Journal of Animal Ecology 2010, 79, 602–609
doi: 10.1111/j.1365-2656.2010.01663.x
Differential effects of experimental increases in
sociality on ectoparasites of free-ranging raccoons
Ryan J. Monello*† and Matthew E. Gompper
Department of Fisheries and Wildlife Sciences, University of Missouri, Columbia, MO 65211, USA
Summary
1. Parasite transmission depends on the rate at which hosts come into contact with one another or
the infectious stages of parasites. However, host contact rates and their influence on parasite transmission are difficult to quantify in natural settings and can fluctuate with host behaviour and the
ecological constraints of parasites.
2. We investigated how experimental increases in rates of contact and social aggregation affected
ectoparasite prevalence and intensity of free-ranging raccoons (Procyon lotor). Twelve independent raccoon populations were subjected to differential resource provisions for 2 years: a clumped
food distribution to aggregate hosts (n = 5 aggregated populations), a dispersed food distribution
to control for the effects of food without aggregating hosts (n = 3) and a no food treatment
(n = 4).
3. Remote cameras indicated that aggregation sizes and rates of contact were three to four times
greater in aggregated compared with that in non-aggregated populations. The number of ticks
(adult Dermacentor variabilis) on raccoons in aggregated populations was 1Æ5–2Æ5 times greater
from May to July, the primary time of tick seasonal occurrence. Conversely, louse (Trichodectes
octomaculatus) populations were c. 40% sparser on male raccoons in aggregated (compared with
that in non-aggregated) populations because of greater overdispersion of lice and a larger number
of male hosts harbouring fewer parasites. No treatment-related differences were found among fleas
(Orchopeas howardi).
4. These results were not consistent with our current understanding of parasite transmission;
greater rates of host sociality led to increases in a parasite that does not rely on host contact for
transmission (ticks) and declines in a parasite that depends on host contact for transmission (lice).
We concluded that D. variabilis increased in aggregated sites because they can detect and seek out
hosts and were more likely to drop off after obtaining a blood meal and re-attach to raccoons in
these locations. Several factors may have contributed to sparser louse populations on male hosts,
including a dilution effect that lowered per capita infestation levels.
5. These results indicate that ectoparasites can interact in unique ways with their hosts that are not
consistent with other types of parasite species or models of their transmission.
Key-words: host, infestation, mammal, parasite, wildlife
Introduction
Parasite transmission depends on the rate at which hosts
come into contact with one another or with the infectious
stages of parasites. This contact rate is generally viewed as
influenced by habitat quality and population density (Anderson et al. 1981). Problematically, contact rates are difficult to
quantify in natural settings and can fluctuate with host
*Corresponding author. E-mail: [email protected]
†Present address: Biological Resource Management Division,
National Park Service, Fort Collins, CO 80525, USA.
behaviour (Anderson & May 1991; Caley & Ramsey 2001).
As a result, those interested in the general relationships
between host characteristics, such as sociality and the probability of being parasitized, have used intra- and interspecific
correlational analyses. For instance, cross-species studies
suggest that parasites can exert selective pressures on host
social structures (Davies et al. 1991; Hochberg 1991) and
intraspecific studies generally show positive correlations
between group size or density and the number of parasite species per host or parasite intensity (Keymer & Read 1991;
Møller, Dufva & Allander 1993; Krasnov, Khokhlova &
Shenbrot 2002). The general cause of these relationships may
2010 The Authors. Journal compilation 2010 British Ecological Society
Sociality of raccoons and ectoparasites 603
be increased contact within groups and, thus, increased
opportunities for parasite transmission (Freeland 1979; Hoogland 1979; Moore, Simberloff & Freehling 1988; Poulin
1991; Côté & Poulin 1995; Porteous & Pankhurst 1998; Tella
2002).
Unfortunately, one cannot assume that contact uniformly
increases the probability of parasite transmission. Sometimes
the relationship is an inverse one. Sociality can result in
decreased parasitism if group living allows more efficient vector-avoidance behaviours such as grooming, or if living with
other potential hosts creates a dilution effect which decreases
the chance of a particular individual being singled out by a
vector or parasite (Duncan & Vigne 1979; Rubenstein &
Hohmann 1989; Mooring & Hart 1992; Côté & Gross 1993;
Gompper 2004; Fauchald et al. 2007; Vicente et al. 2007). In
addition, some species may show density-dependent disease
resistance or acquired immune resistance to particular parasites (Craig et al. 1996; Wilson et al. 2002). Furthermore, the
evidence used to assess the relationship between social contact and parasitism is mainly correlational and the causal
direction of this relationship is typically unknown. As some
researchers suggest that group size is limited by parasite
abundance and impact on host fitness (Alexander 1974; Freeland 1976, 1979), the associated question of whether the shift
from an asocial, low-contact lifestyle to a more social or
high-contact lifestyle carries with it an increased parasite burden has not been examined intraspecifically (Møller et al. 1993).
Here we examined the relationship between ectoparasite
intensity and rates of host contact in a natural setting. Specifically, we manipulated the contact rate and social aggregation size of free-ranging raccoon (Procyon lotor Linnaeus)
populations via differential food provisioning simultaneously monitoring ectoparasite populations. Our efforts
focused on the three most common ectoparasites of raccoons in our study sites, which included the American dog
tick Dermacentor variabilis Say, the louse Trichodectes
octomaculatus Paine and the flea Orchopeas howardi Baker
(Monello & Gompper 2007, 2009). The life-history and ecological constraints of these parasites allow for general predictions of how greater host contact will affect their
infestation levels. Dermacentor variabilis attach and drop off
hosts after feeding. They obtain a single blood meal from
one of a variety of hosts in each of their three life stages (larvae, nymph and adult), but raccoons are the primary host
species for adult D. variabilis in Missouri (Kollars et al.
