https://orcid.org/0000-0003-2133-0220
Figures
Abstract
Expanding geographic distribution and increased populations of ticks has resulted in an upsurge of human-tick encounters in the United States (US), leading to an increase in tickborne disease reporting. Limited knowledge of the broadscale spatial range of tick species is heightened by a rapidly changing environment. Therefore, we partnered with the Forest Inventory and Analysis (FIA) program of the Forest Service, U.S. Department of Agriculture and used passive tick surveillance to better understand spatiotemporal variables associated with foresters encountering three tick species (Amblyomma americanum L., Dermacentor variabilis Say, and Ixodes scapularis L.) in the southeastern US. Eight years (2014–2021) of tick encounter data were used to fit environmental niche and generalized linear models to predict where and when ticks are likely to be encountered. Our results indicate temporal and environmental partitioning of the three species. Ixodes scapularis were more likely to be encountered in the autumn and winter seasons and associated with soil organic matter, vegetation indices, evapotranspiration, temperature, and gross primary productivity. By contrast, A. americanum and D. variabilis were more likely to be encountered in spring and summer seasons and associated with elevation, landcover, temperature, dead belowground biomass, vapor pressure, and precipitation. Regions in the southeast least suitable for encountering ticks included the Blue Ridge, Mississippi Alluvial Plain, and the Southern Florida Coastal Plain, whereas suitable regions included the Interior Plateau, Central Appalachians, Ozark Highlands, Boston Mountains, and the Ouachita Mountains. Spatial and temporal patterns of different tick species can inform outdoorsmen and the public on tick avoidance measures, reduce tick populations by managing suitable tick habitats, and monitoring areas with unsuitable tick habitat for potential missed encounters.
Author summary
The study highlights the significance of passive tick surveillance data collected by Forest Inventory and Analysis crews in providing valuable information into the presence and distribution of ticks in the southeastern United States. We used ecological niche modeling and generalized linear models to assess the geographic regions and temporal periods associated with where and when ticks are likely to be encountered. In the region, ticks remain active throughout the year with their distribution influenced by climatic and topographical factors. Of interest, maximum temperature was a significant environmental variable for all three species suggesting that distributions may be altered as the climate warms. Elevation and landcover were important variables for both Amblyomma americanum and Dermacentor variabilis, whereas Ixodes scapularis populations were correlated with evapotranspiration, vegetation indices, and soil organic matter. The research also identified new tick occurrence records providing data in a region with minimal infrastructure for tick surveillance, but with many ticks and tick-borne diseases. Continued long-term passive surveillance with collaborations with academic and government partnerships will help monitor tick distribution changes resulting from landscape and temperature changes which affect public health risks.
Citation: Butler RA, Papeş M, Vogt JT, Paulsen DJ, Crowe C, Trout Fryxell RT (2024) Human risk to tick encounters in the southeastern United States estimated with spatial distribution modeling. PLoS Negl Trop Dis 18(2): e0011919. https://doi.org/10.1371/journal.pntd.0011919
Editor: Álvaro Acosta-Serrano, University of Notre Dame, UNITED STATES
Received: June 6, 2023; Accepted: January 14, 2024; Published: February 14, 2024
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: Base layers for each map are publicly available from the U.S. census (https://www.census.gov/geographies/mapping-files/time-series/geo/carto-boundary-file.html). Environmental spatial data used in environmental niche modeling were collected from EarthExplorer (http://earthexplorer.usgs.gov) and EarthData (http://earthdata.nasa.gov/). Tick encounter data can be accessed at Dryad (https://doi.org/10.5061/dryad.v41ns1s3n).
Funding: The Forest Inventory and Analysis Program of the Forest Service, United States Department of Agriculture providing funding for the collection and submission of the ticks used in the study; procured by JTV and RTTF. JTV and CC are employed by the USDA Forest Service Southern Research Station. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The geographic distribution of human-tick encounters for medically important tick species is rapidly increasing in the United States (US) [1,2]. Range expansion for commonly encountered tick species (Amblyomma americanum L., Dermacentor variabilis Say, and Ixodes scapularis L.) has resulted in an upsurge of tick-borne disease cases in the last two decades including spotted fever group rickettsiosis, anaplasmosis, ehrlichiosis, alpha-gal syndrome, Powassan virus, and Lyme disease [3–7].
Current distribution range maps for these tick species are commonly based on known occurrences at the continental or county level which is important for surveillance and management of ticks for regional and county health departments [8,9]. These range maps or distribution maps are often based on administrative landmarks (county boundaries) rather than tick and host biological patterns because these maps encompass large spatial areas with environmental factors imprecise for the species’ distribution [10]. Environmental or ecological niche models (ENMs) create maps with estimated distributions based on interactions of the species with environmental variables in time and space [11]. Recent ENMs for A. americanum predicted this species to inhabit regions in the northern, southeastern, and western regions of the US and Mexico, as well as in the Midwest from eastern Texas to Kansas, Oklahoma, and Missouri [12–14]. Similarly, D. variabilis was predicted to occur in parts of Canada and Mexico, as well as northern, southern, and midwestern regions of the US including regions in California [15]. Predictions for I. scapularis were distributed throughout regions in the eastern and midwestern US [16]. Understanding the effects that environmental variables have on each tick species distribution is vital because of the recent and predicted impacts of climate and land-use change on their population dynamics [17–20].
Ticks spend the majority of their lives in the environment compared to time spent on hosts; thus, understanding how the environment influences tick populations will lead to increased understanding of ticks and their associated pathogens which can lead to effective management strategies [21]. For example, soil properties such as percent litter coverage and soil moisture are associated with A. americanum abundance [22,23]. Additionally, land management decisions (e.g., burning) were associated with decreased immature I. scapularis [24] and A. americanum populations [25,26]. Specifically prescribed burns reduce tick abundance by factors such as heat exposure or decrease in soil moisture [26]. Climatic variables (e.g., temperature, vapor pressure, and precipitation) have also been associated with A. americanum, D. variabilis, and I. scapularis (e.g., [27,28]. Landscape variables associated with forests (e.g., land cover, primary productivity) are important for host populations and likely regulate infesting tick populations [e.g., 29,30]. For example, normalized difference vegetation index (NDVI), a measure of greenness, has been associated with the abundance of A. americanum [31]. Knowing if an environmental variable is associated with human-tick encounters can be an important surveillance tool for monitoring and a potential management tool for controlling host species.
As tick populations are expanding in a rapidly changing environment, we collaborated with the Southern Research Station’s Forest Inventory and Analysis (FIA) Program of the Forest Service, U.S. Department of Agriculture (USDA) to evaluate the environmental conditions that could increase likelihood of encountering ticks. We chose to work with FIA because foresters are known to collect data at a variety of sites with varying environmental conditions and their forest crews are consistently exposed to ticks in the environment. Specifically, foresters collected encountered ticks while working on sites around the southeastern US. We used eight years (2014–2021) of tick encounter data to create environmental niche and generalized linear models to understand where and when ticks are likely to be encountered. Here we test the hypothesis that temporal, climatic, physiographic, and soil variables are reliable predictors of human-tick encounters for A. americanum, D. variabilis, and I. scapularis in the southeastern U.S.
Materials and methods
Tick encounter data
Ticks were collected from forest crews employed by FIA. Passive tick collections were opportunistic and occurred when crews worked at plots in the southeastern U.S. between 2014 and 2021. Every year 1/5, 1/7, or 1/10 of the total plots, which are spatially distributed throughout each state (2,000 to 4,800 forest plots per state), are sampled. Crew members visit a single plot each day to inventory each site which takes an entire day. Ticks encountered that day were placed into a single vial containing 80% ethanol and labeled with the date, forestry crew identification number, and GPS coordinates where the crew was working [32]. Encountered ticks were sent to the University of Tennessee, Knoxville Medical and Veterinary Entomology laboratory where they were identified to species and life stage using taxonomic keys [33–36].
