Cuckoos versus hosts in insects and birds: adaptations, counter‐adaptations and outcomes

TL;DR: An adaptive explanation of co‐evolution between brood parasites and their hosts is proposed, which centres on the relative strength of two opposing processes: strategy‐facilitation, in which one line of host defence promotes the evolution of another form of resistance, and strategy‐blocking, which may relax selection on another so completely that it causes it to decay.

Abstract: Avian parents and social insect colonies are victimized by interspecific brood parasites-cheats that procure costly care for their dependent offspring by leaving them in another species' nursery. Birds and insects defend themselves from attack by brood parasites; their defences in turn select counter-strategies in the parasite, thus setting in motion antagonistic co-evolution between the two parties. Despite their considerable taxonomic disparity, here we show striking parallels in the way that co-evolution between brood parasites and their hosts proceeds in insects and birds. First, we identify five types of co-evolutionary arms race from the empirical literature, which are common to both systems. These are: (a) directional co-evolution of weaponry and armoury; (b) furtiveness in the parasite countered by strategies in the host to expose the parasite; (c) specialist parasites mimicking hosts who escape by diversifying their genetic signatures; (d) generalist parasites mimicking hosts who escape by favouring signatures that force specialization in the parasite; and (e) parasites using crypsis to evade recognition by hosts who then simplify their signatures to make the parasite more detectable. Arms races a and c are well characterized in the theoretical literature on co-evolution, but the other types have received little or no formal theoretical attention. Empirical work suggests that hosts are doomed to lose arms races b and e to the parasite, in the sense that parasites typically evade host defences and successfully parasitize the nest. Nevertheless hosts may win when the co-evolutionary trajectory follows arms race a, c or d. Next, we show that there are four common outcomes of the co-evolutionary arms race for hosts. These are: (1) successful resistance; (2) the evolution of defence portfolios (or multiple lines of resistance); (3) acceptance of the parasite; and (4) tolerance of the parasite. The particular outcome is not determined by the type of preceding arms race but depends more on whether hosts or parasites control the co-evolutionary trajectory: tolerance is an outcome that parasites inflict on hosts, whereas the other three outcomes are more dependent on properties intrinsic to the host species. Finally, our review highlights considerable interspecific variation in the complexity and depth of host defence portfolios. Whether this variation is adaptive or merely reflects evolutionary lag is unclear. We propose an adaptive explanation, which centres on the relative strength of two opposing processes: strategy-facilitation, in which one line of host defence promotes the evolution of another form of resistance, and strategy-blocking, in which one line of defence may relax selection on another so completely that it causes it to decay. We suggest that when strategy-facilitation outweighs strategy-blocking, hosts will possess complex defence portfolios and we identify selective conditions in which this is likely to be the case.

Summary

I. INTRODUCTION

• This is especially evident among the cooperative behaviours that centre on the rearing of dependent kin, because they are performed by adults at some personal cost (Bourke & Franks, 1995; Clutton-Brock, 1991) but are exploited by brood parasites seeking to have their offspring raised for free.
• The antagonistic interactions of avian obligate brood parasites and their hosts have therefore become a model system for the study of co-evolution (Rothstein & Robinson, 1998) .
• In some cases, such as with hosts of cuckoos or slave-making ants, the brood parasite reduces host fecundity directly by removing host young from the nest.
• It is therefore interesting to examine just how much these two systems have in common with each other.
• The first question asks how the co-evolutionary arms race proceeds.

(a) Directional selection on traits for defence and attack

• In some cases, the co-evolutionary arms race of defence, counter-attack and counter-defence is reminiscent of coevolution in a classical predator-prey arms race.
• Initially, there is directional selection for the host to defend itself against attack from the parasite.
• There is evidence from the insects for each of these three stages in the arms race.
• Only some hosts of Sphecodes cuckoo bees exhibit defences and fighting behaviour when the parasitic bee attempts to enter the nest, suggesting that these traits do not pre-date an association with brood parasites and that some hosts have only reached the first stage of the arms race (Bogusch et al., 2006) .
• Likewise, some species of Polistes cuckoo wasps battle their way into the host colony and they possess specially enlarged and strengthened head, mandible and leg segments for this purpose (Cervo, 2006) .

