Ethology 110, 851—862 (2004)

2004 Blackwell Verlag, Berlin

ISSN 0179–1613

Selective Defecation and Selective Foraging: Antiparasite Behavior in

Wild Ungulates?

Vanessa O. Ezenwa
Department of Ecology and Evolutionary Biology, Princeton University, Princeton,
NJ, USA

Abstract
Selective defecation and selective foraging are two potential antiparasite
behaviors used by grazing ungulates to reduce infection by fecal–oral transmitted
parasites. While there is some evidence that domestic species use these strategies,
less is known about the occurrence and efficacy of these behaviors in wild
ungulates. In this study, I examined whether wild antelope use selective defecation
and selective foraging strategies to reduce exposure to gastrointestinal nematode
parasites. By quantifying parasite levels in the environment in relation to the
defecation patterns of three species, dik-dik (Madoqua kirkii), Grant’s gazelle
(Gazella granti), and impala (Aepyceros melampus), I found that nematode larval
concentrations in pasture were higher in the vicinity of clusters of feces (dung
middens) compared to single fecal pellet groups or dung-free areas. In addition,
experimental feeding trials in free-ranging dik-dik showed that individuals
selectively avoided feeding near concentrations of feces. Given that increased
parasite contamination was found in the immediate vicinity of fecal clusters, fecal
avoidance could help reduce host consumption of parasites and may therefore be
an effective antiparasite behavior for certain species. On the other hand, while the
concentration of parasite larvae in the vicinity of middens coupled with host
avoidance of these areas during grazing could reduce host contact with parasites,
results showing a positive correlation between the number of middens in a habitat
and larval abundance at control sites suggest that dung middens might increase
and not decrease overall host exposure to parasites. If this is the case, dung
midden formation may not be a viable antiparasite strategy.
Correspondence: Vanessa O. Ezenwa, US Geological Survey, 521 National
Center, Reston, VA 20192, USA. E-mail: vezenwa@usgs.gov

Introduction
Parasites can exact extensive costs on their hosts including reductions in
growth and fecundity and even death. In response, hosts have developed many

