Specimen acquisition

The Mount Allison Tick Bank serves as an archive of ticks and their DNA, collected through passive surveillance, primarily from New Brunswick (NB) and the other Canadian maritime provinces, Nova Scotia (NS) and Prince Edward Island (PEI). Ticks are donated by the general public, primarily via veterinary clinics but also directly from the public. When the tick bank was initiated, ticks were solicited from veterinary clinics by letter and email and presentation at the annual regional veterinary medicine conference (2012 and 2013). No organized promotion of the tick bank occurred in subsequent years although it was mentioned in sporadic media reports; engagement by veterinary clinics was fairly consistent throughout the span of the study.

Upon receipt, Ixodes spp. ticks were identified to species, life stage, sex, and state of engorgement based on [52, 53] (University of Rhode Island, www.tickencounter.org). The initial focus of the tick bank was on Ixodes species ticks, so all non-Ixodes ticks were photographed and archived; identification to species, when possible, occurred at a later date based on images and examination of archived residual specimens when possible. Each sample is accompanied by a submission form which included the date and location where the tick was encountered, instances of recent travel, whether the tick was attached, the species of the host, and contact and consent information. Photographs of the mouthparts, dorsal aspect, and ventral aspect for each tick were taken under a Leica EZ4D dissecting microscope. Ticks were subsequently longitudinally bisected (with the exception of larvae and unengorged nymphs) with half used for DNA extraction and subsequent testing and the other half archived at -20°C. This study selected those ticks which were identified as Dermacentor spp. collected in NB between 2012 and 2021. Any disclosed out-of-province travel in the two weeks prior to tick discovery was noted. These ticks were excluded from mapping (Fig 2), but not molecular analysis. Funding considerations prompted the conversion of free tick testing to a user-pay service in 2020, drastically reducing tick submissions, particularly for non-Ixodes species. An additional four tick records from 2021 were provided by a tick testing company, Geneticks [54], and 58 records of Dermacentor species with no documented travel history submitted for identification to the tick surveillance platform, eTick [48].

Mapping and statistical analysis

Mapping of the location of the tick recoveries used GIS map templates that are the intellectual property of Esri and are used herein under license (Copyright © 2020 Esri and its licensors. All rights reserved).

To assess the probability of the increased D. variabilis recoveries occurring by chance, both over time and in northern versus southern regions, chi squared analysis was used. The four years 2013–2016 (inclusive) were used as the “early” years and the four years 2017–2020 were used as “late” years. Recoveries over multiple years were combined to reduce random fluctuation in the numbers of rare recoveries. The year 2021 was excluded as submissions of all non-I. scapularis ticks decreased with the imposition of cost-recovery for tick testing; data from the year 2012 was excluded for symmetry. For geographic analysis, the province was divided into southern (health zone 1 and 2), middle (health zones 3 and 7) and northern regions (health zones 4 to 6). These regions were based on human health zones [55] rather than ecological variables as anthropic information important for passive surveillance are categorized using these health zones, but they do correspond reasonably well with geographic and climactic parameters [50].

Molecular analysis

All ticks, regardless of species and including those too damaged to identify to species, were tested upon receipt for the presence of 1–3 B. burgdorferi genes, Outer Surface Protein A (OspA), Flagellin B (FlaB) and the 23S ribosomal RNA (rRNA) gene (Table 1; [56, 57]. Presumably due to competition from tick DNA, if the abundance of the B. burgdorferi target DNA was low, in some cases not all Borrelia primers sets would produce an amplicon. This was not specific to D. variabilis as it was also noted for I. scapularis and I. cookei ticks tested in this manner [12, 58]. Extraction of DNA, amplification and visualization were completed as described by [59]. For all nested PCR (nPCR) reactions, the reaction volumes and nPCR protocol were as described in [59] with the annealing temperatures given in Table 1. Negative controls consisting of water in place of DNA were conducted at the start and end of each experimental PCR manipulation in the same workspace, in order to detect reagent contamination prior to or during nPCR or aerosolization of DNA in the PCR workspace. Tick DNA extraction, PCR manipulations and gel electrophoresis were all conducted in separate locations with independent air flow. Amplicons of sizes indicative of either Borrelia spp. or B. burgdorferi were submitted for Sanger sequencing at Génome Québec, McGill University (Montréal, QC). The resulting chromatograms were trimmed and manually curated for ambiguous nucleotides using FinchTV and a Nucleotide Blast of each sequence was performed using the National Center for Biotechnology Information’s (NCBI) GenBank.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Primers and conditions used to amplify Borrelia DNA.

