Fatal attraction in rats infected with Toxoplasma gondii.

Abstract

We tested the hypothesis that the parasite Toxoplasma gondii manipulates the behaviour of its intermediate rat host in order to increase its chance of being predated by cats, its feline definitive host, thereby ensuring the completion of its life cycle. Here we report that, although rats have evolved anti-predator avoidance of areas with signs of cat presence, T. gondii's manipulation appears to alter the rat's perception of cat predation risk, in some cases turning their innate aversion into an imprudent attraction. The selectivity of such behavioural changes suggests that this ubiquitous parasite subtly alters the brain of its intermediate host to enhance predation rate whilst leaving other behavioural categories and general health intact. This is in contrast to the gross impediments frequently characteristic of many other host parasite systems. We discuss our results in terms of their potential implications both for the epidemiology of toxoplasmosis and the neurological basis of anxiety and cognitive processes in humans and other mammals.

Full text

Page 1

Fatal attraction in rats infected with
Toxoplasma gondii
M. Berdoy1,3*, J. P. Webster2and D. W. Macdonald3
1Oxford UniversityVeterinary Services, Oxford OX1 3PT , UK
2WellcomeTrust Centre for the Epidemiology of Infectious Disease, University of Oxford, Oxford OX13FY, UK
3Wildlife Conservation Research Unit, Department of Zoology, University of Oxford, Oxford OX1 3PS, UK
We tested the hypothesis that the parasiteT oxoplasma gondii manipulates the behaviour of its intermediate
rat host in order to increase its chance of being predated by cats, its feline de¢nitive host, thereby ensuring
the completion of its life cycle. Here we report that, although rats have evolved anti-predator avoidance
of areas with signs of cat presence, T.gondii’s manipulation appears to alter the rat’s perception of cat
predation risk, in some cases turning their innate aversion into an imprudent attraction.The selectivity of
such behavioural changes suggests that this ubiquitous parasite subtly alters the brain of its intermediate
host to enhance predation rate whilst leaving other behavioural categories and general health intact.This
is in contrast to the gross impediments frequently characteristic of many other host^parasite systems. We
discuss our results in terms of their potential implications both for the epidemiology of toxoplasmosis and
the neurologicalbasis of anxiety and cognitive processes in humans and other mammals.
Keywords: Rattus norvegicus;T oxoplasma gondii; parasite manipulation; cat odours; anxiety; predation
1. INTRODUCTION
According to the manipulation hypothesis, a parasite
may alter the behaviour of its host for its own bene¢t,
usually by enhancing its transmission rate.The hypothesis
implies that such host behaviour modi¢cation represents
a sophisticated product of parasite evolution aimed at
host manipulation, rather than an accidental side-e¡ect of
infection (Barnard & Behnke 1990; Poulin 1994). Para-
sites that are transmitted through the food chain consti-
tute classic examples of such manipulation: the parasite is
immature in the intermediate host and must be eaten by a
predatory de¢nitive host before it can reach maturity and
complete its life cycle. Unfortunately, however, many
studies have either attached little importance as to
whether the host in question normally carries the parasite