Statistical analysis
Generalized linear models were created from (PROC GLIMMIX) in Statistical Analysis Software (SAS, ver. 9.4, Cary, North Carolina) with two-tailed hypotheses (α = 0.05) to analyze how season affects the presence of each tick species together and by life stage. Date of tick collection was transformed into a categorical season variable (winter, spring, summer, and autumn) based on solstice or equinox to account for daylength. Season was used to determine the probability to detect tick presence for each species separately in the binary logit models. Odds ratios and their 95% confidence intervals were calculated for independent variables that were successful at predicting the presence of each tick species. Relative tick encounters, similar to human-tick encounter phenology, were graphed by summing tick abundance for each month per years collected and transformed as a percentage.
Environmental niche modeling of potential suitability for ticks
FIA crews used hand-held GPS receivers to record the latitude and longitude of each FIA plot. The coordinates used for niche modeling are within 1.6 km (1 mile) of the actual encounter sites because (1) many of the plots are located on private lands and these exact locations are kept confidential (https://www.fia.fs.usda.gov/tools-data/spatial/Policy/default.asp) and (2) ticks may have been encountered traveling to or from plots or while working at the plot. We coarsened the spatial resolution of the model to account for differences between plots and uncertainty of collecting locations. We used these approximate locations as human-tick encounter data to train environmental niche models from 258 sites where A. americanum were encountered, 90 sites where D. variabilis were encountered, and 36 sites where I. scapularis were encountered.
We used 20 environmental rasters (gridded data) as predictor variables in niche models acquired from EarthData (https://earthdata.nasa.gov/) or EarthExplorer (http://earthexplorer.usgs.gov) using NAD 1983 geographic coordinate system (Table 1). Each encounter was matched temporally to raster values with the same or nearest date available. We averaged rasters from every date each tick was collected from human hosts to encompass all environmental conditions when tick species were active on human hosts (FIA crews). To compensate for potential spatial sampling errors of ticks, for ticks collected while walking to and from plots and GPS recorded outside property boundaries, we aggregated (coarsened) the original spatial resolution of rasters from 500 meters to 4.8 km. All raster processing steps were completed in ESRI ArcMap 10.7.
Environmental niche models were fitted using human-tick encounter data and environmental variables in the maximum entropy algorithm Maxent (Version 3.4.0) that estimates species’ potential geographic distributions [47]. Maxent was selected due to its robustness in handling presence-only model fitting with limited occurrences [48]. We removed duplicate presence records and ran independent models for each tick species with cross-validation and five replicates, each based on a maximum of 5,000 iterations. For each model, the algorithm selected 10,000 random background samples (or pseudo-absences) to contrast their environmental conditions to those at presence locations. A 10% training presence threshold (allowing 10% of training presence data to be predicted unsuitable) was applied to convert the model outputs of continuous probability of suitability to binary maps of suitable-unsuitable values. Because Maxent is a stable machine-learning platform for controlling correlated variables, environmental raster images were not assessed for autocorrelation beforehand but left to the algorithm to discern [49] (S1 Table). Highly correlated environmental variables do not impact model performance when model transfer is not being used in Maxent [50]. Due to lower sample size for I. scapularis only the linear, quadratic, and product features were used in Maxent. Additionally, categorical variables (land cover) were removed from the environmental niche model for I. scapularis because the variable contribution was not balanced. For A. americanum and D. variabilis an addition of the hinge feature was used to build a piecewise linear exponent because there were more encounters (i.e., larger presence dataset) available for model training [47].
To quantify model performance we used area under the curve (AUC) of the receiver operating characteristics; AUC ≥ 0.9 indicate excellent models, 0.8–0.9 good, 0.7–0.8 fair, 0.6–0.7 poor, and < 0.6 failed models [51]. We also observed the omission error for each model at a 10% training threshold. In addition, we overlaid an U.S. ecoregions boundaries shapefile to evaluate risk based on regional similarities or differences (e.g., geology, soils, climate, vegetation) of distinct geographic areas [52–56].
Independent model testing
To test each species niche model in addition to the internal cross-validation of Maxent algorithm, we used targeted field work and an independent dataset of previously reported tick records for each tick species by county. We selected three USDA Southern Research Station Experimental Research Forests near the University of Tennessee, Knoxville to confirm the presence or absence of human-tick encounters in this area in 2021. The three experimental forests included Coweeta, Blue Valley, and Bent Creek in western North Carolina and northern Georgia in southern Appalachian Mountains. Each collection site in western North Carolina is representative of forests previously disturbed by logging and mining [57–59]. The Coweeta Hydrologic Laboratory has an area of 1,600 ha and is dominated by mixed-oak forests with understories of noncontinuous mountain laurel and rhododendron [60]. Soils in this region are comprised of Inceptosols and Ultisols [58]. The Blue Valley experimental forest has an area of 526 ha dominated by eastern white pine and stand of oak-hickory in the Blue Ridge Mountains [58]. This area contains acidic, infertile, well drained soils with vegetation primarily consisting of buckberry shrubs [58]. Bent Creek experimental station is comprised of 2,550-ha of upland hardwood forests primarily dominated by oak-hickory stands and understory vegetation of rhododendron [58]. The Asheville Basin region contains soils with low organic matter contents, clay layers, and reduced fertility whereas the Mountain Highland are Inceptisols and acidic [58].
We provided vials to crews to collect encountered ticks, but we also conducted active surveillance in these forests. We set dry ice traps for overnight trapping once monthly (March-June 2021) at five different sites within each forest [61]. The use of dry ice traps were determined by rough terrain at the collection site and limited funding, and the timing of our surveillance was informed by prior knowledge that most ticks are active March through June in the region. Collected ticks were stored in ethanol filled vials, then identified to species and sex as described above.
Additionally, we compared county level records with our predicted distributions. County level records for I. scapularis, A. americanum, and D. variabilis were downloaded on May 14, 2022, from the Center for Disease Control (https://www.cdc.gov/ticks/surveillance/TickSurveillanceData.html). True positives, used to calculate sensitivity, were identified as counties with available tick records and predicted suitable by the niche models. Here, we considered the CDC datasets to be the standard, so false positive counties were those predicted suitable but lacking tick presences, indicating model commission error, and false negative counties were those with known tick presences but predicted unsuitable by the niche models, representing omission error. Although the true distribution for each tick species may not be fully represented due to differences in collection, the CDC datasets represent the most comprehensive comparison for our models because they include surveillance by ArboNET or literature published each year throughout the Southeast. Sensitivity tests were used to determine the proportion of true positives and were calculated in SAS using PROC FREQ. County-level validation of models calibrated with environmental variables with 4.8 km resolution and GPS records of ticks represents a compromise between the need to validate the models and the lack of higher resolution presence data readily available for each species.
Results
A total of 1,901 ticks were collected and submitted to us during the study period from 384 sites: 1,720 (90.5%) A. americanum from 258 sites, 136 (7.2%) D. variabilis from 90 sites, and 45 (2.3%) I. scapularis from 36 sites. Of the A. americanum records, 669 were adults, 851 were nymphs, and 200 were larvae. There were 135 D. variabilis adults and one D. variabilis nymph. Finally, there were 45 I. scapularis adults. Ticks were primarily collected and submitted by crews in Kentucky, South Carolina, and Tennessee (Table 2).
Total number of tick-encounter sites are separated by state in the southeastern United States (2014–2021). In total there were 1720 Amblyomma americanum (200 larvae, 851 nymph, 310 males, and 359 females), 136 Dermacentor variabilis (1 nymph, 73 females, and 62 males) and 45 Ixodes scapularis (27 females and 18 males).
Statistical analysis
Season was significantly associated with presence for all three species: A. americanum (F3, 349 = 13.69; p < 0.0001), D. variabilis (F3, 349 = 6.07; p = 0.0005), and I. scapularis (F3, 349 = 22.14; p < 0.0001). The likelihood of A. americanum being present was highest in spring compared to any other season, and higher in summer compared to autumn and winter. Amblyomma americanum larvae were most likely to be present on human hosts in summer months, nymphs were more likely to be present in spring compared to autumn months, and adults were more likely to be present in the spring compared to any other season. Dermacentor variabilis adult encounters were greater in summer months compared to spring and winter months; a single nymph was collected in the spring. Finally, encounters of adult I. scapularis were more likely to occur in autumn and winter months compared to spring and summer months. A detailed breakdown of each species monthly percent encounter throughout the eight-year study period is shown in Fig 1.
