The University of Chicago
Foraging on Prey that are Modified by Parasites
Author(s): Kevin D. Lafferty
Source: The American Naturalist, Vol. 140, No. 5 (Nov., 1992), pp. 854-867
Published by: The University of Chicago Press for The American Society of Naturalists
Stable URL: http://www.jstor.org/stable/2462792
Accessed: 05-06-2015 20:27 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].
The University of Chicago Press, The American Society of Naturalists and The University of Chicago are
collaborating with JSTOR to digitize, preserve and extend access to The American Naturalist.
http://www.jstor.org
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
Vol. 140, No. 5
The American Naturalist
FORAGING ON PREY THAT ARE MODIFIED
November 1992
BY PARASITES
KEVIN D. LAFFERTY
Departmentof Biological Sciences, Universityof California,Santa Barbara, California93106
SubmittedFebruaty 25, 1991; Revised November 6, 1991; Accepted November 11, 1991
Abstract.-A model that weighs the energetic cost of parasitismfor a predator against the
energeticvalue of prey items that transmitthe parasite to the predatorsuggeststhat there is
oftenno selectivepressureto avoid parasitizedprey.This offersan explanationforwhyparasites
so frequentlyexploit predators and prey as definitiveand intermediatehosts, respectively.
Furthermore,predatorsmay actuallybenefitfromtheirparasitesifenergeticcosts of parasitism
are moderateand preycapture is facilitatedby parasites. Parasite species thatbenefitpredators
throughmodificationof prey are not mutualistic,however.
Althoughthey are often ignored, parasites can affectpredator-preyinteractions. Many parasites exploit trophictransmission(wherebyinfectivestages are
ingestedby the host). Frequently,the definitivehost is a predatorthatpreys on
the intermediatehost. Some protozoans, a few nematodes, many trematodes,
thisway. Althoughthese
mostcestodes, and all acanthocephalansare transmitted
parasitesextracta cost fromtheirdefinitivehosts, and manyothercosts of foraginghave been suggested(Stephens and Krebs 1986),theriskof acquiringparasites
is not oftenconsideredforforagers(Moore 1983). Whyshould predatorscontinue
to supporttrophictransmission?Avoidingparasitizedprey would appear to be a
forpredatorsto recognize parasitized
convenientsolution. Perhaps it is difficult
prey; alternatively,there may be no fitnessadvantages for predatorsthatavoid
parasitizedprey. In this article,I presenta foragingmodel thatconsiders trophically transmittedparasites. The model compares the rate of energygained fora
predatorif some preyare parasitizedand the predatoravoids parasitizedprey,if
some prey are parasitized and the predatoringests parasitized prey, and if no
preyare parasitized.This model suggeststhatpredatorsshould not avoid parasitized prey and thattheymay actually benefitfromthe presence of parasites.
Predatorsoftentake odd or unusual preyindividuals(Temple 1987). Parasitized
prey can be odd and are oftenfound more frequentlythan expected in the diet
ofdefinitivehostpredators(Dobson and Keymer 1985). In fact,thereis increasing
evidence thatlarval parasites modifythe behavioror appearance of intermediate
hosts (see reviews in Holmes and Bethel 1972, Moore 1984, and Dobson and
Keymer 1985). For example, parasitized prey may be more conspicuous, disoriented, less able to flee, or less likelyto show an escape response (Holmes and
Bethel 1972).
Sometimes,it may be possible to quantifythe effectof a parasiteon its intermeAm. Nat. 1992. Vol. 140, pp. 854-867.
All rightsreserved.
C 1992 by The Universityof Chicago. 0003-0147/92/4005-0008$02.00.
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
FORAGING ON PARASITIZED PREY
855
diate host. Dobson and Keymer (1985) definethe degree of parasite-inducedbea, as theincreasedrateat whichpreyare eaten ifparasitized.
haviormodification,
For parasitizedprey withno behavior modification,c = 1. Withc = 2, parasitized preyare capturedtwice as oftenas unparasitizedprey. I have interpretedcx
as the forage ratio (Ivlev 1961) for parasitized prey divided by the forage ratio
forunparasitizedprey. Assumingthateach parasitizedprey carriesone parasite
(Dobson and Keymer [1985] allow multipleinfections),definecxas
hilHi
0
holHo
9
where hi and ho representparasitized (infected)and unparasitizedprey eaten by
predators,and Hi and Ho representparasitized and unparasitizedprey in the
environment.For example, Feare (1971) found that 13% of the dogwhelksconsumed by oystercatcherswere parasitized by larval trematodes,compared with
a prevalence of 5% in the dogwhelk population as a whole (prevalence is the
proportionof hosts thatare parasitized [Margoliset al. 1982]). In thiscase, a = 2.8. Selectivityforparasitizedpreycan reach impressivelevels.