2000). However, because ticks are not transmitted between
animals, we predicted that increases in host sociality would
have little impact on adult tick infestation levels of raccoons.
Conversely, lice are host species specific and cannot survive
off-host for more than a few hours (Durden 2001). The ability of lice to infest new individuals depends on host contact,
and so we predicted that increased host sociality should
result in greater prevalence and louse population size.
Finally, fleas have been shown to be sensitive to the density
of their primary hosts (Krasnov et al. 2002), but they also
have the ability to live on multiple hosts and survive off-host
for extended periods of time. Thus, we predicted that flea
populations would show a response to increased contact
among hosts that is intermediate between the predicted
responses of ticks and lice.
Materials and methods
STUDY SYSTEM
Study sites (n = 12) were located on state or federal public lands
within 60 km of Columbia, MO, USA. Sites consisted of second
growth oak (Quercus spp.) and hickory (Carya spp.) forest with a
maple (Acer spp.) and cedar (Juniperus virginiana) understorey. We
considered all sites to be independent of one another, as we captured
>700 individuals with >500 recapture events from 2005 to 2007,
and no animals were observed to move between sites (Monello 2009).
Each site was assigned to one of three treatments in January 2006:
a permanent feeding station receiving 35 kg ⁄ week-1 of dog food at a
single location to aggregate raccoons (n = 5 populations), a highly
dispersed distribution of the same quantity of dog food to control for
the effects of food addition but not aggregate hosts (control with
food, n = 3 populations) and a no food treatment (control without
food, n = 4 populations). Treatments were assigned randomly
within geographically defined subsets that were used to assure that
treatments were interspersed and to control for unidentified regional
effects. The control sites with food were provisioned by placing food
in 0Æ25-kg piles (c.140 per week) that were randomly moved each
week throughout a 4-km2 portion of each study site. All provisions
were maintained from January to September in 2006 and 2007.
HOST SAMPLING
Traplines of 15 pairs of Tomohawk box traps (30 in total) were established at each site. A single trap was placed c.50 m on each side of a
1-km transect, with adjacent traps spaced 75–100 m apart. Raccoons
were trapped for ‡10 days at each site two to three times per year
between March and November. Traps were baited with mackerel
and checked daily. Raccoons were immobilized with ketamine
hydrochloride and xylazine (Evans 2002), tagged, weighed, sexed and
aged by body size, genital morphology and tooth eruption and wear
(Grau, Sanderson & Rogers 1970). This resulted in five age classes:
cub <5 months; I: 5–14 months; II: 15–38 months; III: 39–
57 months and IV+: >57 months. Data from cubs, which were generally free of ectoparasites (Monello & Gompper 2007, 2009), and
recaptured animals were not included in analyses.
Remote infrared film cameras (DeerCam DC-300; Non-Typical
Inc., Green Bay, WI, USA) were used to monitor feeding station use,
rates of contact and raccoon aggregation sizes in experimental (hereafter referred to as aggregated sites) and control sites. One camera
was maintained from January to August at each permanent feeding
station in aggregated sites. Five cameras were maintained in control
sites with and without food for 10–15 days during both spring and
summer. Cameras were placed in front of a small pile of food (control
with food) or bait (control without food) and relocated every 5 days.
All cameras were placed at c.0.5-m height on a tree near the food or
bait. Aggregation size was recorded as the maximum number of raccoons, excluding cubs, observed per night in a single photograph at
each site. Instantaneous contact rates (%) were assessed for each site
by dividing the number of camera days with photographs that had ‡2
raccoons (excluding cubs) by the total number of camera days. We
assigned a categorical variable for short-term contact while at the
feeding station of 0 (no contact within the c.16–m2 camera range) or
2010 The Authors. Journal compilation 2010 British Ecological Society, Journal of Animal Ecology, 79, 602–609
604 R. J. Monello & M. E. Gompper
1 (contact with conspecifics within the range) (Gompper & Wright
2005).
As food additions may influence population density and body condition, and thus ectoparasites, we quantified population density and
relative body condition. We calculated population density of each
site by estimating adult population size in 2007 using closed population models in Popan-5 (Arnason & Schwarz 1999), and divided these
values by the effective trapping area (Kenward 1985). We estimated
the effective trapping area by multiplying the linear length of the
trapline by a buffer of three fourths of the median summer home
range of raccoons (Kenward 1985) from rural sites in Illinois (Prange
et al. 2004), which were comparable to preliminary home range estimates from sites in this study (M. Wehtje, unpublished data). We
used the residuals from a linear regression of body mass on body size
to assess the relative body condition of each individual (SchulteHostedde et al. 2005), with higher values indicating better body condition.
PARASITE QUANTIFICATION
Adult D. variabilis are relatively large (‡3 mm in length), occur seasonally in our study sites and are readily found on animals in the field
without magnification (Monello & Gompper 2007). We quantified
adult D. variabilis by a thorough search of the entire body. Analyses
included only those tick counts that were conducted between April
and August, which corresponds to the time period when adult ticks
are present at all of our study sites and able to obtain a complete
blood meal (Monello & Gompper 2007).
We quantified the relative abundance of lice and fleas via 10
strokes with a flea comb from the base of the neck to the base of the
tail on the dorsal region of raccoons (Monello & Gompper 2009).
