Comparison of phenology and pathogen prevalence, including infection with the Ehrlichiamuris-like (EML) agent, of Ixodes scapularis removed from soldiers in the midwestern and the northeastern United States over a 15 year period (1997-2012)

Background Since 1997, human-biting ticks submitted to the Department of Defense Human Tick Test Kit Program (HTTKP) of the US Army Public Health Command have been tested for pathogens by PCR. We noted differences in the phenology and infection prevalence among Ixodes scapularis ticks submitted from military installations in different geographic regions. The aim of this study was to characterize these observed differences, comparing the phenology and pathogen infection rates of I. scapularis submitted from soldiers at two sites in the upper Midwest (Camp Ripley, MN, and Ft. McCoy, WI) and one site in the northeastern US (Ft. Indiantown Gap, PA). Methods From 1997 through 2012, the HTTKP received 1,981 I. scapularis from the three installations and tested them for Anaplasma phagocytophilum, Babesia microti, Borrelia burgdorferi and the Ehrlichiamuris-like (EML) agent using PCR; pathogen presence was confirmed via sequencing or amplification of a second gene target. Pathogen and co-infection prevalence, tick engorgement status, and phenology were compared among installations. Results Greater rates of A. phagocytophilum and Ba. microti infections were detected in ticks submitted from installations in Minnesota than in Wisconsin or Pennsylvania, and the EML agent was only detected in ticks from Minnesota and Wisconsin. Midwestern ticks were also more likely to be co-infected than those from Pennsylvania. Both adult and nymphal ticks showed evidence of feeding on people, although nymphs were more often submitted engorged. Adult I. scapularis were received more frequently in June from Minnesota than from either of the other sites. Minnesota adult and nymphal peaks overlapped in June, and submissions of adults exceeded nymphs in that month. Conclusions There were clear differences in I. scapularis phenology, pathogen prevalence and rates of co-infection among the three military installations. Seasonal and temperature differences between the three sites and length of time a population had been established in each region may contribute to the observed differences. The synchrony of adults and nymphs observed in the upper Midwest has implications for pathogen infection prevalence. The EML agent was only detected in Minnesota and Wisconsin, supporting the previous assertion that this pathogen is currently limited to the upper Midwest.