Percent monthly encounter for Amblyomma americanum, Dermacentor variabilis, and Ixodes scapularis encountered by Forestry Inventory and Analysis crews in the southeastern United States between 2014 and 2021.
Environmental niche modeling of potential suitability for ticks
Lambda file results for all environmental variables in our ENM’s can be found in (S2 Table). Three of the five replicate niche models for A. americanum had AUC values greater than 0.8 (Table 3). We averaged the raster outputs (probability of suitability) of the three Maxent models to create a single map of potential suitability for A. americanum (Fig 2). Model 2 had the highest AUC and elevation, landcover, minimum temperature, and vapor pressure were the best predictors of habitat suitability (Table 3). In models 1 and 3, the same three contributing variables as for model 1 predicted environmental suitability except maximum temperature that had a higher contribution than minimum temperature in these two models. Resulting models indicate potential suitable environments where A. americanum are likely to be encountered; thus, crews working in forests throughout Tennessee, Kentucky, northern Florida, Arkansas, North Carolina, and eastern Virginia have an increased likelihood of encountering A. americanum (Fig 2).
The model was based on tick encounters collected by Forest Inventory and Analysis Program of the Forest Service, US Department of Agriculture foresters in the southeastern United States (2014–2021). Landcover, elevation, maximum temperature, minimum temperature, and vapor pressure contributed the most to predicting this tick’s geographic suitability. The following publically available link was used as the base layer of the map (https://www.census.gov/geographies/mapping-files/time-series/geo/carto-boundary-file.html).
Values for each variable represent that variables percent contribution to each model, bolded values were those identified as significant contributing variables to the model (greater than 0.1 or 10%).
Three D. variabilis models had AUC values greater than 0.8 (Table 2) so we averaged the outputs of the three models to generate the predicted suitability map for D. variabilis (Fig 3). High contribution variables to all three models included elevation, landcover, and maximum temperature. Differences between models were represented by precipitation and vapor pressure in the first model, vapor pressure in the second model, and dead belowground biomass and precipitation in the third model (Table 2). Areas of potential suitability and thus concern for crews encountering D. variabilis included regions throughout Tennessee and Kentucky, northern North Carolina, Alabama, Arkansas, and eastern Virginia (Fig 3).
The model was based on tick encounters collected by Forest Inventory and Analysis Program of the Forest Service, US Department of Agriculture foresters in the southeastern United States (2014–2021). Variables landcover, elevation, precipitation, vapor pressure, and dead belowground biomass contributed the most to predicting this tick’s geographic suitability. The following publically available link was used as the base layer of the map (https://www.census.gov/geographies/mapping-files/time-series/geo/carto-boundary-file.html).
Only one model for I. scapularis had an AUC value greater than 0.8 (Table 2). The high contribution environmental variables were evapotranspiration, gross primary productivity, maximum temperature, vegetation indices, and soil organic matter (Table 3). The area of potential suitability for I. scapularis included all ecoregions in the southeastern US; however, there was a reduced risk to encounter I. scapularis ticks in southern Florida, eastern Tennessee and Mississippi, and western Arkansas (Fig 4).
The model was based on tick encounters collected by Forest Inventory and Analysis Program of the Forest Service, US Department of Agriculture foresters in the southeastern United States (2014–2021). Variables soil organic matter, vegetation indices, maximum temperature, gross primary productivity, and evapotranspiration contributed the most to predicting this tick’s geographic suitability. The following publically available link was used as the base layer of the map (https://www.census.gov/geographies/mapping-files/time-series/geo/carto-boundary-file.html).
Independent model testing
We tested each resulting niche model with data from passive and active surveillance at three Experimental Research Forests and by comparing predicted suitability maps with published or available county-level tick records. Passive and active tick collections at these targeted forests resulted in a total of five D. variabilis provided by a camper in the area, but the location of encounter was not recorded. There were no ticks collected in any of the 15 traps that were placed in the three experimental research forests, nor were ticks encountered on the individual walking to and from each site. The CDC tick surveillance datasets indicated D. variabilis to be reported but there were no county records for A. americanum and I. scapularis in Macon County North Carolina where The Coweeta and Blue Valley Experimental Forests are located. Similarly, CDC surveillance data showed that D. variabilis was established but there were no county records for A. americanum and I. scapularis in Buncombe County North Carolina where The Bent Creek Experimental Forest is located. For all three tick species, our potential distribution maps suggested suitable habitats were present in regions scattered throughout Macon County; however, suitable habitats were found scattered throughout Buncombe County for D. variabilis and I. scapularis but not for A. americanum. Our models correctly predicted sites where dry ice traps were located to be environmentally unsuitable for all three tick species; these absences and unsuitable sites were found in a landscape of predicted suitability by ENMs (Fig 5).
Maps indicating unsuitable habitat for sites used as active surveillance with dry ice traps for tick species (a) Amblyomma americanum, (b) Dermacentor variabilis, and (c) Ixodes scapularis in Coweeta, Blue Valley, and Bent Creek experimental forests in western North Carolina and northern Georgia in southern Appalachian Mountains (2021). Eco regions included Piedmont (45), Blue ridge (66), Ridge and Valley (67), Southwestern Appalachians (68), and Central Appalachians (69). The following publically available link was used as the base layer of the map (https://www.census.gov/geographies/mapping-files/time-series/geo/carto-boundary-file.html).
Overall, the CDC website identified 390 I. scapularis-positive counties and 674 no county records, 718 D. variabilis-positive counties and 346 no county records, and 561 A. americanum -positive counties and 503 no county records. When we compared our resulting maps at the county level to the previously recorded data, sensitivity values and associated confidence intervals for A. americanum were 0.9465 (0.9279–0.9651), for D. variabilis were 0.8175 (0.7893–0.8458), and for I. scapularis were 0.9956 (0.9905–1.00). This indicates that the models predicted suitable habitats (~encounter areas) similar to those reported by others. There was no observed difference between county status (established, reported, or new county record) and the number of A. americanum or D. variabilis collected from foresters. Counties noted as established in the CDC database for I. scapularis were found to have multiple ticks collected compared to reported or new county records.
Additionally, as a part of the project, we report several new county records for each species. New county records for A. americanum included: Kemper (MS), Madison (FL), Taylor (FL), Gilchrist (FL), Williamsburg (SC), Clarendon (SC), Lexington (SC), Edgefield (SC), Saluda (SC), Newberry (SC), Fairfield (SC), York (SC), Cherokee (SC), Anderson (SC), Halifax (NC), Franklin (NC), Amherst (VA), Henderson (KY), Hancock (KY), Trimble (KY), Kenton (KY), Scott (KY), Carter (KY), Johnson (KY), Perry (KY), Knox (KY), Jackson (KY), Lincoln (KY), Marion (KY), Casey (KY), Adair (KY), Russell (KY), Cumberland (KY), Monroe (KY), Pike (KY) (Fig 6). New county records for D. variabilis included: Pickens (SC), Wayne (TN), Lawrence (TN), Polk (TN), Daviess (KY), Hancock (KY), Butler (KY), Larue (KY), Green (KY), Taylor (KY), Cumberland (KY), Anderson (KY), Knox (KY), Clay (KY), Anderson (KY), Trimble (KY), Carter (KY), and Johnson (KY) (Fig 6). New county records for I. scapularis included: West Feliciana (LA), Dade (GA), Union (TN), Simpson (KY), Nelson (KY), Rockcastle (KY), Leslie (KY) (Fig 6).
The following publically available link was used as the base layer of the map (https://www.census.gov/geographies/mapping-files/time-series/geo/carto-boundary-file.html).
Discussion
Three-host ticks reside in the environment for the majority of their two- to three-year life cycle, leaving them susceptible to biotic and abiotic factors [62]. Consequently, ecological niches and habitat suitability often differ for each species spatially and temporally in a rapidly changing environment [63]. In this study, we used ecological niche and generalized linear models to describe geographic regions (where) and temporal periods (when) that represent a risk for humans to tick encounters in the southeastern US. We identified environmental variables associated with the human-tick encounters of three common human biting tick species, which are important indicators for tick encounters and could be targeted for management. The only common environmental variable significantly contributing to the human-tick encounters of all three species was maximum temperature, which supports the expectation that tick distributions will change in warming climates [64–66]. Additionally, long-term passive surveillance is an important tool for obtaining broad-extent occurrence records which can be used to monitor geographic distribution shifts of ticks and public health risk [67,68]. This dataset highlights the potential power of passive tick surveillance by trained scientists from federal agencies to generate accurate distributional information.
