout the rest of the HVTR. Additionally, the only woodrats PCR-positive for B. burgdorferi s. s. throughout the HVTR were from the southeastern cluster. This is an important finding because B. burgdorferi s. s. seldom has been observed in woodrats [27, 31, 56] as compared with B. bissettiae [27, 31, 34]. A lack of B. bissettiae-positive rodents in the southeastern section of the HVTR could be the result of limited sampling. This seems unlikely, however, and we posit instead that environmental differences between the two clusters are more likely to account for the absence of B. bissettiae. Although beyond the scope of this study, the use of population genetics approaches may provide deeper insights into the distribution and possible evolution of borreliae within rodents and ticks at fine spatial scales.
A concurrent study at the HVTR identified a relatively high seroprevalence of Anaplasma phagocytophilum, the causative agent of human granulocytic anaplasmosis, in dusky-footed woodrats, Allen’s chipmunks, western gray squirrels (Sciurus griseus) and Douglas squirrels (Tamiasciurus douglasii) . Further, areas west of the Trinity River were more likely to have seropositive animals than east of the river . The cluster of B. bissettiae-positive rodents observed in the current study overlaps the cluster of A. phagocytophilum seropositive samples. Hence, there may be similar environmental processes at play that influence the distribution of B. bissettiiae and A. phagocytophilum. We therefore examined some local environmental and host-specific variables to determine their possible effects on the prevalence and spatial clustering of B. burgdorferi s. l.
Habitat and host specific analyses
The coefficients and predicted probabilities from the top ranked model demonstrate a clear association between woodrats and B. bissettiae, and chipmunks and B. burgdorferi s. s. (Tables 3 and 4). The low prevalence of B. bissettiae in chipmunks and B. burgdorferi s. s. in woodrats underscores the complexity of this system in northern California, and indicates that these transmission cycles are interdependent because of shared hosts and vector ticks. In an earlier study at the HVTR, Allen’s chipmunks were more often parasitized by I. pacificus than by any other tick species , which may partially explain the high prevalence of B. burgdorferi s. s. we detected in that rodent. Prior research also has demonstrated low prevalences of B. bissettiae in I. pacificus and B. burgdorferi s. s. in I. spinipalpis , likely accounting for the occasional presence of B. burgdorferi s. s. in woodrats and B. bissettiae in chipmunks observed in our study. The high prevalence of B. burgdorferi s. s. in Allen’s chipmunks and B. burgdorferi s. l. in other chipmunk species suggests that chipmunks may serve as primary reservoir hosts of such spirochetes [40, 58–59].
The spatial clustering of B. burgdorferi s. l. at the HVTR appears to be driven by the abundance and distribution of woodrats and chipmunks. However, similar proportions of both rodents were trapped in the northwestern and southeastern sections of the HVTR. If clustering of either genospecies is determined solely by the abundance or density of hosts, then similar proportions of infected individuals would be expected in both areas. However, B. burgdorferi s. s.–infected woodrats were only trapped in the southeastern section of the HVTR, suggesting a link between the increased prevalence in local chipmunks and its occasional presence in woodrats. Conversely, no woodrats infected with B. burgdorferi s. s. were trapped in the northwestern section of the HVTR suggesting that woodrats are not a primary reservoir for B. burgdorferi s. s. there. The observed spatial clustering cannot be explained adequately by differences in the relative abundance of hosts and, therefore it must be driven, at least partly, by other factors.
