Department of Biology, University of South Dakota, Vermillion, SD 57069, U.S.A.
Habitat correlates with the spatial distribution of ectoparasites on Peromyscus leucopus in southern Michigan
Article first published online: 30 NOV 2011
© 2011 The Society for Vector Ecology
Journal of Vector Ecology
Volume 36, Issue 2, pages 308–320, December 2011
How to Cite
Mize, E. L., Tsao, J. I. and Maurer, B. A. (2011), Habitat correlates with the spatial distribution of ectoparasites on Peromyscus leucopus in southern Michigan. Journal of Vector Ecology, 36: 308–320. doi: 10.1111/j.1948-7134.2011.00171.x
- Issue published online: 30 NOV 2011
- Article first published online: 30 NOV 2011
- Received 9 February 2011; Accepted 11 July 2011
- spatial distribution;
- Peromyscus leucopus;
- disease ecology
- Top of page
- MATERIALS AND METHODS
- REFERENCES CITED
The goal of this study was to evaluate the role of habitat in determining ectoparasite distribution of Peromyscus leucopus. We tested the hypothesis that ectoparasite occurrence is associated with particular host environments and this association is stronger for ectoparasites with limited interactions (i.e., ticks) than those with frequent interactions (i.e., lice). Ectoparasites from three different groups (Acari, Siphonaptera, and Phthiraptera) were collected from P. leucopus inhabiting a number of forested habitats in southern Michigan. Measurements of plant species structure and composition were collected and models were developed using quadratic discriminant function analysis to determine if habitats associated with ectoparasite presence were different from those associated with their absence. Mice parasitized by ticks were more likely to be found in areas having undergone a recent disturbance. Mice parasitized by ticks, fleas, and lice were more likely to be found in areas having tree species associated with dry soils. Our results show there is a distinct difference in habitats associated with the presence of ectoparasites, though we did not observe a stronger association of host habitat for ticks than for fleas or lice. This implies habitat should be included as an important component of assessments of the spatial distribution of ectoparasites.
- Top of page
- MATERIALS AND METHODS
- REFERENCES CITED
Rodents are reservoir hosts for many pathogens that cause human diseases, i.e., zoonotic pathogens (Weber 1982). The biological vectors associated with mice include ticks (Lyme disease, Rocky Mountain spotted fever, and babesiosis), fleas (plague, sylvatic and murine typhus), and lice (sylvatic typhus). White-footed mice, Peromyscus leucopus, are competent reservoir hosts for all these pathogens. Common species, such as Peromyscus mice, generally have higher abundance and species diversity of parasites than rare species (Arneberg et al. 1998). Lice, fleas, mites, ticks, and botfly larvae are common ectoparasites of Peromyscus leucopus (Whitaker 1968). Furthermore, surveying P. leucopus directly can be more sensitive to picking up tick presence than other survey methods aimed at collecting parasites directly from the environment, such as dragging (Hamer et al. 2010). Thus, P. leucopus is an appropriate study organism for examining the distribution of these ectoparasites.
Ectoparasite distributions among host populations are influenced by the characteristics of the host organism, such as sex (Wilson et al. 2002), age (Anderson and May 1991, Hudson and Dobson 1995), body condition (Wilson et al. 2002) and host density (Tompkins et al. 2002). Hosts are often considered “biological islands” for parasites, providing the habitat necessary to fulfill basic biological needs such as food, shelter, and opportunities for mating (Krasnov et al. 1997, Krasnov et al. 2006). However, ectoparasites are also affected by the external environment (Krasnov et al. 1997, Guerra et al. 2002, Krasnov et al. 2006). For instance, in Wisconsin blacklegged ticks (Ixodes scapularis) were associated with abiotic factors such as soil texture, soil order, forest type, land cover, and bedrock within its hosts’ range, possibly restricting the distribution of blacklegged ticks (Guerra et al. 2002).
If presence of parasites is mainly a function of host characteristics, then the expected distribution of blacklegged ticks in Michigan would coincide with the statewide Peromyscus distribution. However, these ticks are limited in distribution to areas in Menominee County in the Upper Peninsula and along the west coast of the Lower Peninsula in Berrien, Van Buren, and Allegan counties (Walker et al. 1998, Hamer et al. 2007). One possible reason some mice have few or no parasites may be because the host's environment is inhospitable to certain life stages of potential parasites. Another reason is the parasite may not have the opportunity to feed from mice because they are not yet present or have not yet invaded into the host's environment. Conversely, high ectoparasite loads may be experienced in environments that are conducive for ectoparasite survival during phases of their life cycle when they are not on the host. Thus, ectoparasite presence and parasite species assemblages are not just a function of host-parasite relationships but also parasite-habitat and host-habitat relationships. To summarize, a parasite's distribution among its hosts is dependent on finding the right host in the right habitat (Krasnov et al. 1997, Krasnov et al. 2006).