Amblyomma americanum
According to many published papers and citizen scientist reports, the most abundant and encountered tick in the southeastern US is A. americanum and the species is most likely to be encountered in the spring [69,70]. The likelihood of encountering A. americanum at a site was determined by five environmental variables (elevation, landcover, maximum temperature or minimum temperature, and vapor pressure). Previously, others reported variables such as precipitation, vapor pressure, diurnal range, and temperature as important contributors to predicting A. americanum suitability and potential range [17,12,14]. Our environmental niche model highlighted regions at higher risk of encountering A. americanum (Fig 2) that align with the Interior Plateau, Interior River Valleys and Hills, Ozark Highlands, Boston Mountains, Arkansas Valley, Ouachita Mountains, northern region of the Southeastern Plains, northeastern Piedmont, Northern Southern Coastal Plain, Southwestern Appalachians, and western regions of the Central Appalachian ecoregions. These ecoregions suitable for A. americanum contain upland hardwood forests, grassy plateaus, and low elevations in between the Appalachian Mountains and regions in the Southern Plains [71]. A similar pattern to our results was observed based on ticks collected from wildlife hosts in Florida where A. americanum was more likely to be present in the northern region of the Southern Coastal Plain [72].
Importantly, our results also predicted when and where A. americanum may not be encountered: fall and winter months and locations typically thought to have standing water. Less suitable areas were comprised of wetlands and bottomland hardwoods (e.g., Mississippi Alluvial plain, Southern Florida Coastal Plain), in agreement with previous reports [14,73]. These areas have flat plains with wet soils, marshlands, and swampy land cover [74]. Ticks were less likely to be found in the Southeastern Plains, a region that has historically experienced various land use changes such as plantation forestry and agriculture; these land management decisions could have impacts on tick abundance [75,76]. The Southeastern Plains in Tennessee are known to be occupied by A. americanum and their associated tickborne diseases such as human ehrlichiosis and spotted fever group Rickettsioses [77,78].
Similar to our ENM, A. americanum populations are less likely to occur in regions with high elevations such as in the Smoky Mountains [17]. Importantly, because A. americanum was associated with both elevation and temperature variables, it is possible to investigate these relationships in more detail as higher elevations begin to warm. Future models could predict that, as areas become warmer, more tick encounters will occur at higher elevations where this species was previously absent [17].
Dermacentor variabilis
Foresters frequently encountered D. variabilis adults in summer months, as previously reported [79]. Six environmental variables contributed to the environmental niche models with consistent variables including elevation and landcover; and the remaining four variables (maximum temperature, precipitation, vapor pressure, and dead belowground biomass) varied in contribution for each model. Similarly, previous research found precipitation, temperature, and elevation associated with D. variabilis [15]. Ticks were most likely to be present throughout Kentucky and Tennessee, and the predicted suitability was fragmented in northern Arkansas, North Carolina, and Virginia. Like our A. americanum model, the model for D. variabilis predicted suitable areas in ecoregions with upland hardwood forests and fewer extreme elevation changes (e.g., Ozark Highlands, Boston Mountains, Arkansas Valley, Interior Plateau, Interior River Valleys and Hills, Southwestern Appalachians, Piedmont, Mississippi Valley Loess Plain, Ridge and Valley, and the Central Appalachians) [71].
Prior to the range expansion of A. americanum, D. variabilis was the most commonly encountered tick in southeastern US [14,80–82]. Areas where D. variabilis was least likely to be encountered included more southern areas of the study region such as the Southern Coastal Plain and Southeastern Plains, typically characterized as subtropical, low-elevation, sandy, and areas with marshlands in Florida [71]. Historically D. variabilis were reported throughout the Midwest and Southeast but not regions in the upper Northeast and Southwest [83]. Recently, D. variabilis encounters were predicted by warm temperatures and low precipitation, thus climate change could improve environmental conditions for this species, increasing their range into the upper Northeast US [15].
The Southeastern Plains region is characterized by drought-prone and nutrient-poor soils, with most land use practices reversing to industrial forests which has resulted in extreme clear cutting [84]. These conditions could lead to unsuitable environmental conditions for this tick species due to its association with precipitation, landcover, and vapor pressure. Of interest is the potential role that dead belowground biomass has on D. variabilis because this variable can be manipulated with land and forest management decisions. Intense forest harvesting methods could increase dead belowground biomass due to remaining root structures underground [85]. Dead belowground biomass in the form of dead roots has the ability to hold large amounts of moisture [86] for water storage [87]; which could be important for adult D. variabilis survival because they possess greater survival in dry environments compared to other ixodid tick species [88]. Immature D. variabilis often quest below leaf litter [89] making them difficult to collect with active surveillance and easier to collect with passive surveillance, from small sized hosts [59,69,77,90].
Ixodes scapularis
Historical encounters of I. scapularis in the Southeast are receiving more attention as recent reports are confirming more I. scapularis in the region [91]. This study confirms these more recent publications, with adult I. scapularis being encountered in the fall and winter and nearly the entire environmentally suitable region [16,92,93]. Although adult I. scapularis were encountered less frequently than D. variabilis and A. americanum, their predicted geographic range was larger. Specifically, I. scapularis was predicted to be present in all southeastern states and all ecoregions except the floodplains around the Mississippi River (Mississippi Alluvial Plains and Mississippi Valley Loess Plains). The environmental variable landcover was removed from our I. scapularis model because the available presence records did not represent well the variation in landcover categories within the model. Environmental variables that contributed to this ENM included evapotranspiration, gross primary productivity, maximum temperature, vegetation indices, and soil organic matter. Interestingly, the antagonistic effects of transpiration and relative humidity indicate that I. scapularis could withstand questing when evapotranspiration is high because it is buffered by drastic changes in relative humidity in forests [94]. Vegetation indices contributing to adult I. scapularis populations when they quest in colder months is likely associated with plant dormancy traits [95]. Previous research found variables vapor pressure, elevation, forest cover, isothermality, and temperature associated with their ENMs [16,96]. Similar to our I. scapularis model, ticks were less likely to be found at high elevations in the Appalachian Mountains or the Blue Ridge, or southern Florida or the Southern Florida Coastal Plain, and the Louisiana Coast or the Mississippi Alluvial Plain ecoregions [16,71]. In these ecoregions it could be difficult for I. scapularis to survive off host because they are dominated by wetland habitats [71].
Studies of immature I. scapularis reported southern populations questing in leaf litter while genetically distinct northern populations questing higher on the vegetation, out of leaf litter [91,97]. These behavioral and genetic differences might also be reflected in our ecological niches because variables associated with questing and surviving in an environment (e.g., temperature, evapotranspiration, vegetation indices, etc.) contributed to model calibration. Importantly, the I. scapularis submitted by foresters in this study were adults and were not genotyped. Follow up studies should compare the ecological niches of these genetically and behaviorally distinct populations.
Model testing
Our models could predict the presence of each tick species but had difficulties predicting the absence of each tick species. This was confirmed with sensitivity and specificity testing as well as conducting targeted surveillance. These results are to be expected because the Maxent algorithm was only trained with presence records that had latitude and longitude coordinates (GPS records). Presence records are readily available for this study because some crews were more vigilant tick collectors and absences could be due to lack of reporting and/or submitting. Nevertheless, our models were able to sufficiently predict tick absence when compared to current literature and active surveillance. Here, few encounters occurred in areas near the foothills of the Appalachian Mountains and our active surveillance supported that as well. Campers in the Bent Creek area helped us confirm D. variabilis, indicating ticks may be present in the region, but the likelihood of encounters at these sites is rare compared to other locations, a finding that is similar to other studies in the area [98]. These data support the overall findings that higher elevation areas are often unsuitable tick habitat, which could explain the decreased number of human-tick encounters in high elevation areas like Appalachia [99–101]. Previous research documented that air temperature and density decrease as elevation increases [102]. A possible future research direction could be to monitor these sites as temperature and precipitation change in response to climate change [103], because all three tick species in this study were associated with temperature and none were solely dependent on elevation. Utilizing community science and passive surveillance, this study identified 60 new county records for tick occurrence in eight states. Data generated from this research could be an important tool for raising awareness of increased tick species distributions and used for tick avoidance and management regimes.
