J. evol.
Biol.
3: 319-328
(1990)
10l0-061x/90/060319-09
0 1990 Birkhluser
Parasites and sexual selection: Current
Hamilton and Zuk hypothesis
IE 1.50 +0.20/o
Verlag, Base1
status of the
Anders Pape Moller
Dept of Zoology,
Uppsala University,
Box 561, S-7.51 22 Uppsala, Sweden
Key words: Displays; host-parasite cycles; mate choice; parasite avoidance;
sites; resistance to parasites; sexual ornaments; sexual selection.
para-
Abstract
Hamilton and Zuk (1982) suggested that females could continuously base their
choice of mates on heritable resistance to parasites, as generated in a process of
coadaptational
cycles of hosts and their parasites and advertized in terms of
individual
differences in plumage features or displays of the chosen sex. This
hypothesis was based on three assumptions: (1) the full expression of secondary
sexual traits of individual hosts is dependent on their overall health and vigour;
(2) hosts are coevolving with their parasites, thereby continuously
generating
heritable resistance to parasites; and (3) parasites have negative effects on host
viability. Given these assumptions one may then make the following intra- and
interspecific predictions: (1) females will obtain mates with few or no parasites by
choosing males with fully expressed sexual ornaments, (2) males with inferior
ornaments will be less desirable partners and therefore take longer to become
mated, (3) assortative mating occurs with respect to parasite loads, (4) females
which mate with males with poorly developed secondary sexual characters will
copulate outside their pair bond with males showing more fully expressed omaments, and (5) in species especially vulnerable to parasite invasions, sexual selection will lead to more extreme development of health-certifying
traits, allowing
females to more accurately assess the parasite load of a male. Here I shall evaluate
the empirical evidence for these assumptions and predictions and discuss altemative hypotheses advanced, in order to explain the importance of parasites for the
sexual selection theory. An open question not yet studied satisfactorily is whether,
and how often, coadaptational
cycles exist, and, if they exist, how they should
affect mate choice.
319
320
Msller
Introduction
Female choice is composed of two major groups of models, arbitrary traits and
‘good genes’ (Bradbury and Andersson, 1987). One particular kind of ‘good genes’
model suggests that females prefer mates with most extravagant ornaments, because
such males, by wearing the brightest plumage and performing the most vigorous
displays, prove their relative resistance to and therefore freedom from devastating
parasites. Females will thus obtain ‘good’ genes for resistance to parasites for their
offspring (Hamilton
and Zuk, 1982). Compared with other hypotheses, this hypothesis has the great advantage that heritable fitness is generated continuously, and
that strong directional selection by choosy females will not deplete genetic variance
for resistance and for the trait chosen. Freeland (1976) had previously suggested
that parasites and pathogens strongly influenced the evolution of mate choice.
Because hosts continuously evolve genetically determined defence mechanisms to
minimize the effects of parasites, and because parasites continuously evolve new
modes of exploiting the hosts, coevolution between hosts and parasites by frequency-dependent selection will generate stable limit cycles in the gene frequencies
of both hosts and parasites, and thus maintain heritable genetic variation in fitness
(Hamilton,
1982; Eshel and Hamilton, 1984). Not all parasites are equally likely to
generate the kind of host-parasite cycles involved in the Hamilton
and Zuk
hypothesis. For example, neither host nor parasite should be driven to extinction
during the parasite-host interaction. Small differences in viability among both hosts
and parasites will prevent fixation of genotypes, and such variability
may be
particularly likely when many different parasites are coevolving independently with
a single host species, and when many host alleles coding for host resistance are
involved (Eshel and Hamilton,
1984). Ideal parasites, therefore, are those which
have severe effects on juvenile host viability, but persist as chronic infections or as
prolonged after-effects in survivors (Hamilton and Zuk, 1982). The presence of two
or more species of parasites, which differ in their virulence to host genotypes, would
not qualitatively alter the dynamics of the parasite-host system (Hamitlon and Zuk,
1982). Hamilton
and Zuk’s hypothesis was originally supported by a comparative
analysis of the relationship between plumage brightness and infestations by blood
parasites in North American passerine birds. A positive association between incidence of blood infections and extravagance in terms of plumage displays and song
was observed (Hamilton
and Zuk, 1982).