Parasites that spend a portion of their life or whole life stages off their host should have stronger habitat associations than parasites whose life cycles are restricted solely to the host. Ticks, fleas, and lice represent three different modes of interaction with their host: very little host-parasite interaction in the form of a few feeding opportunities (tick), moderate amount of host-parasite interaction through repeated short feeding opportunities (flea), and permanent interaction in the form of lifelong constant association with the host (louse). By including species from different taxonomic groups, this study compares the association of vegetation attributes to different degrees of host interaction.
As Peromyscus abundance does not necessarily correspond to parasite presence and abundance, mapping Peromyscus habitat and distribution is insufficient when determining distributions of their parasites. As explained above, parasite distribution may be correlated with environmental factors such as land cover, vegetation presence and distribution, soil, and weather conditions. Hence, parasite communities of Peromyscus may also vary between different habitat types. The purpose of this study was to investigate parasite occurrence among different vegetation communities across the southern half of the Lower Peninsula of Michigan. We tested the hypothesis that ectoparasite occurrence is associated with particular host environments and that the association for ectoparasites with limited interactions with hosts (i.e., ticks) would be stronger than for those with constant interactions (i.e., lice).
MATERIALS AND METHODS
- Top of page
- MATERIALS AND METHODS
- REFERENCES CITED
Six state game areas (SGAs) in southern Michigan were studied (Figure 1). The SGAs surveyed include Sharonville State Game Area (Jackson and Washtenaw Counties), Flat River State Game Area (Ionia and Montcalm Counties), Three Rivers State Game Area (Cass and St. Joseph Counties), Deford State Game Area (Tuscola County), Verona State Game Area (Huron County), and Barry and Yankee Springs State Game Area (Barry County). These areas were chosen because they span different habitats including forested, lowland, and agricultural land cover types and all locations are in the same ecoregion. GIS data and imagery were available for each of these areas. Furthermore, Integrated Forest Monitoring, Assessment, and Prescription (IFMAP) stand-level surveys were available from the Michigan Department of Natural Resources.
Twelve 50 m circular plots were located in each SGA except Three Rivers and Sharonville, which had 7 and 11 plots respectively, for a total of 66 plots. Plots were randomly located based on proportions of land cover types at each SGA based on satellite imagery. Vegetation data were collected at each plot following the guidelines established by the Michigan Department of Natural Resources. The following vegetation attributes were measured: tree species presence, percent canopy cover, average basal area, height of subcanopy species, ground cover density, and the IFMAP cover class. GPS coordinates were taken at the center of each plot.
Small mammals were collected from one 36-hour trapping event on each plot during the time period from June 22 to August 5, 2007. Plots were sampled using 24 Sherman live traps (H.B. Sherman Traps, Tallahassee, FL) baited with rolled oats and placed 10 m apart in three parallel 80 m transects. Traps were checked in the early morning and evening at 10–12 h intervals for 36 consecutive h. All animals collected were identified to genus and species when possible, sexed, weighed, and marked to recognize recaptures by removing a small tuft of fur from the rear thigh. Non-Peromyscus species were then released. Age class (juvenile or adult (Baker 1983)) and right ear and tail length were measured on each specimen. Length measurements were used to distinguish between P. leucopus and P. maniculatus bairdii (Baker 1983).
Peromyscus caught on the first trap night (hours 12–24) were examined for parasites and released. To incapacitate fleas and allow collection, mice received a dose of 0.2 cc isoflurane (Isoflo, Abbott Laboratories, Chicago, IL) to induce anesthesia, which was maintained with a dose of 0.1 cc isoflurane while monitoring breathing continuously. Once anesthetized, the animal was examined for fleas and ticks, which were collected using #5 watchmakers’ forceps. Engorged ticks were carefully removed from the epidermis taking special care to remove the mouth parts for identification. Fleas and unattached ticks were removed using forceps or by brushing the mouse's body with a hard bristle toothbrush over a white pan. The collected ectoparasites were placed in vials filled with 100% ethanol and labeled with the animal identification number and SGA. Fully recovered animals were released near site of capture, then traps were immediately reset.
A partial lethal take was conducted to assess louse burden. Mice caught on the second trap night (hour 36), including recaptured mice, were administered 0.3 cc of isoflurane to induce a deep sleep and were euthanized by cervical dislocation. After examination for ticks and fleas as above, each mouse was individually wrapped in multiple layers of cheese cloth to prevent cross-contamination of parasites, as multiple animals were stored in the same collection jar in 100% ethanol.
Louse specimens were collected post-mortem in the laboratory by examining each mouse under a dissecting scope. The mouse and cheese cloth were then washed with dish detergent and rinsed with water over a 1 gal jar; the washings were strained in a 200 mm opening 75 μm mesh sieve (U.S.A. Standard Sieve Series, Newark Wire Cloth Co., Newark, NJ) for lice missed during initial inspection. Lice were collected using forceps, and stored using the same method as described above for the fleas and ticks.