and/or studied hosts maintained under highly unnatural
laboratory conditions. The transferability of such studies
and their applicability to the epidemiology and evolution
of disease in the wild may thus be open to question
(Moore & Gotelli 1990;Webster et al. 2000).
The host^parasite system Rattus norvegicus^T oxoplasma
gondii provides a convenient model in which to examine
such questions. T.gondii is an intracellular protozoan
(Beverley 1976) capable of infecting all mammals. Its
associatedisease,toxoplasmosis,
economic, veterinary and medical importance (Luft &
Remington 1986; Schmidt & Roberts 1989) and has
sparked renewed interest due to its debilitating reactiva-
tion in AIDS and other immunosuppressed patients (Luft
& Remington 1986). T .gondii has an indirect life cycle,
where members of the cat family are the de¢nitive hosts
of the parasites and the only mammals known to shed
T .gondii oocysts with their faeces (Hutchinson et al. 1969).
If the oocysts are ingested by another mammal such as a
wild rodent (the intermediate host) small thin-walled
cysts form in various tissues, most commonly the brain.
isofsigni¢cant
Such cysts remain viable for the life of the host
(Remington & Krahenbuhl 1982). A cat can therefore
become infected by either of two routes: it may directly
ingest oocysts shed from another cat in the environment,
or it may ingest cysts when eating infected intermediate-
host prey (Hutchinson et al.1969).
Previous ¢eld and experimental studies demonstrated
that wild rats represent a signi¢cant and persistent
intermediate-host reservoir for T.gondii, with a mean
prevalence of 35% across all populations irrespective of
environmental conditions and maintained, at least in
part, through congenital transmission (Webster 1994a). It
may thus be feasibly expected to bene¢t theT.gondii para-
site if it could somehow enhance the transmission rate
from this large intermediate-host reservoir to the cat de¢-
nitive host, and so complete its life cycle. Moreover, since
sexual reproduction of T.gondii can be accomplished only
in the feline, there might be strong selective pressure on
the parasite to evolve such a mechanism.
Indeed, there are several reasons to predict that the
T .gondii parasite may be able to achieve this. Principally,
the formation of parasitic cysts in the brain of its host
places T .gondii in a privileged position to manipulate
behaviour (Werner et al. 1981). Accordingly, recent studies
on both wild and wild^laboratory hybrid rats have
demonstrated that T.gondii causes an increase in activity
(Webster 1994b) and a decrease in neophobic (fear of
novelty) behaviour (Webster et al. 1994; Berdoy et al.
1995b), both of which can be argued to facilitate trans-
mission to the felid de¢nitive host. In contrast, other
costly behavioural patterns such as competition for mates
and social status (Berdoy et al. 1995a), which do not have
any obvious impact upon cat predation rate, are left un-
altered by the parasite (Berdoy et al. 1995b).
For any small mammal under heavy predation pres-
sure, the capacity to detect and avoid areas associated
with high predation risk is likely to be of strong selective
advantage. Rats have evolved an innate and pronounced
defensive reaction to predator odours, including cat
Proc. R. Soc. Lond. B (2000) 267, 1591^1594
Received 2 March 2000
1591
© 2000 The Royal Society
Accepted 7 April 2000
*Author for correspondence (manuel.berdoy@vet.ox.ac.uk) .
doi 10.1098/rspb.2000.1182