Genetic models of the Hamilton and Zuk hypothesis have clearly demonstrated
that heritable variation in host resistance to parasites increases the probability that
traits revealing the ability of a male to resist parasites will become established and
elaborated in a population (Kirkpatrick,
1986; Charlesworth, 1988; Pomiankowski,
1988). However, the male mating advantage of having a sexual ornament is only
slight or even absent, when female preference for it is rare, and an additional
mechanism is therfore required for increasing the gene frequency of female preference above a certain threshold level and for getting the elaboration process started
within a population. This is a problem common to both the Hamilton and Zuk
hypothesis and several other models of sexual selection (Bradbury and Andersson,
Parasites
and sexual
321
selection
1987). Finally, the parasite hypothesis may work in concert with one or more
additional hypotheses such as other honest advertisement (or handicap) hypotheses
or the run-away process, and if that is the case it would be difficult to make mutually
exclusive predictions able to distinguish between different sexual selection models.
The idea that parasites play an important role in sexual selection has generated
a lot of interest in the scientific community,
and I shall here briefly review the
empirical evidence for the Hamilton
and Zuk hypothesis, first the comparative
interspecific evidence and then the evidence from intraspecific studies. I shall
conclude by briefly discussing some alternative hypotheses about the importance of
parasites for mate choice.
Evidence from comparative
studies
Comparative studies are very important because, like repeated observational or
experimental intraspecific studies, they provide a way to make generalizations about
adaptations and about limits to hypotheses.
According to the Hamilton
and Zuk hypothesis, secondary sexual characters
should be more exaggerated in species particularly vulnerable to parasite invasion,
because extraordinarily
developed ornaments would allow females to evaluate the
parasite burden of males (Hamilton
and Zuk, 1982). A positive association was
found between the prevalence of chronic blood infections and three types of host
displays (male brightness, female brightness, song complexity), which was taken as
evidence in favour of the hypothesis (Hamilton
and Zuk, 1982). At least four
methodological
problems, however, compromised that study: species were used as
independent observations although they were, in many cases, non-independent
because of common ancestry, developmental constraints, and adaptations to common environments (Page1 and Harvey, 1988). Furthermore confounding variables
may have led to spurious correlations, parasite prevalences were affected by sampling
bias, and the reliability of host display scores was not assessed (Read, 1988; Read
and Harvey, 1989).
Reanalysis of the North American and a European sample of passerine birds
controlling for these effects revealed that the associations between parasite loads and
host displays were no longer present (see Read, 1987, 1988; Read and Harvey, 1989;
Zuk, 1989).
One comparative data set on birds still shows a significant positive association
between plumage brightness and blood parasite prevalence. Zuk (1990) controlled
for sampling problems by scoring whether or not each species was above or below
the overall median infection level of all species and then employed a median test to
see whether there was an association between parasite prevalence and plumage
brightness. Effects of phylogeny were controlled by performing within-taxon analyses. Reliability of plumage scores and effects of confounding variables, except from
effects of phylogeny, were not checked. Plumage brightness of males was strongly
positively related to parasite prevalence, particularly among resident species which,
322
MaUer
contrary to migrants, live all the year round with the same set of parasites (Zuk,
1990).
A reanalysis of song complexity in North American passerine birds and the
inclusion of European species, taking phylogeny and habitat structure into account,
clearly demonstrated that phylogeny rather than blood parasite loads was the
important predictor (Read and Weary, 1990).
Comparative analyses of European freshwater fish species and their parasites
revealed a significant positive relationship after controlling for various confounding
variables (Ward, 1988, 1989). However, these analyses suffer from many of the
same methodological
problems as some of the earlier bird studies, and analysis of
an extended data set on fish from North America revealed that the correlation
recorded in the European data set was due to a large extent to uncontrolled
phylogenetic effects (Cabana and Chandler, 1990).
In conclusion, stringent interspecific comparisons have not provided much evidence for the Hamilton and Zuk hypothesis. Since only few parasite and host taxa
have been investigated in these studies, there is still ample space left for the
hypothesis to work (as shown in the next section on intraspecific studies).