Parasite species identification
Each parasite was prepared for identification according to taxon-specific standards. Wet mount tick specimens were identified to species and appropriate life stage by examination under a dissecting microscope using Sonenshine's (1979) dichotomous key. Fleas and lice were cleared based on guidelines from Fox (1940), Kim et al. (1986) and Ferris and Stojanovich (1951) in 10% KOH overnight to view informative internal anatomical features. After clearing, each organism was rinsed in deionized water and allowed to soak for 30 min to end the clearing process at room temperature before they were dehydrated for mounting. Dehydration was achieved by incubating the specimens in the following alcohol concentrations: 30 min each in 70%, 90%, and 100% ethanol and a final soaking for 30 min in 100% ethanol. All specimens were then mounted on slides in Canada balsam and allowed to dry on the bench top overnight before examination. Each specimen was examined to determine the species, life stage, and sex when possible. Fleas were identified using Fox's (1940) key and lice were identified using the keys of Kim et al. (1986) and Ferris and Stojanovich (1951).
Ixodes scapularis and Dermacentor variabilis were analyzed together because the sample sizes for both species were very low and there were three plots (4% 3/66) where both species were collected from parasitized mice. Furthermore, previous studies suggested both species may have similar vegetation associations (Sonenshine et al. 1972, Campbell and MacKay 1979). Flea species were also combined for analysis as there were only two observations of C. pseudagyrtes.
The total number of parasites collected across all mice, prevalence (number of infested hosts/ total number of hosts) (Margolis et al. 1982), average intensity of infestation per infested mouse (number of parasites/ number of infested hosts) (Margolis et al. 1982), number of plots in which the parasite occurred, degree of aggregation, k, and variance to mean ratio of the parasite distribution were calculated for each parasite taxon. Aggregation was estimated using a corrected moment estimator, k, which represents the degree of aggregation of a population of organisms where k<1 is considered an aggregated distribution (Wilson et al. 2002). The variance to mean ratio was used as a measure of aggregation, with values greater than 1 representing a more aggregated distribution (Wilson et al. 2002).
Two hundred ten different vegetation variables were measured at each of the 66 plots. We first needed to reduce the number of variables to maintain degrees of freedom and to meet the condition for discriminant function analysis that the number of variables must be smaller than the number of observations. To reduce the number of variables, we examined correlations within the canopy variables and subcanopy variables (i.e., canopy basal area vs canopy closure) to remove highly correlated variables and select a subset of the vegetation variables. The resulting 29 variables retained were the average basal area of 11 canopy tree species, average height of ten subcanopy species, and percent ground cover for eight vegetation types. Vegetation data were transformed to meet the assumption of multivariate normality by square root transforming the canopy and subcanopy variables and arcsine transforming the ground cover variables.
Discriminant function analysis (DFA) is often used in ecological studies to assess how different two or more groups are based on a consistent set of variables collected for each group (McGarigal et al. 2000, McCune and Grace 2002). Quadratic DFA was conducted to assess the relationship between each parasite group (ticks, fleas, and lice) and the environment. Linear DFA could not be used because the data violated the assumption of equal variance/covariance matrices across groups. Each parasite group was evaluated separately by dividing the 66 plots into three groups: 1) plots where no Peromyscus were found, 2) plots with Peromyscus but no parasites, and 3) plots with Peromyscus parasites. The relative contribution of each variable to a given discriminant axis was determined by comparing the absolute value of its standardized coefficient to the mean of the absolute values of all coefficients for that axis. Discriminant functions were used to calculate posterior probabilities for each plot. Overall accuracy of the classification and the kappa coefficient of similarity (k) were calculated as an assessment of the model's ability to separate the groups (Cohen 1968, Hudson and Ramm 1987, McGarigal et al. 2000). If the kappa value is close to 0 then the assignment of plots to groups is no better than chance, while values close to 1 indicate the discriminate function was able to statistically distinguish between the groups. All analyses were performed using R software (R Development Core Team) with the exception of the DFA, which was conducted using PROC Discrim in SAS software (SAS Institute v9.1).
- Top of page
- MATERIALS AND METHODS
- REFERENCES CITED
Three hundred thirty-one small mammals were captured in the field (Table 1); 164 were identified as P. leucopus and 21 juveniles could only be identified to the genus Peromyscus. These 185 mice were checked for ticks and fleas in the field, of which 104 mice, including 22 recaptured animals, were euthanized and additionally inspected in the lab for louse infestations. Parasites from three taxa were collected: 69 larval and nymphal ticks (Acari), 98 adult fleas (Siphonaptera), and 91 adult lice (Phthiraptera) (Table 2). Of the 69 ticks collected, 46 were Ixodes scapularis (blacklegged tick) and 23 were Dermacentor variabilis (dog tick). Of the 98 fleas collected, 95 were Orchopeas leucopus, two were Ctenophthalmus pseudagyrtes, and one was unknown. All 91 lice collected were Hoplopleura hesperomydis and both sexes were present. The average intensity of infestation across taxa ranged from 1.8 to 4.1 parasites per infested mouse (Table 2). While fleas had the lowest intensity of infestation, they were present on the most plots (28/66) and had the highest prevalence of the taxa examined. The tick species were combined for further analysis because the observations for both species were too low to analyze separately.