(b) Evading front-line defences through secrecy

• In some cases, parasites switch from attempting to out-gun host front-line defences to evading hosts simply by avoiding further confrontation.
• Having bludgeoned its way past host front-line defences, or circumvented them by more subtle means, the parasite sets about commandeering host resources for its own reproduction.
• Most of the variation in egg appearance is controlled genetically and individual females lay eggs of a consistent phenotype throughout their lives, whether they are hosts or parasites (reviewed by Kilner, 2006) .
• If cuckoos can keep up with their hosts as they chase through signature space, and cuckoo egg mimicry becomes more and more refined, hosts are more likely to make discrimination errors and mistakenly reject their own eggs instead of the cuckoo's, sometimes removing eggs from clutches that are not even parasitized (Marchetti, 1992) .
• Social insect parasites are versatile mimics of these hydrocarbon signatures (e.g. Lenoir et al., 2001; Martin et al., 2010a) , and in some cases mimicry is due to the biosynthesis of host-specific signatures prior to parasitism (e.g. Martin et al., 2010a, b) .

(b) Forgery of the host signature after parasitism: generalist parasites

• It is common for insect brood parasites to adopt a strategy of chemical camouflage and acquire the colony-specific hydrocarbon signature after parasitism.
• The principal co-evolutionary consequence of forging the host's signature after parasitism, rather than expressing it beforehand, is that parasites can be individual generalists, capable of flexibly adapting to exploit any of their hosts.
• Insect hosts place parasites under selection to refine their mimicry of the host hydrocarbon signature which, in some cases, gives rise to parasites that become more and more chemically invisible themselves, effectively presenting a blank slate to be daubed with their hosts' particular hydrocarbons (Brandt et al., 2005a; D'Ettore & Errrard, 1998; Lenoir et al., 2001) .
• Whereas closely related ant species usually have similar hydrocarbon profiles, hosts L. muscorum and L. acervorum are unusually distinct, suggesting that they have diversified under selection from their slave-making parasite.
• Newly hatched Horsfield's bronze-cuckoo nestlings beg like nestlings of their primary malurid hosts, but can modify their calls if they find themselves being raised by a secondary host (Langmore et al., 2008) .

(c) Escaping host recognition through crypsis

• A third way in which parasites can evade the host recognition system is effectively to become invisible.
• Marchant (1972) speculated that this egg colouration had been selected for crypsis in the dark domed nests of the typical Chalcites cuckoo host.
• If parasites reduce the complexity of their hydrocarbon signatures to evade host detection, then hosts can render parasites detectable only if they themselves have simplified their signature to a greater extent.
• The authors classified species as either 'co-evolved' (current hosts and parasites) or 'not exposed to co-evolution' (current non-hosts) and, using Levene's tests, compared the variance in six types of cuticular hydrocarbons (Table 1 ) between the two categories of species.

(d) Traits that prevent rejection, rather than recognition

• If parasites fail to fool their host's recognition system, they could still persist in the host nest if they nevertheless manage to avoid being rejected.
• Shell strengthening thus seems to have evolved in direct response to host egg rejection behaviour, a conclusion further bolstered by intraspecific analyses of common cuckoo eggs.
• The relationship between co-evolution and the extent of diversity in different components of the ant cuticular hydrocarbon signature.
• The greater the loss in fecundity sustained by 'old cowbird hosts' as a consequence of parasitism, the more likely the incidence of clutch desertion (Hosoi & Rothstein, 2000) .
• In several systems, parasites seem to employ multiple strategies for evading recognition and resisting rejection (see also Section III.2).

III. OUTCOMES OF CO-EVOLUTION

• Having identified ways in which co-evolution might proceed, the authors now turn to the second aim of this review, namely to identify common outcomes of co-evolution.
• Four alternatives are apparent and they cut across the specific co-evolutionary arms races identified above.

(1) Successful resistance by hosts

• Further indirect evidence that avian host recognition systems can defeat cuckoos is provided by the error-free rejection of oddly marked experimental eggs by hosts that are currently not exploited by brood parasites.
• Slave rebellion benefits the host, not through any improvement in direct fitness but because Temnothorax spp. populations are highly kin-structured, and relatives at neighbouring freeliving colonies are spared from attack by the slavemaker as a consequence.
• Co-evolutionary arms races between brood parasites and their hosts can yield four possible outcomes for the host (successful resistance, the evolution of defence portfolios, acceptance, and tolerance; cf. Davies, 2000, p.119) .
• Empirical work suggests that the particular outcome is not necessarily contingent on the type of preceding co-evolutionary arms race (as identified in Section II).