(13/5)/(87/95)
In Argentinealpine lakes, for example, amphipods in the guts of fishare commonlyparasitized by acanthocephalan larvae, but living,parasitizedamphipods
have notbeen observed despitepersistenteffort(A. M. Kuris, personalcommunication).
The modificationof parasitized prey potentiallyhurtspredatorsby increasing
theirexposure to parasites. Since hosts can learn to avoid parasitizedfood that
theyassociate with a certaintaste (Keymer et al. 1983), it may be possible for
some predators to avoid parasitized prey that they associate with a modified
behavioror appearance (Lozano 1991). Avoidingparasitizedprey,however,carries a cost because it reduces the numberof prey itemsaccepted by a predator.
In fact,because the modificationof prey by parasites may lead to an increase in
predationrate, Holmes (Holmes and Bethel 1972; Holmes and Price 1986) sugbetweenthe cost of parasiteacquisitionand easier predageststhereis a trade-off
tion. It is plausible, sometimes,thatthe benefitis greaterthanthe cost (Holmes
and Bethel 1972) and that the predatorwill obtain more energywithparasitism
thanwithoutparasitism.
FORAGING MODEL
The followingmodel compares the costs and benefitsof avoiding or ingesting
parasitized prey and examines whethermodificationof prey by parasites can
benefita predator.The net rate of energygain,E, acquired by a predatorconsists
of unparasitizedprey and parasitizedprey,less the cost of parasitism,such that
Elt = k(Ho + as-Hi) - energeticcost of parasites/t,
(2)
where t is time,k is the energyassimilatedfroma singleprey item,and r is the
rate of predationon unparasitizedprey. This assumes thata predator'sforaging
rateforunparasitizedpreyis independentof whetherthe predatoris parasitized,
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
856
THE AMERICAN NATURALIST
parasitizedpreyhave the same energeticvalue as unparasitizedprey,and modification resultsin an increase in preyavailability(or "catchability").
In this model, the cost of parasitismto the predatoris considered to be the
combined cost of the parasites withinthe predator(the parasites withina host
are definedas an "infrapopulation" [Margolis et al. 1982]). Parasite fecundity
and mortalityare assumed to be dependenton the numberof parasiteswithinthe
predator("intensity" is the numberof parasitesin a host [Margoliset al. 1982]).
is dependenton the predator's
The rate of change of a parasite infrapopulation
ingestionof parasites, their successful establishmentin the predator,and the
parasite mortalityrate (R. M. Anderson 1974), such that
dildt = oxrqHi- W",
(3a)
where q is the proportionof parasites that establish withinthe host, u is the
initial,instantaneous,per capita parasite mortalityrate, i is the intensityof the
and in is a coefficient
of an intensity-dependent
increase
parasiteinfrapopulation,
in parasite mortalityrate. At equilibrium,therefore,the parasite infrapopulationis
u
arqHi)
I/(3b
and the cost of parasitismis the parasite intensitytimes the per parasite cost
(adjusted forcrowding),such that
(3c)
Elt = -gilf,
where g is the initialrate of energyremoved per parasite and f is a coefficient
decrease in the energyremovedfrom
fromzero to one of the intensity-dependent
the host by an individualparasite.
Incorporatingthe cost of parasitisminto equation (2) yields equations forenergygained by a predatorunder the followingthreeconditions.If the parasiteis
presentand the predatordoes not avoid parasitizedprey,
Elt = k(rHO + etrHi) - g
(4a)
If the parasite is presentand the predatoravoids parasitizedprey,
Elt = krHo.
(4b)
system,
Finally,if the parasite is absent fromthe predator-prey
Elt = krH.
(4c)
The model cannot be interpretedin its presentstate, because the population
equilibriaof parasitizedand unparasitizedpredatorsand preyvary withcx(Dobson and Keymer 1985; Hadeler and Freedman 1989). Therefore,the following
model was used to obtainpredator-prey
popuLotka-Volterra-style
predator-prey
lationequilibriaforsubstitutioninto equations (4a)-(4c):
dH/dt = bH - dH2
-
rPH
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
(Sa)
FORAGING ON PARASITIZED PREY
857
and
dP/dt = criPH-
(Sb)
jP'9
where b is the initial,instantaneous,per capita prey birthrate, dH is the initial,
instantaneous,per capita prey death rate, P is the numberof predators,c is the
conversionof ingestedprey into new predators,andjP is the instantaneous,per
capita predator death rate. In addition to the familiarassumptions of Lotkamortalityfor both prey
Volterramodels, this model assumes density-dependent
and predatorand no handlingtime(forgeneratingstable equilibria). To incorpoequations were includedforparasitizedpredators(Pi)
rate parasites, differential
and prey. These equations are outlined schematicallyin figure1. It is assumed
that all predatorsare born unparasitized(PO) and there is no immunityto new
infections.