U. S. Copyright Clearance Center Code Statement: 0179-1613/2004/11011–851/$15.00/0 www.blackwell-synergy.com

852 V. O. Ezenwa

sophisticated methods of combating parasitic invasion. While immunological

defenses are the most commonly cited parasite defense strategies, behavioral
defenses against parasites also occur widely and may in fact be the first line of
defense for many hosts (Hart 1990, 1994; Moore 2002). Antiparasite behaviors
can range from commonplace behaviors that remove or repel parasites such as
grooming, fly-repelling and comfort movements (see Hart 1994 and Moore 2002
for review) to fairly complex behaviors that alleviate or prevent infection such as
leaf swallowing in chimpanzees (Huffman et al. 1996) and cannibalism avoidance
in larval salamanders (Pfennig et al. 1998).
Hart (1992) outlined two criteria that must be met for a behavior to be
considered a viable antiparasite strategy. First, the parasite involved must have
negative fitness effects on the host, and second, the relevant behavior must
effectively reduce or eliminate the parasite. Selective defecation and selective
foraging are two behaviors that have been discussed as potential defense strategies
used by grazing ungulates to reduce infestation by fecal–oral transmitted parasites
(Hart 1990, 1992). While both behaviors are known to occur in domestic animals,
evidence that they actually reduce levels of parasitism is scant. For example,
domestic horses create permanent defecation sites and it has been suggested that
this selective clustering of fecal eliminations in space serves to sequester parasites
allowing subsequent avoidance during foraging (Taylor 1954; Ödberg & Francis-
Smith 1976). Similar observations have also been made in wild primates (Gilbert
1997), but direct links between selective defecation and reduced parasitism are
lacking in both groups. Selective foraging in the form of fecal avoidance has been
documented in cattle (Taylor 1954), sheep (Crofton 1957; Cooper et al. 1977,
2000; Hutchings et al. 1998), horses (Ödberg & Francis-Smith 1977), and captive
reindeer (Moe et al. 1999). By selectively avoiding grazing near feces, animals are
thought to reduce consumption of parasites, thereby lowering their infection
rates. Unlike selective defecation, selective foraging has been associated with
reduced parasite intake (Michel 1955; Cooper et al. 2000), which suggests that this
behavior may serve as an effective antiparasite strategy for grazing ungulates.
Despite the many examples of selective foraging and defecation in domestic
animals, there is almost no information about the extent to which these behaviors
may be used by wild ungulates as antiparasite strategies. The general assumption
seems to be that wild ungulates use similar behaviors to control parasitic
infections (e.g. Hart 1990; Møller et al. 1993). The fact that many wild ungulates
deposit feces in heavily used dung middens, for example, indicates that selective
defecation does occur. However, because middens often serve an important scent-
marking role for many species (Leuthold 1977; Estes 1991), an antiparasite
function of this behavior cannot be presupposed. Furthermore, since there has
been no clear demonstration of fecal avoidance in a wild ungulate, it remains
unknown whether this form of selective foraging, which is relatively widespread
among domestic species, even occurs in the wild.
To understand more about antiparasite behavior in wild ungulates, I studied
selective foraging and defection behaviors in African antelope in relation to
gastrointestinal nematode parasitism. Gastrointestinal nematodes are common
 Antiparasite Behavior in Ungulates 853
fecal–oral transmitted parasites of ruminants, and in this study I focused on the
ÔstrongyleÕ nematodes (Order: Strongylida; Superfamily: Trichostrongyloidea), a
group which includes highly pathogenic genera such as Haemonchus, Ostertagia,
Oesophagostomum, Trichostrongylus, and Cooperia. In the typical life cycle of
these parasites, eggs are shed in the feces and hatch within 1–2 days as larvae.
First and second stage larvae feed on microorganisms in the feces and within days
to weeks infective third stage larvae migrate out of the fecal mass onto
surrounding vegetation to await ingestion by a susceptible host (Bowman 1999).
Strongyle nematodes cause significant morbidity and mortality among domestic
livestock worldwide, and the negative effects of these parasites on production in
domestic systems are well-documented (Sykes 1994; Urquhart et al. 1996).
Increasing evidence also suggests that these parasites negatively impact the
condition, fecundity, and survival of wild ungulates (Gulland 1992; Stein et al.
2002).
Given the potential fitness costs of strongyle nematode parasitism and the
possible benefits a host could gain from reducing exposure to these parasites, I
examined whether selective defecation and foraging serve as important defense
strategies in three antelope species: dik-dik (Madoqua kirkii), Grant’s gazelle
(Gazella granti), and impala (Aepyceros melampus). Since all three species are
territorial and use fecal scent-marking to varying degrees to demarcate territory
ownership (Estes 1991), I investigated whether this form of selective defecation
reduces host exposure to strongyle nematodes by concentrating infective larvae
around dung middens and facilitating parasite avoidance during foraging. In
addition, since it is unknown whether free-ranging antelope engage in selective
foraging (fecal avoidance), I also tested whether individuals selectively avoid
foraging near large concentrations of feces in order to reduce parasite intake rates.

Methods
Study Location and Species

This study was conducted at the Mpala Research Center, Kenya (0017¢N,
3653¢E). The Center is located in the semi-arid region of central Kenya in the
Laikipia district. Annual rainfall at Mpala ranges from 400 to 500 mm, and the
vegetation is dominated by Acacia bushland/grassland. Approximately 20 large
herbivore species occur at Mpala, 15 of which are bovids. The three study species,
dik-dik, Grant’s gazelle and impala, are among the most numerous bovid species
at the study site. Dik-dik live as monogamous pairs. Both males and females are
territorial and both sexes deposit feces in dung middens (Estes 1991). Grant’s
gazelle and impala are polygynous and individuals can be divided into three
distinct classes: territorial males, non-territorial or bachelor males, and females
and juveniles. Only territorial males actively defend territories and they create and
maintain dung middens within their territory boundaries (Estes 1991). In all three
species, territoriality and midden use occur year round at the study site (Ezenwa,
pers. obs.).
854 V. O. Ezenwa