Page 2

(Vernet-Maury et al. 1984; Blanchard et al. 1990; Berdoy
& Macdonald 1991; Klein et al. 1994). Even naive labora-
tory rats that have not been in contact with cats for
several hundred generations still show strong aversive
reactions when confronted with cat odours. Such innate
anti-predator behaviour and the inherent anxiety that
signs of cat presence seem to engender (Blanchard et al.
1990) is, from the parasite’s point of view, an obvious
obstacle militating against its successful transmission to its
cat de¢nitive host. Here we investigate whether the para-
site is able to interfere with the rat’s innate reaction to
potential predation risk by cats.
2. MATERIAL AND METHODS
Observations were carried out on adult Lister^hooded
laboratory rats, which were outbred four generations previously
with male rats trapped from rural UK farms. Laboratory^wild
hybrids, rather than pure wild rats, were used so as to ensure
known parasitic and social histories of individuals, whilst still
obtaining behavioural patterns comparable to those of their
wild counterparts. The Lister^hooded laboratory strain was
chosen because of its reported behavioural similarity to wild
rats (Mitchell 1976). The laboratory rat population was serolo-
gically and parasitologically T .gondii negative. All rats were
also treated with ivermectin anthelmintic (MSD-Agvet Ltd,
Hoddesdon, UK) in order to ensure freedom from helminthic
or ectoparasitic infections that could bias the data (Ostlind
et al. 1985).
Experimental rats (n ˆ 32) were orally infected with 20 cysts
of the low-virulence cyst-forming RRA (Beverley) strain in
isotonic saline. This strain had been maintained by continuous
passage of infective brain homogenate in outbred AA strain
mice bred in house at the University of Strathclyde (precise
details are published in Webster 1994b). Control rats (n ˆ 32)
were sham inoculated with isotonic saline. At the end of the
study the rats were killed with carbon dioxide. T .gondii anti-
bodies were determined by the IgG indirect latex agglutination
test (T oxoreagent; Eiken, T okyo, Japan; Tsubota et al. 1977).
Titres51:32 were considered positive (Webster 1994a,b; Webster
et al. 1994).T .gondii brain cysts were determined by microscopic
examination of macerated brains in phosphate-bu¡ered saline.
Data from any exposed rat found to be serologically or parasito-
logicallyT .gondii negative at the end of the study were excluded
from analysis.Thus the ¢nal sample size for analysis consisted of
23 infected rats and 32 uninfected rats.
T o test the potential e¡ect of T .gondii on the rat’s perception
of predation risk we observed the nocturnal exploratory
behaviour of rats in outdoor pens (2m£2m). The ground was
covered with a layer of white sand to provide a homogeneous
and neutral surface that could be cleaned between each test.
The pens were enriched with a labyrinth of bricks dividing the
area into an array of 16 cells. Each corner contained seven
drops of one of four distinct odours deposited on and within
wooden nest-boxes: the rat’s own smell (own straw bedding),
neutral smell (fresh straw bedding treated with water), cat
odour (fresh bedding treated with undiluted cat urine) and
rabbit odour (fresh bedding treated with undiluted rabbit urine).
Rabbit odour served as a control for a mammalian non-
predator.The position of the four smells (own, water, rabbit and
cat) was changed between each test in order to avoid positional
biases. Each of the scented areas also contained a water and
food bowl coveredby a transparentplastic cover.
Each rat was tested singly and videotaped from dusk to dawn
with a low-intensity camera ¢xed on a sca¡olding 3m above the
test pens. The pens were illuminated from above with two 1kw
halogen lamps to which the rats had completely habituated
(Berdoy1994).
The e¡ect of infection status on visits to the four scented
areas was tested using a pro¢le analysis in the General Linear
Model procedure in SAS (SAS 1988) to take into account the
fact that responses to the four areas are linked. Since the
number of cells visited is proportional to rat activity (only rats
who emerge from their nest-boxes will show a preference or
avoidance to smells) the test of parallelism was carried out on
means weighed by overall cell use after checking that there was
no di¡erence between infected and uninfected rats (F1,54ˆ 0.85,
p ˆ 0.4). Residuals were tested for normality. The level of aver-
sion or preference to cat areas was tested by comparing (t-test)
the relative visits to cat versus rabbit areas (cat minus rabbit).
3. RESULTS
The rats’ nocturnal behaviour in the outdoor pens
(total of 670 rat-hours of observation) revealed a signi¢-
cant divergence between infected and uninfected rats in
their overall response to the smells (GLM repeated
measures, F3,159ˆ 9.19, p ˆ 0.0001), which was caused by a
di¡erentialresponse to
p ˆ 0.0001; ¢gure 1). Uninfected rats exhibited a healthy
aversionofcat-scented
p ˆ 0.002). Infected rats, however, were signi¢cantly less
averse (n ˆ 23, t ˆ 2.36, p ˆ 0.002) and showed no overall
avoidance of areas with signs of cat presence (t ˆ 0.21,
p ˆ 0.8). Alterations induced by T.gondii infection were
con¢ned to the predator’s odour, as both types of rats
behaved similarly with respect to areas containing their
own smell (which was preferred by both), neutral smell
and rabbit odour (¢gure 1).
Since the number of cells visited is proportional to
exploratory activity, the impact of T .gondii was predict-
ably more visible amongst rats who explored the pen
more intensively (n ˆ 55, F1,54ˆ 27.38, p ˆ 0.0001). Thus,
amongst the most active animals (top 25%, n ˆ14/55;
seveninfectedandseven
catodours(F1,54ˆ 22.03,
areas(n ˆ 32,
t ˆ 73.33,
uninfected),control rats
1592 M. Berdoyandothers
Fatalattraction inrats infected withT oxoplasmagondii
Proc. R. Soc. Lond. B (2000)
30
25
15
20
0
means (±s.e.m.) visits (weighted by activity)
own
n.s.
10
5
neutral
n.s.
non-infected rats (n = 32)
infected rats (n = 23)
n.s.
p = 0.0001
rabbitcat
Figure 1. Mean ( § s.e.m.) numbers of visits (weighted by
overall rat activity) to the four scented areas in the outdoor
pens over one night. Uninfected and T.gondii-infected rats
di¡er only in their response to areas associated with high
predation risk (F1,54ˆ 22.03, p ˆ 0.0001).