Background
Ixodes scapularis is an important vector for several human pathogens, including Borrelia burgdorferi, Anaplasma phagocytophilum, Babesia microti, and the newlydescribed Ehrlichia muris-like (EML) agent [1], agents of Lyme disease, anaplasmosis, babesiosis, and ehrlichiosis, respectively. In order to facilitate the care of military personnel, dependents and Department of Defense (DOD) civilian employees that have been exposed to I. scapularis and other ticks, the US Army Public Health Command (USAPHC) developed the DOD Human Tick Test Kit Program (HTTKP) in 1989 which identifies ticks removed from these individuals and tests them for associated pathogens [2]. Results of tick species identification, tick engorgement status and pathogen analysis are reported to the health care provider for the individual reporting the tick bite in order to aid in the assessment of tick-borne disease risk. The results of tick identification and pathogen analyses are also used in combination with US Army entomological field studies (field-collected tick and mammal host infection frequencies) to assess the threat of tickborne diseases on specific military installations and guide decisions for tick-borne disease management based on national guidelines and published studies [3,4]. Starting in 1997, the HTTKP used polymerase chain reaction (PCR) to test ticks for human pathogens. From 1997-2012, 40,917 ticks (26,613 Amblyomma americanum,7,193 Dermacentor variabilis, 6,860 I. scapularis and 251 other spp.) were received and tested from approximately 100 military installations primarily in the eastern half and upper midwestern section of the US. During this time, several differences in the phenology and pathogen prevalence among I. scapularis from different geographic regions were noted based upon tick submissions. Specifically, the adult life stage of I. scapularis was received more frequently in summer months (June and July) from military medical treatment facilities in the upper midwestern US than from military installations in other regions that routinely submit ticks, and the EML agent was only detected in I. scapularis removed from soldiers at Camp Ripley, MN and Ft. McCoy, WI and not in I. scapularis from any other military installations within the geographic range of the tick, despite the testing of >2,000 I. scapularis ticks from all areas (E.S., unpublished data).
The aim of this study was to further characterize these observed differences, comparing the phenology and prevalence of A. phagocytophilum, Ba. microti, B. burgdorferi and the EML agent in I. scapularis submitted from soldiers at two sites in the Upper Midwest, Camp Ripley, MN, (46°10'32"N, 94°22'23"W), and Ft. McCoy, WI, (44°0'36"N, 90°40'60"W), and one site in the northeastern US, Ft. Indiantown Gap, PA (40°26'13"N, 76°34'34"W), from 1997 through 2012. These three sites were chosen because they share several important environmental features: all comprise large acreages of brush and woods that provide excellent habitat for ticks and their vertebrate hosts, notably Peromyscus leucopus, the white-footed mouse, an important reservoir of several human pathogens including A. phagocytophilum, Ba. microti, B. burgdorferi and the EML agent. The distribution of tick species there are very similar. The dominant anthropophilic tick species are I. scapularis and Dermacentor variabilis, and they are very abundant at the three installations. Interestingly, sympatric A. americanum populations have not been reported from these installations [5,6], and military personnel from these locations have not submitted this species to the HTTKP (E.S., unpublished data). At each site, the predominant tick-borne diseases are Lyme and anaplasmosis. In addition, soldier populations at the three installations are standardized in several ways that facilitate comparisons. All three installations are field training facilities used primarily by Reserve and National Guard units that deployed there for 2-wk sessions. During deployment periods, soldiers left the installations infrequently and dressed in identical uniforms while training. The ticks submitted to the HTTKP were from these trainees and rarely from dependents or DOD civilians. Only ticks that were attached to humans were submitted. All soldiers were instructed to use insect repellents and received similar training for personal tick surveillance. The HTTKP is a long-standing and wellutilized program at all three installations, and the regulations controlling and standardizing soldier activities at these field-training installations enabled us to have a more precisely regimented human population, which is unique and valuable for passive tick surveillance studies.

Methods
Morphologic assessment of ticks I. scapularis removed from military personnel and submitted to the HTTKP from Camp Ripley, Ft. McCoy and Ft. Indiantown Gap from 1997 through 2012 were identified using standard taxonomic keys. Only ticks that had actually bitten humans were accepted by the HTTKP. Only female adult and nymphal I. scapularis totals were included in this study since they comprise the vast majority (>98%) of I. scapularis submitted to the HTTKP. The engorgement status of each female and nymphal tick was estimated as unengorged, partly engorged, or fully engorged at the time of examination. The date of tick removal reported by the patient and health care provider on the submission form was used as a surrogate marker for when the tick was active. Ticks were only included in the phenology comparison if the date of tick removal was available.

DNA isolation
From 1997 through 2004, DNA isolation was performed using the IsoQuick nucleic acid extraction kit (ORCA Research, Bothell, WA) as described previously [2]. Beginning in 2005, the Zymo Genomic DNA II Kit (Orange, CA) was adopted to improve workflow while maintaining DNA quality and quantity [7].

PCR
Individual I. scapularis removed from soldiers were tested for A. phagocytophilum, Ba. microti, B. burgdorferi and EML using PCRs as described below. All initial positive results were confirmed by testing the DNA extract with the same PCR or second PCR for a different genetic region (Table 1), or DNA sequencing for selected pathogens. Positive specimens were defined as samples that produced at least two separate PCR positive results.