All three study species have been found to be infected with strongyle
parasites at the study site. Adult worms collected and identified from study
animals included the genera Agriostomum, Cooperia, Cooperioides, Gazello-
strongylus, Haemonchus, Oesophagostomum, Ostertagia, and Longistrongylus
(Ezenwa 2003). Mean strongyle fecal egg counts measured over an 18-month
period ranged from 843 ± 90 eggs/g feces (EPG) in dik-dik to 963 ± 39 EPG in
impala and 2560 ± 97 EPG in Grant’s gazelle (Ezenwa 2003). For comparison,
egg counts between 600 and 2000 EPG are considered to be indicative of
moderate (4000–10 000) worm burdens in domestic sheep, and counts over 2000
EPG are considered indicative of high (>10 000) worm burdens (McKenna
1987). A sheep with a burden of 5000 Haemonchus contortus worms can lose up to
250 ml of blood per day resulting in anemia, weight loss, and weakness (Urquhart
et al. 1996). If these parasites have similar effects on wild antelope, then the
parasite loads observed for the study species could have important fitness
consequences.

Defecation Patterns and Parasite Abundance

To determine the relationship between dung midden use and strongyle
abundance in the external environment, I performed strongyle larval counts on
pasture vegetation surrounding different dung formations. Counts were done on
vegetation surrounding three different feces formations: dung middens (defined as
fecal clusters with four or more individual fecal pellet groups deposited on top of
or adjacent to each other), single fecal pellet groups, and control areas (areas at
least 1.0 m away from any dung pellets). One midden/single/control set of larval
counts was done in each of three distinct territories for Grant’s gazelle and
impala, and in each of four territories for dik-dik. For Grant’s gazelle and impala,
locations of territories were determined by the presence and behavior of territorial
males. Sampling locations within territories were selected by locating active dung
middens in areas heavily used by territory occupants (Ezenwa, pers. obs.). These
middens were selected as the midden sites, and then appropriate single and control
sites were selected in the same vicinity. For dik-dik, the locations of territories
were determined by mapping the locations of dik-dik pairs along a road in a
densely occupied habitat (Ezenwa, unpubl. data). Clusters of sightings were
considered to be likely territories and these areas were searched for dung middens.
The distribution of middens was then used to define probable territories. Midden
sites used for sampling were selected based on the frequency with which they
contained fresh feces. Because dik-dik rarely produce single-pellet groups
(Ezenwa, pers. obs.), larval counts were done exclusively on middens and
controls for this species. Control sites were selected as described above. For all
species, sampling locations (midden, single, or control) within each territory were
no more than 20 m apart to control for differences in vegetation structure that
might affect larval dispersal. All larval counts were completed between May and
August 2001.
 Antiparasite Behavior in Ungulates 855
To sample nematode larvae on vegetation, grass clippings were taken from a
half-meter circle around each sampling point (midden, single-pellet group or
control area) until a sandwich-sized (16.51 cm · 8.25 cm) Ziploc bag was filled.
All grass clippings were collected between 09:00 and 11:00 h, soaked in 1.0 l of
water with detergent (Tween 80) for 24 h and then nematode larvae were
extracted following Lancaster (1970). Lugol’s iodine was added to the larval
suspension after extraction and all third stage infective (L3) strongyle nematode
larvae were counted using a binocular microscope at 40· magnification. After the
extractions, all grass clippings were air-dried and weighed to calculate L3/kg dry
weight herbage for each sampling location.

Midden Use Frequencies

At each Grant’s gazelle and impala territory where nematode larvae counts
were performed, I also estimated the proportion of fecal pellet groups deposited in
middens in that territory. Based on behavioral observations of territory
inhabitants, I determined the areas of each territory which were most heavily
used and then surveyed fecal pellet groups on the ground in these areas. Surveys
began in the center of the heavily used sites and extended outward until at least 10
distinct fecal pellet formations were located. I classified all fecal pellet formations
encountered during each survey as being either a midden (‡ 4 pellet groups) or
non-midden (1–3 pellet groups), and then used these counts to calculate a midden
use frequency for each territory: frequency ¼ number of middens encountered/
total number of fecal pellet formations encountered. Between 10 and 25 fecal
pellet formations were used to calculate midden use frequencies for each territory.
Midden use was not estimated for dik-dik because the overwhelming majority of
fecal pellet groups for this species occur in middens.