What is badly needed at the moment is methodologically
stringent comparative
analyses, preferably encompassing entire parasite communities of host species and
using some independent measure of the overall effects of parasites on their hosts. A
clearly defined and soundly based score of showiness or display conspicuousness is
also needed before any further analyses can be made.
Evidence from intraspecific
studies
In veiw of the large number of methodological
problems affecting comparative
analyses, observational and experimental intraspecific tests of the assumptions and
the predictions of the hypothesis may seem more promising. A number of such tests
have been published, are in press or in progress.
The assumption that the expression of sexual ornaments will reflect parasite
infestation levels was supported in eight out of ten intraspecific tests (Table I),
suggesting that ornament size may in fact reveal male parasite loads to females.
Lack of association between parasite loads and expression of ornaments could be
due to assessment of the ‘wrong’ variable; for example, researchers may have
looked for changes in colouration, while parasites instead affected the size, the
symmetry or the fine structure of sexual ornaments. It has only rarely been
investigated how often females actually obtain parasite-free males by using male
displays as cues (Endler and tyles, 1989), but in one study females only marginally
improved their chances of obtaining a non-infected male by using male colouration
as a mate choice cue (Ressel and Schall, 1989). The assumption of a relationship
between conspicuousness and parasite loads also requires that only traits affected
by the parasites are used by females as mate choice cues, and this appears indeed
to be the case in at least two studies (Zuk et al., 1990; A. P. Msller, unpublished
data).
I. Results
Zuk (1987a,
1987b, 1987~
1987d, 1988)
e
Kennedy
et al.
(1987)
0
( 1990)
Milinski
and
Bakker
E
E
E
Hausfater
et al.
(1990)
0
0
Gray tree
frogs
(1989)
Ressel
and
Schall
0
E
Fence
lizards
and predictions
Hillgarth
(l,m
E
E
E
E
Pheasants
Zuk et al
(19w
E
E
E
E
hypothesis.
or lab
Borgia (1986).
Borgia and
Collis ( 1989)
0
0
0
0
Satin
bowerbirds
based on field
Meller (199Oa.
1990b)
0
E
E
E
Barn swallows
(E or e), when
Clayton
(1990)
O/E?
Rock doves
and small case letters when the evidence is contradictory.
or as experimental
Johnson and
Boyce (@IO),
Spurrier et
al. (1990)
E
Sage grouse
and Zuk
Junglefowl
of the Hamilton
Evidence is listed as either observational
(0 or o), when based on correlational
evidence,
experiments.
Capitals are used when the evidence confirms the assumptions
or the predictions
References
Females choose
males with
fewer parasites
Predictions:
0
0
Negative effect
of parasites
on host fitness
0
Three-spined
sticklebacks
tests of the assumptions
Guppies
O?
o
Field
crickets
of intraspecific
Heritable
resistance
Expression of
ornaments
depends on
parasite load
Assumptions:
Table
324
Maller
Parasites and host should show coadaptational
cycles in order to generate
continuously heritable resistance (Hamilton and Zuk, 1982). It is still premature to
judge whether such coadaptational cycles exist except in the models of theoreticians,
the only known example being the resistance cycle of a crop to a parasitic fungus
over two periods (Barrett, 1988). Field demonstrations of the existence of parasitehost cycles depending on their period will be difficult but are crucial; ideally mate
choice patterns should be examined at different stages of the cycle. Other possible
ways, in which additive genetic variation in resistance is continuously generated, are
discussed by Hamilton ( 1990).
The second assumption, namely, that host resistance to parasitism is partly
heritable, has been studied five times, and in all studies evidence was found for
significantly repeatable or heritable resistance (Table 1). Heritable resistance to
parasites is widespread among domestic animals and includes immunological,
anatomical and physiological mechanisms (Wakelin and Blackwell, 1988). However, the few studies reported in Table 1 represent some of the first field-derived
estimates of genetic resistance to parasites, and confounding factors such as
common environments have been successfully controlled for in some cases.
The third assumption, that parasites have a negative effect on host fitness, was
supported in all of the eight studies where it has been investigated (Table 1).
Although parasitologists often state that evidence for negative fitness effects of
parasites on their hosts is relatively scarce, results of many field studies clearly
indicate that parasites often have severely detrimental effects (e.g. review in Msller
et al., 1990). The fitness advantages accruing to hosts as a result of their mate
choice have not been calculated, but they must be large if host resistance to
parasites is partially heritable.