Anaplasma phagocytophilum/Ehrlichia spp. PCR
From 1997 through 2003, I. scapularis were tested for A. phagocytophilum using nested primers that amplify a portion of the 16S rRNA gene [8] as previously described [2]. Positive samples were confirmed using a second PCR reaction with the same assay.
From 2004 through 2012, I. scapularis were tested for A. phagocytophilum and the EML agent in a multiplex PCR for a conserved region of the Anaplasma/Ehrlichia groEL gene [9]. With this assay, base pair mismatches in regions of the detection probe hybridization produce species-specific melting temperature curves for organism differentiation of four tick-borne pathogens, Ehrlichia chaffeensis, E. ewingii, A. phagocytophilum and the EML agent [1,9]. PuRe Taq Ready-To-Go beads (GE Healthcare Life Sciences, Piscataway, NJ) were reconstituted in a 20 μl volume containing 2 μl of tick DNA as template, 0.5 μM each primer ESp-F and ESp-R, 0.2 μM probes Ec/e-FL, Aph-FL and ESp-RD (LightCycler-Red 640), and an additional 2.25 mM MgCl 2 to bring the final reaction volume of MgCl 2 to 4.05 mM. PCR was performed in a Roche LightCycler 1.2 or 2.0. (Roche Applied Science, Indianapolis, IN) using cycling parameters as described in Bell and Patel [9]. Samples positive in this initial PCR A. phagocytophilum were reconfirmed using the nested 16S rRNA primers from Massung et al. [8]. Positive control was nucleic acid of A. phagocytophilum, kindly provided by Dr. Stephen Dumler, (at that time) Johns Hopkins School of Medicine. Samples positive for the EML agent in the initial PCR were tested again in the groEL multiplex, and those positive in that test were sent to Mayo Clinic, Rochester, MN, for sequence confirmation.
The groEL multiplex that detects the EML agent was used by the HTTKP from 2004 to 2012. However, this pathogen was not described until 2009 [1], and at that time the HTTKP was able to back-check melting curve records only until 2007; therefore, only EML agent PCR results from 2007 to 2012 are reported.

Babesia microti PCR
From 2002 through 2008, I. scapularis were tested for Ba. microti in a conventional PCR using universal primers targeting the ss-rDNA gene [10]. PCR was performed in 25 μL reaction volumes prepared with Ready-To-Go PCR Beads and 1 μL of tick DNA as template using cycling parameters as described. Samples positive in the initial PCR were then tested in a conventional PCR with primers targeting the 18S rRNA gene of Ba. odocoilei and Ba. microti [11]. PCR was performed in 25 μL reaction volumes prepared with Ready-To-Go PCR Beads and 1 μL of tick DNA using cycling parameters as described. Positive control was nucleic acid of Ba. microti, kindly provided by Dr. Peter Krause, (at that time) University of Connecticut School of Medicine.

Borrelia burgdorferi PCR
From 1997 through 2003, I. scapularis were tested for B. burgdorferi in a nested PCR using primers that amplify a portion of the p66 gene [13]. Positive samples were tested with a second conventional PCR using primers targeting the fla gene of B. burgdorferi [14]. Use of both methods by the HTTKP has been previously described [2]. From 2004 through 2012, I. scapularis were tested for B. burgdorferi in a real-time PCR with primers and a probe for the ospA gene of B. burgdorferi [15] prepared with Ready-To-Go beads in a 20 μl volume containing 2 μl of tick DNA and using cycling parameters as described. Positive samples were tested by a second realtime PCR targeting the inner part of the fla gene [16] using Ready-To-Go beads in a 20 μl volume containing 2 μl of tick DNA as template and using cycling parameters as described. The positive control for both reactions was B. burgdorferi strain B31 (gift of Dr. Robert Wirtz, CDC Atlanta).
B. burgdorferi PCR of I. scapularis received in 2012 from Ft. McCoy and Camp Ripley was reconfirmed at Texas A&M using nested PCR for the 16S-23S rRNA intergenic spacer region [17].