https://doi.org/10.1371/journal.pone.0292703.t001

Dermacentor recoveries

Between 2012 and 2021, 668 Dermacentor spp. ticks were collected through community-based passive surveillance in New Brunswick and submitted to the MTA Tick Lab. For 2020 and 2021, ticks and or records of 4 additional D. variablis ticks from the tick testing company, Geneticks [54] and 58 from the tick surveillance platform, eTick [48] were obtained. 141 specimens were retained after removal of submissions where the tick was acquired outside of New Brunswick, submissions that did not explicitly exclude travel outside of New Brunswick, ticks too damaged for species-level identification and D. albipictus (of which 13 were found from one moose and three dogs; Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Summary of Dermacentor variabilis recoveries from New Brunswick by year.

https://doi.org/10.1371/journal.pone.0292703.t002

Dermacentor variabilis collected in New Brunswick

Of the submitted D. variabilis from New Brunswick, 98.5% (139/141) were adults. In comparison, from the neighboring province of Nova Scotia with extensive and well-established populations, 99.2% were adults (S1 Table in S1 File). Of the adult ticks, approximately 10–20% were male in the first years (2012–2016); from 2017 onwards, the proportion of males recovered increased to approximately 30–60% (Table 2). Overall, dogs were the most common host (48%), with humans also well represented (41%), although canine-derived samples were, as expected, under-represented after 2020 when free submissions were discontinued. D. variabilis were also readily found on wildlife, or walking in the outside or inside environment (Table 2), consistent with their wide host range [7]. The major peak of D. variabilis tick recoveries was from May to August (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Locally-acquired Dermacentor variabilis recoveries by passive surveillance, by month, in New Brunswick from 2013 to 2020.

https://doi.org/10.1371/journal.pone.0292703.g001

There is a trend of increasing D. variabilis recoveries over time evident by inspection and supported by comparing the proportion of D. variabilis to total tick recoveries for years 2017–2020 to 2013–2016 by Chi squared analysis (p = 0.014). To determine if this might be because D. variabilis were establishing populations in the province, the geographical origins of locally acquired ticks were plotted (Fig 2). Dermacentor variabilis specimens collected in 2012 were all were associated with travel to regions with well-established and abundant D. variabilis populations (primarily Nova Scotia) [12]. Starting in 2013, locally-acquired D. variabilis were recovered in New Brunswick, although close to the New Brunswick-Nova Scotia border. This was followed in 2014 with locally-acquired ticks found in the southwest of the province in the St. John region, and by 2016, stretching along the southern portion of the province. In 2017, multiple disperse local recoveries were reported. These included recoveries from the southern and middle portions of the province and along the Maine, US border. Widely distributed and expanding populations of D. variabilis had been previously documented in Maine, including near the Maine-New Brunswick border, by 2006 [45]. By 2019, locally-acquired D. variabilis were additionally being reported along the coastal region. By 2020 D. variabilis were also recovered in the northwest of the province at the New Brunswick-Maine-Quebec border; D. variabilis having also been reported from this part of Quebec in 2017 [60]. Of note, this widespread recovery occurred during a time when non-essential travel across provincial borders was severely restricted in response to the COVID-19 pandemic. By 2021, D. variabilis, while still not abundant, was being reported throughout most of the province of New Brunswick. Chi squared analysis indicated that for both the “early expansion” period of 2013–2016 and the “later expansion” period of 2017–2020, proportion of the D. varibilis relative to total tick recoveries in the middle and northern parts of the province were greater than for the southern part of the province (p = 3.9 x e^-7 and 1.8 x e^-5, respectively), where I. scapularis was well established and a strong contributor to total tick recoveries.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. The distribution of locally-acquired D. variabilis ticks from New Brunswick from 2012–2021.

Areas of concentration, 50km in diameter, of collection lasting > 2 years, are indicated with red circles. The GIS map template includes human health zones and scale. Map templates were obtained from ArcGIS Online maps hosted by Esri.

https://doi.org/10.1371/journal.pone.0292703.g002

Borrelia burgdorferi and B. miyamotoi in D. variabilis ticks

All ticks, regardless of species and including those too damaged to identify to species, were tested upon receipt for 1–3 B. burgdorferi genes. Twenty D. variabilis ticks showed an amplicon consistent with the potential presence of B. burgdorferi. Sequence conformation of re-amplified amplicons from the four of these ticks with at least two amplicons consistent with B. burgdorferi genes, produced only one, a D. variabilis male (S1 Fig in S1 File), obtained in 2021 from a dog from southeastern New Brunswick without a recent travel history, with a sequence consistent with the B. burgdorferi 16-23S intergenic spacer (IGS) region; sequence analysis of the B. burgdorferi Outer Surface Protein A gene segment produced uninterpretable sequence (S2 Fig in S1 File). The other three ticks produced uninterpretable sequences.

