Field surveys were conducted at 16 sites with contiguous forest in Ontario and Quebec, Canada in July and August 2019 ( Figure 1 ). Sites were selected based on their different degrees of B. burgdorferi risk related to the abundances and life stages of I. scapularis present locally as defined by the Institut national de santé publique du Québec (2018) and Public Health Ontario (2018), which ranged from possible to significant risk ( Supplementary Table 1 ). At each site, three grids of 40 m by 70 m were set up for sampling ticks and mammal hosts, which were maximally separated by 100 meters due to geographic barriers (e.g., streams or park trails).

Within each grid, four 70-meter long transects were used to sample ticks one time by dragging a 1 m² white flannel over low-lying vegetation. Flannels were checked every 10 meters, and questing ticks were removed. All ticks were kept in microvials with 95% ethanol, and larvae were pooled while nymphs and adults were kept individually. Tick specimens were identified to the species using dichotomous keys (Lindquist et al., 2016).

At each site, 84 Sherman live traps (H.B. Sherman Traps, Inc., Florida, United States) were placed along four parallel transects within each grid for three consecutive nights, representing a total of 4032 trap nights in our study. We targeted mouse (P. leucopus and P. maniculatus), shrew (B. brevicauda and S. cinereus), vole (Microtus pennsylvanicus and Myodes gapperi), and jumping mouse (Napaeozapus insignis and Zapus hudsonius) species. In the afternoon, a bait mixture of peanut butter and oatmeal, an apple piece, and a cotton ball were placed in each trap. Traps were checked the following morning. Juveniles and non-targeted rodent species were immediately released at the site of capture. Individuals of targeted species were euthanized via isoflurane inhalation followed by cervical dislocation. One red squirrel (Tamiasciurus hudsonicus) and two hairy-tailed moles (Parascalops breweri) were also euthanized due to severe injuries. Small mammals were searched for feeding ticks, and mammalian liver tissues were dissected and placed into microvials with 95% ethanol. Liver tissues were selected for pathogen testing, as they have been used for B. burgdorferi detection in wild rodents (Zinck and Lloyd, 2022). As in Tessier et al. (2004), a nested PCR using species-specific COIII primers was used to identify Peromyscus species ( Supplementary Methods ). All samples were accessioned in the collections of the Redpath Museum, McGill University ( Supplementary Table 2 ). Ethical approval and permits were issued by McGill University (AUP No. 2019-8086), the Ministère des Forêts, de la Faune et des Parcs (SEG permit No. 2019-06-04-008-00-S-F), and the Ministry of Natural Resources and Forestry (WSCA No. 1093495).

Concurrently, nine trail cameras (Force-10, SpyPoint Inc., Quebec, Canada) were placed 1 meter above the ground facing inside our grids and set to take three consecutive photos without delay for each detection. Host species were identified from photographs taken by camera traps. Birds, domestic pets, humans, and unidentified individuals were not included in our dataset.

For each site, total I. scapularis abundance was estimated by the sum of questing and feeding ticks collected from tick dragging and small mammal trapping, respectively. We used the total number of collected mammal individuals as a proxy for the abundance of small mammals locally. The relative abundance of P. leucopus was quantified as the number of collected P. leucopus individuals divided by the local abundance of collected small mammals. The number of mammal host species was estimated as the number of distinct species collected via small mammal trapping and detected in camera photographs.

DNA extractions and nested PCRs conducted by Geneticks, Inc. targeted five pathogens in our tick and small mammal specimens ( Supplementary Methods ; Supplementary Table 3 ). Adults and nymphs were tested individually, while larvae were pooled by grid if questing (2-10 larvae per pool) and by host if feeding (1-10 larvae per pool). All I. scapularis and small mammal specimens were tested for Anaplasma phagocytophilum, Babesia species, and Borrelia species. If a band was visible (i.e., positive PCR), we then tested twice more for false positives. The p44 gene was targeted to test for A. phagocytophilum (Holden et al., 2003). Babesia odocoilei and B. microti were targeted with the 18S rRNA region using the mic494 and odo563 inner primers, respectively. An additional primer set targeting the 18S rRNA of each Babesia species was used for confirmation (Persing et al., 1992). We also tested for B. burgdorferi sensu stricto and B. miyamotoi using the 5S-23S intergenic space region and the 18S rRNA region, respectively (Dibernardo et al., 2014; Zinck et al., 2021). An additional test using the flaB gene confirmed the presence of B. burgdorferi sensu lato (Wodecka, 2011). Bio Basic DNA Sequencing (Ontario, Canada) completed Sanger DNA sequencing of positive samples, with sequences assessed for quality control, ambiguous base calls, and end-reading errors using 4Peaks software. Pathogen species were confirmed with GenBank using a MEGABLAST search in the nucleotide BLAST database.