Expandingequations (Sa) and (Sb) yields
dHoIdt = bH - dHHo - rPHo - f3PiHo,
dHildt
dP /dt
PiHo
c(rPHO + orPH;
-
dHHi - curPHi,
Pigil-)
-
jPP0 - ctrqPH;,
(6a)
(6b)
(6c)
and
dPi/dt = aarPoHi- jPPi,
(6d)
where X is the transmissionrate frompredatorto prey (parasitized predators
excreteparasiteeggs thatare eaten by prey). Equations (6a)-(6d) are not solvable
by analytical techniques, and thereforecomputersimulationwas employed to
findvarious predator-preyequilibria. Predator-preyequilibria,in the absence of
parasites, were recorded by settingPi and Hi to zero. These values were then
incorporatedinto equation (4c) to indicatethe rate of energygained by predators
in theabsence of parasites. This value acts as a pointof referenceforcomparisons
withthe followingsituationsin which parasites are included.
Predator-preyequilibria in the presence of parasites were recorded according
to equations (6a)-(6d) over a range of cx.These values were then incorporated
into equation (4a) to indicate the rate of energygained by predatorsthat ingest
parasitizedprey and into equation (4b) to show the rate of energygained by an
individualpredatorthatavoids parasitizedprey.
RESULTS
Increases in modificationof prey by parasites resultin a decrease in the prey
equilibria and a less dramatic increase in the predator population (fig. 2). In
addition,increasingbehavior modificationcauses an asymptoticincrease in the
prevalence of parasites in predators(fig.3). This yieldsan increase in the prevalence of the parasite in the prey population (because the infectionrate from
predator to prey is increased) followed by an eventual decline in prevalence
(because parasitizedpreyare rapidlyremovedfromthe population).These population-leveleffectsare similarto the theoreticalresultsof Dobson and Keymer
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
THE AMERICAN NATURALIST
858
_
death,
j
birfi,c
PREDATORS,P
u
_death,
I
Pi
Infected,
f,
infection,
q
Po
Uninfected,
1
i
Parasites,
ingestion,r
Ho
UIninfected,
ingestion,ar
Infected,Hi
birth,b
FIG. 1.-Flow
d
death,
chartforpredator,prey,and parasite populations
(1985) and Hadeler and Freedman(1989). Incorporatingthesepopulationdensities
intothe energygain functions(eqq. [4a]-[4c]) reveals thatan individualpredator
can benefitfromparasites if costs of parasites are moderateand prey are sufficientlymodifiedby parasites (fig.4). Avoidance of parasitizedprey is an appropriatestrategyonly if parasite cost is highand modificationof prey by parasites
is low (fig.4).
DISCUSSION OF ASSUMPTIONS
Violations of key assumptions that may have importantimplicationsfor the
outcome of this and other models have not been addressed. The most crucial
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
FORAGING ON PARASITIZED PREY
859
CD)
CZ
,>
Pry
0
E
a:i
Predator
z_
1 2 3 4 5 6 7 8 9 10
Modification
ofpreybyparasites,oc
FIG.. 2.-The
effectof a on predatorand prey equilibriumpopulatioin size
>
redator
0
ci)
Ci-
~~Prey
1 2 3 4 5 6 7 8 9 10
Modification
ofpreybyparasites,cc
FIG.
3.-The prevalence of parasitizedprey and predatorsas a functionof a
assumptionis that values of ax greaterthan one must reflectan increase in the
catchabilityof parasitized prey (Moore and Gotelli 1990). Otherwisethe model
is irrelevant.For example, selectivityof parasitizedpreyby predatorsmay occur
in the absence of modificationby parasites if a parasite has a constanttransmission rate to the intermediatehost. In thiscase, parasite prevalence and intensity
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
THE AMERICAN NATURALIST
860
-
Ingest:nocost
111
s
co
/
CD
CD
m_
CD
QD
4 :=
/
0
: Noparasite
preymortality
Ingest:nocost,linked
Ingest:highcost
CD
Cl)
cost
Ingest:moderate
1('S
0
('S
cost
~~~~~~Avoid:
high
Avoid:lowcost
1
2 3 4 5 6 7 8 9 10
ofpreybyparasites,ca
Modification
rateofenergy
gainforcases in
ofa andparasitecoston a predator's
FIG.4.-The effect
whicha predator
avoidsor ingestsparasitized
preyor in whichno preyare parasitized.
in the intermediatehost may be correlatedwith age, and, if a predatorprefers
prey on the basis of a variable associated with age, such as size, it will appear
thatthe predatoris selectingparasitizedprey.
Furthermore,if all sources of nonpredationmortalityin the intermediatehost
increase in directproportionwithac (prey mortality= aedHHi), no benefitforthe
predatoris possible underany circumstances(fig.4). This is because preydensities would decrease witha and the predatorwould encounterless preythanifthe
parasitewere absent. This would not necessarilybe thecase ifnon-definitive-host
predatorsenjoyed increased prey capture of parasitizedprey but did not reduce
preypopulationlevels substantially.