IV. CONCLUSIONS

• (1) Despite their considerable taxonomic disparity, there are striking parallels in the way that co-evolution proceeds between brood parasites and their hosts in the insects and the birds.
• (2) Five types of co-evolutionary arms race can be identified from the empirical literature, which are common to both systems.
• (3) Evidence from the better studied brood parasites and their hosts suggests that several types of co-evolutionary arms race, each focused on different modes of host defence, can play out concurrently between a single host and its brood parasite.
• Nevertheless, there is considerable interspecific variation in the complexity and depth of host defence portfolios.
• Whether this variation is adaptive, or merely reflects evolutionary lag, is unclear.

Figures

Fig. 2. The relationship between co-evolution and the extent of diversity in different components of the ant cuticular hydrocarbon signature. ‘Co-evolved species’ includes 26 species of host and parasitic ant, ‘species that have not co-evolved’ includes 52 ant species not currently described as host to a brood parasite of any sort (see text and Table 1 for details). For each species, Martin & Drifjhout (2009) describe the number of different sorts of hydrocarbon present in the cuticle. The standard deviation of this number, calculated using the species in each category, is shown on the y-axis.

Table 1. The number of different n alkanes, alkenes, dienes, monomethylalkanes, dimethylalkanes and trimethylalkanes exhibited by 78 species of ant, with respect to their exposure to co-evolution resulting from social parasitism. Hydrocarbon data were taken from Martin & Drijfhout (2009). ‘Ant status’ refers to the mode of reproduction shown by each species, and was collated from information presented in Hölldobler & Wilson (1990); Lenoir et al. (2001) and Buschinger (2009). Species not currently described as hosts (non-hosts) were classified as ‘not co-evolved’. All other species were classified as ‘co-evolved’

Table 2. The results of a Levene’s test comparing variance in the number of different hydrocarbon types in species that have co-evolved as a result of social parasitism with species that have no (known) history of co-evolution with brood parasites. Statistically significant differences are shown in bold and are caused by lower variances in the ‘co-evolved’ species group (see Fig. 2)

Fig. 1. How co-evolution between insect brood parasites and their hosts might reduce cuticular hydrocarbon diversity. Hosts start by possessing a unique hydrocarbon signature for identifying members of their own species (the bar symbolizes the insect cuticle, while each shaded circle is a different sort of hydrocarbon). The parasite’s hydrocarbon signature does not initially match the host’s but the parasite can evade detection by simplifying its signature so that no part of it stands out when compared with the host. By simplifying its own signature in response, the host can expose the parasite. This process continues until the parasite has no signature of its own and is chemically invisible or ‘insignificant’. The net result is reduced cuticular hydrocarbon diversity in both the parasite and the host.

Full text

Biol. Rev. (2011), 86, pp. 836–852. 836
doi: 10.1111/j.1469-185X.2010.00173.x
Cuckoos versus hosts in i nsects and birds:
adaptations, counter-adaptations
and outcomes
Rebecca M. Kilner
1∗
and Naomi E. Langmore
2
1
Department of Zoology, University of Cambridge, CB2 3EJ, United Kingdom
2
Research School of Biology, Australian National University, Canberra 0200, Australia
ABSTRACT
Avian parents and social insect colonies are victimized by interspecific brood parasites—cheats that procure costly
care for their dependent offspring by leaving them in another species’ nursery. Birds and insects defend themselves
from attack by brood parasites; their defences in turn select counter-strategies in the parasite, thus setting in motion
antagonistic co-evolution between the two parties. Despite their considerable taxonomic disparity, here we show striking
parallels in the way that co-evolution between brood parasites and their hosts proceeds in insects and birds. First, we
identify five types of co-evolutionary arms race from the empirical literature, which are common to both systems. These
are: (a) directional co-evolution of weaponry and armoury; (b) furtiveness in the parasite countered by strategies in the
host to expose the parasite; (c) specialist parasites mimicking hosts who escape by diversifying their genetic signatures;
(d) generalist parasites mimicking hosts who escape by favouring signatures that force specialization in the parasite;
and (e) parasites using crypsis to evade recognition by hosts who then simplify their signatures to make the parasite
more detectable. Arms races a and c are well characterized in the theoretical literature on co-evolution, but the other
types have received little or no formal theoretical attention. Empirical work suggests that hosts are doomed to lose
arms races b and e to the parasite, in the sense that parasites typically evade host defences and successfully parasitize
the nest. Nevertheless hosts may win when the co-evolutionary trajectory follows arms race a, c or d. Next, we show
that there are four common outcomes of the co-evolutionary arms race for hosts. These are: (1) successful resistance;
(2) the evolution of defence portfolios (or multiple lines of resistance); (3) acceptance of the parasite; and (4) tolerance
of the parasite. The particular outcome is not determined by the type of preceding arms race but depends more on
whether hosts or parasites control the co-evolutionary trajectory: tolerance is an outcome that parasites inflict on hosts,
whereas the other three outcomes are more dependent on properties intrinsic to the host species. Finally, our review
highlights considerable interspecific variation in the complexity and depth of host defence portfolios. Whether this
variation is adaptive or merely reflects evolutionary lag is unclear. We propose an adaptive explanation, which centres
on the relative strength of two opposing processes: strategy-facilitation, in which one line of host defence promotes the
evolution of another form of resistance, and strategy-blocking, in which one line of defence may relax selection on
another so completely that it causes it to decay. We suggest that when strategy-facilitation outweighs strategy-blocking,
hosts will possess complex defence portfolios and we identify selective conditions in which this is likely to be the case.
Key words: social parasite, co-evolution, arms race, cowbird, slave-making ant, Polistes, virulence, chemical insignificance,
hydrocarbon, recognition system.
CONTENTS
I. Introduction ................................................................................................ 837
II. Types of co-evolutionary arms race ........................................................................ 838
(1) Front-line parasite attack and host defence ............................................................ 838
* Address for correspondence E-mail: rmk1002@cam.ac.uk
Biological Reviews 86 (2011) 836–852 © 2011 The Authors. Biological Reviews © 2011 Cambridge Philosophical Society