For each site, we calculated pathogen presence, prevalence, and diversity in I. scapularis and small mammal hosts. Pathogen presence indicated whether pathogens were present (1) or absent (0) locally in I. scapularis or in small mammal hosts. Pathogen prevalence was calculated as a proportion for questing I. scapularis by dividing the number of infected individuals and larval pools of I. scapularis by the total number of I. scapularis (individuals and larval pools). Feeding I. scapularis were excluded from this calculation, as they better represent the pathogens circulating in hosts and may artificially increase local infection prevalence. Pathogen diversity was defined as the total number of pathogen species found in I. scapularis and small mammal hosts.

All statistical analyses were performed in R v4.2.2. (R Core Team, 2022). We assessed the effect of I. scapularis and mammal hosts on pathogen presence, prevalence, and diversity across our sites in Central Canada. Using the rcorr function in the Hmisc package (Harrell Jr, 2021), we first calculated the correlation coefficients between small mammal abundance, the relative abundance of P. leucopus, mammal species richness, questing I. scapularis abundance, and total I. scapularis abundance. Small mammal abundance was highly correlated with mammal diversity (r = 0.54, p < 0.05), and was not included in further analyses. Using the scale function, biotic factors were centered by subtracting the variable average from each value and standardized. Spatial autocorrelation among biotic factors was assessed with Moran’s I with an inverse distance weights matrix using the moran.test function in the spdep package (Bivand and Wong, 2018).

We first evaluated the effect of I. scapularis abundance and mammal hosts on pathogen presence with two binomial generalized linear mixed models with a cloglog link function using the glmer function in the lme4 package (Bates et al., 2015). Our binary response variable was pathogen presence (1) or absence (0) in I. scapularis and small mammal hosts. Two separate models were run to determine the independent impacts that I. scapularis and mammal hosts have on pathogen presence. The first model used the total I. scapularis abundance as an independent variable. The independent variables of a second model included the relative abundance of P. leucopus and mammal species richness. We subsequently analyzed the impact of mammals hosts on pathogen prevalence in questing I. scapularis with a binomial generalized linear mixed model and a cloglog link function using the glmer function. A binomial model was used, as pathogen prevalence was calculated as a proportion. The independent variables in this third model were the relative abundance of P. leucopus and mammal species richness. Site was included as a random factor in all three models to account for spatial autocorrelation. Model selection was based on AIC values, with a smaller AIC indicating a better model fit, and the variance and standard deviation of Site.

Finally, we ran an ordinal logistic regression with the polr function in the MASS package (Venables and Ripley, 2002) to assess if pathogen diversity across our sites was affected by mammal hosts. Our independent variables included the relative abundance of P. leucopus and the number of mammal host species. We used the stepAIC function in the cAIC4 package (Säfken et al., 2021) to determine if additional models should be assessed.

The abundance of questing and feeding I. scapularis ranged from 0 to 164 individuals across our sites ( Tables 1 , 2 ). We collected a total of 382 questing I. scapularis including 255 larvae (29 pools), 126 nymphs, and one adult male, as well as 65 feeding I. scapularis including 57 larvae (17 pools) and 8 nymphs.

We collected a total of 105 small mammal individuals ( Table 2 ; Supplementary Table 4 , 5 ). The most abundant species was P. leucopus (31.4% of collected individuals; n = 33), which was present at 9 of 16 sites. Other collected small mammals, in decreasing order of abundance, were N. insignis (22.8%; n = 24), B. brevicauda (16.2%; n = 17), M. gapperi (15.2%; n = 16), P. breweri (1.9%; n = 2), S. cinereus (1.0%; n = 1), and M. pennsylvanicus (1.0%; n = 1). At each site, small mammal abundance ranged from 1 to 18 individuals, while the relative abundance of P. leucopus ranged from 0 to 1 ( Table 2 ). We identified 5 mammal species in photographs taken by cameras, which included squirrels (Sciurus carolinensis), chipmunks (T. striatus), white-tailed deer (O. virginianus), raccoons (P. lotor), and coyotes (Canis latrans). Between 2 and 8 mammal species were detected at each site via small mammal trapping and in camera photographs.