If males advertize their genetical resistance to parasites through their conspicuousness, and if parasites have negative fitness effects on their hosts, then females
should be able to choose mates which harbour fewer parasites than do the average
male. This very important prediction has been investigated in nine studies, all of
which demonstrated clear female choice of males with few parasites (Table 1). If
females prefer conspicous males with few parasites, the duration of male pre-mating
periods should be directly related to their parasite loads. This was in fact the case
in one observational study of swallows Hirundu rustica (Msller, 1990b). Females of
many animal species mate with a single male, but regularly copulate with additional
males. If females participate in such extra-pair copulations, because preferred males
resistant to parasites are unavailable at the time of mate acquisition, one may
predict that they would preferentially copulate with parasite-free males and particularly so if mated to parasitized males (Hamilton,
1990; Marller, 1990b). This was
actually the case in the monogamous swallow (Mraller, 1990b). Finally, the relative
extent of choosiness in the two sexes should be related to their relative parental
investments (Trivers, 1972). Thus both males and females of monogamous species
should choose mates on the basis of resistance to parasites as revealed by conspicuousness. Individuals therefore should mate assortatively with respect to parasite
loads, as was observed in the swallow (Msller, 1990b).
There is a number of additional predictions, which, however, are not specific for
Parasites
and sexual
selection
325
the Hamilton and Zuk hypothesis. For instance, Borgia and Collis (1989) suggested
that it was possible to predict from the hypothesis that there would be an
age-related decrease in the parasite load of males, if heavily parasitized hosts failed
to survive. However, any such age-related differences can just as well be attributed
to age-related acquired immunity to parasites (Wakelin and Blackwell, 1988) a
phenomenon unrelated to the Hamilton and Zuk hypothesis.
Alternative
hypotheses
An obvious and frequently proposed alternative to the Hamilton
and Zuk
hypothesis is the parasite avoidance hypothesis, suggesting that females simply
avoid mating with infected males because they thereby acquire a direct fitness
advantage which can evolve by natural selection (e.g., Borgia, 1986; Borgia and
Collis, 1989; Hamilton,
1990; Spurrier et al., 1990). This latter hypothesis particularly applies to ectoparasites, venereal diseases and microparasites directly transmitted from one host to another. Clearly, the parasite avoidance hypothesis does not
predict an effect of parasites on the expression of ornaments or displays, neither
does it predict female choice based on such parasite-mediated sexual traits. Data on
satin bowerbirds Ptilonorhynchus violaceus and perhaps other species suggest that
uninfected females, by choosing parasite-free males, may avoid novel pathogens
transmitted by parasites (Borgia and Collis, 1989). Direct avoidance of parasitized
males would still lead to rapid changes in the frequency of resistance genes among
hosts because of sexual selection, althouth this would not explain the presence and
variability in the expression of sexual ornaments. One study has tested whether
females were able to assess loads of ectoparasites, but experimental removal of
feather lice from their sage grouse Cenrrocercus urophasianus hosts revealed that
females did not use the presence of ectoparasites when choosing mates (Spurrier et
al., 1990).
Borgia and Collis (1989) suggested another alternative, the correlated infection
hypothesis, according to which females choose mates with few ectoparasites because
of a correlation between the level of ectoparasite infestation and heritable resistance
to endoparasites. Simultaneous negative effects of many different kinds of parasites
on hosts and interactions and even synergisms between parasites are integral parts
of the Hamilton
and Zuk hypothesis (see Hamilton
and Zuk, 1982), and the
correlated infection hypothesis is therefore not a real alternative to it.
A third alternative is the efficient parent hypothesis suggesting that individuals
mating with a heavily parastized conspecific will acquire a mate unable to provide
high quality parental care (Hamilton,
1990). This idea is not very probable because
it is unlikely that male showiness other than plain cleanliness plays a role in mate
choice; there are plenty of opportunities to assess overall health and vigour at close
distance and during hard work rather than using plumage brightness and song
displays as evidence. Bright males playing an important
role in the rearing of
offspring would also attract the attention of predators to the offspring rather than
increase their probability of survival.