Anothercriticalassumptionis thatthe rate of predationmust be independent
of whethera predatorfeeds on parasitized prey. If a predatorbecomes satiated
because modificationhas increased prey availability,the actual rate of ingestion
will be less thanthe model predicts.In thiscase, however,the predatormay gain
otherbenefitsfromreduced foragingtime,such as decreased exposure to its own
predators(McNamara and Houston 1987). If thepredatordevelops anorexia in response to parasitism,its foragingrate will decline, and the model will no longer
be valid. Althoughanorexia is noted as a response to parasitism(Symons 1989),
itis infrequently
reportedforhosts thatbecome infectedby eatingpr.ey.Nonetheless, ifthe predatorbecomes sick because of parasites and can no longerexploit
prey at the same level as a healthyindividual,the model is no longervalid. In
this case, selectivityon parasitized prey could be a resultof a reduced capacity
to capture unparasitizedprey.
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
FORAGING ON PARASITIZED PREY
861
The assumptionthat there is only one larval parasite per parasitized prey is
effectson behavior
violated in many, but not all, systems. Intensity-dependent
are likelyto be seen in "typical parasites" (sensu Kuris 1974),whereastheeffects
of parasitic castratersare likelyto be independentof intensity(Kuris 1974). In
theformercase, the intensityof larval parasites in parasitizedintermediatehosts
is oftendistributedas a negative binomial (Crofton1971). For prey containing
more than one larval parasite, Dobson and Keymer (1985) assume the effecton
behavioris additive. Under thiscondition,theconclusionsof the model will hold.
If the effectis not additive,however, heavily parasitizedintermediatehosts will
not be vulnerablein proportionto the potentialcosts of parasitism.In this case,
a- should be based on the mean intensityin an intermediatehost, and the model
can predictonly an average rate of energygain.
The assumptionthatparasitizedpreyhave the same energycontentas unparasitizedprey can be violated by (1) a positive correlationbetween size and exposure to parasites, (2) a negative correlationbetween size and parasite-induced
mortality,(3) a reductionin growthassociated withparasitism,or (4) an increase
in growthassociated with parasitism. If parasitized prey have a lower energy
contentthan unparasitizedprey (mechanism 2 or 3), predatorsthat forage on
parasitizedprey will ingestless energythan predictedby the model. Of course,
if 1 or 4 is correct,predatorsthatforageon parasitizedpreywillgain moreenergy
thanthe model predicts.
If otherprey itemsare included in the predator'sdiet, the qualitativeoutcome
offoragingon thisparasitizedspecies will notbe affected.However, theenergetic
returnof the parasitizedpreyspecies could be devalued to theextentthatswitching to anotherprey species would make a more efficientuse of the predator's
timeand resources.
Finally, the regulationof parasite infrapopulationsis not well understood.In
supportof the model, high parasite intensitiescan result in increased parasite
mortalityand reduced parasitefecundity(a logical correlateof per parasitecost)
regula(Read 1951;Jonesand Tan 1971; Keymer 1982). If parasiteinfrapopulation
tionis based on an immuneresponse,however,thecost ofparasitismis discontinuous (it stops when the host becomes immune),whereas the benefitof foraging
on parasitized prey is continuous (it continuesthroughthe predator's lifetime).
Therefore,if immunityis permanent,a benefitwill occur given enough time. A
benefitmay also occur ifimmunityis concomitant(immunityrequirescontinuous
antigenicstimulation),dependingon thecost oftheparasite.If parasiteinfrapopulations are not regulated,a benefitwill be less likelyand avoidance more plausible.
DISCUSSION
Avoidance and Benefit
In figure4, where the energyrate curve for ingestioncrosses above the No
parasite line, a predatorcan benefitfromparasites. A benefitis always possible,
given some modificationof the intermediatehost by parasites,ifthereis no cost
of parasitism.Even withmoderatecosts of parasitism,a benefitforpredatorsis
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
862
THE AMERICAN NATURALIST
possible if modificationis highenough. The rate of energygain is not a positive
linearfunctionof the degree of modification,however,because, at highlevels of
modification,prey become rare and the numberof prey ingesteddeclines.
The effectsof parasiteson intermediatehosts and theresultantbenefitto definitive hosts may play an importantrole in predator-preydynamicsand foraging
strategy.The parasiteprovidesa deliveryserviceforhard-to-get
prey.If parasites
allow a more efficientexploitationof otherwisedifficult-to-capture
prey items,
they may be an importantfactor determiningdiet breadth and the impact of
predationas a force in structuring
prey populations.This stands in markedcontrastto thetraditionalview of a predatoras an agentthatweeds out sick individuals and bringsabout the increased healthof the preypopulation(Slobodkin 1974;
Holmes 1982). Althoughthinningoccurs, its eventual impact is to perpetuate
parasitetransmissionand futurepredatorsuccess.