The inclusion of “month” in the top model indicates a significant temporal association between the presence of B. burgdorferi s. l. and rodents. This may be the result of temporal activity patterns of rodent hosts (Fig 4) and their associated tick vectors. Chipmunk activity generally increases in the late spring and summer months, whereas woodrats are active throughout the year [60–61]. However, it would be difficult to discern the difference between temporal activity patterns of hosts, and a pattern where hosts acquire infections of different genospecies at different times of the year. The temporal activity patterns of I. pacificus and I. spinipalpis also are important in the transmission of B. burgdorferi s. l. to rodent hosts [35, 49–50, 62–67]. Ixodes pacificus nymphs feed primarily on lizards and rodents and are active from late winter to early summer [30, 32–33, 50, 68], matching the prevalence of B. burgdorferi s. s. observed in rodents in this study (Fig 4). Activity of I. spinipalpis nymphs and larvae on rodent hosts has been shown to peak in the fall to spring months [28, 35], coinciding with the peak prevalence of B. bissettiae observed in rodents in this study (Fig 4). The diel activity patterns of woodrats (nocturnal) and chipmunks (diurnal) may more closely correspond to differing diel patterns of questing I. spinipalpis and I. pacificus and may determine the species of tick most likely to be attached to a given host [38, 60–61, 65].
The elevation at which a rodent was captured also was a significant predictor of the presence of B. burgdorferi s. l. in rodents. Similar numbers of woodrats and chipmunks were captured in most sampled areas at the HVTR. Therefore, the increased probability of infection with B. burgdorferi s. s. at higher elevations and B. bissettiae at lower elevations (Table 4, Fig 3) cannot be adequately explained by the distribution of hosts, dominance of a host at a site, or a greater number of samples at a given elevation.
Forest units west of the Trinity River generally are lower in elevation and more humid than eastern units that tend to be higher in elevation and drier (Fig 1). Both B. burgdorferi s. s. and B. bissettiae-infected rodents were found in areas west of the Trinity River, but only B. bissettiae was found to significantly cluster in rodents west of the river at a lower average elevation than the eastern side of the HVTR. Similarly, only B. burgdorferi s. s.-positive rodents were found to cluster on the eastern portion of the HVTR at a higher average elevation than those sampled throughout the HVTR. This pattern may be multi-factorial including differences in local climate [66–67], habitat characteristics [27–28, 69], and their effect on the distribution and abundance of reservoir hosts and primary vectors. In Alameda County, California, B. burgdorferi s. s. was more prevalent in ticks from warmer and drier areas, whereas B. bissettiae-infected I. pacificus ticks were more prevalent in the cooler, more humid environments influenced by the maritime conditions in San Francisco Bay . A significant positive association exists between elevation and I. pacificus abundance and B. burgdorferi s. l. prevalence in California, likely due to the effect of temperature and humidity on tick abundance at different elevations [66–67]. This association may help explain the presence and clustering of B. burgdorferi s. s. in woodrats and chipmunks observed in the southeastern cluster.
Although “forest unit” was not included in the top model, it was included in four of the top six models suggesting that it had some predictive value. Furthermore, the fourth most competitive model that contained the “forest unit” variable had the greatest model likelihood, indicating that it best fit the data, but was penalized by AICc because of small sample sizes and additional parameters resulting from the classification of forest unit structural stages. Overall, the abundance of woodrats and chipmunks caught in each forest unit was expected because of each host’s known habitat preferences [38, 43, 46, 47, 60–61]. Preliminary data demonstrated that the highest prevalence of B. burgdorferi s. s. occurred in chipmunks in younger forest units (brushy pole and young multi-storied forests) and the highest prevalence of B. bissettiae was in woodrats occupying young multi-storied forests and non-forested grassland (G. M. Hacker, unpubl. data). A longer-term sampling effort within these forest units is needed to bolster sample sizes and better evaluate the preliminary habitat associations observed in this study. The abundance and distribution of I. pacificus and I. spinipalpis immatures within these forest units is also undoubtedly contributing to the variation observed in prevalence of B. burgdorferi s. l. in woodrats and chipmunks and offer fertile ground for future research.