|State Game Area||Trap nights||P. leucopus||Recaptures||Non-target captures*||Recaptures|
|3) Three rivers||487.5||39||5||10||0|
|Species||Total||Prevalence||Average Intensity*||Plots||k‡||Variance to Mean Ratio|
|Acari (Ticks)||69||13% (24/185)||2.8||12/66||0.919||6.868|
|Ixodes scapularis||46||6% (12/185)||3.8||6/66||0.826||7.812|
|Dermacentor variabilis||23||7% (13/185)||1.8||8/66||2.611||1.667|
|Siphonaptera (Fleas)||98||29% (54/185)||1.8||28/66||1.863||3.062|
|Orchopeas leucopus||95||28% (52/185)||1.8||27/66||1.863||3.062|
|Ctenophthalmus pseudagyrtes||2||1% (2/185)||1||2/66||–||–|
|Hoplopleura hesperomydis||91||12% (22/185)||4.1||14/66||0.682||11.336|
Ticks (Acari). Vegetation characteristics were significantly different between the plots having mice parasitized with ticks and the plots with clean mice or no mice as determined by the separation of these three groups in the DFA (Table 3 and Figure 2). The first discriminant axis (Table 4) had a strong positive association with primary seedling ground cover, primary barren ground cover, secondary forb ground cover, black ash (Fraxinus nigra) canopy basal area, black oak (Quercus velutina) canopy basal area, and red pine (Pinus resinosa) canopy basal area and a strong negative association with primary grass ground cover, black cherry (Prunus serotina) subcanopy height, and secondary seedling ground cover. Thus, the first axis functionally represents a gradient from unsuitable to suitable mouse habitat. The second discriminant axis (Table 4) had a strong positive association with secondary leaf ground cover and secondary seedling ground cover, quaking aspen (Populus tremuloides) subcanopy height, black ash subcanopy height, and white oak (Quercus alba) canopy basal area and a strong negative association with red oak (Quercus rubrum) canopy basal area, big tooth aspen (Populus grandidentata) canopy basal area, sassafras (Sassafras albidum) subcanopy height, elm (Ulmus americana) subcanopy height, and primary forb ground cover. The second axis represents a gradient from dry and disturbed to wet and undisturbed vegetation associations.
|Data set||Eigen Value||Variation‡||F*||P|
|Tick||1st Eigen Value||2.0154||66%||1.7852||0.009|
|2nd Eigen Value||1.0384||34%|
|Flea||1st Eigen Value||3.1553||76%||2.2528||<0.001|
|2nd Eigen Value||0.9776||24%|
|Louse||1st Eigen Value||2.1800||67%||1.9015||0.004|
|2nd Eigen Value||1.0859||33%|
|Variable||DFA 1||DFA 2||DFA 1||DFA 2||DFA 1||DFA 2|
|Silver maple CB||0.78||0.63||0.51||0.86||0.85||0.52|
|White oak CB||-0.14||0.99||-0.73||0.68||-0.62||0.78|
|Quaking aspen CB||0.78||0.63||0.90||-0.43||0.92||0.38|
|Black ash CB||0.98||-0.21||-0.26||0.97||0.66||-0.75|
|Big tooth aspen CB||0.56||-0.83||0.99||0.15||0.58||0.81|
|White pine CB||0.87||–0.49||0.42||0.91||1.00||0.08|
|Red maple CB||0.93||–0.37||0.94||0.35||0.87||0.49|
|Black oak CB||0.97||–0.25||0.76||0.65||0.99||–0.14|
|Red pine CB||0.95||–0.32||0.95||–0.33||0.57||0.82|
|Red oak CB||0.19||-0.98||0.92||–0.38||0.99||–0.11|
|Black cherry CB||0.84||–0.54||0.96||0.27||0.96||0.27|
|Black ash SH||0.71||0.70||0.30||0.95||0.93||0.37|
|White pine SH||0.89||–0.45||0.86||0.51||0.99||–0.16|
|Red oak SH||0.86||–0.51||0.93||0.37||0.30||0.95|
|Quaking aspen SH||–0.08||1.00||0.04||-1.00||-0.73||0.68|
|Black cherry SH||-0.92||0.38||-0.99||–0.17||–0.05||-1.00|
|Red maple SH||0.83||–0.55||0.87||0.49||0.86||0.50|
The discriminant function accurately discriminated between plots with no mice, mice, and mice parasitized by ticks. Classification accuracy was 97% (64/66 correctly classified); this represents a classification power roughly 95% better than random assignment (kappa = 0.95) (Table 5). Not only were the three groups different, but the model was able to discriminate among those groups with a high level of accuracy, indicating the centroids (mean in multivariate space) of each group were distinctly different. Therefore, habitat characteristics can be used to describe the presence of P. leucopus and ticks on plots.