Sequencing
Sequence determination of a subset of samples from Camp Ripley found to be co-infected with one or more pathogens was performed at University of North Texas Health Science Center. Residual amplification primers were removed from PCR products using ExoSAP-IT® (USB Corporation, Cleveland Ohio) before sequence determination. DNA sequence was determined for both strands of all amplicons using a BigDye® Terminator v.3.1 Cycle Sequencing kit (Applied Biosystems, Foster City, CA). Unincorporated dye terminators were removed prior to electrophoresis using Performa® DTR Gel Filtration Cartridges (Edge BioSystems, Gaithersburg, MD). Capillary electrophoresis was performed using an ABI PRISM® 310 Genetic Analyzer (Applied Biosystems, Foster City, CA). Final sequence analysis and editing was performed using Sequencher 4.1.4 (Gene Codes Corporation, Ann Arbor MI). Pathogen identification was determined by comparison of edited DNA sequence using the BLASTN version 2.2.10 (National Center for Biotechnology Information) with sequence data in GenBank.
Sequencing was also performed at Mayo Clinic, Rochester, MN, for confirmation of a subset of EML PCR positive samples. The amplification product was purified as described above. DNA sequences for both strands of all amplicons were determined using a BigDye® Terminator v.1.1 Cycle Sequencing kit (Applied Biosystems, Foster City, CA). Unincorporated dye terminators were removed prior to electrophoresis using Performa® Ultra 96-well Plate (Edge BioSystems, Gaithersburg, MD). Capillary electrophoresis was performed using an ABI PRISM® 3730 xl DNA Analyzer (Applied Biosystems, Foster City, CA). Final sequence analysis and editing was performed using Sequencher 5.0 (Gene Codes Corporation, Ann Arbor MI). Pathogen identification was also determined as described in above paragraph.

Pathogen prevalence and pathogen coinfection comparison
Prevalence of infection with A. phagocytophilum, Ba. microti, B. burgdorferi and the EML agent in I. scapularis adults and nymphs was compared among the three installations using pairwise Fisher's exact tests combining across all years. Overall coinfection rates of adult and nymphal I. scapularis were also compared among all three sites, combining all years, using pairwise Fisher's exact tests. To assess whether the frequency of coinfections of A. phagocytophilum with B. burgdorferi, and Ba. microti with B. burgdorferi in individual ticks differed from what would be expected due to chance under an independence assumption, an expected coinfection probability was first calculated by multiplying the individual infection probabilities. These expected probabilities were then compared to those that were observed. This was further investigated with kappa statistics to measure the degree to which the infection status agreed between the two types above and beyond chance alone.

Investigation of the time of establishment of I. scapularis at Ft. Indiantown Gap
A factor which may contribute to pathogen diversity and coinfection in I. scapularis is the length of time that a population has been established at a given location [18].

Phenology comparison
The relative abundance of I. scapularis adults vs. nymphs was compared among the three installations in the month of June, from 1997-2012, using pairwise chi-square tests. In all analyses, for each pairwise comparison between the three installations, p-values less than 0.017 were considered statistically significant using a Bonferroni adjustment. For non-pairwise comparisons, p-values of 0.05 were considered statistically significant. All analyses were performed using SAS version 9 (SAS Institute Inc., Cary, NC).
The locations of Camp Ripley, Ft. McCoy and Ft. Indiantown Gap were mapped using the calculations of the magnitude of the difference between summer and winter temperature, or the "amplitude of the annual cycle of maximum temperature" as presented in Gatewood et al. [19]. This study demonstrated that seasonal synchrony of the immature stages of I. scapularis is significantly related to average maximum and minimum temperatures, among other seasonal climate variables.

Engorgement comparison
The numbers of fed (partly and fully engorged) and unengorged female I. scapularis were compared to the numbers of fed and unengorged nymphs collected from 1997 through 2012 using a chi-square test to investigate the presumption that tick-bite victims are more likely to detect and remove females than nymphs, due to their larger size.
This manuscript reports no research performed on human or animals. PCR testing reported throughout this article was conducted solely on ticks from the environment. The ticks had bitten humans, but they were in no way human samples. Personal Protected Information of the person bitten by the tick is not disclosed to our lab when a tick is sent for identification and testing.  Table 2. In the subset of ticks that was tested for B. burgdorferi at two laboratories, a 97% (69/71) concurrence was observed between the two laboratories. The Anaplasma/Ehrlichia groEL gene multiplex is designed to also detect E. chaffeensis and E. ewingii; however these agents were not detected in any I. scapularis.

Sequencing
Sequencing confirmed the presence of the suspected pathogenic agents in a subset of samples PCR-positive for polymicrobial infections and a second subset of samples positive for the EML agent. In addition, three A. phagocytophilum sequences, from Camp Ripley were compared with the Ap-ha 16S RNA gene sequence reported by Chen et al. [20] (GenBank accession number U02521) via BLAST sequence alignment using the optimized algorithm for highly similar sequences, Megablast. Each of these three sequences shared 100% identity with the Ap-ha sequence, the strain associated with human illness, rather than the Ap-v1 variant that is nonpathogenic to humans but common in Odocoileus virginianus, the white-tailed deer [21].