Potential for sampling bias
All three species were absent on human hosts in areas bordering the Mississippi river, coastlines around Louisiana, and southern Florida. This resulted in us reporting that ticks were absent from areas comprising wetlands, analogous to previous studies [99,104]. The absence of ticks on human hosts in these areas could also indicate sampling bias because fewer records were submitted by forest crews working in these areas. This is a known sampling limitation of this study: all data presented here are dependent on forestry crews submitting ticks. Future research should focus on confirming these findings by monitoring ticks in wetland habitats of these regions. Sensitivity and specificity tests indicated our niche models were efficient in their ability to detect known presences of all three species but had poor abilities in detecting a tick absence. Regardless, our relative activity and niche models can be used as a first approximation of risk for encountering ticks and their pathogens, as well as helping with tick management by forestry personnel.
Land management
Our results indicated that the burned area variable was not a good predictor of human-tick encounters; which suggests that burned locations are not a suitable habitat for tick encounters. Prescribed burning may remove tick and mammal habitats that promote the abundance of tick populations [105], but there are several studies in the Southeast that both support and reject these findings. Previously, prescribed burns reduced the abundance of ticks and the prevalence of tickborne pathogens in an area [106,107] which may be due to increased soil temperatures, decreased soil moisture, or destruction of tick habitats [108,109]. Conversely, a weak negative association between prescribed burns and the number of ticks in an area could be due to increased host use of burned areas [110] or little impact on tick species that quest and dwell within the soil [111]. Others propose that the methods for controlled burns, such as season, burning intensity, and frequency could be important factors for managing ticks in prescribed burns [112]. Future research to address the potential of controlled burns in large land management areas to reduce recreational and work-related tick encounters is warranted. Tick encounter data can be accessed at Dryad (https://doi.org/10.5061/dryad.v41ns1s3n). [113].
Conclusions
Passive tick surveillance data provided by FIA crews were valuable in delivering descriptions for when and where ticks were likely to be encountered and contributed to supporting our hypothesis. Data confirmed that ticks are active year-round in the southern US and habitats suitable for encounters can be explained by climatic and topographical features. Regions in the Southeast that were least suitable for all three tick species included the Blue Ridge, Mississippi Alluvial Plain, and the Southern Florida Coastal Plain, whereas suitable regions for all tick species included the Interior Plateau, Central Appalachians, Ozark Highlands, Boston Mountains, and the Ouachita Mountains. Temporal periods associated with encountering each tick species varied; for example, I. scapularis was associated with cold seasons (autumn and winter) whereas D. variabilis and A. americanum were associated with warm seasons (spring and summer). This research provides important information for isolating regions for management of ticks when they are more seasonally active. This study also helps forestry managers alert field crews about tick activity and new county records of tick species. Additionally, this study provides crews with the opportunity to test management decisions against tick encounters. Future studies will aim to combine forest management strategies at high-risk geographic regions to understand associations on human-tick encounters.
Supporting information
S1 Table. Correlation values for environmental variables that contributed 10% or more to each environmental niche model.
https://doi.org/10.1371/journal.pntd.0011919.s001
(DOCX)
S2 Table. Lambda file results for environmental variables in each Maxent environmental niche model.
https://doi.org/10.1371/journal.pntd.0011919.s002
(DOCX)
Acknowledgments
We thank the Forest Inventory and Analysis Program of the Forest Service, United States Department of Agriculture for organizing the collection of ticks in this study. This manuscript was prepared in part by United States government employees as part of their official duties and therefore is in the public domain. We also want to thank Jennifer Chandler, Corey Day, and Katy Smith for the initial manuscript review. The Department of Entomology and Plant Pathology at the University of Tennessee helped support RAB’s graduate stipend and tuition.
References
- 1. Jordan RA, Egizi A. The growing importance of lone star ticks in a Lyme disease endemic county: passive tick surveillance in Monmouth County, NJ, 2006–2016. PLoS One. 2019;14: 2006–2016. pmid:30753233
- 2. Eisen RJ, Paddock CD. Tick and tickborne pathogen surveillance as a public health tool in the United States. J Med Entomol. 2021;58: 1490–1502. pmid:32440679
- 3. Hamer SA, Tsao JI, Walker ED, Hickling GJ. Invasion of the Lyme disease vector Ixodes scapularis: implications for Borrelia burgdorferi endemicity. Ecohealth. 2010;7: 47–63. pmid:20229127
- 4. Dahlgren FS, Mandel EJ, Krebs JW, Massung RF, McQuiston JH. Increasing incidence of Ehrlichia chaffeensis and Anaplasma phagocytophilum in the United States, 2000–2007. Am J Trop Med Hyg. 2011;85: 124–131.
- 5. Dahlgren FS, Paddock CD, Springer YP, Eisen RJ, Behravesh CB. Expanding range of Amblyomma americanum and simultaneous changes in the epidemiology of spotted fever group rickettsiosis in the United States. Am J Trop Med Hyg. 2016;94: 35–42. pmid:26503270
- 6. Mitchell CL, Lin FC, Vaughn M, Apperson CS, Meshnick SR, Commins SP. Association between lone star tick bites and increased alpha-gal sensitization: evidence from a prospective cohort of outdoor workers. Parasites and Vectors. 2020;13: 1–4. pmid:32928302
- 7. Álvarez-López DI, Ochoa-Mora E, Heitman KN, Binder AM, Álvarez-Hernández G, Armstrong PA. Epidemiology and clinical features of Rocky Mountain spotted fever from enhanced surveillance, Sonora, Mexico: 2015–2018. Am J Trop Med Hyg. 2021;104: 190–197. pmid:33146112
- 8. Springer YP, Eisen L, Beati L, James AM, Eisen RJ. Spatial distribution of counties in the continental United States with records of occurrence of Amblyomma americanum (Ixodida: Ixodidae). J Med Entomol. 2014;51: 342–351. pmid:24724282
- 9. Lehane A, Parise C, Evans C, Beati L, Nicholson WL, Eisen RJ, et al. Reported county-level distribution of the American dog tick (Acari: Ixodidae) in the contiguous United States. J Med Entomol. 2020;57: 131–155. pmid:31368492
- 10. Collins SD, Abbott JC, McIntyre NE. Quantifying the degree of bias from using county-scale data in species distribution modeling: can increasing sample size or using county-averaged environmental data reduce distributional overprediction? Ecol Evol. 2017;7: 6012–6022. pmid:28808561
- 11. Peterson AT. Uses and requirements of ecological niche models and related distributional models. Biodivers informatics. 2006;3: 59–72. https://doi.org/10.17161/bi.v3i0.29.