Page 3

continued to exhibit a stable avoidance of cat-scented
areas throughout the night, whereasT.gondii-infected rats
showed a preference for areas with signs of cat presence
(¢gure 2).
4. DISCUSSION
Inherent within the parasite manipulation hypothesis
is the premise that behavioural modi¢cation represents a
sophisticated product of parasite evolution rather than an
accidental side-e¡ect of infection (Barnard & Behnke
1990). However, in the few cases where the relationship
between physiology and behaviour has been investigated,
clinical parasitism is usually evident and has caused the
complete loss of a particular behaviour rather than a
modi¢cation of a speci¢c complex behavioural pattern as
illustrated here (e.g. Rau 1983, 1984). Even studies indi-
cating that parasites can a¡ect host learning and spatial
performance (e.g. Stretch et al.1960; Kvalsvig1988; Nokes
et al. 1992) have been confounded by parasite-induced
disruptions of overall host health status (Thompson &
Kavaliers 1994). The same does not appear to be true of
subclinical T .gondii infection. We found that infected indi-
viduals show no di¡erence from uninfected individuals in
terms of general health status (Webster 1994b; Berdoy
et al. 1995b), and behavioural categories unlikely to
in£uence predation rate, even when energetically costly,
appear unaltered (Berdoy et al. 1995a). Moreover, we
found here that the alterations induced by T .gondii infec-
tion were con¢ned to the predator’s odour, as both types
of rats behaved similarly with respect to areas containing
their own smell (which was preferred by both), neutral
smell and rabbit odour (¢gure 1). This suggests that the
potentially fatal attraction exhibited by infected rats was
not caused by a gross impairment of olfactory faculties.
Instead, manipulation by T .gondii appears to alter subtly
the cognitive perception of the host in the face of
predation risk. As with any evidence of host behavioural
alterations, further investigations should now ideally
incorporate the outcome of real predation rates by the
appropriate de¢nitive host as the yardstick of advantage
to the parasite (Webster et al. 1994, 2000; Poulin 1992;
Moore & Gotelli 1990). Nevertheless, whilst direct preda-
tion studies are fraught with practical as well as some
ethical di¤culties, we have shown previously that
T .gondii-infected rats are indeed more likely to be caught
by traps in the wild (Webster et al.1994).
In addition to the implications raised here for the
epidemiology of T .gondii in the wild in terms of increased
transmission rates, the results of this study may also have
causal and functional implications.
From a causal view point, our ¢ndings may have impli-
cations for the study of the neurological basis of beha-
viour. Indeed, the reaction by potential prey to cat
stimuli is used to study the neurological basis of anxiety
and the mechanisms of anxiolytic (anxiety relieving)
drugs. Such studies have found, for example, that
blocking the normally anxiogenic NMDA receptors in
the amygdala causes rats to approach cats `fearlessly’
(Adamec et al. 1999) in much the same way as our
infected rats approached the areas treated with cat urine.
One could speculate that such an e¡ect might imply an
anxiolytic action of T .gondii. Likewise, exposure of
laboratory rats to predator odours, but not other noxious
odours, induces fast wave activity in the dentate gyrus of
the hippocampus (File et al. 1993; Hogg & File 1994).
Such a response can be blocked by serotonin (5-HT)
antagonists (Blanchard et al. 1990; Kavaliers & Colwell
1991) or even by the presence in these mice of another
protozoan, Eimeria vermiformins (Kavaliers & Colwell
1994). Such observations could suggest that some parasitic
infections, such as T.gondii and E.vermiformins, may be
able to attenuate the 5-HT-sensitive predator-induced
response, thereby reducing the accompanying anxiety-
related anticipatory defence reactions of a host to a
predator.
Finally, we believe that these results may also provide a
functional explanation of the altered brain function in
infected humans, where T.gondii prevalence has been
found to range from 22% in the UK to 84% in France
(Desmonts & Couvreur 1974). Although humans repre-
sent a dead-end host for the parasite, our results could
suggest that the reports of altered personality and IQ
levels inT.gondii-infected patients (Burkinshaw et al. 1953;
Flegr & Hrdy 1994) represent the outcome of a parasite
evolved to manipulate the behaviour of another mammal.
It is noteworthy that rat behaviour is often viewed as the
outcome of a con£ict between pronounced neophobic
reactions and strong exploration tendencies characteristic
Fatalattraction inrats infected withT oxoplasmagondiiM. Berdoyandothers 1593
Proc. R. Soc. Lond. B (2000)
10
5
-5
0
-10
21+
non-infected
sorties
T. gondii infected
cumulative preference (cat - rabbit)
220 1816 1412 1086
Figure 2. Development of preference or avoidance throughout
the night exhibited by the 25% most active rats (n ˆ 14,
seven infected rats, seven uninfected). Results are shown as
the mean cumulative number of cat cells minus the number of
rabbit cells visited during each sortie. The data above the x-
axis therefore represent a relative preference for the cat areas
whilst data below the x-axis indicate avoidance. Vertical bars
describe 95% con¢dence intervals. Time on the x-axis is
represented in terms of sorties within the night. Sorties are
characterized by bursts of rat activity separated by intervals
when the rats shelter into a nest-box for a minimum of 1 min.
The rising line for uninfected rats indicates a prolonged, and
sensible, avoidance of cat-scented area that is essentially main-
tained throughout the night. In contrast, T.gondii-infected
rats tend to exhibit a preference for predator-scented areas.
The di¡erence between uninfected and T.gondii-infected rats
is signi¢cant from the third sortie onwards.