This is the first study demonstrating the importance of Allen’s chipmunks in the ecology of B. burgdorferi s. l. in the western United States. The spatial clustering of B. burgdorferi s. l. observed in this study can be explained, in part, by the spatial and temporal distribution of host species, time-of-year, and elevation. Spatial clustering and environmental associations of B. burgdorferi s. l. in vector ticks as well as in rodents should be evaluated in other forested areas of northern California to validate our findings and to clarify the vector-pathogen-host interrelationships. Most importantly, the public generally and foresters in particular should be informed about the potential risk of exposure to spirochete-infected vector ticks in certain subtypes of forests that harbor significant populations of woodrats and chipmunks.
S1 Table. All host-related information used for statistical analyses in this study.
Data include all relevant host species measurements, date of capture, location of capture (UTM coordinates), related forest structure attributes, elevation, and PCR test results.
S2 Table. All parameter coefficients, associated standard errors, and p-values for the top six models ranked via AICc.
S3 Table. Summary information on all multinomial log-linear models analyzed including the number of parameters, log likelihood, AICc score, delta AIC, model weight, and cumulative weight.
We thank the Hoopa Tribe, the Hupa people, and Hoopa Tribal Forestry for access to their lands, as well as the technicians from Hoopa Tribal Forestry who assisted with rodent trapping. We also thank Dr. Robert Van Kirk, Dr. Jianmin Zhong, and Dr. Matthew Johnson from Humboldt State University for technical assistance. This research was supported by a National Institutes of Health grant AI22501 to R. S. Lane. This work was conducted as partial fulfillment of the requirements for a MS degree in Natural Resources, Department of Wildlife, Humboldt State University by the lead author.
- 1. Schwartz AM, Hinckley AF, Mead PS, Hook SA, Kugeler KJ. Surveillance for Lyme disease—United States, 2008–2015. Atlanta, GA: Department of Health & Human Services, Centers for Disease Control and Prevention; 2017 Nov 10.
- 2. Dolan MC, Maupin GO, Panella NA, Golde WT, Piesman J. Vector competence of Ixodes scapularis, I. spinipalpis, and Dermacentor andersoni (Acari: Ixodidae) in transmitting Borrelia burgdorferi, the etiologic agent of Lyme disease. J. med. entomol. 1997 Mar 1; 34(2):128–35. pmid:9103755
- 3. Lane RS, Piesman J, Burgdorfer W. Lyme borreliosis: relation of its causative agent to its vectors and hosts in North America and Europe. Annu. Rev. Entomol. 1991 Jan; 36(1):587–609.
- 4. Margos G, Hojgaard A, Lane RS, Cornet M, Fingerle V, Rudenko N, et al. Multilocus sequence analysis of Borrelia bissettii strains from North America reveals a new Borrelia species, Borrelia kurtenbachii. Ticks Tick Borne Dis. 2010 Dec 31; 1(4):151–8. pmid:21157575
- 5. Clark KL, Leydet BF, Hartman S. Lyme borreliosis in human patients in Florida and Georgia, USA. Int. J. Med. Sci. 2013 Jan 1; 10(7):915–31 pmid:23781138
- 6. Clark KL, Leydet BF, Threlkeld C. Geographical and genospecies distribution of Borrelia burgdorferi sensu lato DNA detected in humans in the USA. J. Med. Microbiol. 2014 May 1; 63(5):674–84.
- 7. Fedorova N, Kleinjan JE, James D, Hui LT, Peeters H, Lane RS. Remarkable diversity of tick or mammalian-associated Borreliae in the metropolitan San Francisco Bay Area, California. Ticks Tick Borne Dis. 2014 Oct 31; 5(6):951–61. pmid:25129859
- 8. Rudenko N, Golovchenko M, Ruzoek D, Piskunova N, Mallatova N, Grubhoffer L. Molecular detection of Borrelia bissettii DNA in serum samples from patients in the Czech Republic with suspected borreliosis. FEMS Microbiol. Letts. 2009 Mar 1; 292(2):274–81.