Animals that were wet or muddy could not be combed and were
excluded. Hair and ectoparasites were immediately placed in a plastic
bag, sealed and frozen within 8 hours. In the laboratory, lice and fleas
were identified to species and quantified under a dissecting scope.
Measures of parasitism follow Bush et al. (1997); prevalence was
the percentage of hosts infested and intensity represents the population size of a parasite species among infested animals (i.e. animals
without parasites are not included in measures of intensity). We used
generalized linear models with a binomial distribution and logit
transformation to determine whether the presence or absence (i.e.
prevalence) of each ectoparasite differed between treatment categories. We included month · treatment and sex · treatment interactions because prior research indicated their importance and potential
to interact with our study design (Monello & Gompper 2007, 2009).
Pairwise comparisons were used for all-level comparisons if a significant difference was found (a = 0Æ05).
MODEL SELECTION – PARASITE INTENSITY
We used information-theoretic model selection to evaluate the ability
of treatments, host characteristics and month to explain the intensity
of each ectoparasite species. Model factors included aggregation,
food supplementation, host age and sex and month of capture. A priori hypothesis and model formulation was based in part on the bestfit models from previous research that was conducted at these sites in
the absence of experimental manipulations (Monello & Gompper
2007, 2009). We included aggregation and food supplementation as
separate factors with the previously identified variables from the
best-fit models and conducted separate analyses for each ectoparasite
species (Table S1; available online). Intercept-only and global models were also included in model selection procedures for each ectopar-
asite. Year was not included as a factor because no significant model
effects of year were found for any of the ectoparasite species
(P ‡ 0Æ086 in all cases).
We used generalized linear models with a negative binomial distribution and log link for model selection (Wilson & Grenfell 1997;
Hardin & Hilbe 2007). Ectoparasite intensity of individual raccoons
was the sample unit. We conducted a goodness-of-fit test to assess the
ability of model factors to explain ectoparasite intensity by comparing the global model against the intercept-only model (Franklin et al.
2000). For each ectoparasite species, we proceeded with model selection procedures only if global models provided a better fit than intercept-only models (a = 0Æ05). We calculated Akaike’s Information
Criterion (corrected for small sample size; AICc) and ranked the
models based on differences between the best approximating model
and all other models (DAICc) in the candidate set. Models with a
DAICc value £2 of the best-fitting model were considered to have substantial empirical support as a best-fitting model. To better assess
model structure, we calculated the overdispersion parameter of the
best-fit model for each ectoparasite species, model weight (i.e. the relative likelihood that a particular model is the best-fit model) and
model-averaged estimates and odds ratios (±95% CI) for each variable in the 90% confidence set of models (Rmodel weights ‡0Æ90;
Burnham & Anderson 2002).
Results
RACCOON AGGREGATION, BODY CONDITION AND
DENSITY
Aggregation size of adult raccoons was the greatest in sites
with the permanent feeding stations (Kruskal–Wallis
H¢ = 230Æ948, d.f. = 2, P < 0Æ001; aggregated > control
sites with food > control sites without food, all comparisons
P £ 0Æ006 based on Mann–Whitney U-test), with up to 10
individuals simultaneously visiting the food plots in aggregated sites (mean number of individuals per photo ± 95%
CI – control without food: 1Æ00 ± 0Æ00; control with food:
1Æ16 ± 0Æ06; aggregated: 3Æ22 ± 0Æ19). Instantaneous contact rates for adults were 80% in aggregated sites (n = 358
photos), 15% in control with food sites (n = 149) and 0% in
control without food sites (n = 43). Photograph capture
effort in the control with and without food sites (215–330
photograph nights per year per site) exceeded the aggregated
sites (70–105 photograph nights per year per site) because
cameras in the aggregated sites always ran out of film (but
not food) within 3 days.
Raccoon density did not differ among treatments (Kruskal–Wallis H¢ = 0Æ50, d.f. = 2, P = 0Æ778), averaging
29Æ4 ± 7Æ3 (SE) in control without food sites (range 22Æ4–
33Æ6, n = 4), 34Æ8 ± 7Æ6 in control with food sites (range
29Æ3–35Æ2, n = 3) and 32Æ32 ± 8Æ50 in aggregated sites
(range 12Æ9–46Æ4, n = 5). Supplemental food increased the
weight and improved the relative body condition of female
raccoons (mean body condition ± 95% CI; control without
food: )0Æ54 ± 0Æ21; control with food: )0Æ17 ± 0Æ16; aggregated: 0Æ03 ± 0Æ15), but no differences were observed among
males (control without food: 0Æ26 ± 0Æ25; control with food:
0Æ40 ± 0Æ27; aggregated: 0Æ35 ± 0Æ17).
2010 The Authors. Journal compilation 2010 British Ecological Society, Journal of Animal Ecology, 79, 602–609
Sociality of raccoons and ectoparasites 605
ECTOPARASITE PREVALENCE
ECTOPARASITE INTENSITY
Infested animals in aggregated populations harboured more
adult D. variabilis ticks (mean intensity: 36Æ4 ± 5Æ8, range 1–
157, n = 127 animals) than those in control without food
(22Æ6 ± 4Æ4, range 1–103, n = 81) and control with food
populations (24Æ1 ± 4Æ8, range 1–122, n = 87). Maximum
tick intensities were maintained from May to July in aggregated populations, whereas intensity in control populations
with and without food increased through summer, peaked in
July and consistently displayed a lower intensity than aggregated populations (Fig. 1).