Screening of ticks for the relapsing fever spirochete, B. miyamotoi was not routinely performed upon receipt of the ticks so a total of 181 D. variabilis ticks (144 ticks collected in 2017, 34 ticks collected in 2018 and 23 from 2012–2016) were tested. In contrast to the positive control (DNA from the liver of a B. miyamotoi-infected Peromyscus leucopus) only two D. variablis ticks, both adult females from a canine host collected in 2014, produced amplicons consistent with the size expected for B. miyamotoi, however, the presence of B. miyamotoi could not be confirmed by Sanger sequencing.

Does the presence of locally acquired ticks mean that self-sustaining populations D. variabilis have been established?

The presence of ticks could be explained by adventitious ticks translocated from adjacent high tick population areas. If these ticks were able to survive but were unable to find hosts from which to feed, find a mate or have surviving progeny, then local populations would not be able to establish. There are several lines of evidence that argue for the ticks recovered in this study representing newly established populations. The most obvious argument is that the number of tick recoveries increased throughout the time span captured in this study; populations of D. variabilis were both high and stable in the adjacent region of northern Nova Scotia since at least 2014 [12], so if the ticks recovered in New Brunswick were purely adventitious ticks, their number should have also remained stable. Additional evidence that D. variabilis recoveries represent the newly established populations include that the ticks were recovered from wildlife hosts and the recovery of immature stages in at least one year. Furthermore, D. variabilis were still recovered from multiple locations throughout the province in 2020 when provincial borders were closed to non-essential travel, and these locations were the same regions from which ticks were recovered in previous and subsequent years. The recovery of increased numbers of ticks and recoveries from dispersed locations are all consistent with the establishment of new populations, as well as with the biology of the species. With a broad host range [7] finding hosts should not be a challenge. Finding a mate is a problem that is alleviated one generation after an adult female is translocated while feeding on a host, as long as her eggs are able to hatch and the progeny thrive. As this species is a hardy one with relatively few climactic barriers [61, 62], propagation should not be a limiting consideration. To determine if the D. variabilis recoveries reported here do represent newly established populations, it would be helpful to recover additional immature developmental stages, ideally on wildlife, through active surveillance. As [63] did not recover D. variabilis specimens during active surveillance in the region during the initial time span covered by this study, 2014, repeated active surveillance would provide valuable conformation. [7] report a good correlation between active and passive surveillance in detecting northward population spread for D. variabilis and [8] document population expansion of D. variabilis in central Canada using passive surveillance, so the increasing tick recoveries in New Brunswick, reported here is likely to represent the establishment of new D. variabilis populations.

Recoveries of D. variabilis occurred from multiple locations within the province, initially in the southeast and southwest. By 2014 recoveries started to occur between these foci and in the north of the province. The pattern of D. variabilis recoveries, in the milder south and coastal regions where I. scapularis first established [50], is suggestive of establishing, and possibly established, populations in the more climactically moderate regions. However, the locally acquired D. variabilis did not all originate from the more climactically moderate regions of the province, unlike what was found for I. scapularis [50]. Climate modelling by [1] predicts the northward expansion of suitable conditions for D. variabilis populations under all climate change scenarios; these suitable regions encompass both Nova Scotia and New Brunswick. Nevertheless, their climate model indicates that both provinces are only marginally suitable for D. variabilis populations. This prediction is seemingly at odds with the abundant recoveries of D. variabilis in Nova Scotia [12, 48] coupled to the growing recoveries of D. variabilis in New Brunswick documented here. However, the model used captured climactic variables but not biotic factors such as host movement that can lead to tick introductions. Additionally, the sensitivity of such models is greater for predicting where tick populations would not be sustainable than where populations would be sustainable, neither does the model predict tick abundance [1]. D. variabilis is a tough and versatile tick, capable of survival even in ecosystems not thought to be suitable such as the challenging climate of northern Canadian Prairies [7, 61, 62] so climactic conditions may not constrain population establishment as significantly as for other tick species.

How are D. variabilis populations being established?