Fecal Avoidance Experiments

To investigate the occurrence of fecal avoidance behavior in wild ungulates, I
performed feeding trials on free-ranging dik-dik. Trials consisted of placing two
piles with three cups each of calf feed (composition: lucerne, bran, low-grade
maize, wheat) 1.0 m apart near heavily used dik-dik walkways. One pile of feed
was designated as a Ôno dungÕ treatment and the other as a ÔdungÕ treatment. The
ÔdungÕ treatment had two cups of completely dried dik-dik fecal pellets placed at
three points adjacent to, but not mixed in with the feed, while nothing was done to
the Ôno dungÕ treatment. The experimental site was then watched for up to 10
continuous hours to monitor all visits by dik-dik. The duration of all feeding
bouts (designated as total time spent by an individual near the experiment site
after at least one bite had been taken, but before moving off) and the number of
bites taken from either treatment during a feeding bout were recorded, as were the
age, sex, and identity of the feeder when possible. For each trial, after the first 50
bites were taken from either treatment, the ÔdungÕ and Ôno dungÕ sides were
reversed to control for possible side bias, and observations continued until at least
856 V. O. Ezenwa

50 more bites were recorded in the new configuration. All feeding trials were
conducted between 7 February and 4 May 2000. The feeding behaviors of at least
seven individually distinguishable dik-dik were recorded during the trials, but the
actual number of distinct individuals involved is unknown because many were not
individually identifiable.

Statistical Analysis
I used statview 5 for Windows (SAS Institute, Cary, NC, USA) for all
statistical comparisons, and significance was accepted at p £ 0.05 for all tests. To
compare nematode larval counts around different feces types, I used an analysis of
variance (anova) corrected for multiple comparisons using the Fisher’s Protected
Least Significant Difference (PLSD) test. Counts were log transformed for the
analysis. To test for associations between midden use frequencies and larval
abundance at control sites, I calculated a standardized larval abundance for all
Grant’s gazelle and impala territories [e.g. Grant’s gazelle, site 1 ¼ (L3/g dry
weight)/(Max L3/g dry weight) · 100] and compared this to midden use frequency
using a Spearman’s rank correlation test. Lastly, I used a paired t-test to test for
fecal avoidance in dik-dik.

Results
Dung Formation and Larval Abundance
Estimates of infective larvae abundance on vegetation surrounding dung
middens, single-pellet groups, and control areas indicate that larval abundance
varied with dung formation. For Grant’s gazelle feces there was significant
variation in abundance of infective larvae across dung types (anova: F2,6 ¼ 6.1,
p ¼ 0.04; Fig. 1). Middens produced significantly more larvae than did single-
pellet groups (Fisher’s PLSD: p ¼ 0.02) or control areas (p ¼ 0.03), but there was
no difference in larval abundance around single-pellet groups and control areas
(p ¼ 0.87). For impala, differences across all dung types were only marginally
significant (anova: F2,6 ¼ 3.6, p ¼ 0.09), however middens did produce signifi-
cantly more larvae than did controls (Fisher’s PLSD: p ¼ 0.04). There was no
difference between middens and singles (p ¼ 0.13), or singles and controls (p ¼
0.40; Fig. 1). Dik-dik dung middens also produced significantly more larvae than
did control areas (anova: F1,6 ¼ 11.1, p ¼ 0.02; Fig. 1).