326
MaUer
In conclusion, intraspecific studies of the Hamilton and Zuk hypothesis generally
confirm its assumptions and predictions, whereas most of the comparative evidence
is equivocal. It is therefore crucial that all intraspecific evidence, both negative and
positive, is reported in order to make generalizations possible.
Acknowledgements
I would like to thank N. Hillgarth, M. Milinski, L. Partridge, A. Read, P. S&mid-Hempel and S.
Ulfstrand for discussions or comments on earlier versions of this paper. The study was supported by a
grant from the Swedish Natural Science Research Council.
References
Barrett, J. A. 1988. Frequency-dependent selection in plant-fungal interactions. Phil. Trans. R. Sot.
Lond. B319: 473-483.
Borgia, G. 1986. Satin bowerbird parasites: a test of the bright male hypothesis. Behav. Ecol. Sociobiol.
19: 355-358.
Borgia, G., and K. Collis. 1989. Female choice for parasite-free male satin bowerbirds and the evolution
of bright male plumage. Behav. Ecol. Sociobiol. 25: 445-454.
Bradbury, J. W., and M. B. Andersson. 1987. Sexual s&&ion: testing the alternatives. John Wiley, New
York, USA.
Cabana, G., and M. Chandler. 1990. Lack of evidence for a relationship between parasite load and
sexual dichromatism in North American fish. Anim. Behav. (in press).
Charlesworth, B. 1988. The evolution of mate choice in a fluctuating environment. J. theor. Biol. 130:
I91 -204.
Clayton, D. H. 1990. Mate choice in parasitized rock doves: lousy males lose. In J. E. Loye and M Zuk
(eds.), Ecology, behavior and evolution of bird-parasite interactions. Oxford University Press,
Oxford, UK.
Endler, J. A., and A. M. Lyles. 1989. Bright ideas about parasites. Trends Ecol. Evol. 4: 246-248.
Eshel, I., and W. D. Hamilton. 1984. Parent-offspring correlation in fitness under fluctuating selection.
Proc. R. Sot. Lond. B122: I - 14.
Freeland, W. J. 1976. Pathogens and the evolution of primate sociality. Biotropica 8: 12-24.
Hamilton, W. D. 1982. Pathogens as causes of genetic diversity in their host populations, pp. 269-296.
In R. M. Andersson and R. M. May (eds.), Population biology of infectious disease agents.
Verlag-Chemie, Weinheim.
Hamilton, W. D. 1990. Mate choice near and far. Am. Zool. (in press).
Hamilton, W. D., and M. Zuk. 1982. Heritable true fitness and bright birds: a role for parasites? Science
218: 384-387.
Hausfater, G., H. C. Gerhardt and G. Klump. 1990. Parasites and mate choice in gray treefrogs, Hyla
oersicolor. Am. Zool. (in press).
Hillgarth, N. 1990. Parasites and female choice in the ring-necked pheasant, Phasianus colchicus. Am.
Zool. (in press).
Johnson, L. J., and M. S. Boyce. 1990. Female choice of males with low parasite loads in sage grouse.
In J. E. Loye and M. Zuk (eds.), Ecology, behavior and evolution of bird-parasite interactions.
Oxford University Press, Oxford, UK.
Kennedy, C. E. J., J. A. Endler, S. L. Poynton and H. McMinn. 1987. Parasite load predicts mate choice
in guppies. Behav. Ecol. Sociobiol. 21: 291-295.
Parasites and sexual selection
321
Kirkpatrick, M. 1986. Sexual selection and cycling parasites: a simulation study of Hamitons’s
hypothesis. J. theor. Biol. 119: 263-271.
Milinski,
M., and T. C. M. Bakker.
1990. Female sticklebacks use male coloration in mate choice and
hence avoid parasitized males. Nature. 344: 330-333.
Moller, A. P. 1990a. Effects of an haematophagous mite on the barn swallow (Hirundo rustica): a test
of the Hamilton and Zuk hypothesis. Evolution (in press).
Moller, A. P. 199Ob. Parasites, sexual ornaments and mate choice in the barn swallow Hirundo rusfica.
In J. E. Loye and M. Zuk (eds.), Ecology, behavior and evolution of bird-parasite interactions.