Intensity of lice was lower in aggregated populations
(control without food = 6Æ7 ± 2Æ1, n = 55; control with
food = 7Æ0 ± 2Æ3,
n = 56;
aggregated = 5Æ0 ± 0Æ8,
n = 94). However, there were sex-specific treatment differences; female raccoons exhibited little difference in louse
intensity among treatments, whereas males in aggregated
populations had a lower intensity of lice than those in control with and without food populations (Fig. 2). The intensity of lice on male raccoons in aggregated treatments
exhibited greater overdispersion (variance : mean = 12Æ99)
than the intensity in control with food (variance : mean = 9Æ61) and control without food (variance : mean = 9Æ30) categories, with a greater number of
male hosts in aggregated treatments harbouring fewer lice
Intensity of ticks
60
Control (with and without food, n = 168)
Aggregated (n = 127)
40
20
0
April
May
June
July
August
Fig. 1. Mean intensity (±95% CI) of adult ticks (Dermacentor variabilis) observed on raccoons in aggregated and non-aggregated sites
(control with and without food).
12
10
Intensity of lice
We captured and assessed parasites for 414 raccoons during
2006 and 2007 (mean number of individuals per site type ±
SE – control without food: 25Æ8 ± 3Æ3, n = 103; control
with food: 40Æ3 ± 3Æ3, n = 121; experimental: 38Æ0 ± 7Æ6,
n = 190). The male : female capture ratio was similar in
each treatment category during both years of the experiment
(mean of both years ± SE; control without food: 1Æ1 ± 0Æ3,
control with food: 1Æ1 ± 0Æ2, aggregated: 1Æ2 ± 0Æ1; contingency table Chi-squared test, P ‡ 0Æ521 across both years or
within individual years). Prevalence of D. variabilis averaged
93 ± 2% (mean ± 95% CI; all estimates of data dispersion
hereafter represent 95% CI) across all sites and did not differ
between treatments (P ‡ 0Æ450 for all comparisons). The
prevalence of lice did not differ because of treatment or treatment · month (P ‡ 0Æ196 in both cases), but did exhibit a
sex · treatment
interaction
(v2 = 71Æ886,
d.f. = 3,
P < 0Æ001). Pairwise comparisons of the sex · treatment
interaction term indicated that male raccoons had a greater
prevalence of lice than females regardless of treatment category (P < 0Æ001 for all comparisons). Overall, almost 3·
more male raccoons (71 ± 6%, n = 217) harboured lice
than females (28 ± 7%; n = 197). However, even though
prevalence of lice was lower among males in aggregated sites,
there were no within-sex treatment differences detected for
males or females (P ‡ 0Æ185 for all within sex treatment comparisons). Prevalence of the flea O. howardi was 14 ± 3%
overall and did not differ between treatments (P ‡ 0Æ079 for
all comparisons).
Control without food (n = 55)
Control with food (n = 56)
Aggregated (n = 94)
8
6
4
2
0
Female
Male
Fig. 2. Mean intensity (±95% CI) of lice (Trichodectes octomaculatus) observed on male and female raccoons in control with and without food and aggregated sites.
(Fig. 3). The intensity and distribution of lice on females
were similar among treatments (Fig. 3).
Intensity of fleas did not exhibit any patterns among treatments (control without food: 1Æ8 ± 0Æ5, n = 16; control with
food: 1Æ7 ± 1Æ2, n = 17; aggregated: 2Æ0 ± 0Æ6, n = 25) or
host sex (males = 2Æ0 ± 0Æ7, n = 37; females = 1Æ7 ± 0Æ6,
n = 21).
MODEL SELECTION
Goodness-of-fit tests that compared the global and interceptonly models found a significant difference for D. variabilis
ticks (likelihood ratio Chi-square = 52Æ090, d.f. = 10,
P < 0Æ001) and lice (Likelihood ratio v2 = 32Æ188,
d.f. = 14, P = 0Æ004). There were no differences between
the global and intercept-only models for fleas (Likelihood
ratio v2= 6Æ635, d.f. = 13, P = 0Æ920) and so model selection analyses were not conducted for this species.
The best-fit models for D. variabilis tick intensity were
aggregation + month and aggregation + month + sex
(Table 1; overdispersion parameter of best-fit model =
0.732). The combined weight of evidence for these two mod-
2010 The Authors. Journal compilation 2010 British Ecological Society, Journal of Animal Ecology, 79, 602–609
606 R. J. Monello & M. E. Gompper
male raccoons
(a)
No. of individual raccoons
20
Control without food
Control with food
Aggregated
15
10
5
0
10
30
40
female raccoons
(b)
40
No. of individual raccoons
20
Control without food
Control with food
Aggregated
30
20
food supplementation and age and sex of the host had no
effect on tick intensity when compared with aggregation and
month (Table 2).
Louse intensity was best explained by the aggregation + sex and sex-only models (Table 1; overdispersion
parameter of best-fit model = 1.348). All models with a
weight of evidence ‡0Æ01 included sex as a factor. The weight
of evidence that the aggregation + sex was the best model
was 0Æ54 and the sex-only model had a weight of 0Æ24.
Although food was included in some models, these consistently underperformed compared with models with aggregation (Tables 1 and 2). Model-averaged estimates indicated
that sex was the most important variable, with the intensity
of lice on females being 0Æ32–0Æ64 times that of males when
included in the model. Aggregation had the opposite relationship to that found with ticks, with non-aggregated populations having louse populations that were denser than that
on aggregated populations (Table 2).