Even forcases in which a predatordoes not benefitfromparasites,theremay
be no selectionforavoidingparasitizedpreyunless thereis a highcost of parasitism and little modificationof prey by parasites. For example, oystercatchers
rejectclams thatare heavilyparasitizedby trematodemetacercariae,but it is not
evidentthatrejected clams are easier preythanless parasitizedclams (Hulscher
1973). Avoidance, assumingit is a heritabletrait,can spread in a predatorpopulationonly ifit allows individualsto increase theirfitness(expressed in the model
as Elt). For the cases in which avoidance is profitable,the spread of avoidance
and perfectrecognitionof parasitizedprey) would eventu(assumingheritability
ally lead to the local extinctionof the parasiteand an increase in the energygain
forall predators.At a moderatedegree of modificationand a highcost of parasitism, avoidance by all predatorswould lead to an increase in the energygain of
the population; however, avoidance cannot be selected for because it results,
initially,in a decrease in an individual's fitness.Furthermore,if the abilityof a
predatorto recognize parasitizedpreyis positivelycorrelatedwithhow different
the prey appears, the abilityto avoid parasitizedpreywill be least when the net
costs are triehighest.These resultshelp explain, even withoutrelyingon arguments about constraintsof recognitionor heritability,
why the modificationof
intermediatehosts is such a successful and pervasive strategyfor trophically
transmitted
parasites.
Adaptations of Prey
Intermediatehosts appear to suffergreaterfitnesscosts due to parasites than
definitivehosts. Parasitic castrationand increased mortalitydue to modification
by parasites can be consequences of parasitismforintermediatehosts. In many
cases, the prey intermediatehost becomes parasitized by eating eggs or larval
parasites in food. With such high costs of parasitism,why don't intermediate
hosts avoid food resources thatcontainparasite eggs?
Moore (1983) found no significantdifferencein the feedingrate of terrestrial
isopods (which serve as the intermediatehost for the acanthocephalanPlagiorhynchus)when she presentedthe pill bugs with starling(definitivehost) feces
withand withoutparasite eggs. She suggeststhatthe ingestionof food-richbird
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
FORAGING ON PARASITIZED PREY
863
feces mightbe worththe risk of parasitic castrationfor isopods. Therefore,for
prey,the costs of avoidance may also outweighthe risksof parasitism.
Altruistichost suicide was firstsuggested as an explanation for the altered
behavior of parasitized hosts by Shapiro (1976). In this case, altered behaviors
are assumed to be host adaptations against parasitismsuch that a host's intentionaldeath reduces the riskof parasitismforits kin. Holmes (1982) has suggested
that if prey are parasitically castrated, increased predation, due to behavior
modification,mightbenefitthe prey population by removingunproductive,resource-consumingindividuals.Althoughthepresence of castratedindividualscan
negativelyaffectuninfectedindividualsthroughcompetition(Lafferty1991), incorporatingparasiticcastrationinto equation (6a) (changingbH to bHO) does not
supportthepredictionthatsuicide is adaptive. This is because thepreypopulation
equilibriumcontinuesto decrease as modificationby parasites is increased since
predationon castratedprey eventuallyfeeds back, via transmission,to a higher
proportionof parasitizedprey.
-
Adaptations of Parasites
Trophicallytransmitted
parasites should evolve to increase a- and limitpathology for definitivehosts. Natural selection will favor parasites with traits that
increase the probabilitythat the death of the intermediatehost will result in
transmission(Wright1966) and/orshortentheirgenerationtimeby increasingthe
rate at whichtransmissionoccurs. Not all parasitetraitsthatchange preybehavior are necessarilyadaptive, however, especially iftheyare constrainedby phylogeny(Moore and Gotelli 1990). In addition,by reducingits impacton the definitive host, a parasite mightreduce the likelihoodthatthe predatorwill choose to
avoid parasitizedpreyin the future(thisrequiresgroupselection,however). The
parasite mightalso gain the immediatebenefitof not inducinga hostile immune
response (Sprent 1969). In fact,parasites thatexploittrophictransmissiongenerally cause littlepathologyfor theirdefinitivehosts (Bailey 1975; Kennedy 1975;
Geraci and St. Aubin 1987), especially when compared with the major effects
thatintermediatehosts suffer.For example, over a wide rangeof infectionintensities with the tapewormHymenolepis citelli,the energybudget of white-footed
mice was reduced by only 2% (Mungerand Karasov 1989). These adaptationsdo
not mean thatparasites gain a fitnessadvantage because theybenefittheirdefinitive hosts. Instead, host benefitis an incidentalresult of natural selection for
parasitetraitsthatincrease transmissionand survival.