For questing I. scapularis, only nymphs tested positive for our pathogens of interest, while the 255 questing larvae (29 pools) and one adult male that we collected all tested negative. Of 126 I. scapularis nymphs, five tested positive for B. odocoilei, with one nymph at Site 2, Site 10, and Site 11 and two nymphs at Site 13 ( Tables 1 , 2 ). Similarly, I. scapularis nymphs infected with B. burgdorferi were found at Site 2 (one nymph), Site 10 (18 nymphs), and Site 11 (three nymphs). Local infection prevalence in questing I. scapularis ranged from 0% to 14.3% for B. odocoilei and from 0% to 25.4% for B. burgdorferi ( Supplementary Table 6 ). No questing ticks tested positive for A. phagocytophilum, B. microti, or B. miyamotoi.

For feeding I. scapularis, 2 of 17 larval pools and 1 of 8 nymphs were infected ( Tables 1 , 2 ; Supplementary Table 6 ). One larva at Site 11 was infected with B. odocoilei, but it was found feeding on a P. leucopus infected with B. microti. A larva from Site 13 feeding on an uninfected P. maniculatus and a nymph from Site 8 feeding on an uninfected P. leucopus were both infected with B. burgdorferi. No feeding ticks tested positive for A. phagocytophilum, B. microti, or B. miyamotoi.

Only two of the 105 small mammals that we collected were infected ( Supplementary Tables 4 , 5 ). At Site 11, one P. leucopus was infected with B. microti and at Site 16, one B. brevicauda also tested positive for B. odocoilei. Local infection prevalence of small mammal hosts ranged from 0% to 10% for B. odocoilei and from 0% to 16.7% for B. microti ( Supplementary Table 7 ). No small mammals tested positive for A. phagocytophilum, B. burgdorferi, or B. miyamotoi.

We detected pathogens in I. scapularis and small mammals hosts at 6 of our 16 sites, where up to 3 different pathogen species were present locally ( Figure 1 ; Tables 1 , 2 ). Local pathogen prevalence in questing I. scapularis adults, nymphs, and pools of larvae ranged from 0% to 26.8% ( Table 2 ). Pathogen diversity was found to be highest in areas with long-established populations of I. scapularis in southern Ontario and Quebec (Sites 2, 10, and 11). These sites had over 100 I. scapularis from at least two different life stages, where 2 or 3 pathogen species were detected locally ( Figure 1 ; Tables 1 , 2 ).

Small mammal abundance was significantly correlated to mammal species richness (r = 0.54, p < 0.05), questing I. scapularis abundance (r = 0.54, p < 0.05), and total I. scapularis abundance (r = 0.55, p < 0.05). Therefore, we excluded small mammal abundance from further analyses. We detected spatial autocorrelation in questing I. scapularis abundance (Moran’s I = 0.206, p < 0.05), but none was detected for the relative abundance of P. leucopus (Moran’s I = 0.072, p = 0.207) and mammal species richness (Moran’s I = 0.138, p = 0.115).

There was no effect of pathogen presence with total I. scapularis abundance (p = 0.144; Supplementary Table 8 ), the relative abundance of P. leucopus (p = 0.874; Supplementary Table 9 ), or mammal species richness (p = 0.192; Supplementary Table 9 ). Similarly, pathogen prevalence in questing I. scapularis was not affected by the relative abundance of P. leucopus (p = 0.824) or mammal species richness (p = 0.767; Supplementary Table 10 ). Feeding ticks were not included in this model, as to not artificially increase the local pathogen prevalence. Finally, the relative abundance of P. leucopus (p = 0.822) did not significantly predict pathogen diversity and was subsequently removed from our model ( Supplementary Table 11 ). Only mammal species richness (OR = 11.826, p < 0.05) was found to significantly predict pathogen diversity, with higher odd ratios of pathogen detection with greater mammal species richness ( Figure 2 ; Supplementary Table 11 ).