Midden Use and Larval Abundance

When midden use frequencies within Grant’s gazelle and impala territories
were compared against larval abundances at control sites, standardized larval
abundance was positively correlated with midden frequency (Spearman rank:
rs ¼ 0.93, p ¼ 0.04; Fig. 2).
 Antiparasite Behavior in Ungulates 857
control
single
midden
4 b
3.5 b
b
Log nematode larvae/kg
a a a
3 a
2.5
2
1.5
1
0.5
0
dik– dik Grant’s gazelle impala
Species

Fig. 1: Infective nematode larvae ± SE on vegetation surrounding control, midden, and single-fecal
pellet group sites for dik-dik, Grant’s gazelle and impala fecal pellets. Letters (a vs. b) indicate
significant differences among treatments within a species

100
90
Standardized larval abundance

80
70
at control sites

60
50
40
30
20
10
0
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Midden frequency in territory

Fig. 2: Relationship between the proportion of middens encountered in Grant’s gazelle and impala
territories and standardized larval abundance measured at control sites (Spearman’s rank correlation:
rs ¼ 0.93, p ¼ 0.04)

Fecal Avoidance Experiments

Dik-dik feeding trials showed that individuals preferentially avoided feces
when feeding. Individuals presented with a choice of calf feed in the vicinity of
858 V. O. Ezenwa

either ÔdungÕ or Ôno dungÕ took significantly more bites from the Ôno dungÕ
treatment (Paired t-test: t ¼ )3.57, df ¼ 30, p ¼ 0.001; Fig. 3).

Discussion
For African antelope, midden formation is thought to serve an important
scent marking function (Leuthold 1977). Dung middens have also been
hypothesized as having an antiparasite function (see Hart 1990 for review)
because they potentially help sequester parasite infective stages to limited areas of
pasture reducing host exposure to parasites. In the current study, larval
abundance counts around different feces formations showed that dung middens
do have higher infective larvae concentrations in their immediate vicinity when
compared to single fecal pellet groups and control areas. However, it is unclear
whether these high concentrations of larvae actually remain confined to midden
sites. The observed positive correlation between the proportion of dung middens
in a host territory and larval abundance at control sites suggests that middens
may somehow contribute to larval contamination in surrounding areas of pasture.
Although this result is correlational and other unexplored factors may influence
the observed relationship, the idea that dung middens could increase pasture
larval concentrations on a broader scale is feasible under certain conditions.

3.5
Number of bites taken

3
per feeding bout

2.5

1.5

0.5

0
Dung No dung
Treatment

Fig. 3: Number of bites ± SE taken by dik-dik given a choice between ÔdungÕ and Ôno dungÕ food
sources. Significantly more bites were taken from Ôno dungÕ treatments (p ¼ 0.001)
 Antiparasite Behavior in Ungulates 859
Because high concentrations of feces and moist conditions inside middens
might foster the production of large numbers of infective larvae, dung middens
could act as important source populations of infective larvae. While under most
conditions larvae actively disperse only a maximum distance of about 1 m (Durie
1961), during wet periods, rain can facilitate the passive dispersal of larvae on a
much wider scale. In tropical climates, pasture larvae levels have been shown to be
highest during the rainy season when larval dispersal from fecal pats is aided by
rain (e.g. Waruiru et al. 1998, 2001). Similarly, rain could be an important
mechanism moving large numbers of larvae from midden hotspots to surrounding
areas of a host territory. The period during which the larval counts were
performed in this study (May–August 2001) was fairly wet at the study site (see
Ezenwa 2004a), thus it is possible that rain dispersal could account for the
increased larval abundance at control sites in those territories that had more
middens. The fact that any larvae were detected at control sites, which were all at
least 1 m away from any visible fecal pellets, suggests that active dispersal alone
cannot account for the observed dispersion of infective larvae at the study
locations. However, since the current study did not include a spatial analysis of
dung middens in relation to larval abundance at the controls, additional work is
needed to determine if, and to what extent, dung middens affect pasture larval
counts on a broad scale.
If dung middens do contribute to larval contamination of non-midden areas
as preliminary results suggest, then midden formation is unlikely to be an effective
antiparasite strategy. In fact, increased parasitism may be a cost of midden use,
and as such, midden-making species might suffer from higher infection rates than
species that do not create middens. Results of an associated study comparing
strongyle nematode infection rates in territorial and non-territorial bovids
support this hypothesis. The study showed that territorial species, a majority of
which form dung middens, had significantly higher strongyle infection rates than
non-territorial species that do not create middens (Ezenwa 2004b). Furthermore,
territorial male Grant’s gazelles were found to have significantly higher strongyle
egg counts than did either non-territorial males or females, possibly because these
males spend more time on territories and in closer proximity to dung middens
(Ezenwa 2004b). Both of these findings support the idea that midden-making and
increased exposure to dung middens may be associated with higher infection rates.
If this is the case, then parasitism could be a major factor limiting the use of dung
middens in territorial antelope. This could explain why, in species like dik-dik
where both males and females defecate almost exclusively in middens, dung
middens are generally placed only on the periphery of territories whereas
glandular secretions are used for scent-marking in more central locations (see
Hendrichs 1975) despite the fact that these secretions are potentially more costly
to produce. To gain a better understanding of the potential costs of midden
formation, future studies will explore whether there is in fact a causal relationship
between midden use frequency within a territory, overall larval abundance in the
environment, and host infection rates. An examination of the relative costs and
benefits of midden use will also shed light on the functions of this behavior.
860 V. O. Ezenwa