Page 4

of opportunistic omnivores. The uneasy balance between
these con£icting motivations, very pronounced in rats but
also visible in humans (`the omnivores paradox’, Rozin
1976), may thus provide a particularly fertile ground for
manipulation byT.gondii.
We thank M. Dowie, T. McFadden and Karen Williams for
their help with observations, Dr P. Johnson for statistical advice
and Dr J. Alexander and members of his research group at
Strathclyde University for supplying theT .gondii cysts.This work
was funded by the Natural Environment Research Council.
REFERENCES
Adamec, R. E., Burton, P., Shallow, T. & Budgell, J. 1999
NMDA receptors mediate lasting increases in anxiety-like
behaviour produced by the stress of predator exposureö
implications for anxiety associated with post-traumatic stress
disorder. Physiol. Behav. 65,723^737.
Barnard, C. J. & Behnke, J. M. (ed.) 1990 Parasitism and host
behaviour. London, UK:Taylor and Francis.
Berdoy, M. 1994 Making decisions in the wild: constraints,
con£icts and communication in foraging wild rats. In
Behavioral aspects of feeding (ed. B. G. Galef Jr, M. Mainardi &
P. Valsechi), pp. 289^313. London, UK: Harwood Academic
Publishers.
Berdoy M. & Macdonald D.W. 1991Factors a¡ecting feeding in
wild rats. Acta Oecologica12, 261^279.
Berdoy, M., Smith, P., & MacDonald, D. W. 1995a Stability of
social status in wild rats: age and the role of settled domi-
nance. Behaviour132,193^212.
Berdoy, M., Webster, J. P. & Macdonald, D. W. 1995b Parasite-
altered behaviour: is the e¡ect of T oxoplasma gondii on Rattus
norvegicus speci¢c? Parasitology111, 403^409.
Beverley, J. K. A. 1976 T oxoplasmosis in animals. Vet. Rec. 99,
123^127.
Blanchard, R. J, Blanchard, D. C., Rodgers, J. & Weiss, S. M.
1990 The characterisation and modelling of antipredator
defensive behavior. Neurosci.Biochem. Rev.14, 463^472.
Burkinshaw, J., Kirman, B. H., Sorsby, A. 1953 T oxoplasmosisin
relationto mental de¢ciency. Br. Med. J. 1,702^704.
Desmonts, G., Couvreur, J. 1974 Congenital toxoplasmosis. A
prospective study of 378 pregnancies. New Engl. J. Med. 290,
1110^1116.
File, S. E., Zangrossi, H. & Andrews, N. 1993 Novel environ-
ment and cat odor change GABA and 5-HT release and
uptake in the rat. Pharmacol. Biochem. Behav. 45, 931^934.
Flegr, J. & Hrdy, I. 1994 In£uence of chronic toxoplasmosis on
some human personality factors. Folia Parasitol. 41,122^126.
Hogg, S. & File, S. E.1994 Regional di¡erences in rat benzodia-
zepine binding in response to novelty and cat odour.
Neuropharmacology 33, 865^868.
Hutchison,W. M., Dunachie, J. F., Siim, J. C. & Work, K. 1969
The life cycle ofT oxoplasmagondii. Br. Med. J. 4, 806.
Kavaliers, M. & Colwell, D. D. 1991 Sex di¡erences in opioid
and nonopioid mediated predator-induced analgesis in mice.
Brain Res. 568, 173^177.
Kavaliers, M. & Colwell, D. D. 1994 Parasite infection attenu-
ates nonopioid mediated predator-induced analgesis in mice.
Physiol. Behav. 55, 505^510.
Klein, S. L., Lambert, K. G., Durr, D., Schaefer,T. & Waring,
R. E. 1994 In£uence of environmental enrichment and sex on
predator stress response in rats. Physiol. Behav. 56, 292^297.
Kvalsvig, J. D. 1988 The e¡ect of parasitic infection on cognitive
performance. Parasitol.T oday 41, 854^867 .
Luft, B. J. & Remington, J. S. 1986 T oxoplasmosis in the central
nervous system. Curr. Clin.T op. Infect. Dis. 6, 316^357.
Mitchell, D. 1976 Experiments of neophobia in wild and labora-