Pathogen prevalence and pathogen coinfection comparison
As shown in Table 2 Table 2). Slightly more coinfections of A. phagocytophilum and B. burgdorferi were observed in adults and nymphs from Ft. McCoy than would be expected under an independence assumption. There was not a statistically significant frequency of coinfection of ticks with B. burgdorferi and Ba. microti compared to what would be expected by chance.

Time of establishment of I. scapularis at Ft. Indiantown Gap
Data summarized in Table 3 suggests that I. scapularis was not abundant at Ft. Indiantown Gap until~2000. Seven studies of hunter-killed deer and small mammals and their associated ticks performed by US Army entomologists, 1988-2006, at Ft. Indiantown Gap reported an increase of I. scapularis and B. burgdorferi over these 18 years [22][23][24][25], (Melissa Miller, personal communication). A corresponding increase was observed during the same time period in ticks submitted to the HTTKP, which received only D. variabilis from Ft. Indiantown Gap from 1993 through 1996, followed by very small numbers of I. scapularis in 1998 and 1999, increasing to robust populations after 2003. Human cases of confirmed and probable Lyme disease reported from Lebanon Co.,   (Figure 3). This suggests that variation in the seasonal synchronicity of I. scapularis life stagesin this case, adults and nymphs -may be driven by the greater variation in seasonal climates at the different sites.

Engorgement comparison
Overall, 91% of females and 88% of nymphal I. scapularis submitted to the HTTKP 1997-2012 were removed unengorged, likely reflecting daily tick checks and prompt removal of ticks by soldiers. Unattached ticks were not submitted, and <2% (127/6,678) of I. scapularis received by the HTTKP from 1997-2012 were males, presumably attached mating with the female ticks when removed from the human host. A small but significantly greater proportion of fed nymphs was observed compared to fed adults (142/1149 = 12.4% vs 79/871 = 9.1%, p = 0.019), supporting the presumption that tick-bite victims are more likely to detect and remove the larger female than the tiny nymph before it has time to feed and become engorged.