Selective foraging is a second potential antiparasite strategy used by grazing

ungulates to reduce parasitism. Several studies show that domestic sheep use fecal
avoidance as a cue for parasite avoidance (Hutchings et al. 1998; Cooper et al.
2000), which suggests that selective foraging in relation to dung serves a distinct
antiparasite function in these animals. In this study, free-ranging dik-dik exhibited
fecal avoidance when given a choice between food placed near clusters of dung
and dung-free food. Because gastrointestinal parasites can have extensive negative
effects on their hosts (e.g. Gulland 1992; Urquhart et al. 1996; Stein et al. 2002)
and these negative effects increase with increasing parasite loads, any reduction in
the number of parasites ingested probably directly influences host fitness. Since
nematode larval counts surrounding dik-dik dung middens were significantly
higher than larval counts in control areas, the avoidance of clusters of dung by
individuals in this species could serve to significantly reduce consumption of
infective parasite stages. Selective foraging behavior has been shown to reduce
parasite intake rates in both domestic sheep (Cooper et al. 2000) and cattle
(Michel 1955), thus it is possible that similar behavior in dik-dik also reduces
parasite intake rates.
Even if fecal avoidance is an effective antiparasite strategy used by some wild
ungulates, there are significant costs associated with this behavior that could lead
to individual- and species-level variation in responses to feces. Because selective
foraging may, under certain conditions, reduce host nutrient intake rates or
increase total foraging time necessary to fulfill nutritional requirements, the costs
of avoidance counterbalanced by the benefits in terms of parasite reduction can
result in foraging trade-offs between nutrient intake and parasite avoidance
(Hutchings et al. 1999, 2000, 2001a, 2001b). Since adequate host nutrition is
critical to parasite resistance and resilience (Van Houtert & Sykes 1996), and
because parasites themselves disrupt host metabolism and digestive efficiency
(Coop & Kyriazakis 2001), an animal’s physiological state, including current
nutritional status, existing parasite load and immune function are all important
determinants of foraging selectivity (Hutchings et al. 1998, 1999, 2000, 2001a,
2001b). Studies on domestic sheep, for example, show that parasitized sheep are
more selective than non-parasitized sheep; highly food motivated sheep are less
selective than unmotivated sheep; and immune sheep are less selective than non-
immune sheep (Hutchings et al. 1998, 1999, 2000, 2001a, 2001b). For wild
ungulates, the opportunity costs of selective foraging can be very high, especially
when resources are limited or predators are abundant. As a consequence,
differences in immune function, nutritional status and parasitism, as well as a
variety of ecological factors will contribute to variation in the expression of fecal
avoidance behavior.
Differences in speciesÕ defecation behaviors will also influence fecal avoidance
behavior since defecation patterns directly affect the distribution of parasites in
the environment. Species that form dung middens may be more likely to engage in
fecal avoidance behaviors if they run a higher risk of infection. Also, if rainfall
does play an important role in dispersing larvae away from dung middens, then
selective foraging behavior may vary with season and may also occur on a larger
 Antiparasite Behavior in Ungulates 861
scale than was examined in the current study. Future work will investigate the
existence of fecal avoidance behavior in a wider range of species, examine host
foraging behavior in relation to feces and larval distributions at both fine (patch-
level) and coarse (territory-level) scales, and explore relationships between
selective foraging behavior and host infection rates. This will help improve our
understanding of the behavioral mechanisms wild animals use to combat
parasitism.