10
Discussion
0
10
20
Number of lice
30
40
Fig. 3. The observed distribution of lice (Trichodectes octomaculatus)
among (a) male and (b) female raccoons. Equal sample sizes were
used to estimate kernel density curves for each treatment category by
randomly selecting the same number of animals within each sex and
category (n = 42 males, n = 42 females).
els was 0Æ71. No other model was within two AICc units of
these best-fit models, and all models with a weight of evidence
>0Æ01 included month. The best-fit model with food supplementation was the global model, which had a weight of evidence of 0Æ08. Model-averaged estimates and odds ratios
indicated that aggregation increased tick intensity, as tick
intensity in non-aggregated populations was 0Æ48–0Æ78 times
that of aggregated populations (Table 2). Tick intensity
increased from May to July, with maximum tick intensities
occurring in July (Table 2 and Fig. 1). The model factors
We observed large shifts in ectoparasite intensity because of
increased host contact and aggregation. The direction and
magnitude of these effects were parasite specific and inconsistent with our original predictions. Tick intensity increased to
maximum levels earlier in the season and was 1Æ5–2Æ5 times
higher throughout their seasonal occurrence on animals in
aggregated sites. Conversely, louse intensity was lower in
aggregated sites and fleas exhibited no differences among
treatments. Such findings suggest that the outcome of
increases in host contact or aggregation is highly variable
and depends on the host–parasite association in question.
Other studies on ecto- and endoparasites have also attributed
increases (Arneberg et al. 1998; Morand & Poulin 1998;
Krasnov et al. 2002; Wright & Gompper 2005) and declines
(Mooring & Hart 1992; Fauchald et al. 2007; Vicente et al.
2007) in macroparasite abundance or intensity to population
density and group size, but to our knowledge this is the first
study to examine such questions in an experimental manner
across ectoparasite taxa within the same system and host.
Table 1. Ranking of models estimating intensity of ticks (adult Dermacentor variabilis) and lice (Trichodectes octomaculatus) on raccoons from
sites subjected to aggregation and resource treatments
Model
Ticks (n = 295 raccoons)
Iaggregation + month
Iaggregation + month + sex
Iaggregation + month + age
Iaggregation + food + month + age + sex
Ifood + month + sex
Lice (n = 205 raccoons)
Iaggregation + sex
Isex
Ifood + sex
Log-likelihood
No. parameters
DAICc
AICc weight
)1270Æ31
)1269Æ87
)1268Æ87
)1266Æ82
)1271Æ23
6
7
9
11
7
0Æ00
1Æ21
3Æ46
3Æ65
3Æ93
0Æ46
0Æ25
0Æ08
0Æ08
0Æ07
)577Æ50
)579Æ33
)578Æ87
3
2
3
0Æ00
1Æ61
2Æ75
0Æ54
0Æ24
0Æ14
Rankings are based on generalized linear models with a negative binomial distribution and log identity, and only models in the 90% confidence
interval are shown.
2010 The Authors. Journal compilation 2010 British Ecological Society, Journal of Animal Ecology, 79, 602–609
Sociality of raccoons and ectoparasites 607
Table 2. Model-averaged estimates and odds ratios (lower, upper 95% CI) of parameters included in the 90% confidence set of models used to
estimate the intensity of ticks (adult Dermacentor variabilis) and lice (Trichodectes octomaculatus)
Parametera
Ticks (n = 295 raccoons)
No aggregation
April
May
June
July
Sex (female)
No food
Age I: 5–14 months
Age II: 15–38 months
Age III: 39–57 months
Lice (n = 205 raccoons)
No aggregation
Sex (female)
No food
Model-averaged
estimate
Unconditional
standard error
Odds ratio
(95% CI)
P-value
)0Æ42
0Æ18
0Æ75
0Æ80
0Æ97
)0Æ05
)0Æ06
)0Æ04
)0Æ03
)0Æ01
0Æ13
0Æ25
0Æ21
0Æ21
0Æ20
0Æ06
0Æ06
0Æ05
0Æ04
0Æ03
0Æ61 (0Æ48, 0Æ78)
1Æ20 (0Æ72, 2Æ01)
2Æ17 (1Æ42, 3Æ32)
2Æ29 (1Æ53, 3Æ44)
2Æ75 (1Æ88, 4Æ03)
0Æ89 (0Æ71, 1Æ13)
0Æ74 (0Æ53, 1Æ03)
0Æ75 (0Æ53, 1Æ08)
0Æ81 (0Æ59, 1Æ12)
0Æ91 (0Æ65, 1Æ28)
<0.001
0.260
<0.001
<0.001
<0.001
<0.001
0.078
0.158
0.257
0.586
0Æ16
)0Æ74
0Æ02
0Æ11
0Æ18
0Æ03
1Æ34 (0Æ99, 1Æ81)
0Æ45 (0Æ32, 0Æ64)
1Æ18 (0Æ84, 1Æ64)
0.054
<0.001
0.940
a
The parameters, aggregation, month (August), sex (male), food, and age (IV: >57 months), were redundant and set to 0 in model-averaged
estimates and 1 in odds ratios.
These findings indicate that even in the absence of density
alterations, host social behaviour, contact and resource distribution can alter the infestation levels of directly transmitted ectoparasites. Thus, the interaction between host contact
and ectoparasites appears to be considerably more complex
than the current understanding, which is primarily based on
models of endoparasite abundance (Anderson & May 1978;
Dobson 1990).