Mutualism
Mutualismcan be categorizedby whethera thirdparty(outside host and parasite) is requiredforthe host to benefit(Abrams 1987). A directbenefitmay occur
iftheparasiteprovides some kindof nutritivesupplementforthehost. An indirect
benefitoccurs when the parasite mediatesinteractionswitha thirdparty(usually
host enemies) to the advantage of the host (Boucher et al. [1982] describe direct
and indirectinteractionsas symbioticand nonsymbioticmutualisms).Through
indirectbenefit,a parasite may have the net impactof a mutualist.
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
864
THE AMERICAN NATURALIST
Althoughmutualismhas been described as an interactionthat increases the
rate of growthof populations of two interactingspecies (discussion in Boucher
et al. 1982 and Freedman et al. 1987 for three-speciesinteractions),from an
evolutionaryperspective,mutualismmustbe based at the level of the individual
(Kuris 1980; Janzen 1985). In otherwords, individualsthatprovidea benefitmust
also obtain a benefit.This may or may not be consistentwithinteractionsat the
populationor species level (Abrams 1987).
The only evidence of a directbenefitforhosts fromparasites (excludingdigestive symbionts)is fromLincicome (1971) who found increased weightgains in
ratsinfectedwiththe protozoan Trypanosomalewisior the nematodeTrichinella
spiralis under special circumstances.Under normalcircumstances,T. lewisi can
cause arthritis,
abortion,and, in youngrats,death (Duca 1939; Shaw and Dusanic
1973),and thefitnessof individualsinfectedwithT. spiralisis apparentlyreduced,
since female mice show reductionsin fecundityproportionalto the intensityof
infection(Weatherly1971). Therefore,in nature,T. lewisi and T. spiralis should
not be considered beneficialfortheirrodenthosts.
Examples of parasites thatindirectlyprovide a benefitfortheirhosts are more
clearly substantiated.A host, because of parasites, may enjoy freedomfrom
competitors.For example, the nematodeparasiteParelophostrongylustenuishas
littleeffecton white-taileddeer but causes severe morbidityin moose, which
frees deer from competition(Barbehenn 1969; R. C. Anderson 1972; but see
Nudds 1990). In Africa,native grazers are protectedfromcompetitionwithlivestock because the latter develop wasting disease (nagana) afterinfectionwith
sylvatictrypanosomes.This disease has historicallydeterminedpatternsof human settlementand is responsibleforpreservingvast areas of unspoiled wilderness (Ford 1971). These examples do not indicate a mutualisticrelationshipbetween host and parasite individuals. Althoughhost species A individualsmay
benefitfroma parasite species that reduces the level of competitionwith host
species B, the individualparasites that provide the benefitfor host species A
individualsdo not receive a benefitfromthe same host species A individuals;
these parasite individualsare inside host species B individuals.
Anothertypeof indirectbenefitoccurs when parasitesprotecthostsfromother
parasites. For example, it has been suggested that the presence of a cowbird
brood parasite may generate a net benefitfor host nestlingsby eating botflies,
which,undersome circumstances,may be the major source of nestlingmortality
(Smith 1968). This phenomenonis even more likelyto occur between parasites
in similarhost niches in whichcompetitionbetweenparasites is strong.Parasites
will increase both theirown and theirhost's fitnessby helpingto preventsubsequentparasitism(Schad 1966; Holmes 1983; Freeland 1986). Heterologousimmunity,in particular(see review in Christensenet al. 1987), is a plausible benefitof
forexample, a host may
parasitism.Under conditionsof concomitantimmunity,
be betteroffretainingestablishedparasites ifthisaffordsprotectionagainstnew,
more pathogenic,infections.These interactionscan appropriatelybe viewed as
indirectlymutualistic,because parasites benefitthe host individualthattheyestablishin.
of preyby parasites,
The benefitreceived by predators,due to the modification
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
FORAGING ON PARASITIZED PREY
865
is also indirectbecause it requiresthe involvementof a thirdparty(the intermediate host preyitem). Freedman (1990) considersthatthe predatorand parasiteare
obligatemutualistsif the persistenceof the predatorpopulationis dependenton
thepresence of theparasite. In the presentmodel, althoughthe predatorbenefits,
the parasite is not necessarilymutualisticin the evolutionarysense. The benefit
thatthe host receives is based on the ingestionof parasitizedprey,not the establishmentof parasites. Individual parasites that become established may have a
net negativeimpacton the predator,whereas parasites thatare ingested,but fail
to establish, clearly benefitthe predator. This is clearly a case in which one's
definitionof mutualismhas an impacton how the relationshipis classified.
ACKNOWLEDGMENTS
Thanks to P. Abrams,A. Bush, A. Dobson, J. Moore, S. Rothstein,J. Shields,
and R. Warnerforhelpfuldiscussions, E. Diaz de Leon, J. Endler, T. Huspeni,
C. Osenberg, G. Rosenqvist, C. Sandoval, and T. Stevens forcommentson the
manuscript,and ArmandKuris forthe above, as well as supportand enthusiasm.