- 9. Rudenko N, Golovchenko M, Grubhoffer L, Oliver Jr. JH. Updates on Borrelia burgdorferi sensu lato with respect to public health. Ticks Tick Borne Dis. 2011 Sep 30; 2(3):123–8. pmid:21890064
- 10. Seinost G, Dykhuisen DE, Dattwyler RJ, Golde WT, Dunn JJ. Four clones of Borrelia burgdorferi sensu stricto cause invasive infection in humans. Infect. Immunol. 1999 Jul 1; 67(7):3518–24.
- 11. Pritt BS, Mead PS, Johnson DKH, Neitzel D., Respicio-Kingry LB, Davis JP, et al. Identification of a novel pathogenic Borrelia species causing Lyme borreliosis with unusually high spirochaetaemia: a descriptive study. Lancet Infect. Dis. 2016 May 31; 16(5):556–64. pmid:26856777
- 12. Margos G, Lane RJ, Fedorova N, Koloczek J, Piesman J, Hojgaard A., et al. Borrelia bissettiae sp. nov. and Borrelia californiensis sp. nov. prevail in diverse enzootic transmission cycles. Int. J. Syst. Evol. Microbiol. 2016 Mar 1; 66(3):1447–52. pmid:26813420
- 13. Postic D, Ras NM, Lane RS, Hendson M, Baranton G. Expanded diversity among Californian Borrelia isolates and description of Borrelia bissettii sp. novel (Formerly B. Group DN127). J. Clin. Microbiol. 1998 Dec 1; 36(12):3497–504. pmid:9817861
- 14. Postic D, Garnier M, Baranton G. Multilocus sequence analysis of atypical Borrelia burgdorferi sensu lato isolates-description of Borrelia californiensis sp. novel, and genomospecies 1 and 2. Int. J. Med. Microbiol. 2007 Jul 2; 297(4):263–71. pmid:17374507
- 15. Rudenko N, Golovchenko M, Grubhoffer L, Oliver JH. Borrelia carolinensis sp. nov., a new (14th) member of the Borrelia burgdorferi sensu lato complex from the southeastern region of the United States. J. clin. microbiol. 2009 Jan 1; 47(1):134–41. pmid:19020062
- 16. Rudenko N, Golovchenko M, Lin T, Gao L, Grubhoffer L, Oliver JH. Delineation of a new species of the Borrelia burgdorferi sensu lato complex, Borrelia americana sp. nov. J. Clin. Microbiol. 2009 Dec 1; 47(12):3875–80. pmid:19846628
- 17. Margos G, Piesman J, Lane RS, Ogden NH, Sing A, Straubinger RK, et al. Borrelia kurtenbachii sp. nov.: A widely distributed member of the Borrelia burgdorferi sensu lato species complex in North America. Int. J. Systematic Evolut. Microbiol. 2014 Jan 1; 64(1):128–30.
- 18. Margos G, Fedorova N, Kleinjan JE, Hartberger C, Schwan TG, Sing A, et al. Borrelia lanei sp. nov. extends the diversity of Borrelia species in California. Int. J. Systematic Evolut. Microbiol. 2017; 67(10):3872–3876
- 19. Postic D, Assous MV, Grimont PAD, Baranton G. Diversity of Borrelia burgdorferi sensu lato evidenced by restriction fragment length polymorphism of rrf (5S)-rrl (23S) intergenic spacer amplicons. Int. J. Systematic Evolut. Microbiol. 1994 Oct 1; 44(4):743–52.
- 20. Girard YA, Fedorova N, Lane RS. Genetic diversity of Borrelia burgdorferi and detection of Borrelia bissettii-like DNA in serum of north-coastal California residents. J. Clin. Microbiol. 2011 Mar 1; 49(3):945–54. pmid:21177909
- 21. Rudenko N, Golovchenko M, Vancova M, Clark K, Grubhoffer L, Oliver Jr., JH. Isolation of live Borrelia burgdorferi sensu lato spirochaetes from patients with undefined disorders and symptoms not typical for Lyme borreliosis. Clin. Microbiol. Infect. 2016 Mar 31; 22(3):267–e9.