Cuckoos versus hosts in insects and birds 837
(a) Directional selection on traits for defence and attack ............................................... 838
(b) Evading front-line defences through secrecy ........................................................ 838
(2) Host recognition systems ............................................................................... 839
(a) Forgery of the host signature before parasitism: specialist parasites ................................. 839
(b) Forgery of the host signature after parasitism: generalist parasites .................................. 840
(c) Escaping host recognition through crypsis .......................................................... 841
(d) Traits that prevent rejection, rather than recognition ............................................... 842
(3) Do co-evolutionary arms races follow predictable trajectories? ........................................ 845
III. Outcomes of co-evolution .................................................................................. 845
(1) Successful resistance by hosts .......................................................................... 845
(2) Further resistance by hosts: defence portfolios ......................................................... 846
(3) Acceptance of the parasite ............................................................................. 847
(4) Tolerance of the parasite ............................................................................... 848
(5) Do co-evolutionary arms races yield predictable outcomes? ........................................... 849
IV. Conclusions ................................................................................................ 849
V. Acknowledgements ......................................................................................... 850
VI. References .................................................................................................. 850
I. INTRODUCTION
Cooperation of any sort is usually costly and is therefore
vulnerable to cheating. This is especially evident among
the cooperative behaviours that centre on the rearing of
dependent kin, because they are performed by adults at some
personal cost (Bourke & Franks, 1995; Clutton-Brock, 1991)
but are exploited by brood parasites seeking to have their
offspring raised for free. The incentive to cheat is so strong
that brood parasitism has arisen both within and between
species on many independent occasions. The interspecific
brood parasites in particular are taxonomically diverse as
well as numerous and span the birds, frogs, fish and insects
(Brown, Morales & Summers, 2009; Davies, 2000; Davies,
Bourke & Brooke, 1989; Sato, 1986).
Obligate interspecific brood parasitism is particularly well
known among the birds, where it has evolved independently
in seven different clades (Sorenson & Payne, 2005), yielding
roughly 100 parasitic species that are scattered across the
world (Davies, 2000). These brood parasites typically lay an
egg in a nest belonging to another species and then abandon
it, first to be incubated by the hosts, and then reared to
independence after hatching. From the host’s perspective,
brood parasitism is costly because at the very least it reduces
their current fecundity to some extent (but see Lyon &
Eadie, 2004). The costs of parasitism select hosts that can
defend themselves against attack by the parasite, and host
defences reciprocally select counterstrategies in the brood
parasite. The antagonistic interactions of avian obligate
brood parasites and their hosts have therefore become a
model system for the study of co-evolution (Rothstein &
Robinson, 1998).
Interspecific brood parasitism is also well documented
among the insects (where it is often referred to as ‘social
parasitism’). Here the parasites target their attack especially
on the social insects, such as wasps, bees, bumblebees and
ants, and they appropriate a colony’s workforce to rear
their own young (Brandt et al., 2005a; Buschinger, 2009;
Cervo, 2006; Davies et al., 1989; Dronnet et al., 2005; Pierce
et al
., 2002). Among the slavemaker ants, and some species
of parasitic wasp (Cervo, 2006) this involves taking over a
host colony and then launching multiple secondary raids
on neighbouring colonies to steal host pupae, who are
enslaved to rear yet more parasitic young (Brandt et al.,
2005a). Although the natural history differs markedly, there
are strong conceptual similarities between brood parasitism
in birds and insects. In each case, at the very least, the victims
of the brood parasite are forced to divert costly care away
from kin towards rearing unrelated parasitic young. In some
cases, such as with hosts of cuckoos or slave-making ants, the
brood parasite reduces host fecundity directly by removing
host young from the nest. There is now evidence from
diverse social insect systems that victims defend themselves
against parasitism, and that their defences have selected
counter-adaptations in the parasite (e.g. Bogusch, Kratochvil
& Straka, 2006; Brandt et al., 2005a; Cervo, 2006; Martin,
Helanter
¨
a & Drijfhout, 2010b). Just as in birds, insect brood
parasites and their hosts co-evolve.
Despite their taxonomic disparity, co-evolution with
brood parasites exposes social insects and avian parents
to convergent selective pressures. It is therefore interesting to
examine just how much these two systems have in common
with each other. The aim of this review is to address two
broad questions, which have not been considered in previous
comparisons of avian and insect brood parasites. The first
question asks how the co-evolutionary arms race proceeds.
Do certain sorts of host defences predictably select certain
sorts of counter-adaptations in the parasite, for example,
and therefore do we see the same types of host defences
and parasite counter-adaptations in both the birds and the
insects? The second question addresses the outcome of co-
evolution. Are there predictable endpoints, common to both
insects and birds, and are some arms races more likely
to favour victory for the parasite rather than the host (or
vice versa)? Although we draw on diverse studies from both
insects and birds to answer these questions, here we have
not attempted an exhaustive survey of the vast literature on
co-evolution in each taxonomic system. Our focus instead
Biological Reviews 86 (2011) 836–852 © 2011 The Authors. Biological Reviews © 2011 Cambridge Philosophical Society