- 12. Raghavan RK, Townsend Peterson A, Cobos ME, Ganta R, Foley D. Current and future distribution of the lone star tick, Amblyomma americanum (L.) (Acari: Ixodidae) in North America. PLoS One. 2019;14: 1–13. pmid:30601855
- 13. Ma D, Lun X, Li C, Zhou R, Zhao Z, Wang J, et al. Predicting the potential global distribution of Amblyomma americanum (Acari: Ixodidae) under near current and future climatic conditions, using the maximum entropy model. Biology (Basel). 2021;10. pmid:34681156
- 14. Rochlin I, Egizi A, Ginsberg HS. Modeling of historical and current distributions of lone star tick, Amblyomma americanum (Acari: Ixodidae), is consistent with ancestral range recovery. Exp Appl Acarol. 2022;8: 1–19. pmid:36482230
- 15. Minigan JN, Hager HA, Peregrine AS, Newman JA. Current and potential future distribution of the American dog tick (Dermacentor variabilis, Say) in North America. Ticks Tick Borne Dis. 2018;9: 354–362. pmid:29275873
- 16. Peterson AT, Raghavan RK. The geographic distribution of Ixodes scapularis (Acari: Ixodidae) revisited: the importance of assumptions about error balance. J Med Entomol. 2017;54: 1080–1084. pmid:28591858
- 17. Sagurova I, Ludwig A, Ogden NH, Pelcat Y, Dueymes G, Gachon P. Predicted northward expansion of the geographic range of the tick vector Amblyomma americanum in North America under future climate conditions. Environ Health Perspect. 2019;127: 1–14. pmid:31670575
- 18. Boorgula GDY, Townsend Peterson A, Foley DH, Ganta RR, Raghavan RK. Assessing the current and future potential geographic distribution of the American dog tick, Dermacentor variabilis (Say) (Acari: Ixodidae) in North America. PLoS One. 2020;15: 1–13. pmid:32776959
- 19. Ginsberg HS, Rulison EL, Miller JL, Pang G, Arsnoe IM, Hickling GJ, et al. Local abundance of Ixodes scapularis in forests: effects of environmental moisture, vegetation characteristics, and host abundance. Ticks Tick Borne Dis. 2020;11: 101271. pmid:31677969
- 20. Tran T, Prusinski MA, White JL, Falco RC, Vinci V, Gall WK, et al. Spatio-temporal variation in environmental features predicts the distribution and abundance of Ixodes scapularis. Int J Parasitol. 2021;51: 311–320. pmid:33359203
- 21. Estrada-Peña A., Climate, niche, ticks, and models: what they are and how we should interpret them. Parasitol Res. 2008;103: 87–95. pmid:19030890
- 22. Semtner PJ, Hair JA. The ecology and behavior of the lone star tick (Acarina: Ixodidae) V. Abundance and seasonal distribution in different habitat types. J Med Entomol. 1971;8: 329–335. pmid:4779928
- 23. Raghavan RK, Goodin DG, Hanzlicek GA, Zolnerowich G, Dryden MW, Anderson GA, et al. Maximum entropy-based ecological niche model and bio-climatic determinants of lone star tick (Amblyomma americanum) niche. Vector-Borne Zoonotic Dis. 2016;16: 205–211. pmid:26824880
- 24. Mather TN, Duffy DC, Campbell SR. An unexpected result from burning vegetation to reduce Lyme disease transmission risks. J Med Entomol. 1993;30: 642–645. pmid:8510127
- 25. Gleim ER, Conner LM, Berghaus RD, Levin ML, Zemtsova GE, Yabsley MJ. The phenology of ticks and the effects of long-term prescribed burning on tick population dynamics in southwestern Georgia and northwestern Florida. PLoS One. 2014;9: 13–16. pmid:25375797
- 26. Gallagher MR, Kreye J, Machtinger E, Everland A, Schmidt N, Skowronski NS. Can restoration of fire-dependent ecosystems reduce ticks and tick-borne disease prevalence in the eastern United States? Ecol Appl. 2022; 1–22. pmid:35426200
- 27. Diuk-Wasser MA, Hoen AG, Cislo P, Brinkerhoff R, Hamer SA, Rowland M, et al. Human risk of infection with Borrelia burgdorferi, the Lyme disease agent, in eastern United States. Am J Trop Med Hyg. 2012;86: 320–327.
- 28. Bacon EA, Kopsco H, Gronemeyer P, Mateus-Pinilla N, Smith RL. Effects of climate on the variation in abundance of three tick species in Illinois. J Med Entomol. 2022;59: 700–709. pmid:34875079
- 29. Ferrell AM, Brinkerhoff RJ. Using landscape analysis to test hypotheses about drivers of tick abundance and infection prevalence with Borrelia burgdorferi. Int J Environ Res Public Health. 2018;15: 18–20. pmid:29649156
- 30. Titcomb G, Allan BF, Ainsworth T, Henson L, Hedlund T, Pringle RM, et al. Interacting effects of wildlife loss and climate on ticks and tick-borne disease. Proc R Soc B Biol Sci. 2017;284. pmid:28878055
- 31. Spare M, Boorgula GD, Thomson D, Bai J, Goodin D, Anderson G, et al. Surveillance of host-seeking ticks in the Flint Hills region (USA) and associations with environmental determinants. Parasitologia. 2021;1: 137–147.
- 32. Trout Fryxell RT, Vogt JT. Collaborative-tick surveillance works: an academic and government partnership for tick surveillance in the southeastern United States (Acari: Ixodidae). J Med Entomol. 2019;56: 1411–1419. pmid:31049584
- 33. Yunker CE, Keirans JE, Clifford CM, Easton ER. Dermacentor ticks (Acari: Ixodoidae: Ixodidae) of the new world: a scanning electron microscope atlas. Proc Entomol Soc Washingt. 1986;88: 609–627.
- 34.
Durden LA. Nymphs of the genus Ixodes (Acari: Ixodidae) of the United States: taxonomy, identification key, distribution, hosts, and medical/veterinary importance (Thomas Say publications in entomology). Lanham, MD: Entomological Society of America; 1996.
- 35. Keirans JE, Durden L. Illustrated key to nymphs of the tick genus Amblyomma (Acari: Ixodidae) found in the United States. J Med Entomol. 1998;35: 489–495.
- 36. Egizi AM, Robbins RG, Beati L, Nava S, Evans CR, Occi JL, et al. A pictorial key to differentiate the recently detected exotic Haemaphysalis longicornis Neumann, 1901 (Acari, Ixodidae) from native congeners in North America. Zookeys. 2019;818: 117–128. pmid:30766418
- 37. Jenson SK, Domingue JO. Extracting topographic structure from digital elevation data for geographic information system analysis. Photogramm Eng Remote Sensing. 1988;54: 1593–1600. Available: https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=06a20725ae38b4dce81951bbb230b197dd346daa.
- 38. Running S, Mu Q, Zhao M. MYD17A2H MODIS/Aqua Gross Primary Productivity 8-Day L4 Global 500m SIN Grid V006. In: NASA EOSDIS Land Processes DAAC. 2015.
- 39. Myneni R., Knyazikhin Y. VIIRS/NPP Leaf Area Index/FPAR 8-Day L4 Global 500m SIN Grid V001. In: NASA EOSDIS Land Processes DAAC [Internet]. 2018 [cited 2 Apr 2022]. Available: https://doi.org/10.5067/VIIRS/VNP15A2H.001.
- 40. Friedl M., Sulla-Menashe D. MCD12Q1 MODIS/Terra+Aqua Land Cover Type Yearly L3 Global 500m SIN Grid V006. In: NASA EOSDIS Land Processes DAAC [Internet]. 2019 [cited 2 Apr 2022]. Available: https://doi.org/10.5067/MODIS/MCD12Q1.006.
- 41. Wan Z., Hook S., Hulley G. MOD11A2 MODIS/Terra Land Surface Temperature/Emissivity 8-Day L3 Global 1km SIN Grid V006. In: NASA EOSDIS Land Processes DAAC [Internet]. 2015 [cited 2 Apr 2022]. Available: https://doi.org/10.5067/MODIS/MOD11A2.006.
- 42. Running S, Mu Q, Zhao M. MOD16A2 MODIS/Terra Net Evapotranspiration 8-Day L4 Global 500m SIN Grid V006. In: NASA EOSDIS Land Processes DAAC [Internet]. 2017 [cited 2 Apr 2022]. Available: https://doi.org/10.5067/MODIS/MOD16A2.006.
- 43.
Thornton MM, Shrestha R,Wei Y, Thornton PE, Kao S, Wilson BE. Daymet: Daily Surface Weather Data on a 1-km Grid for North America, Version 4. In: ORNL DAAC, Oak Ridge, Tennessee, USA. [Internet]. 2020 [cited 4 Dec 2022]. Available: https://doi.org/10.3334/ORNLDAAC/1840.
- 44. Giglio L., Justice C., Boschetti L., Roy D. MCD64A1 MODIS/Terra+Aqua Burned Area Monthly L3 Global 500m SIN Grid V006. In: NASA EOSDIS Land Processes DAAC [Internet]. 2015 [cited 6 Jun 2023]. Available: https://doi.org/10.5067/MODIS/MCD64A1.006.