- 22. Bunikis J, Tsao J, Luke CJ, Luna MJ, Fish D, Barbour AG. Borrelia burgdorferi infection in a natural population of Peromyscus leucopus mice: a longitudinal study in an area where Lyme borreliosis is highly endemic. J. Infect. Dis. 2004 Apr 15; 189(8):1515–23. pmid:15073690
- 23. LoGuidice K, Ostfeld RS, Schmidt KA, Keesing F. The ecology of infectious disease: effects of host diversity and community composition on Lyme disease risk. Proc. Natl. Acad. Sci. 2003 Jan 21; 100(2):567–71. pmid:12525705
- 24. Mather TN, Wilson ML, Moore SI, Ribeiro JMC, Spielman A. Comparing the relative potential of rodents as reservoirs of the Lyme disease spirochete (Borrelia burgdorferi). J. Epidemiol. 1989 Jul 1; 130(1):143–50.
- 25. Brown RN, Lane RS. Lyme disease in California: a novel enzootic transmission cycle of Borrelia burgdorferi. Sci. 1992; 256:1439–42.
- 26. Lane RS, Peavey CA, Padgett KA, and Hendson M. Life history of Ixodes (Ixodes) jellisoni (Acari: Ixodidae) and its vector competence for Borrelia burgdorferi sensu lato. J. Med. Entomol. 1999; 36:329–340. pmid:10337104
- 27. Brown RN, Peot MA, Lane RS. Sylvatic maintenance of Borrelia burgdorferi (Spirochaetales) in northern California: untangling the web of transmission. J. Med. Entomol. 2006; 43 (4):743–51. pmid:16892634
- 28. Eisen L, Eisen RJ, Mun J, Salkeld DJ, Lane RS. Transmission cycles of Borrelia burgdorferi and Borrelia bissettii in relation to habitat type in northwestern California. J. Vect. Ecol. 2009 Jun 1; 34(1):81–91.
- 29. Peavey CA, Lane RS, Kleinjan JE. Role of small mammals in the ecology of Borrelia burgdorferi in a peri-urban park in north coastal California. Exp. Appl. Acarol. 1997 Aug 1; 21(8):569–84. pmid:9291589
- 30. Salkeld DJ, Lane RS. Community ecology and disease risk: Lizards, squirrels, and the Lyme disease spirochete in California, USA. Ecol. 2010 Jan 1; 91(1):293–8.
- 31. Vredevoe LK, Stevens JR, Schneider BS. Detection and characterization of Borrelia bissettii in rodents from the central California coast. J. Med. Entomol. 2004; 41(4):736–45. pmid:15311469
- 32. Burgdorfer W, Lane RS, Barbour AG, Gresbrink RA, Anderson JR. The western black-legged tick, Ixodes pacificus: a vector of Borrelia burgdorferi. Am. J. Trop. Med. Hyg. 1985 Sep 1; 34(5):925–30. pmid:3898886
- 33. Clover JR, Lane RS. Evidence implicating nymphal Ixodes pacificus (Acari: Ixodidae) in the epidemiology of Lyme disease in California. Am. J. Trop. Med. Hyg. 1995 Sep 1; 53(3):237–40. pmid:7573703
- 34. Sholty KE. Strain distribution of Borrelia burgdorferi and Anaplasma phagocytophilum in sciurids and woodrats in northwestern California. 2015. Masters Thesis. Wildlife Department. Humboldt State University. Arcata, California, USA.