Acknowledgements
I thank the Mpala Research Center staff for logistical support and the Office of the President of
Kenya for permitting this research to be conducted in Kenya. D. Rubenstein and three anonymous
reviewers provided valuable comments on earlier versions of this manuscript. This work was supported
by a NSF pre-doctoral fellowship, a Fulbright fellowship, an EPA STAR graduate fellowship, and the
Department of Ecology and Evolutionary Biology, Princeton University.

Literature Cited
Bowman, D. D. 1999: GeorgisÕ Parasitology for Veterinarians, 7th edn. W. B. Saunders, Philadelphia,
PA.
Coop, R. L. & Kyriazakis, I. 2001: Influence of host nutrition on the development and consequences of
nematode parasitism in ruminants. Trends Parasitol. 17, 325—330.
Cooper, J., Gordon, I. J. & Pike, A. W. 2000: Strategies for the avoidance of faeces by grazing sheep.
Appl. Anim. Behav. Sci. 69, 15—33.
Crofton, H. D. 1957: Nematode parasite populations in sheep on lowland farms. VI. Sheep behavior
and nematode infections. Parasitology 48, 251—260.
Durie, P. H. 1961: Parasitic gastro-enteritis of cattle: the distribution and survival of infective strongyle
larvae on pasture. Aust. J. Agr. Res. 12, 1200—1211.
Estes, R. D. 1991: The Behavior Guide to African Mammals. Univ. California Press, Berkeley, CA.
Ezenwa, V. O. 2003: Habitat overlap and gastrointestinal parasitism in sympatric African bovids.
Parasitology 126, 379—388.
Ezenwa, V. O. 2004a: Interactions among host diet, nutritional status and gastrointestinal parasite
infection in wild bovids. Int. J. Parasitol. 34, 535—542.
Ezenwa, V. O. 2004b: Host social behavior and parasitic infection: a multifactorial approach. Behav.
Ecol. 15, 446—454.
Gilbert, K. A. 1997: Red howling monkey use of specific defecation sites as a parasite avoidance
strategy. Anim. Behav. 54, 451—455.
Gulland, F. M. D. 1992: The role of nematode parasites in Soay sheep (Ovis aries L.) mortality during
a population crash. Parasitology 105, 493—503.
Hart, B. L. 1990: Behavioral adaptations to pathogens and parasites: 5 strategies. Neurosci. Biobehav.
Rev. 14, 273—294.
Hart, B. L. 1992: Behavioral adaptations to parasitism: an ethological approach. J. Parasitol. 78,
256—265.
Hart, B. L. 1994: Behavioral defense against parasites: interaction with parasite invasiveness.
Parasitology 109, S139—S151.
Hendrichs, H. 1975: Changes in a population of Dikdik, Madoqua (Rhynchotragus) kirki (Gunther
1880). Z. Tierpsychologie 38, 55—69.
Huffman, M. A., Page, J. E., Sukhdeo, M. V. K., Gotoh, S., Kalunde, M. S., Chandrasiri, T. &
Towers, G. H. N. 1996: Leaf-swallowing by chimpanzees: a behavioral adaptation for the control
of strongyle nematode infections. Int. J. Primatol. 17, 475—503.
Hutchings, M. R., Kyriazakis, I., Anderson, D. H., Gordon, I. J. & Coop, R. L. 1998: Behavioral
strategies used by parasitized and non-parasitized sheep to avoid ingestion of gastro-intestinal
nematodes associated with faeces. Anim. Sci. 67, 97—106.
862 V. O. Ezenwa