- 45.
Yu Y, Saatchi SS, Walters BF, Ganguly S, Li S, Hagen S., Melendy L, Nemani RR, Domke GM, Woodall CW. Carbon Pools across CONUS using the MaxEnt Model, 2005, 2010, 2015, 2016, and 2017. In: ORNL DAAC, Oak Ridge, Tennessee, USA [Internet]. 2021. Available: https://doi.org/10.3334/ORNLDAAC/1752.
- 46.
Ross CW, Prihodko L, Anchang J, Kumar S., Ji W., Hanan NP. Global hydrologic soil groups (HYSOGs250m) for curve number-based runoff modeling. In: ORNL DAAC, Oak Ridge, Tennessee, USA [Internet]. 2018. Available: https://doi.org/10.3334/ORNLDAAC/1566.
- 47. Phillips SJ, Dudík M. Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography. 2008;31: 161–175.
- 48. West AM, Kumar S, Brown CS, Stohlgren TJ, Bromberg J. Field validation of an invasive species Maxent model. Ecol Inform. 2016;36: 126–134.
- 49. Elith J, Phillips SJ, Hastie T, Dudík M, Chee YE, Yates CJ. A statistical explanation of MaxEnt for ecologists. Divers Distrib. 2011;17: 43–57.
- 50. Feng X, Park DS, Liang Y, Pandey R, Papeş M. Collinearity in ecological niche modeling: confusions and challenges. Ecol Evol. 2019;9: 10365–10376. pmid:31624555
- 51. Safari S, Baratloo A, Elfil M, Negida A. Evidence based emergency medicine; part 5 receiver operating curve and area under the curve. Emerg (Tehran, Iran). 2016;4: 111–3. Available: http://www.ncbi.nlm.nih.gov/pubmed/27274525%0Ahttp://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=PMC4893763.
- 52. Omernik J. Ecoregions of the conterminous United States. Map (scale 1:7,500,000). Ecoregions conterminous United States Map (scale 17,500,000). 1987;77: 118–125.
- 53.
Omernik JM. Ecoregions: a spatial framework for environmental management. In: Davis WS and TPS, editor. In: Biological Assessment and Criteria: Tools for Water Resource Planning and Decision Making. Boca Raton, FL.: Lewis Publishers; 1995. pp. 49–62.
- 54.
Cooperation C for E. Ecological regions of North America: toward a common perspective. Montreal, Quebec, Canada: Commission for Environmental Cooperation; 1997.
- 55. Omernik JM, Griffith GE. Ecoregions of the conterminous United States: evolution of a hierarchical spatial framework. Environ Manage. 2014;54: 1249–1266. pmid:25223620
- 56. Omernik JM. Perspectives on the nature and definition of ecological regions. Environmental management. 2004. pp. 27–38. pmid:16044553
- 57. McNab WH. Predicting forest type in Bent Creek Experimental Forest from topographic variables. Proceedings of the Sixth Biennial Southern Silvicultural Research Conference. Mmephis, Tennessee; 1990. pp. 496–504. Available: https://www.ptonline.com/articles/how-to-get-better-mfi-results.
- 58. Adams MB, Loughry L, Plaugher L. Experimental forests and ranges of the USDA forest service. USDA For Serv. Newtown Square, PA; 2004.
- 59. Swank WT, Crossley DA. Forest hydrology and ecology at Coweeta. vol. 66. Suparyanto dan Rosad (2015. Springer Science & Business Media; 2012.
- 60. Clinton BD. Light, temperature, and soil moisture responses to elevation, evergreen understory, and small canopy gaps in the southern Appalachians. For Ecol Manage. 2003;186: 243–255.
- 61. Mays SE, Houston AE, Trout Fryxell RT. Comparison of novel and conventional methods of trapping ixodid ticks in the southeastern U.S.A. Med Vet Entomol. 2016;30: 123–134. pmid:26801319
- 62.
Sonenshine DE, Roe MR. Biology of ticks. New York, NY: Oxford University Press.; 1993.
- 63. Cuervo PF, Flores FS, Venzal JM, Nava S. Niche divergence among closely related taxa provides insight on evolutionary patterns of ticks. J Biogeogr. 2021;48: 2865–2876.
- 64. Ogden NH, Lindsay LR, Beauchamp G, Charron D, Maarouf A, O’Callaghan CJ, et al. Investigation of relationships between temperature and developmental rates of tick Ixodes scapularis (Acari: Ixodidae) in the laboratory and field. J Med Entomol. 2004;41: 622–633. pmid:15311453
- 65. Ogden NH, Bigras-Poulin M, O’Callaghan CJ, Barker IK, Lindsay LR, Maarouf A, et al. A dynamic population model to investigate effects of climate on geographic range and seasonality of the tick Ixodes scapularis. Int J Parasitol. 2005;35: 375–389. pmid:15777914
- 66. Estrada-Peña A, Ayllón N, de la Fuente J. Impact of climate trends on tick-borne pathogen transmission. Front Physiol. 2012;3: 1–12. pmid:22470348
- 67. Schillberg E, Lunny D, Lindsay LR, Nelder MP, Russell C, Mackie M, et al. Distribution of Ixodes scapularis in northwestern Ontario: results from active and passive surveillance activities in the northwestern health unit catchment area. Int J Environ Res Public Health. 2018;15. pmid:30314334
- 68. Kopsco HL, Xu G, Luo CY, Rich SM, Mather TN. Crowdsourced photographs as an effective method for large-scale passivetick surveillance. J Med Entomol. 2020;57: 1955–1963. pmid:32812635
- 69. Merten H, Durden LA. A state-by-state survey of ticks recorded from humans in the United States. J Vector Ecol. 2000;25: 102–113. pmid:10925803
- 70. Petry WK, Foré SA, Fielden LJ, Kim HJ. A quantitative comparison of two sample methods for collecting Amblyomma americanum and Dermacentor variabilis (Acari: Ixodidae) in Missouri. Exp Appl Acarol. 2010;52: 427–438. pmid:20585839
- 71. Sayler KL, Acevedo W, Taylor JL. Status and trends of land change in the eastern United States—1973 to 2000 U.S. Geological Survey Professional Paper 1794–D. Reston, Virginia:; 2016.
- 72. Hertz JC, Ferree Clemons BC, Lord CC, Allan SA, Kaufman PE. Distribution and host associations of ixodid ticks collected from wildlife in Florida, USA. Exp Appl Acarol. 2017;73: 223–236. pmid:29110170
- 73. Faulkner S, Barrow W, Keeland B, Walls S, Telesco D. Effects of conservation practices on wetland ecosystem services in the Mississippi Alluvial Valley. Ecol Appl. 2011;21: 31–48.
- 74. Kambly S, Moreland TR. Land cover trends in the southern Florida Coastal Plain: U.S. Geological Survey Scientific Investigations Report 2009–5054. Reston, Virginia; 2009.
- 75. Napton DE, Auch RF, Headley R, Taylor JL. Land changes and their driving forces in the southeastern United States. Reg Environ Chang. 2010;10: 37–53.
- 76. Kerr SM, Rayner JO, Wood RR, Schultze S, McCreadie J. Ticks of Alabama: the fauna and spatial distribution of medically important species across the state. J Vector Ecol. 2022;47: 38–50. pmid:35366678
- 77. Cohen SB, Yabsley MJ, Freye JD, Dunlap BG, Rowland ME, Huang J, et al. Prevalence of Ehrlichia chaffeensis and Ehrlichia ewingii in ticks from Tennessee. Vector-Borne Zoonotic Dis. 2010;10: 435–440. pmid:19877819
- 78. Moncayo AC, Cohen SB, Fritzen CM, Huang E, Yabsley MJ, Freye JD, et al. Absence of Rickettsia rickettsii and occurrence of other spotted fever group rickettsiae in ticks from Tennessee. Am J Trop Med Hyg. 2010;83: 653–657.