Discussion
This study demonstrates clear differences in the tick phenology, pathogen prevalence and rates of co-infection among three study sites across the midwestern and northeastern US. We present evidence to support our initial observation that adult I. scapularis were active in the summer months in the Upper Midwest, especially at Camp Ripley, and that the EML agent was detected only in from Camp Ripley and Ft. McCoy. In this detailed comparison of our tick data from Camp Ripley, Ft. McCoy in the Upper Midwest, and Ft. Indiantown Gap in the Northeast, we also describe several other differences in prevalence of infection/coinfection of A. phagocytophilum and Ba. microti. It is likely that numerous factors and interactions are contributing to the abundance and distribution of I. scapularis and its associated pathogens at these sites. Several of these are discussed below with reference to several contemporary studies of environmentally-collected I. scapularis that investigate regional transmission cycles and examine hypotheses to explain the mechanisms driving these observed differences.
An overlap of the larval and nymphal stages of I. scapularis has been reported at midwestern sites, specifically Camp Ripley [5,6,19], and our data revealed an overlap of the nymphal and adult stages there, suggesting that all three life stages are active simultaneously. Gatewood et al. [19] observed that there is a greater seasonal synchrony at sites with colder temperatures, specifically, the amplitude of the annual cycle of maximum temperature. We have used a temperature gradient map Table 3 Comparison of confirmed and probable Lyme disease cases from Lebanon Co., PA, and tick surveillance data from Ft. Indiantown Gap, Lebanon Co., PA, 1988-2006Year 19881990199119931999  from this publication to demonstrate the gradient our three study sites (Figure 3). This overlap of life stages may impact pathogen distribution. Gatewood et al. [19] hypothesized that the overlap of I. scapularis immature life stages in the Upper Midwest may enhance maintenance of certain short-lived strains of B. burgdorferi, and this phenology may also enhance maintenance of other organisms vectored by I. scapularis, because synchrony of life stages has been shown to favor both systemic (transmission between host and tick feeding on infectious host) and non-systemic (transmission between infected and non-infected ticks feeding in close proximity on an uninfected host) transmission [5,27]. Evidence of tick pathogen infection from our data sets indicates further differences in pathogen prevalences that may be associated with synchrony of life stages. We found greater prevalence of A. phagocytophilum and Ba. microti at I.  (Table 2), nor in 1,970 I. scapularis adults and nymphs received by the HTTKP during the same time period and tested using the PCR methods described above from > 40 locations in Alabama, Connecticut, Delaware, Florida, Indiana, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, North Carolina, Pennsylvania, Rhode Island, South Carolina and Virginia (E.S., unpublished data). The presence of the EML agent in only I. scapularis from cold climates with seasonal synchrony of life stages suggests that these conditions may be permissive for this bacterium, and suggests that the EML agent might also be characterized by shorter periods of infectivity in vertebrate hosts.
Summer synchronous adult activity of I. scapularis may not play an important role in systemic pathogen maintenance in the reservoir hosts (mice, small mammals) because adult ticks are not found on these animals; adult I. scapularis were not collected from mice and small mammals in an extensive study of ticks at Camp Ripley [5]. However, I. scapularis adults also feed readily on medium-sized mammals, as do I. scapularis nymphs and larvae [28], so non-systemic transmission of pathogens by adults and immatures co-feeding on the  larger mammal hosts may contribute to pathogen prevalence in areas where the three life stages overlap.
Recent research has revealed that length of time that a population of I. scapularis is present in a given location may also contribute to the pattern of pathogen diversity. Detailed studies dating from 2004 of the invasion and establishment of I. scapularis in lower Michigan suggested that newly-invading populations of I. scapularis had less pathogen diversity (specifically A. phagocytophilum and Ba. microti) but equivalent B. burgdorferi prevalence when compared with long-established I. scapularis populations in northern Michigan, Wisconsin and Minnesota [18]. Greater pathogen diversity presages coinfections in I. scapularis, and the pattern of pathogen infection/coinfection described from the newly-invaded as compared to the long-established Michigan sites resembles the differences in pathogen/coinfection prevalences that we observed at Ft. Indiantown Gap as compared to Camp Ripley and Ft. McCoy.
I. scapularis populations have been documented from Minnesota and Wisconsin for 40 years [18], and the increasing abundance and prevalence of infection of I. scapularis at Camp Ripley and Ft. McCoy have been the subject of surveillance since Lyme disease emerged in the late 1970s [5,[29][30][31][32][33]. Reports of tick surveillance at Ft. Indiantown Gap are fewer but suggest that I. scapularis was not collected there until much more recently. Several parallel lines of inquiry suggest that I. scapularis established at Ft. Indiantown Gap in the late 1990s (Table 3). From these data sets, it appears that I. scapularis was absent or rare in the area of Ft. Indiantown Gap in the early 1990s, with numbers of ticks and cases of Lyme disease increasing in the late 1990s, then increasing sharply until reaching a plateau about 2003.