|Actual Class||No Peromyscus||Without ticks||With ticks||Totals|
Fleas (Siphonaptera). Vegetation characteristics were significantly different between the plots having mice parasitized with fleas and plots with clean mice or no mice as determined by the separation of these three groups in the DFA (Table 3 and Figure 3). The first discriminant axis (Table 4) had a strong positive association with big tooth aspen canopy basal area, black cherry canopy basal area, red pine canopy basal area, red maple canopy basal area, and secondary forb ground cover and a strong negative association with primary grass ground cover, secondary seedling ground cover, and black cherry subcanopy height; thus functionally the first axis represents a gradient from unsuitable to suitable mouse habitat. The second discriminant axis (Table 4) had a strong positive association with dogwood (Cronus spp) subcanopy height, elm subcanopy height, black ash subcanopy height, black ash canopy basal area, and white pine (Pinus strobus) canopy basal area and a strong negative association with primary grass ground cover, secondary leaf ground cover, and quaking aspen subcanopy height. The second axis represents a gradient from dry to wet vegetation associations.
The discriminant function accurately discriminated between plots with no mice, mice, and mice parasitized by fleas. Classification accuracy was 97% (64/66 correctly classified); this represents a classification power roughly 95% better than random assignment (kappa = 0.95) (Table 6). Not only were the three groups different, but the model was able to discriminate among those groups with a high level of accuracy; this indicates the centroids of each group were distinctly different. Therefore, habitat characteristics can be used to describe the presence of P. leucopus and fleas on plots.
|Actual Class||No Peromyscus||Without fleas||With fleas||Totals|
Lice (Phthiraptera). Vegetation characteristics were significantly different between the plots having mice parasitized with lice and plots with clean mice or no mice as determined by the separation of these three groups in the DFA (Table 3 and Figure 4). The first discriminant axis (Table 4) was strongly positively associate with white pine canopy basal area, red oak canopy basal area, black oak canopy basal area, white pine subcanopy height, and primary forb ground cover and a strong negative association with white oak canopy basal area, quaking aspen subcanopy height, and primary grass ground cover; thus functionally the first axis represents a gradient from unsuitable to suitable mouse habitat. The second discriminant axis (Table 4) had a strong positive association with secondary forb ground cover, red oak subcanopy height, dogwood subcanopy height, elm subcanopy height, and red pine canopy basal area and a strong negative association with black ash canopy basal area, secondary leaf ground cover, secondary seedling ground cover, and black cherry subcanopy height. The second axis represents a gradient from dry to wet vegetation associations.
The discriminant function accurately discriminated between plots with no mice, mice, and mice parasitized by lice. Classification accuracy was 97% (64/66 correctly classified); this represents a classification power roughly 95% better than random assignment (kappa = 0.95) (Table 7). Not only were the three groups different, but the model was able to discriminate among those groups with a high level of accuracy; this indicates the centroids of each group were distinctly different. Therefore, habitat characteristics can be used to describe the presence of P. leucopus and lice on plots.
|Actual Class||No Peromyscus||Without lice||With lice||Totals|
- Top of page
- MATERIALS AND METHODS
- REFERENCES CITED
Contrary to our hypothesis, we did not observe a stronger association of host habitat for ticks than fleas or lice. The strength of association with parasites and habitat is no different among the three taxa.
Differences among studies in parasite prevalence may occur for many reasons (e.g., collection methods, timing of collections, geographic area, etc.). Surveying parasites via their hosts was ideal for this study as ectoparasites are attracted to and hone in on the presence of their host (Marshall 1981); hence collection of host organisms would be more sensitive than directly sampling the environment alone (Hamer et al. 2010). Because we only trapped mammals during a short period, it is possible not all of the potential parasite species were collected, limiting the implications of this study to those species of parasites found mid-June to August. One of our objectives, however, was to implement this approach and see if it would be able to discern habitat-ectoparasite relationships even on a small dataset and we were successful in that regard, so that future, larger, and more frequently trapped studies could employ this method. With respect to common tick species in Michigan, this time period should have allowed us to detect both blacklegged ticks and dog ticks, even though their activity patterns may have varied over the period (Hamer 2009), though nymphs and larvae of I. scapularis peak in June (S. Hamer, Michigan State University, personal communication). Flea species collections were biased toward fur or body fleas, as nidicolous (nest associated) fleas were not collected. Therefore, it is possible both ticks and fleas’ spatial distributions and vegetation associations are incomplete. A regular, year-long trapping protocol would help discern any temporal relationships between parasite presence and vegetation variables, in addition to any uncertainty concerning the presence and distribution of parasites of P. leucopus. Due to limitations in resources, tradeoffs must be made when covering a wide range of habitats with proper replications and time spent at each site. For this study, we determined that our sampling in southern Michigan was appropriate for establishing a baseline relationship between ectoparasites and their host habitat.