- 35. Burkot TR, Maupin GO, Schneider BS, Denatale CH, Happ CM, Rutherford JS, et al. Use of a sentinel host system to study the questing behavior of Ixodes spinipalpis and its role in the transmission of Borrelia bissettii, human granulocytic ehrlichiosis, and Babesia microti. Am. J. Trop. Med. Hyg. 2001 Oct 1; 65(4):293–9. pmid:11693872
- 36. Lane RS, Mun J, Eisen RJ, Eisen L. Western gray squirrel (Rodentia: Sciuridae): a primary reservoir host of Borrelia burgdorferi in Californian oak woodlands? J. Med. Entomol. 2005; 42(3):388–96. pmid:15962792
- 37. Salkeld DJ, Leonhard S, Girard YA, Hahn N, Mun J, Padgett KA, et al. Identifying the reservoir hosts of the Lyme disease spirochete Borrelia burgdorferi in California: the role of the western gray squirrel (Sciurus griseus). Am. J. Trop. Med. Hyg. 2008 Oct 1; 79(4):535–40. pmid:18840740
- 38. Hall ER. The mammals of North America. 2nd. Wiley & Sons. 1981.
- 39. Waters JR, Zabel CJ. Abundances of small mammals in fir forests in northeastern California. J. Mammal. 1998 Dec 3; 79(4):1244–53.
- 40. Roy RN, Straub MH, Stephenson N, Sholty KE, Foley J. Distribution and diversity of Borrelia burgdorferi sensu lato group bacteria in Sciurids of California. Vector Borne Zoonotic Dis. 2017 Nov 1; 17(11): 735–742. pmid:28976820
- 41. Mayer KE, Laudenslayer WF. A guide to wildlife habitats of California. California Department of Forestry and Fire Protection, Sacramento, California. USA. 1988.
- 42. O’hara KL, Latham PA, Hessburg P, Smith BG. A structural classification for inland northwest forest vegetation. West. J. Appl. Forest. 11: 97–102.
- 43. Hughes KD. Habitat associations of dusky-footed woodrats in managed douglas-fir/hardwood forests of northern California. Masters Thesis. Wildlife Department. Humboldt State University. Arcata, California, USA. 2006
- 44. Klenner W, Sullivan TP. Partial and clearcut harvesting of dry Douglas-fir forests: implications for small mammal communities. For. Ecol. Manag. 2009 Feb 10; 257(3):1078–86.
- 45. Urban NA, Swihart RK. Small mammal responses to forest management for oak regeneration in southern Indiana. Forest Ecol. Manag. 2011 Feb 1; 261(3):353–61.
- 46. Whitaker DA. Relation of thin and release timber management practices to abundance of woodrats, chipmunks, mice, and ticks within the Hoopa Valley Indian Reservation. Masters Thesis. Wildlife Department. Humboldt State University. Arcata, California, USA. 2008
- 47. Waldien DL, Hayes JP, Huso MM. Use of downed wood by Townsend's chipmunks (Tamias townsendii) in western Oregon. J. Mammal. 2006; 87(3):454–60.
- 48. Allan BF, Keesing F, Ostfeld RS. Effect of forest fragmentation on Lyme disease risk. Conserv. Biol. 2003 Feb 1; 17(1):267–72.
- 49. Tälleklint-Eisen L, Lane RS. Variation in the density of questing Ixodes pacificus (Acari: Ixodidae) nymphs infected with Borrelia burgdorferi at different spatial scales in California. J. Parasitol. 1999 Oct; 1:824–31.
- 50. Tälleklint-Eisen L, Lane RS. Spatial and temporal variation in the density of Ixodes pacificus (Acari: Ixodidae) nymphs. Environ. Entomol. 2000 Apr; 29(2):272–80.
- 51. Swei A, Briggs CJ, Lane RS, Ostfeld RS. Impacts of an introduced forest pathogen on the risk of Lyme disease in California. Vector Borne Zoonotic Dis. 2012 Aug 1; 12(8):623–32. pmid:22607076
- 52. Lane RS, Steinlein DB, Mun J. Human behaviors elevating exposure to Ixodes pacificus (Acari: Ixodidae) nymphs and their associated bacterial zoonotic agents in a hardwood forest. J. Med. Entomol. 2004; 41(2):239–48. pmid:15061284
- 53. Swofford D. PAUP 4.0: phylogenetic analysis using parsimony. Smithsonian Institution; 1998.