Hutchings, M. R., Kyriazakis, I., Gordon, I. J. & Jackson, F. 1999: Trade-offs between nutrient intake
and fecal avoidance in herbivore foraging decisions: the effect of animal parasitic status, level of
feeding motivation and sward nitrogen content. J. Anim. Ecol. 68, 310—323.
Hutchings, M. R., Kyriazakis, I., Papachristou, T. G., Gordon, I. J. & Jackson, F. 2000: The
herbivoresÕ dilemma: trade-offs between nutrition and parasitism in foraging decisions. Oecologia
124, 242—251.
Hutchings, M. R., Gordon, I. J., Kyriazakis, I. & Jackson, F. 2001a: Sheep avoidance of faeces-
contaminated patches leads to a trade-off between intake of forage and parasitism in subsequent
foraging decisions. Anim. Behav. 62, 955—964.
Hutchings, M. R., Kyriazakis, I. & Gordon, I. J. 2001b: Herbivore physiological state affects foraging
decisions between nutrient uptake and parasite avoidance. Ecology 82, 1138—1150.
Lancaster, M. B. 1970: The recovery of infective nematode larvae from herbage samples. J. Helminthol.
44, 219—230.
Leuthold, W. 1977: African Ungulates: A Comparative Review of their Ethology and Behavioral
Ecology. Springer-Verlag, Berlin.
Møller, A. P., Dufva, R. & Allander, K. 1993: Parasites and the evolution of host social behavior. Adv.
Stud. Behav. 22, 65—102.
McKenna, P. B. 1987: The estimation of gastrointestinal worm burdens in young sheep flocks: a new
approach to the interpretation of faecal egg counts. I. Development. N. Z. Vet. J. 35, 94—97.
Michel, J. F. 1955: Parasitological significance of bovine grazing behaviour. Nature 175.
Moe, S. R., Holand, O., Colman, J. E. & Reimers, E. 1999: Reindeer (Rangifer tarandus) response to
feces and urine from sheep (Ovis aries) and reindeer. Rangifer 19, 55—60.
Moore, J. 2002: Parasites and the Behavior of Animals. Oxford Univ. Press, New York.
Ödberg, F. O. & Francis-Smith, K. 1976: A study on eliminative and grazing behavior-the use of the
field by captive horses. Equine Vet. J. 8, 147—149.
Ödberg, F. O. & Francis-Smith, K. 1977: Studies on the formation of ungrazed eliminative areas in
fields used by horses. Appl. Anim. Ethol. 3, 27—34.
Pfennig, D. W., Ho, S. G. & Hoffman, E. A. 1998: Pathogen transmission as a selective force against
cannibalism. Anim. Behav. 58, 1255—1261.
Stein, A., Irvine, R. J., Ropstad, E., Halvorsen, O., Langvatn, R. & Albon, S. D. 2002: The impact of
gastrointestinal nematodes on wild reindeer: experimental and cross-sectional studies. J. Anim.
Ecol. 71, 937—945.
Sykes, A. R. 1994: Parasitism and production in farm animals. Anim. Prod. 59, 155—172.
Taylor, E. L. 1954: Grazing behaviour and helminthic disease. Br. J. Anim. Behav. 2, 61—62.
Urquhart, G. M., Armour, J., Duncan, J. L., Dunn, A. M. & Jennings, F. W. 1996: Veterinary
Parasitology, 2nd edn. Blackwell Science, Oxford, UK.
Van Houtert, M. F. J. & Sykes, A. R. 1996: Implications of nutrition for the ability of ruminants to
withstand gastrointestinal nematode infections. Int. J. Parasitol. 26, 1151—1168.
Waruiru, R. M., Munyua, W. K., Thamsborg, S. M., Nansen, P., Bogh, H. O. & Gathuma, J. M. 1998:
Development and survival of infective larvae of gastrointestinal nematodes of cattle on pasture in
Central Kenya. Vet. Res. Commun. 22, 315—323.
Waruiru, R. M., Thamsborg, S. M., Nansen, P., Kyvsgaard, N. C., Bogh, H. O., Munyua, W. K. &
Gathuma, J. M. 2001: The epidemiology of gastrointestinal nematodes of dairy cattle in Central
Kenya. Trop. Anim. Health Prod. 33, 173—187.

Received: March 3, 2004

Initial acceptance: May 12, 2004

Final acceptance: June 26, 2004 (S. K. Sakaluk)