There are several possible reasons for the increase in tick
intensity in aggregated sites: ticks may seek out the established feeding areas; stationary raccoons may be more likely
to be found and infested by questing ticks; and raccoons may
be transporting ticks to the site where nymphs have a greater
probability of dropping off, moulting and later re-infesting
raccoons at the food plots. These hypotheses are non-exclusive and supported by other studies. Adult D. variabilis concentrate and focus questing behaviour in areas that exhibit
greater wildlife and human activity (Burg 2001), and can
move up to 100 m to reach such areas (Carroll & Nichols
1986). Dermacentor spp. are capable of detecting and seeking
out sources of CO2 (e.g. Garcia 1965) and are probably
attracted to CO2 plumes from the large, consistent host
aggregation at the food plots. Adult D. variabilis can overwinter or complete their entire life cycle in one summer in
Missouri (Kollars et al. 2000), which probably contributes to
the extended period of high infestation levels observed in
aggregated sites. The importance of aggregation relative to
other variables in this experiment suggests that changes in
host social behaviour or space use can overshadow other factors that typically influence tick intensity, such as host age or
sex (Monello & Gompper 2007). Higher tick intensities that
are maintained at their maximum level for a longer period of
time could also alter pathogen transmission between ticks
and hosts, with the direction and magnitude of effect dependent on whether the host is an efficient reservoir for the pathogen (Brunner & Ostfeld 2008).
The intensity of lice was lower on males in aggregated
sites, but no treatment differences were observed among
females. Fewer lice on males in aggregated sites was
because of greater overdispersion, which resulted in a larger proportion of hosts with sparser louse populations in
comparison with non-aggregated sites (Figs 2 and 3).
Smaller per capita parasitism in denser social groups (i.e.
parasite dilution) has been observed in other studies, but
these observations are primarily from ungulates and
mobile ectoparasites that have a negligible population
response, to greater host numbers (Côté & Poulin 1995;
Fauchald et al. 2007). Dilution has not been observed
among host-specific parasites that require close host contact for transmission and cannot survive off-host (Mooring & Hart 1992). Indeed, higher contact and aggregation
are predicted to increase the prevalence and intensity of
lice, which complete their life cycle within 30–40 days
(Durden 2001) and should benefit from increased transmission opportunities. However, dilution could be responsible for the patterns observed in this study if it results in
a greater number of smaller louse populations (because of
increased transmission) that are less likely to persist or be
detected with the methods used here.
It is surprising that treatment effects were sex specific.
Females may not have exhibited a decline in lice because of
initial lower intensity. Thus, the ability to cause a measurable
decline via dilution or any other route among female hosts is
reduced. It is also possible that treatment-induced alterations
in louse intensity among females were too small to detect with
the methods used here. Overall, the tick and louse burden of
males were affected by aggregation in a disproportionate
manner compared with that of females. Sex-biased use of the
food plots in aggregated sites might explain this, but several
lines of evidence suggest that this was not the case. The proportion of males and females captured was similar among
treatment categories, and a separate telemetry-based study
2010 The Authors. Journal compilation 2010 British Ecological Society, Journal of Animal Ecology, 79, 602–609
608 R. J. Monello & M. E. Gompper
found no sex-related differences in the use of food plots (M.
Wehtje, unpublished data).
No relationships were observed between fleas and treatment categories. One likely reason for this is that Sciurus spp.
are the primary hosts of O. howardi (Durden 1980) and alterations to raccoon ecology may have little or no impact on the
relatively low levels of prevalence and intensity observed in
this study. This is consistent with previous work from these
sites in the absence of experimental manipulations that found
peaked flea abundance during the spring but no other abiotic
or biotic correlates of flea infestation (Monello & Gompper
2009; see also Durden 1980). It is also possible that fleas primarily reside and infest hosts in dens and resting sites, and
thus increased social contact outside these settings has little
influence on flea burden.
Results from this study suggest that parasite life-history
characteristics – including route of transmission, ability to
survive and travel off-host, and host specificity – are critical
in predicting how ectoparasite populations will respond to
alterations in host demographics or behavioural ecology.
Models of directly transmitted endoparasites that predict a
positive relationship between parasite abundance and host
contact (Anderson & May 1978; Dobson 1990) have been
found to fit some host-flea interactions (Krasnov et al. 2002).
However, our results suggest that additional host-ectoparasite relationships need to be investigated prior to assuming
such relationships for other ectoparasites. Ectoparasites are
subjected to both environmental and host influences and
exhibit vastly different constraints between taxa in terms of
movement abilities, survival off-host, behaviour and routes
of reproduction. Yet, even when such characteristics are
taken into account, it is difficult to accurately predict the general response to host aggregation or contact rates. Ticks are
not reliant on host contact or aggregation for successful
infestation, yet in this study they displayed the greatest treatment-related differences, perhaps by their ability to seek out
hosts and a life cycle that includes bouts of host detachment
and attachment that probably moved them closer to aggregation sites. To elucidate such patterns among diverse host and
parasite types, we emphasize the need for more empirical
research on this topic in a manner that directly compares the
responses of multiple parasite species.
Acknowledgements
This work was supported by the NSF grant DEB-0347609. We thank B. Cook,
D. Kuschel, K. McBride, K. Mottaz, E. O’Hara, S. Pennington, K. Purnell, M.
Wehtje, A. Wiewel and J. Wisdom for assistance in the field and laboratory.
References
Alexander, R.D. (1974) The evolution of social behavior. Annual Review of
Ecology and Systematics, 5, 325–383.
Anderson, R.M. & May, R.M. (1978) Regulation and stability of host–parasite
population interactions. I. Regulatory processes. Journal of Animal Ecology,
47, 219–247.
Anderson, R.M. & May, R.M. (1991) Infectious Diseases of Humans: Dynamics
and Control. Oxford University Press, Oxford.