Species ranges are dynamic as their distributions may change between seasons and from year to year depending on the species (Lomolino et al. 2005). This is especially true for species that have rapidly changing ranges such as those colonizing an area post disturbance or invasive species (Lomolino et al. 2005). Studies examining species range will always suffer from these problems. Therefore, it is not possible to fully describe the habitat associations of blacklegged ticks in Michigan, as it is unknown if their absence is because the habitat was unsuitable or they had not yet had the opportunity to invade that area of the southern peninsula (Guerra et al. 2002, Hamer et al. 2007, Hamer et al. 2009). Nevertheless, suitable habitat affects the ability of ticks to become established along with host availability and movement (Manangan et al. 2007). It is also worth noting that the vegetation associations described in this study may be characteristic of invading tick populations and not necessarily characteristic of established populations. Furthermore, there is the question of the host species for each of the parasites. Both blacklegged and dog ticks are generalist species. As generalists, the full extent of their habitat distributions cannot be fully discerned by examining only one of several host species. Mice are considered to be one of the most important host species for Ixodes scapularis (Shaw et al. 2003). In contrast, Orchopeas leucopus and Hoplopleura hesperomydis are both specific to mice of the genus Peromyscus (Fox 1940, Kim et al. 1986).
We found habitats without mice were characterized by a general lack of canopy cover, with grass being the primary ground cover, whereas presence of canopy cover was a strong indicator of the occurrence of Peromyscus, which is in agreement with King (1968). Primary grass ground cover was the key variable that indicated a lack of mice on plots for all three of the parasite taxa. Though each analysis did not weigh the same variables as indicators of plots without mice, they were all in agreement that these factors indicated a lack of cover for these animals.
Mice parasitized by ticks or lice are more likely to be found in areas that have undergone a recent disturbance. Plots with ticks were characterized by colonizers such as black cherry, sassafras, and elm which are indicators of disturbance (Szafoni 1990, Barnes and Wagner 2002). Plots with lice were characterized by the presence of black cherry and high density of leaf cover and seedlings as secondary ground cover which are also indicators of disturbance. Overall, species that are indicators of disturbance, such as these primarily early colonizers, consistently weighted heavily to separate plots with parasites from plots without parasites for both ticks and lice. Our findings generally support those found by others. Lubelczyk et al. (2004) found tick abundance increased when invasive shrub species were present, indicating a change from the natural vegetation in Maine. The authors concluded disturbances leading to the introduction and successful establishment of invasive species were positive indicators of tick abundance.
Similar to Guerra et al. (2002), we found ticks to be present in forest plots characterized by high densities of oak and maple species in the canopy. Plots without parasites were characterized by tree species associated with wet soils (Szafoni 1990, Barnes and Wagner 2002). For example, plots without ticks were associated with silver maple, quaking aspen, and big tooth aspen; plots without fleas were characterized by dogwood, elm, and black ash, while plots without lice were characterized by dogwood, elm, and aspen. Supporting these results are findings by Guerra et al. (2002) and Manangan et al. (2007) who found sites without ticks were dominated by clay soils, which retain water and support wetland vegetation species. Vegetation species strongly associated with dry soil types were also strongly associated with the presence of all parasite taxa, though there were subtle differences in the composition associated with each. This demonstrates that a lack of water tolerant tree species may also be an indicator for parasite presence, contradicting previous work that suggests high relative humidity (RH) is important for the presence and survival of both I. scapularis (≤93% RH (Stafford 1994)) and D. variabilis (85% RH (Knülle and Devine 1972)). Given the high relative humidity required by these species, it is counterintuitive that the habitats most predictive of their presence appear to be dryer vs wetter mesic habitats; therefore, relative humidity may not be the most important factor in determining the presence of these species of ticks.
Past studies have used a priori defined habitat classifications to determine parasite species assemblages (Krasnov et al. 1997), not focusing on the potential effects of vegetation on flea or louse presence, but rather on the effects of microclimate (i.e., temperature and humidity) (Eskey 1938, Marshall 1981, Christie 1982, Krasnov et al. 2001, Adjemian et al. 2006) and host species assemblages (Krasnov et al. 2005). However, Krasnov and associates (2002, 2004, 2006) have established that environmental influences such as vegetation and soil attributes affect flea species assemblages and influence flea species richness far more than host body parameters. Our work tends to support those findings of Krasnov and associates where vegetation and percent cover of various ground vegetation influenced presence or absence of flea species. Our results demonstrate vegetation species assemblages are a strong indicator of the presence or absence of ectoparasites. This study is the first to look at the relationship between individual vegetation species and the presence of lice, where vegetation indicative of disturbance and dry soils were descriptive of louse presence on mice; although Calvete et al. (2003) suggested environmental factors influencing red-legged partridge lice presence by examining the influence of mean environment temperature and Normalized Difference Vegetation Index (NDVI), which is highly correlated with environmental humidity. This empirical study demonstrates that the collection of vegetation data and subsequate analysis using DFA is an effective method of describing the habitat of ectoparasites of Permoyscus mice.