- 54. Kulldorff M. A spatial scan statistic. Communications in Statistics-Theory and methods. 1997 Jan 1; 26(6):1481–96.
- 55. Brown RN, Lane RS. Reservoir competence of four chaparral-dwelling rodents for Borrelia burgdorferi in California. Am. J. Trop. Med. and Hyg. 1996 Jan 1; 54(1):84–91.
- 56. Swei A, Bowie VC, Bowie RCK. Comparative genetic diversity of Lyme disease bacteria in northern Californian ticks and their vertebrate hosts. Ticks Tick Borne Dis. 2015 Mar 9; 6(3):414–423. pmid:25843810
- 57. Foley JE, Clueit SB, Brown RN. Differential exposure to Anaplasma phagocytophilum in rodent species in northern California. Vector Borne Zoonotic Dis. 2008 Feb 1; 8(1):49–56. pmid:18047398
- 58. Marsot M, Chapuis JL, Gasqui P, Dozières A, Masséglia S, Pisanu B, et al. Introduced Siberian chipmunks (Tamias sibiricus barberi) contribute more to Lyme borreliosis risk than native reservoir rodents. PLOS One 2013 Jan 31; 8(1):e55377. pmid:23383170
- 59. Slajchert T, Kitron UD, Jones CJ, Mannelli A. Role of the eastern chipmunk (Tamias striatus) in the epizootiology of Lyme borreliosis in northwestern Illinois, USA. J. Wildl. Dis. 1997 Jan; 33(1):40–6. pmid:9027689
- 60. Carraway LN, Verts BJ. Neotoma fuscipes. Mammal. Species. 1991 Nov 6; (386):1–0.
- 61. Gannon WL, Forbes RB. Tamias senex. Mammal. Species. 1995 Jun; 23(502):1–6.
- 62. Eisen RJ, Eisen L, Castro MB, Lane RS. Environmentally related variability in risk of exposure to Lyme disease spirochetes in northern California: effect of climatic conditions and habitat type. Environ. Entomol. 2003 Oct; 32(5):1010–8.
- 63. Salkeld DJ, Castro MB, Bonilla D, Kjemtrup A, Kramer VL, Lane RS, et al. Seasonal activity patterns of the western black-legged tick, Ixodes pacificus, in relation to onset of human Lyme disease in northwestern California. Ticks Tick Borne Dis. 2014 Oct 31;5(6):790–6. pmid:25113980
- 64. Furman DP, Loomis EC. 1984. The ticks of California (Acari: Ixodida). Bulletin of the California Insect Survey. 25, 1–239.
- 65. Alekseev AN, Dubinina HV. Abiotic parameters and diel and seasonal activity of Borrelia-infected and uninfected Ixodes persulcatus (Acarina: Ixodidae). J. Med. entomol. 2000 Jan; 37(1):9–15. pmid:15218900
- 66. Swei A, Meentemeyer R, Briggs CJ. Influence of abiotic and environmental factors on the density and infection prevalence of Ixodes pacificus (Acari: Ixodidae) with Borrelia burgdorferi. J. Med. Entomol. 2011 Jan 31; 48(1):20–8. pmid:21337944
- 67. Eisen RJ, Eisen L, Girard YA, Fedorova N, Mun J, Slikas B, et al. A spatially-explicit model of acarological risk of exposure to Borrelia burgdorferi-infected Ixodes pacificus nymphs in northwestern California based on woodland type, temperature, and water vapor. Ticks Tick Borne Dis. 2010 Mar 31;1(1):35–43. pmid:20532183
- 68. Casher L, Lane R, Barrett R, Eisen L. Relative importance of lizards and mammals as hosts for ixodid ticks in northern California. Exper. App. acarol. 2002 Jan 1;26(1–2):127.
- 69. Hamm KA. Abundance of dusky-footed woodrats in managed forests of north coastal California. Masters Thesis. Wildlife Department. Humboldt State University. Arcata, California, USA. 1995.