Anderson, R.M., Jackson, H.C., May, R.M. & Smith, A.M. (1981) Population
dynamics of fox rabies in Europe. Nature, 289, 765–771.
Arnason, A.N. & Schwarz, C.J. (1999) Using POPAN-5 to analyse banding
data. Bird Study, 46, S157–S168.
Arneberg, P., Skorping, A., Grenfell, B. & Read, A.F. (1998) Host densities as
determinants of abundance in parasite communities. Proceedings of the
Royal Society of London Series B, 265, 1283–1289.
Brunner, J.L. & Ostfeld, R.S. (2008) Multiple causes of variable tick burdens
on small-mammal hosts. Ecology, 89, 2259–2272.
Burg, J.G. (2001) Seasonal activity and spatial distribution of host seeking
adults of the tick Dermacentor variabilis. Medical and Veterinary Entomology, 15, 413–421.
Burnham, K.P. & Anderson, D.R. (2002) Model Selection and Multimodel
Inference: A Practical Information-Theoretic Approach. Springer-Verlag,
New York.
Bush, A.O., Lafferty, K.D., Lotz, J.M. & Shostak, A.W. (1997) Parasitology
meets ecology on its own terms: Margolis et al. revisited. Journal of Parasitology, 83, 575–583.
Caley, P. & Ramsey, D. (2001) Estimating disease transmission in wildlife, with
emphasis on leptospirosis and bovine tuberculosis in possums, and effects of
fertility control. Journal of Applied Ecology, 38, 1362–1370.
Carroll, J.F. & Nichols, J.D. (1986) Parasitization of meadow voles, Microtus
pennsylvanicus (Ord), by American dog ticks, Dermacentor variabilis (Say),
and adult tick movement during high host density. Journal of Entomological
Science, 21, 102–113.
Côté, I.M. & Gross, M.R. (1993) Reduced disease in offspring: a benefit of coloniality in sunfish. Behavioral Ecology and Sociobiology, 33, 269–274.
Côté, I.M. & Poulin, R. (1995) Parasitism and group-size in social animals–a
metaanalysis. Behavioral Ecology, 6, 159–165.
Craig, L.E., Norris, D.E., Sanders, M.L., Glass, G.E. & Schwartz, B.S. (1996)
Acquired resistance and antibody response of raccoons (Procyon lotor) to
sequential feedings of Ixodes scapularis (Acari: Ixodidae). Veterinary Parasitology, 63, 291–301.
Davies, C.R., Ayres, J.M., Dye, C. & Deane, L.M. (1991) Malaria infection
rate of Amazonian primates increases with body weight and group size.
Functional Ecology, 5, 655–662.
Dobson, A.P. (1990) Models for multi-species parasite–host communities. Parasite Communities: Patterns and Processes (eds G. Esch, A.O. Bush & J.M.
Aho), pp. 261–288. Chapman & Hall, London.
Duncan, P. & Vigne, N. (1979) The effect of group size in horses on the rate of
attacks by bloodsucking flies. Animal Behaviour, 27, 623–625.
Durden, L.A. (1980) Seasonal abundance of fleas (Siphonaptera) and sucking
lice (Anoplura) on gray squirrels, Sciurus carolinensis, in central Tennessee,
USA. Journal of Medical Entomology, 17, 483–484.
Durden, L.A. (2001) Lice (Phthiraptera). Parasitic Diseases of Wild Mammals
(eds W.M. Samuel, M.J. Pybus & A.A. Kocan), pp. 3–17. Iowa State University Press, Ames, IA.
Evans, R.H. (2002) Raccoons and relatives (Carnivora, Procyonidae). Zoological Restraint and Anesthesia (ed. D. Heard). International Veterinary Information Service, Ithaca, NY.http://www.ivis.org/special_books/Heard/
evans/ivis.pdf. Accessed January 2010.
Fauchald, P., Rødven, R., Bårdsen, B.-J., Langeland, K., Tveraa, T., Yoccoz,
N.G. & Ims, R.A. (2007) Escaping parasitism in the selfish herd: age, size
and density-dependent warble fly infestation in reindeer. Oikos, 116, 491–
499.
Franklin, A.B., Anderson, D.R., Gutièrrez, R.J. & Burnham, K.P. (2000) Climate, habitat quality, and fitness in northern spotted owl populations in
northwestern California. Ecological Monographs, 70, 539–590.
Freeland, W.J. (1976) Pathogens and the evolution of primate sociality. Biotropica, 8, 11–24.
Freeland, W.J. (1979) Primate social groups as biological islands. Ecology, 60,
719–728.
Garcia, R. (1965) Collection of Dermacentor andersoni (Stiles) with carbon
dioxide and its application in studies of Colorado tick fever virus. American
Journal of Tropical Medicine and Hygiene, 14, 1090–1093.
Gompper, M.E. (2004) Correlations of coati (Nasua narica) social structure
with parasitism by ticks and chiggers. Contribuciones mastozoológicas en
homenaje a Bernardo Villa (eds V. Sánchez-Cordero & R.A. Medellı́n), pp.
527–534. Instituto de biologı́a e Instituto de Ecologı́a, Universidad Nacional
Autónoma de México, México.
Gompper, M.E. & Wright, A.N. (2005) Increased prevalence of raccoon roundworm, (Baylisascaris procyonis) owing to manipulated contact rates of hosts.
Journal of Zoology, 266, 215–219.
Grau, G.A., Sanderson, G.C. & Rogers, J.P. (1970) Age determination in raccoons. Journal of Wildlife Management, 34, 364–372.