Oxford University Press, Oxford, UK.
Msller, A. P., K. Allander and R. Dufva. 1990. Fitness effects of parasites on passerine birds: a
review. In J. Blondel, A. Gosler, J. D. Lebreton and R. McCleery (eds.), Demographical,
physiological, genetical and behavioral aspects of population biology of passerine birds. Springer
Verlag, Berlin, FRG.
Pagel, M., and P. H. Harvey. 1988. Recent developments in the analysis of comparative data. Q. Rev.
Biol. 63: 413-440.
Pomiankowski, A. N. 1988. The evolution of female mate preferences for male genetic quality. Oxford
SUIT. Evol. Biol 5: 136- 184.
Read, A. F. 1987. Comparative evidence supports the Hamilton and Zuk hypothesis on parasites and
sexual selection. Nature 328: 68-70.
Read, A. F. 1988. Sexual selection and the evolution of parasites, Trends Ecol. Evol. 3: 97-102.
Read, A. F. 1990. Sexual selection and parasites: comparative tests of the Hamilton and Zuk
hypothesis. Am. Nat. (in press).
Read, A. F., and P. H. Harvey. 1989. Reassessment of comparative evidence for Hamilton and Zuk
theory on the evolution of secondary sexual characters. Nature 339: 618-620.
Read, A. F., and D. M. Weary. 1990. Sexual selection and the evolution of bird song: a test of the
Hamilton-Zuk hypothesis. Behav. Ecol. Sociobiol. 26: 47-56.
Ressel, S., and J. J. Schall. 1989. Parasites and showy males: malarial infection and color variation in
fence lizards. Oecologla 78: 158-164.
Spurrier, M. F., M. S. Boyce, and B. F. J. Manly. 1990. Effects of parasites on mate choice by captive
sage grouse Centrocercus
urophasianus.
In J. E. Loye and M. Zuk (eds.), Ecology, behavior and
evolution of bird-parasite interactions. Oxford University Press, Oxford, UK.
Trivers, R. L. 1972. Parental investment and sexual selection, pp. 136-179. In B. Campbell (cd.),
Sexual selection and the descent of man, 1871-1971. Aldine Press, Chicago, IL.
Wakelin, D., and J. M. Blackwell (eds.). 1988. Genetics of resistance to bacteria1 and parasitic
infection. Taylor and Francis, London, UK.
Ward, P. I. 1988. Sexual dichromatism
and parasitism
in British and Irish freshwater fish. Anim.
Behav. 36: 1210-1215.
Ward, P. I. 1989. Sexual showiness and parasitism in freshwater fish: combined data from several
isolated water systems, Oikos 55: 428-429.
Zuk, M. 1987a. The effects of gregarine parasites, body size, and time of day on spermatophore
production and sexual selection in field crickets. Behav. Ecol. Sociobiol. 21: 65-72.
Zuk, M. 1987b. Variability in attractiveness of male field crickets (Orthoptera: Gryllidae) to females.
Anim. Behav. 35: 1240-1248.
Zuk, M. 1987~. Seasonal and individual variation in gregarine parasite levels in the field crickets
Gryllus veletti and G. pennsylvanicus.
Ecol. Entomol. 12: 341-348.
Zuk, M. 1987d. The effects of greg%ne parasites on longevity, weight loss, fecundity and developmental time in the field cricket Gryilus Ye/ens and G. pennsylvanicus.
Ecol. Entomol. 12: 349-354.
Zuk, M. 1988. Parasite load, body size, and age of wild-caught male field crickets (Orthoptera:
Gryllidae): effects on sexual selection. Evolution 42: 969-976.
Zuk, M. 1989. Validity of sexual selection in birds. Nature 340: 104-105.
Zuk, M. 1990. Parasites and bright birds: new data and a new prediction. In J. E. Loye and M. Zuk
(cds.), Ecology, behavior and evolution of bird-parasite interactions. Oxford University Press,
Oxford, UK.
328
Zuk,
Msller
M., R. Thomhill,
J. D. Ligon
Am. Zool. (in press).
Received
5 January
1990;
accepted 1 April
1990.
Corresponding
Editor:
G. Wagner
and K. Johnson.
1990. Parasites
and mate choice
in red junde
fowl.