Inspection of the locations of locally-acquired D. variabilis in New Brunswick shows that the ticks are found in distinct and discontinuous foci, in this study, generally separated by 100 ‐ 200 km. This finding is consistent with the presence of disjunct populations in both Canada [42, 46, 47] and in the United States. These disjunct populations have been assumed to represent separate tick introductions via infested livestock or companion animals [42, 46, 47]. Perhaps not surprisingly given the common name of “American dog tick”, adult D. variabilis ticks are the predominant Dermacentor species found on companion animals in the United States [64] and specimens collected in our tick bank were frequently recovered from dogs (Table 2, [12]. Immature D. variabilis have been found feeding from small mammals [65–67] but rarely from birds or humans [68–71]. This would limit the ability of this species to disperse large distances as juveniles. Adult D. variabilis, however, readily feed from larger mammals including humans, companion and agricultural animals. This raises the obvious possibility that the distributed pattern of D. variabilis recoveries represent anthropic introductions mediated by the travel of companion dogs, a mode of introduction that has been previously implicated in the introduction of D. variabilis in other regions [42, 46, 47]. If population establishment and expansion of D. variabilis is more reflective of chance anthropic “seedings” by human actions than large scale climactic factors, risk modeling will be complex. Finer scale population surveillance that captures anthropic variables, biotic and abiotic factors that include human use of habitats, microclimates, quantification of tick density and surveillance of wildlife for the pathogens transmitted by D. variabilis, among other factors, is warranted to monitor and predict tick populations to protect human and animal health in regions of expanding tick populations.

Dermacentor variabilis and Borrelia

While D. variabilis can transmit multiple pathogens of human and veterinary importance, their ability to maintain and transmit Borrelia spirochetes remains contentious with conflicting reports. The possibility of D. variabilis ticks acquiring Borrelia spirochetes from infected hosts is not surprising, and does not necessarily mean that the ticks could transmit the pathogen. There have been studies that have detected Borrelia spirochetes in D. variabilis that fed from infected hosts. D. variabilis larvae removed from infected P. leucopus in a highly endemic area had B. burgdorferi spirochetes identified both serologically and morphologically [30]. Similarly, [32] found B. burgdorferi in adult D. reticulatus in Belarus; B. burgdorferi, B. afzelii, or B. valaisiana were identified in 2.7% of the 226 ticks tested. There also have been isolated reports of Dermacentor spp. ticks being infected with B. burgdorferi elsewhere in North America. An adult D. albipictus found feeding on a canine host in Ontario, Canada had molecularly identified B. burgdorferi, although unfortunately the dog was not tested for B. burgdorferi exposure to investigate prior infection or transmission [31]. The finding of a potentially Borrelia-positive reported D. variabilis may fall into this category of a Dermacentor tick acquiring B. burgdorferi from an infected dog, although again the dog was not tested for Borrelia. While there is evidence of D. variabilis carrying B. burgdorferi, caution needs to be exercised in interpreting these results. In all cases, amplicons suggestive of B. burgdorferi require sequence validation to confirm the presence of B. burgdorferi. By convention, amplification and sequence conformation of two or more Borrelia genes is considered necessary to consider a tick a “true” positive [12, 56], so the D. variabilis identified here would not be considered to be a true positive. This assessment criteria is designed to reduce false positives, although it can also produce false negative results. Competition from the more abundant tick DNA can reduce detection of Borrelia sequences and target DNA in archived samples is subject to DNA degradation. Conversely, PCR, the method used in most recent studies to assess the presence of Borrelia, can detect DNA from both viable and non-viable, but not yet eliminated, Borrelia leading to false positive results. [57] report that B. burgdorferi and B. miymotoi were detected in approximately 4% and 2%, respectively, of wildlife species from southern New Brunswick. If D. variabilis were efficient at acquiring and retaining Borrelia infection, approximately 6 and 4, respectively, of the ticks tested in this study would have been expected to be carrying these Borrelia species. That fewer were recovered may represent degradation of Borrelia DNA in vivo in D. variabilis ticks; assessing ticks collected by active surveillance, so possibly sooner after feeding, would address this question.