J. Holmes and an anonymous refereeprovided valuable comments.This work
was aided by fundsfroma Universityof California,Santa Barbara, General Affiliates Scholarshipand the Ellen SchamburgBurleyGraduate ScholarshipforOutstandingResearch Achievement.
LITERATURE CITED
Abrams,P. A. 1987. On classifyinginteractionsbetween populations.Oecologia (Berlin) 73:272-281.
Anderson,R. C. 1972. The ecological relationshipsof meningealworm and native cervids in North
America. Journalof WildlifeDiseases 8:304-310.
Anderson,R. M. 1974. Mathematicalmodels of host-helminthparasite interactions.Pages 43-69 in
M. B. Usher and M. H. Williamson,eds. Ecological stability.Chapman & Hall, New York.
report.InternationalJournal
Bailey, G. N. A. 1975. Energeticsof a host-parasitesystem:a preliminary
forParasitology5:609-613.
Barbehenn,K. R. 1969. Host-parasiterelationshipsand species diversityin mammals:an hypothesis.
Biotropica 1:29-35.
Boucher, D. H., S. James, and K. H. Keeler. 1982. The ecology of mutualism.Annual Review of
Ecology and Systematics13:315-347.
Christensen.N. O., P. Nansen, B. 0. Fagbemi, and J. Monrad. 1987. Heterologous antagonistic
and synergisticinteractionsbetween helminthsand between helminthsand protozoans in
concurrentexperimentalinfectionof mammalianhosts. ParasitologyResearch 73:387-410.
Crofton,H. D. 1971. A quantitativeapproach to parasitism.Parasitology63:179-194.
Dobson, A. P., and A. E. Keymer. 1985. Life historymodels. Pages 347-384 in D. W. T. Crompton
UniversityPress, Cambridge.
and B. B. Nickol, eds. Acanthocephalanbiology. Camnbridge
Duca, C. J. 1939. Studies on the age resistanceagainsttrypanosomeinfections.II. The resistanceof
age groups to T;ypanosornalewtisi,and the blood response of rats infected
rats of different
withthis parasite. AmericanJournalof Hygiene 29:25-32.
Feare, C. J. 1971. Predation of limpetsand dogwhelks by oystercatchers.Bird Study 18:121-129.
Ford, J. 1971. The role of trypanosomiasesin Africanecology: a study of the tsetse-flyproblem.
Clarendon,Oxford.
Freedman, H. I. 1990. A model of predator-preydynamicsas modifiedby the action of a parasite.
MathematicalBiosciences 99:143-155.
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
866
THE AMERICAN NATURALIST
Freedman, H. I., J. F. Addicot, and B. Rai. 1987. Obligate mutualismwitha predator:stabilityand
persistenceof three-speciesmodels. Theoretical Population Biology 32:157-175.
Freeland, W. J. 1986. Arms races and covenants: the evolutionof parasite communities.Pages 289309 in J. Kikkawa and D. Anderson,eds. Communityecology, patternand process. Blackwell Scientific,Boston.
Geraci, J. R., and D. J. St. Aubin. 1987. Effects of parasites on marine mammals. International
JournalforParasitology5:407-414.
populationswithparasiticinfection.Journal
Hadeler, K. P., and H. I. Freedman. 1989. Predator-prey
of MathematicalBiology 27:609-631.
Holmes, J. C. 1982. Impact of infectiousdisease agents on the populationgrowthand geographical
distributionof animals. Pages 37-51 in R. M. Andersonand R. M. May, eds. Dahlem workshop on populationbiology of infectiousdisease agents. Springer,Berlin.
1983. Evolutionaryrelationshipsbetween parasitichelminthsand theirhosts. Pages 161-185
in D. J. Futuymaand M. Slatkin,eds. Coevolution. Sinauer, Sunderland,Mass.
Holmes, J. C., and W. M. Bethel. 1972. Modificationof intermediatehost behaviour by parasites.
Pages 123-149 in E. U. Canning and C. A. Wright,eds. Behavioural aspects of parasite
transmission.Academic Press, London.
Holmes, J. C., and P. W. Price. 1986. Communitiesof parasites. Pages 187-213 in J. Kikkawa and
D. J. Anderson,eds. Communityecology: patternand process. Blackwell Scientific,Boston.
and trematodeinfectionin Macorna balthica. NetherlandsJourHulscher,J. B. 1973. Burying-depth
nal of Sea Research 6:141-156.
Ivlev, V. S. 1961. Experimentalecology of the feedingof fishes.Yale UniversityPress, New Haven,
Conn.