Hardin, J.W. & Hilbe, J.M. (2007) Generalized Linear Models and Extensions.
Stata Press, College Station, TX.
2010 The Authors. Journal compilation 2010 British Ecological Society, Journal of Animal Ecology, 79, 602–609
Sociality of raccoons and ectoparasites 609
Hochberg, M.E. (1991) Viruses as costs to gregarious feeding behaviour in the
Lepidoptera. Oikos, 61, 291–296.
Hoogland, J.L. (1979) Aggression, ectoparasitism, and other possible costs of
prairie dog (Sciuridae, Cynomys spp.) coloniality. Behaviour, 69, 1–35.
Kenward, R.E. (1985) Ranging behaviour and population dynamics in grey
squirrels. Behavioural Ecology. Ecological Consequences of Adaptive Behaviour (eds R.M. Sibly & R.H. Smith), pp. 319–330. Blackwell Scientific Publications, Oxford.
Keymer, A.E. & Read, A.F. (1991) Behavioural ecology: the impact of parasitism. Parasite–Host Associations (eds C.A. Toft & A. Aeschlimann), pp. 37–
61. Oxford University Press, Oxford.
Kollars, T.M., Jr, Oliver, J.H., Jr, Masters, E.J., Kollars, P.G. & Durden, L.A.
(2000) Host utilization and seasonal occurrence of Dermacentor species
(Acari: Ixodidae) in Missouri, USA. Experimental and Applied Acarology,
24, 631–643.
Krasnov, B., Khokhlova, I. & Shenbrot, G. (2002) The effect of host density on
ectoparasite distribution: an example of a rodent parasitized by fleas. Ecology, 83, 164–175.
Møller, A.P., Dufva, R. & Allander, K. (1993) Parasites and the evolution of
host social behavior. Advances in the Study of Behaviour, 22, 65–102.
Monello, R.J. (2009) Experimentally assessing the influence of resource availability and social aggregation on the parasites of raccoons. PhD dissertation, University of Missouri, Columbia, MO.
Monello, R.J. & Gompper, M.E. (2007) Biotic and abiotic predictors of tick
(Dermacentor variabilis) abundance and engorgement on free-ranging raccoons (Procyon lotor). Parasitology, 134, 2053–2062.
Monello, R.J. & Gompper, M.E. (2009) Relative importance of demographics,
locale, and seasonality underlying louse and flea parasitism of raccoons (Procyon lotor). Journal of Parasitology, 95, 56–62.
Moore, J., Simberloff, D. & Freehling, M. (1988) Relationships between bobwhite quail social-group size and intestinal helminth parasitism. American
Naturalist, 131, 22–32.
Mooring, M.S. & Hart, B.L. (1992) Animal grouping for protection from parasites: selfish herd and encounter-dilution effects. Behaviour, 123, 173–193.
Morand, S. & Poulin, R. (1998) Density, body mass and parasite species richness of terrestrial mammals. Evolutionary Ecology, 12, 717–727.
Porteous, I.S. & Pankhurst, S.J. (1998) Social structure of the mara (Dolichotis
patagonum) as a determinant of gastro-intestinal parasitism. Parasitology,
116, 269–275.
Poulin, R. (1991) Group-living and infestation by ectoparasites in passerines.
Condor, 93, 418–423.
Prange, S., Gehrt, S.D., Wiggers, E.P. & O’Shea, T.J. (2004) Influences of
anthropogenic resources on raccoon (Procyon lotor) movements and spatial
distribution. Journal of Mammalogy, 85, 483–490.
Rubenstein, D.I. & Hohmann, M.E. (1989) Parasites and social behavior of
island feral horses. Oikos, 55, 312–320.
Schulte-Hostedde, A.I., Zinner, B., Millar, J.S. & Hickling, G.J. (2005) Restitution of mass–size residuals: validating body condition indices. Ecology, 86,
155–163.
Tella, J.L. (2002) The evolutionary transition to coloniality promotes higher
blood parasitism in birds. Journal of Evolutionary Biology, 15, 32–41.
Vicente, J., Hoefle, U., Fernandez-De-Mera, I.G. & Gortazar, C. (2007) The
importance of parasite life history and host density in predicting the impact
of infections in red deer. Oecologia, 152, 655–664.
Wilson, K. & Grenfell, B.T. (1997) Generalized linear modelling for parasitologists. Parasitology Today, 13, 33–38.
Wilson, K., Thomas, M.B., Blanford, S., Doggett, M., Simpson, S.J. & Moore,
S.L. (2002) Coping with crowds: density-dependent disease resistance in desert locusts. Proceedings of the National Academy of Sciences, 99, 5471–5475.
Wright, A.N. & Gompper, M.E. (2005) Altered parasite assemblages in raccoons in response to manipulated resource availability. Oecologia, 144, 148–
156.
Received 19 August 2009; accepted 10 January 2010
Handling Editor: Andrew White
Supporting Information
Additional Supporting Information may be found in the online version of this article.
Table S1. A priori hypotheses and corresponding models of ectoparasite intensity of raccoons subjected to aggregation and resource treatments (b0 is the intercept and bi(X) are the parameters of independent
variables). Model selection was conducted for each ectoparasite species and included global and intercept models.
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials may be
re-organized for online delivery, but are not copy-edited or typeset.
Technical support issues arising from supporting information (other
than missing files) should be addressed to the authors.
2010 The Authors. Journal compilation 2010 British Ecological Society, Journal of Animal Ecology, 79, 602–609