- 79. Burg JG. Seasonal activity and spatial distribution of host-seeking adults of the tick Dermacentor variabilis. Med Vet Entomol. 2001;15: 413–421. pmid:11776460
- 80. Lingren M, Rowley WA, Thompson C, Gilchrist M. Geographic Distribution of ticks (Acari: Ixodidae) in Iowa with emphasis on Ixodes scapularis and their infection with Borrelia burgdorferi. Vector-Borne Zoonotic Dis. 2005;5: 219–226. pmid:16187889
- 81. De Jesus CE, Ganser C, Kessler WH, White ZS, Bhosale CR, Glass GE, et al. A survey of tick-borne bacterial pathogens in Florida. Insects. 2019;10: 1–13. pmid:31540253
- 82. Thompson AT, White SA, Doub EE, Sharma P, Frierson K, Dominguez K, et al. The wild life of ticks: using passive surveillance to determine the distribution and wildlife host range of ticks and the exotic Haemaphysalis longicornis, 2010–2021. Parasites and Vectors. 2022;15: 1–13. pmid:36127708
- 83. Bishopp FC, Trembley HL. Distribution and hosts of certain North American ticks. J Parasitol. 1945;31: 1–54.
- 84. Brown DG, Johnson KM, Loveland TR, Theobald DM. Rural land-use trends in the conterminous United States, 1950–2000. Ecol Appl. 2005;15: 1851–1863.
- 85. Butnor JR, Doolittle JA, Kress L, Cohen S, Johnsen KH. Use of ground-penetrating radar to study tree roots in the southeastern United States. Tree Physiol. 2001;21: 1269–1278. pmid:11696414
- 86. Wei X, Kimmins JP. Asymbiotic nitrogen fixation in harvested and wildfire-killed lodgepole pine forests in the central interior of British Columbia. For Ecol Manage. 1998;109: 343–353.
- 87. Zhang D, Dai Y, Wang L, Chen L. Influence of living and dead roots of gansu poplar on water infiltration and distribution in soil. Appl Sci. 2020;10.
- 88. Hair JA, Sauer JR, Durham KA. Water balance and humidity preference in three species of ticks. J Med Entomol. 1975;12: 37–47. pmid:1159729
- 89. Conlon JM, Rockett CL. Ecological investigations of the American dog tick, Dermacentor variabilis (say), in northwest Ohio (Acari: Ixodidae). Int J Acarol. 1982;8: 125–131.
- 90. Butler RA, Trout Fryxell RT, Houston AE, Bowers EK, Paulsen D, Coons LB, et al. Small-mammal characteristics affect tick communities in southwestern Tennessee (USA). Int J Parasitol Parasites Wildl. 2020;12: 150–154. pmid:32547921
- 91. Arsnoe I, Tsao JI, Hickling GJ. Nymphal Ixodes scapularis questing behavior explains geographic variation in Lyme borreliosis risk in the eastern United States. Ticks Tick Borne Dis. 2019;10: 553–563. pmid:30709659
- 92. Eisen RJ, Eisen L. The blacklegged tick, Ixodes scapularis: an increasing public health concern. Trends Parasitol. 2018;34: 295–309. pmid:29336985
- 93. Lockwood BH, Stasiak I, Pfaff MA, Cleveland CA, Yabsley MJ. Widespread distribution of ticks and selected tick-borne pathogens in Kentucky (USA). Ticks Tick Borne Dis. 2018;9: 738–741. pmid:29502988
- 94. Cascone S, Coma J, Gagliano A, Pérez G. The evapotranspiration process in green roofs: a review. Build Environ. 2019;147: 337–355.
- 95. Sundstrom KD, Lineberry MW, Grant AN, Duncan KT, Ientile MM, Little SE. Equine attachment site preferences and seasonality of common North American ticks: Amblyomma americanum, Dermacentor albipictus, and Ixodes scapularis. Parasites and Vectors. 2021;14: 1–7. pmid:34391460
- 96. Hahn MB, Jarnevich CS, Monaghan AJ, Eisen RJ. Modeling the geographic distribution of Ixodes scapularis and Ixodes pacificus (Acari: Ixodidae) in the contiguous United States. J Med Entomol. 2016;53: 1176–1191. pmid:27282813
- 97. Arsnoe IM, Hickling GJ, Ginsberg HS, McElreath R, Tsao JI. Different populations of blacklegged tick nymphs exhibit differences in questing behavior that have implications for human Lyme disease risk. PLoS One. 2015;10: 1–21. pmid:25996603
- 98. Seagle MP, Vierling MR, Almeida RJ, Clary DJ, Hidell W, Scott E V., et al. Low abundance of three tick species in the Piedmont of North Carolina. J Med Entomol. 2021;58: 489–492. pmid:32804202
- 99. Bunnell JE, Price SD, Das A, Shields TM, Glass GE. Geographic information systems and spatial analysis of adult Ixodes scapularis (Acari: Ixodidae) in the Middle Atlantic region of the U.S.A. J Med Entomol. 2003;40: 570–576. pmid:14680128
- 100. Ford K, Nadolny R, Stromdahl E, Hickling G. Tick surveillance and disease prevention on the Appalachian trail. Park Sci. 2015;32: 36–41.
- 101. Springer YP, Jarnevich CS, Barnett DT, Monaghan AJ, Eisen RJ. Modeling the present and future geographic distribution of the lone star tick, Amblyomma americanum (ixodida: Ixodidae), in the continental United States. Am J Trop Med Hyg. 2015;93: 875–890. pmid:26217042
- 102. Zalakeviciute R, López-Villada J, Rybarczyk Y. Contrasted effects of relative humidity and precipitation on urban PM2.5 pollution in high elevation urban areas. Sustain. 2018;10.
- 103. Portmann RW, Solomon S, Hegerl GC. Spatial and seasonal patterns in climate change, temperatures, and precipitation across the United States. Proc Natl Acad Sci U S A. 2009;106: 7324–7329. pmid:19380730
- 104. Stein KJ, Waterman M, Waldon JL. The effects of vegetation density and habitat disturbance on the spatial distribution of ixodid ticks (acari: Ixodidae). Geospat Health. 2008;2: 241–252. pmid:18686272
- 105. Pascoe EL, Plourde BT, Lopéz-Perez AM, Foley JE. Response of small mammal and tick communities to a catastrophic wildfire and implications for tick-borne pathogens. J Vector Ecol. 2020;45: 269–284. pmid:33207067
- 106. Stafford KC, Ward JS, Magnarelli LA. Impact of controlled burns on the abundance of Ixodes scapularis (Acari: Ixodidae). J Med Entomol. 1998;35: 510–513. pmid:9701937
- 107. Gleim ER, Zemtsova GE, Berghaus RD, Levin ML, Conner M, Yabsley MJ. Frequent prescribed fires can reduce risk of tick-borne diseases. Sci Rep. 2019;9: 1–10. pmid:31292479
- 108. Iverson LR, Hutchinson TF. Soil temperature and moisture fluctuations during and after prescribed fire in mixed-oak forests, USA. Nat Areas J. 2002;22: 296–304.
- 109. Gilliam ME, Rechkemmer WT, McCravy KW, Jenkins SE. The influence of prescribed fire, habitat, and weather on Amblyomma americanum (Ixodida: Ixodidae) in west-central Illinois, USA. Insects. 2018;9: 36. pmid:29565805
- 110. Allan BF. Influence of prescribed burns on the abundance of Amblyomma americanum (Acari: Ixodidae) in the Missouri Ozarks. J Med Entomol. 2009;46: 1030–1036. pmid:19769033
- 111. Padgett KA, Casher LE, Stephens SL, Lane RS. Effect of prescribed fire for tick control in California chaparral. J Med Entomol. 2009;46: 1138–1145. pmid:19769046
- 112. Scifres CJ, Oldham TW, Teel PD, Drawe DL. Gulf Coast tick (Amblyomma maculatum) populations and responses to burning of coastal prairie habitats. Southwest Nat. 1988;33: 55–64. https://doi.org/10.2307/3672088.
- 113. Butler R.A., M Papes J Vogt T, Paulsen D J, Crowe C, R T T Fryxell: Human risk to tick encounters in the southeastern United States estimated with spatial distribution modeling Dryad datapackage: https://doi.org/10.5061/dryad.v41ns1s3n