The decisions of wildlife managers can have a lasting impact on disease risk. In the areas examined in this study, disturbance was an indicator for the presence of ticks and lice. This warrants further investigation concerning the impact of disturbance on parasite species in other areas. Lubelczyk et al. (2004) found the presence of ticks was positively associated with the presence of several invasive species and landscape changes which may be creating favorable tick habitats. Management decisions may reduce disease risk by considering the impact of management actions on the host habitats of arthropod vectors of disease. The results of this study could be used to help create risk assessment maps for current or future diseases spread by these species of ticks, fleas, and lice, as similar studies have used environmental data for this purpose such as Wimberly et al. (2008), Carbajal de la Fuentae et al. (2009), and Linard et al. (2009). This is potentially very useful to both wildlife managers and community health professionals alike.
- Top of page
- MATERIALS AND METHODS
- REFERENCES CITED
We thank P. Muzzall, N. Walker, M. Kaufman, and J. Owen for the use of their lab space and equipment, as well as P. Muzzall for helping with flea identification, G. Parsons for helping with louse identification, B. Lundrigan and L. Abraczinskas for helping with the Peromyscus specimens, S. Hamer, M. Cook, and D. Lipp for their help in the field, and E. Monroe and J. Mize for their helpful comments during the revision process. Three anonymous reviewers provided helpful comments. This work was made possible by financial assistance from the Michigan Department of Natural Resources.
All procedures adhered to the Animal Use Guidelines established by Michigan State University Institutional Animal Care and Use Committee (IACUC). This project was authorized by the Animal Use Committee under Animal Use Form (AUF) number 04/07–039–00. Flea and louse voucher specimens were deposited at Michigan State University Entomology Museum accession number MSU 2009–01, East Lansing, MI and Smithsonian National Museum of Natural History transaction number 2051221, Washington, D.C.; tick voucher specimens were deposited at Smithsonian National Museum of Natural History in the U.S. National Tick Collection accession number RML 124449–124453, care of Georgia Southern University, Statesboro, GA; and Peromyscus vouchers were deposited at the Michigan State University Museum Mammal Research Collection accession numbers MSU 37467- 37595, East Lansing, MI.
- Top of page
- MATERIALS AND METHODS
- REFERENCES CITED
- 2006. Analysis of genetic algorithm for Rule-Set Production (GARP) modeling approach for predicting distributions of fleas implicated as vectors of plague, Yersinia pestis, in California. J. Med. Entomol. 43: 93–103. , , , and .
- 1991. Infectious Diseases of Humans: Dynamics and Control. Oxford University Press, U.S.A . and .
- 1998. Host densities as determinants of abundance in parasite communities. Proc. R. Soc. Lond. B. 265: 1283–1289. , , , and .
- 1983. Michigan Mammals. Michigan State University Press, East Lansing , MI .
- 2002. Michigan Trees. A Guide to the Trees of Michigan and the Great Lakes Region. The University of Michigan Press, Ann Arbor , MI . and .
- 2003. Ectoparasite ticks and chewing lice of red-legged partridge, Alectoris rufa, in Spain. Med. Vet. Entomol. 17: 33–37. , , , and .
- 1979. Distribution of the American dog tick, Dermacentor variabilis (Say), and its small-mammal hosts in relation to vegetation types in a study area in Nova Scotia. Canad. J. Zool. 57: 1950–1959. and .
- 2009. The association between the geographic distribution of Triatoma pseudomaculata and Triatoma wygodzinskyi (Hemiptera: Reduviidae) with environmental variables recorded by remote sensors. Infect. Genet. Evol. 9: 54. , , , , and .
- 1982. Plague: review of ecology. Ecol. Dis. 1: 111–115.
- 1968. Weighted kappa: nominal scale agreement with provision for scaled disagreement or partial credit. Psychol. Bull. 70: 213–220.
- 1938. Fleas as vectors of plague. Am. J. Publ. Hlth. 28: 1305.
- 1951. The Sucking Lice. California Academy of Sciences. and .
- 1940. Fleas of Eastern United States. The Iowa State College Press.
- 2002. Predicting the risk of Lyme disease: Habitat suitability for Ixodes scapularis in the north central United States. Emerg. Infect. Dis. 8: 289–297. , , , , , , , , and .
- 2007. Zoonotic pathogens in Ixodes scapularis, Michigan. Emerg. Infect. Dis. 13: 1131–1133. , , , , , , and .
- 2010. Invasion of the Lyme disease vector Ixodes scapularis: Implications for Borrelia burgdorferi Endemicity. EcoHealth 7: 47–63. , , , and .
- 2009. Use of tick surveys and serosurveys to evaluate pet dogs as a sentinel species for emerging Lyme disease. Am. J. Vet. Res. 70: 49–56. , , , , , and .