While Borrelia has been detected in D. variabilis, this does not necessarily indicate that these ticks can transmit the pathogens and previous research suggests that Dermacentor spp. ticks are not competent vectors of Borrelia burgdorferi. Their poor vectoral capacity is thought to result from the action of antimicrobial peptides and lysozymes in the tick’s haemolymph that are effective in eliminating B. burgdorferi [72, 73]. The effectiveness of this immune response was demonstrated in experiments in which B. burgdorferi was injected directly into the haemolymph of D. variabilis, causing a rapid surge in antimicrobial peptides, lysozymes, and haemocytes leading to rapid bacterial elimination [33, 34]. Consistent with this finding, D. variabilis nymphs were found to lack detectable B. burgdorferi despite molting from infected larvae and these nymphs were unable to transmit B. burgdorferi to naïve mice in the laboratory [37]. These findings are mirrored by evidence that Dermacentor ticks appear unable to maintain a B. burgdorferi infection in nature [35, 36]. Both I. scapularis and D. variabilis larvae and nymphs were found to frequently feed on white-footed mice (Peromyscus leucopus), some of which were infected with B. burgdorferi but in contrast to I. scapularis, in which multiple life stages became infected, none of the D. variabilis ticks collected from the same population of mice were found infected with B. burgdorferi [35]. Thus, as a result of this rapid clearance of B. burgdorferi, Dermacentor fails to maintain a B. burgdorferi infection long enough for the bacteria to be transferred between developmental stages and this would be expected to prevent transmission to new hosts.

Nevertheless, some case studies provide indirect evidence that Dermacentor spp. ticks may be able to vector, as well as acquire, Borrelia infections. A case study from Bulgaria described an individual bitten by a Dermacentor marginatus tick, who subsequently developed an erythema migrans (EM) rash and was then found to be positive for B. burgdorferi exposure on serology, which was confirmed positive through a skin biopsy [74]. Similar anecdotal evidence of rashes and ulcers associated with Dermacentor spp. tick bites in France has also been noted [75]. Although these studies were on Dermacentor species other than D. variabilis, these results show that Borrelia can be detected in Dermacentor genus ticks and possibly vectored by them, which has reinvigorated investigation of the vectorial potential of D. variabilis. In a large study, 127 D. variabilis were obtained from the Human Tick Test Kit Program, created for ticks removed from military personnel in Wisconsin, USA. Of those ticks, 11% tested positive for B. burgdorferi by PCR [76]. These ticks were all engorged adults removed from human hosts, implying either that the D. variabilis ticks were infected with B. burgdorferi as nymphs and retained the infection into the adult stage or that the human hosts were infected prior to the tick bite and had a high enough spirochetal load to transmit the infection to the tick. This alternative was deemed unlikely as none had a known history of Lyme disease symptoms [76]. Although D. variabilis transiently acquiring Borrelia from an infected host does not equate to ready transmission of infection, there is a possibility, if not high probability of transmission if the ticks refeed rapidly. Unambiguous evidence of the duration of viable B. burgdorferi persisting in Dermacentor spp. ticks would require culture of bacteria, collected at sequential intervals from the initial acquisition of infection. The persisting ambiguity about the ability of D. variabilis to acquire Borrelia spp. motivated our testing for B. burgdorferi and B. miyamotoi in D. variabilis ticks collected in New Brunswick. While no ticks met the formal criteria for B. burgdorferi or B. miyamotoi presence and both pathogens were under-represented in D. variabilis ticks relative to wild animals suggesting poor retention of Borrelia spp. these ticks are vectors for a variety of other pathogens that warrant surveillance and appear to be extending their range northward into New Brunswick.

Study limitations and mitigations

This study took advantage of the power of community science; the public collected and submitted ticks throughout the region in exchange for identification of the tick and pathogen testing. Despite the many advantages of this community-based passive surveillance approach, interpretation of these findings entails awareness of issues intrinsic to passive surveillance. Passive surveillance relies on public detection and submissions of ticks, but the human population is not evenly distributed. As the tick recoveries documented here were not clustered around cities (Fig 2), using the location of tick encounters rather than donor address mitigated this bias. Motivation of members of public to participate and sample damage in the mail are difficult to quantify but are highly unlikely to be uniform, however, tick donations were primarily from veterinary clinics and participating veterinary clinics were largely constant during most of the ten-year span of this study. The first foci detected was in the southeast of the province, close to the study location so it is difficult to differentiate between the proximity to Nova Scotia’s abundant D. variabilis populations from recovery bias in explaining this tick “hot spot”. Nevertheless, the southwestern foci, which was not close to the study site, appeared only a year later, nor would recovery bias apply to the eTick platform. A bias affecting host data did occur in 2020. Prior to 2020, ticks from all hosts were collected and tested, whereas financial considerations resulted in paid testing in 2020 so that after this date, submissions were human-biased. As the eTick platform does not provide tick testing and involves no cost to the image contributors, this consideration would not apply to that data set.