Infection rates of B. burgdorferi were robust (3/11 = 24% in adults, 4/15 = 27% in nymphs) in ticks submitted from Ft. Indiantown Gap to the HTTKP as soon as they were first tested in 2000, but infection with A. phagocytophilum was rarely found in these ticks, and Ba. microti has not been detected (Table 2), thus supporting the hypothesis that newly-invasive populations have less pathogen diversity [18]. However, the prevalence of infection of B. burgdorferi was robust at all three locations, with a statistical difference in B. burgdorferi prevalence found only in adults from Ft. McCoy (Table 2), and other studies have reported that newly-invaded populations had nearly equivalent B. burgdorferi prevalences with those of long-established populations [18]. Furthermore, the duration of establishment of I. scapularis populations may not influence the establishment of the EML agent in these ticks, given that it has not been detected in 356 I. scapularis adults and nymphs submitted to the HTTKP from northeastern sites of long-established populations, Groton, CT, and Newport, RI (E.S., unpublished data).
As all ticks submitted to the HTTKP are removed from humans, this research reveals that adult I. scapularis at Camp Ripley are biting humans in summer alongside nymphs, and may also be important in the spread of human disease in Minnesota and other areas of synchronous tick activity. The relative abundance of adults to nymphs found biting humans in June at Camp Ripley might be explained by the size difference of adult and nymph, i.e., the larger adult is more likely to be detected; however, comparison with the June data from Ft. McCoy and Ft. Indiantown Gap, where more nymphs than adults were removed from soldiers in June, suggests that the actual distribution of adults and nymphs at that time period is accurately represented. Adult tick activity in June may have human health implications, because of the high pathogen prevalence in these ticks. Adult I. scapularis have higher pathogen prevalence, and a greater chance of coinfection, since the tick has completed two meals (the larval and the nymphal meal) on potentially infected hosts or alongside infected conspecifics. However, as reported above, most (91%) adult female I. scapularis submitted to the HTTKP were unengorged, so hopefully B. burgdorferi infection was forestalled in many cases by prompt tick removal, because the time of attachment of I. scapularis nymphs until B. burgdorferi transmission is thought to be >36 hours, perhaps as long as 72 hrs. [4], at which time both nymphal and adult ticks show obvious signs of engorgement [3]. The time of attachment of I. scapularis nymphs until transmission of A. phagocytophilum and Ba. microti in animal systems appears to be shorter, within the first 24 h of tick, when the tick does not yet appear partly engorged [3,34,35]. Transmission of Powassen virus, a life-threatening human pathogen vectored by I. scapularis but not investigated in the study, is thought to occur far more rapidly, within 15 min of attachment of I. scapularis nymphs [36]. Duration of attachment studies are lacking for novel pathogens EML and Borrelia miyamotoi, both associated with the bite of I. scapularis. Studies of the transmission of human pathogens by the adult stage of I. scapularis are also lacking and there is a need for better understanding of the timing from attachment until transmission occurs for the numerous pathogens vectored by this tick.
There are several potential limitations of the data collected in this study. One limitation is that the data are passively collected through detection of ticks on military personnel. The tick's habit of attachment to the host makes it an ideal subject for passive surveillance, and the investigation of ticks removed from human hosts provides an estimate of actual human encounters with infected ticks. This information is especially important in surveillance of I. scapularis ticks, because in some areas of the US, I. scapularis is present but does not typically attack humans. However, capturing accurate temporal and spatial data of ticks removed from humans can be challenging. Ticks usually attach undetected by the host, and people often cannot accurately describe where or when a tick was acquired. Tick-bite victims are more able to recall when a tick was found and removed; therefore, we have used the date of tick removal in our phenology comparisons. Likewise, tick-bite victims may not know where they encountered their tick. Our study overcomes this limitation, in part, through our selection of three installations in which the soldiers are present only for temporary training during which they rarely leave the installations.
Another limitation is that our data were dependent on the sampling effort of the military personnel. The number of soldiers training at these installations far exceeds the numbers that have submitted ticks to the HTTKP. At Camp Ripley, for example, during April -September 2007-2011, the number of soldiers training per month was typically >10,000 (Jay Brezinka, personal communication). It is not known how many ticks fed on soldiers undetected or how many ticks were removed by soldiers and not submitted to the HTTKP. All soldiers are trained to use personal protective measures against tickbite [37], to perform tick checks, and to submit all attached ticks to their military medical treatment facilities for identification and testing by the HTTKP, however, compliance with these instructions is thought to be inadequate [38].

Conclusions
In summary, we demonstrate that tick phenology and pathogen prevalence exhibit statistically significant differences among three Army installations participating in the HTTKP. These observed differences are congruent with other recent studies comparing maximum seasonal temperature differences and the length of tick population establishment at these three sites. Adult I. scapularis ticks are biting humans in summer alongside nymphs, and may also be important in the spread of human disease in Minnesota and other areas of synchronous tick activity. Our data also support the preliminary report of Pritt et al. [1] that the EML agent appears to be limited to Minnesota and Wisconsin. Longitudinal data sets such as presented here will be useful in future studies investigating regional trends in tick species distribution and pathogen presence.