Janzen, D. H. 1985. The natural historyof mutualisms.Pages 40-99 in D. H. Boucher, ed. The
biologyof mutualism:ecology and evolution. OxfordUniversityPress, New York.
Jones,A. W., and B. D. Tan. 1971. Effectof crowdingupon growthand fecundityin the mouse bile
duct tapewormHyrnenolepismicrostoma.Journalof Parasitology57:88-93.
Kennedy, C. R. 1975. Ecological animal parasitology.Blackwell Scientific,Oxford.
Keymer,A. D., D. W. T. Crompton,and B. J. Sahakian. 1983. Parasite-inducedlearnedtaste aversion
involvingNippostrongylusin rats. Parasitology86:455-460.
Keymer,A. E. 1982. Density-dependentmechanismsin the regulationof intestinalhelminthpopulations. Parasitology84:573-587.
Kuris, A. M. 1974. Trophic interactions:similarityof parasitic castratorsto parasitoids. Quarterly
Review of Biology 49:129-148.
1980. An ecological classificationof symbioticassociations. Paper presented at the 146th
annual meetingof the AmericanAssociation forthe Advancementof Science, San Francisco.
Lafferty,K. D. 1991. Effectsof parasitic castrationon the salt marsh snail, Cerithideacalifornica.
Ph.D. thesis. Universityof California,Santa Barbara.
Lincicome, D. R. 1971. The goodness of parasitism:a new hypothesis.Pages 139-227 in T. C. Cheng,
ed. Aspects of the biology of symbiosis. UniversityPark, Baltimore.
Lozano, G. A. 1991. Optimal foragingtheory:a possible role forparasites. Oikos 60:391-395.
Margolis,L., G. W. Esch, J. C. Holmes, A. M. Kuris, and G. M. Schad. 1982. The use of ecological
termsin parasitology(reportof an ad hoc committeeof the AmericanSociety of Parasitologists). Journalof Parasitology68:131-133.
McNamara, J. M., and A. I. Houston. 1987. Starvationand predationas factorslimitingpopulation
size. Ecology 68:1515-1519.
Moore, J. 1983. Responses of an avian predatorand its isopod preyto an acanthocephalanparasite.
Ecology 64: 1000-1015.
1984. Alteredbehavioralresponses in intermediatehosts-an acanthocephalanparasite strategy. American Naturalist123:572-577.
Moore, J., and N. J. Gotelli. 1990. A phylogeneticperspective on the evolution of altered host
behaviors: a critical look at the manipulationhypothesis.Pages 193-233 in C. J. Barnard
and J. M. Behnke, eds. Parasitismand host behavior. Taylor & Francis, London.
Munger,J. C., and W. H. Karasov. 1989. Sublethal parasites and host energybudgets: tapeworm
infectionin white-footedmice. Ecology 70:904-921.
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions
FORAGING ON PARASITIZED PREY
867
Nudds, T. D. 1990. Retroductivelogic in retrospect-the ecological effectsof meningealworms.
Journalof WildlifeManagement54:396-402.
Read, C. P. 1951. The "crowdingeffect"in tapeworminfections.Journalof Parasitology37:174-178.
Schad, G. A. 1966. Immunity,competition,and naturalregulationof helminthpopulations.American
Naturalist100:359-364.
Shapiro, A. 1976. Beau geste? AmericanNaturalist110:900-902.
of pregnancyin the infected
Shaw, G. L., and D. G. Dusanic. 1973. Trypanosomalewisi: termination
rat. ExperimentalParasitology33:46-55.
Slobodkin, L. B. 1974. Prudent predation does not require group selection. American Naturalist
108:665-678.
Smith,N. G. 1968. The advantage to being parasitized. Nature (London) 219:690-694.
Sprent,J. F. A. 1969. Evolutionaryaspects of immunityin zooparasitic infections.Pages 3-62 in
G. J. Jackson, R. Herman, and I. Singer, eds. Immunityto parasitic animals. AppletonNew York.
Century-Crofts,
Stephens,D. W., and J. R. Krebs. 1986. Foragingtheory.PrincetonUniversityPress, Piinceton,N.J.
Symons, L. E. A. 1989. Pathophysiologyof endoparasiticinfection:compared with endoparasitic
infestationand microbialinfection.Academic Press, Sydney.
from
Temple, S. A. 1987. Do predatorsalways capture substandardindividualsdisproportionately
prey populations? Ecology 68:669-674.
Weatherly,N. 1971. Effectson littersize and littersurvivalin Swiss mice infectedwith Trichinella
spiralis duringgestation.Journalof Parasitology57:298-301.
Wright,C. A. 1966. The pathogenesis of helminthsin the Mollusca. HelminthologicalAbstracts
35:201-224.
Associate Editor: Peter Cliesson
This content downloaded from 128.111.205.74 on Fri, 05 Jun 2015 20:27:47 UTC
All use subject to JSTOR Terms and Conditions