- 1995. Macroparasites: observed patterns. In: B. Grenfell and A.P. Dobson, (eds.) Ecology of Infectious Diseases in Natural Populations. pp. 144–176. Oxford University Press, USA . and .
- 1987. Correct formulation of the kappa coefficient of agreement. Photogramm. Eng. Rem. 53421–53422. and .
- 1986. The Sucking Lice of North America: An Illustrated Manual for Identification. Pennsylvania State University Press University Park, PA . , , and .
- 1968. Biology of Peromyscus (Rodentia). The American Society of Mammalogists. 593 pp.
- 1972. Evidence for active and passive components of sorption of atmospheric water vapour by larvae of the tick Dermacentor variabilis. J. Insect Physiol. 18: 1653–1664. and .
- 2005. Spatial variation in species diversity and composition of flea assemblages in small mammalian hosts: geographical distance or faunal similarity? J. Biogeogr. 32: 633–644. , , , , and .
- 2002. The effect of substrate on survival and development of two species of desert fleas (Siphonaptera: Pulicidae). Parasite-Journal De La Societe Francaise De Parasitologie 9: 135–142. , , , and .
- 2001. Effect of air temperature and humidity on the survival of pre-imaginal stages of two flea species (Siphonaptera: Pulicidae). J. Med. Entomol. 38: 629–637. , , , and .
- 2004. Flea species richness and parameters of host body, host geography and host ‘milieu’. J. Anim. Ecol. 73: 1121–1128. , , , and .
- 1997. Host-habitat relations as an important determinant of spatial distribution of flea assemblages (Siphonaptera) on rodents in the Negev desert. Parasitology 114: 159–173. , , , , and .
- 2006. Habitat variation in species composition of flea assemblages on small mammals in central Europe. Ecol. Res. 21: 460–469. , , , and .
- 2009. A multi-agent simulation to assess the risk of malaria re-emergence in southern France. Ecological Modeling 220: 160–174. , , , and .
- 2005. Biogeography 3rd edition. Sinauer Associates. , , and .
- 2004. Habitat associations of Ixodes scapularis (Acari : Ixodidae) in Maine. Environ. Entomol. 33: 900–906. , , , , , and .
- 2007. Habitat factors influencing distributions of Anaplasma phagocytophilum and Ehrlichia chaffeensis in the Mississippi alluvial valley. Vector-Borne Zoonot. Dis. 7: 563–573. , , , , , and .
- 1982. The use of ecological terms in parasitology (report of an ad hoc committee of the American Society of Parasitologists). J. Parasitol. 68: 131–133. , , , , and .
- 1981. The Ecology of Ectoparasitic Insects. Academic Press. London .
- 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach , OR . and .
- 2000. Multivariate Statistics for Wildlife and Ecology Research. Springer, New York , NY . , , and .
- 2003. Factors influencing the distribution of larval blacklegged ticks on rodent hosts. Am. J. Trop. Med. Hyg. 68: 447–452. , , , and .
- 1979. Ticks of Virginia. The Insects of Virginia: No. 13. Polytechnic Institute and State University, College of Agriculture and Life Sciences.
- 1972. Rocky Mountain Spotted Fever in relation to vegetation in the eastern United States, 1951–1971. Am. J. Epidemiol. 96: 59–69. , , and .
- 1994. Survival of Immature Ixodes scapularis (Acari: Ixodidae) at Different Relative Humidities. J. Med. Entomol. 31: 310–314.
- 1990. Vegetation Management Guideline Autumn Olive (Eleaganus umbellata Thunb.). Illinois Nature Preserves Commission.
- 2002. Parasites and host population dynamics. In: P.J. Hudson, A. Rizzoli, B. Grenfell, H. Heesterbeek, and A.P. Dobson, (eds.). The Ecology of Wildlife Diseases. pp. 45–62. Oxford University Press, U.S.A. , , , , , , , , , , , , , and .
- 1998. Geographic distribution of ticks (Acari: Ixodidae) in Michigan, with emphasis on Ixodes scapularis and Borrelia burgdorferi. J. Med. Entomol. 35: 872–882. , , , , , , , , and .
- 1982. Diseases Transmitted by Rats and Mice: Health Hazards to Humans and Domesticated Animals. Thomson Publications, Fresno , CA .
- 1968. Parasites. In: J.A. King (ed.) Biology of Peromyscus (Rodentia). The American Society of Mammalogists. 593 pp.
- 2002. Heterogeneities in macroparasite infections: patterns and processes. In: P.J. Hudson, A. Rizzoli, B. Grenfell, H. Heesterbeek, and A.P. Dobson (eds.). The Ecology of Wildlife Diseases. pp. 6–44. Oxford University Press, U.S.A . , , , , , , and .
- 2008. Spatial heterogeneity of climate and land cover constraints on distributions of tick-borne pathogens. Global Ecol. Biogeogr. 17: 189–202. , , , , and .