Abundances of key hosts, such as white-tailed deer and white-footed mice, have been associated with a greater I. scapularis abundance and greater B. burgdorferi prevalence in Ontario and Quebec (Bouchard et al., 2011; Bouchard et al., 2013; Simon et al., 2014; Werden et al., 2014). However, here, pathogen presence and prevalence were not affected by the abundance of I. scapularis nor the abundance and diversity of mammal hosts across our sites. Pathogen spread and transmission may have been affected by bird hosts, an unexplored factor in our study, due to their ability to feed immature ticks and harbor several tick-borne pathogens (Scott et al., 2022). Migratory birds may facilitate range expansions of tick-borne pathogens by introducing infected adventitious ticks to new locations in Canada (Scott et al., 2020; Scott and Pesapane, 2021). In addition, ground foraging birds were found to be significant contributors to the spread and transmission of tick-borne pathogens at the most northern parts of the distribution range of I. scapularis (Leo et al., 2017; Dumas et al., 2022).

We observed that mammal species richness significantly predicted pathogen diversity, with up to 3 pathogen species being detected locally ( Figures 1 , 2 ). Sites with greater pathogen diversity were associated with more diverse mammal communities. Locally, small mammals play an important role in feeding immature ticks and pathogen maintenance (Mather et al., 1989; LoGiudice et al., 2003). As these small mammals search for food resources, they may disperse ticks and their pathogens over short distances into nearby forest patches (Marrotte et al., 2017; Borgmann-Winter et al., 2021). Larger mammal hosts, such as white-tailed deer, can feed large burdens of ticks and act as key reproductive hosts for adult I. scapularis (LoGiudice et al., 2003; Werden et al., 2014). These mammal hosts may act as important facilitators for the long-range dispersal and establishment of tick populations and tick-borne pathogens, as they expand their ranges poleward in response to climate and land use changes (Dawe and Boutin, 2016; Diuk-Wasser et al., 2021). However, the spread and transmission of tick-borne pathogens may be limited in areas where I. scapularis or reservoir hosts, such as P. leucopus, have not yet established (Roy-Dufresne et al., 2013; Simon et al., 2014; Ripoche et al., 2022; Millien et al., 2023). In these areas, other medically significant tick vectors (e.g., Ixodes cookei with Powassan virus; Gasmi et al., 2018) and reservoir hosts (e.g. chipmunks or shrews) may contribute more strongly to the spread and transmission of tick-borne pathogens (Levi et al., 2016; Gasmi et al., 2018; Dumas et al., 2022).

The co-occurrence of multiple tick-borne pathogens may increase the risk of co-infection in tick and host populations locally (Cutler et al., 2021). Co-infections can occur in adult ticks after feeding on different infected reservoir hosts or when co-feeding with infected ticks on the same host (Voordouw, 2015; Cutler et al., 2021). Although we did not detect any co-infections in our tick and small mammal specimens, co-infections have been detected at varying levels of infection prevalence in I. scapularis in Canada (Dibernardo et al., 2014; Dumas et al., 2022; Wilson et al., 2022). The majority of these co-infections occurred in areas near our sites in southern Ontario and Quebec, where long-established I. scapularis populations are located (Dumas et al., 2022; Wilson et al., 2022). If more tick-borne pathogens are co-occurring locally, it may lead to an increased risk of co-infections of tick-borne diseases in human populations. As a result, humans co-infected with multiple tick-borne pathogens may display complex clinical manifestations that present diagnostic challenges (Cutler et al., 2021). Although we detected no pathogen species at some sites, there does not appear to be any environmental or host suitability limitations in these areas that will prevent those tick-borne pathogens from spreading there in the future.

Our study demonstrates that comprehensive surveillance efforts targeting questing and feeding I. scapularis of all life stages and small mammal hosts is required to detect the geographic extent and co-occurrence of tick-borne pathogens in Canada. Concurrent testing of multiple tick-borne pathogens is necessary to better detect the risk of co-infections, especially as the co-occurrence of pathogens become more prevalent in areas with increased tick abundances and more diverse host communities. These results show expanding ranges of certain tick-borne pathogens transmitted by I. scapularis, especially in areas where B. burgdorferi has not yet established. It would also be relevant to test questing larval I. scapularis for tick-borne pathogens with known transovarial transmission, such as B. odocoilei (Zembsch et al., 2021). Emerging tick-borne pathogens are advancing poleward in Canada with the expanding ranges of tick and host populations, where risks for pathogen transmission will rise. Proactive surveillance efforts outside the known distributions of pathogens of concern for wildlife and human health will improve our ability to better anticipate the risk for tick-borne diseases in these regions.