tory rats: a revaluation.J. Comp. Phys. Psych. 90,190^197.
Moore, J. & Gotelli, N. J. 1990 A phylogenetic perspective on
the evolution of altered host behaviours: a critical look at the
manipulation hypothesis. In Parasitism and host behaviour (ed.
C. J. Barnard & J. M. Behnke), pp. 193^233. London, UK:
T aylor and Francis.
Nokes, C., Gratham-McGregor,S. M., Sawyger, A. W., Cooper,
E. S. & Bundy, D. A. P. 1992 Parasitic helminth infection and
cognitive function in school children. Proc. R. Soc. Lond. B247,
77^81.
Ostlind, D. A., Nartowicz, M. A. & Mickle,W. G.1985 E¤cacy
of ivermectin against Syphacia obvelata (nematoda) in mice. J.
Helminthol. 59, 257^261.
Poulin, R. 1992 Altered behaviour in parasitized bumblebees:
parasite manipulation or adaptive suicide? Anim. Behav. 44,
172^176.
Poulin, R. 1994 The evolution of parasite manipulation of host
behaviour: a theoretical analysis. Parasitology 109(Suppl.),
S109^S118.
Rau, M. E. 1983 Establishment and maintenance of behavioural
dominance in male mice infected with T richinella spiralis.
Parasitology 86, 319^322.
Rau, M. E. 1984 Loss of behavioural dominance in male mice
infected withT richinella spiralis. Parasitology 88, 371^373.
Remington, J. S. & Krahenbuhl, J. L. 1982 Immunology of
T oxoplasma gondii. In Immunology of human infection, vol. 2
(ed. A. J. Nahmias & J. O’Reilly), pp. 327^371. New Y ork:
Plenum Publishing Corporation.
Rozin, P. 1976 The selection of foods by rats, humans and other
animals. Adv. Study Behav. 6, 21^76.
SAS Institute, Inc. 1988 SAS/STATTMUser’ s guide, release 6.03 edn.
Cary, NC: SAS Institute, Inc.
Schmidt, G. D. & Roberts, L. S. 1989 Foundations of parasitology,
4th edn. St Louis, MO: Times Mirror/Mosby College
Publishing.
Stretch, R. G. A., Leytham, G. W. H. & Kershaw, W. E. 1960
The e¡ect of acute schistosomiasis upon learning in rats
under di¡erent levels of motivation. AnnlsTrop. Med. Parasitol.
54, 487^92.
Thompson, S. N. & Kavaliers, M. I. 1994 Physiological bases
for parasite-induced alterations of host behaviour. Parasitology
109(Suppl.), S119^S139.
Tsubota, N., Hiraoka, K., Sawada, Y., Ohshima, S. &
Ohshima, M. 1977 Studies on the latex agglutination test for
toxoplasmosis: evaluation of the microtiter test as a serologic
test for toxoplasmosis in some animals. Jpn. J. Parasitol. 26,
291.
Vernet-Maury, E., Polak, E. H. & Demael, A. 1984 Structure^
activity relationship of stress-inducing odorants in the rat. J.
Chem. Ecol. 10, 1007^101 7.
Webster, J. P. 1994a Prevalence and transmission of T oxoplasma
gondii in wild brown rats, Rattus norvegicus. Parasitology 108,
407^411.
Webster, J. P. 1994b The e¡ect of T oxoplasma gondii and other
parasites on activity levels in wild and hybrid Rattus norvegicus.
Parasitology109, 583^589.
Webster, J. P., Brunton, C. F. A. & Macdonald, D. W. 1994
E¡ect of T oxoplasma gondii on neophobic behaviour in wild
brown rats, Rattus norvegicus. Parasitology109, 37^43.
Webster, J. P., Gowtage-Sequeira, S., Berdoy, M. & Hurd, H.
2000 Predation of beetles (T enebrio molitor) infected with tape-
worms (Hymenolepis diminuta): a note of caution for the
Manipulation Hypothesis. Parasitology120, 313^318.
Werner, H., Masihi, K. N. & Senk, U. 1981 Latent T oxoplasma
infection as a possible risk factor for CNS disorders.
Zentralblatt fu « r Bakteriologie Mikrobiologie und Hygiene A250,
368^375.
1594M. Berdoyandothers
Fatalattraction inrats infected withT oxoplasmagondii
Proc. R. Soc. Lond. B (2000)