838 Rebecca M. Kilner and Naomi E. Langmore
is on common concepts. We apologise at the outset to the
many researchers whose work we were unable to include in
this review simply for reasons of brevity.
II. TYPES OF CO-EVOLUTIONARY ARMS RACE
(1) Front-line parasite attack and host defence
Bird and social insect nurseries (see Mock & Parker, 1997
for a definition of ‘nursery’) are extremely well defended by
their owners, so the parasite’s first task in appropriating this
resource commonly involves breaching the various physical
lines of defence that protect the nest.
(a) Directional selection on traits for defence and attack
In some cases, the co-evolutionary arms race of defence,
counter-attack and counter-defence is reminiscent of co-
evolution in a classical predator-prey arms race. Initially,
there is directional selection for the host to defend itself
against attack from the parasite. Host defences then select for
improved armoury in the parasite which, in turn, places hosts
under directional selection to overcome this improvement
in the parasite (Barrett, Rogers & Schluter, 2008; Nuismer,
Ridenhour & Oswald, 2007). There is evidence from the
insects for each of these three stages in the arms race. For
example, only some hosts of Sphecodes cuckoo bees exhibit
defences and fighting behaviour when the parasitic bee
attempts to enter the nest, suggesting that these traits do
not pre-date an association with brood parasites and that
some hosts have only reached the first stage of the arms
race (Bogusch et al., 2006). Evidence for the second stage
comes from Bombus (Psithyrus) cuckoo bumblebees, which
now possess thicker cuticles and longer stings than their hosts
to enable the parasite to breach host defences effectively
(Fisher & Sampson, 1992). Likewise, some species of Polistes
cuckoo wasps battle their way into the host colony and they
possess specially enlarged and strengthened head, mandible
and leg segments for this purpose (Cervo, 2006). Finally,
Polistes dominulus hosts of the cuckoo wasp Polistes sulcifer,
have reached the third stage of this arms race. In response
to parasitism, these hosts have apparently increased their
body size because wasps from parasitized populations are
larger in almost every respect than those from unparasitized
populations (Ortolani & Cervo, 2010).
(b) Evading front-line defences through secrecy
In some cases, parasites switch from attempting to out-gun
host front-line defences to evading hosts simply by avoiding
further confrontation. For example, a queen of the slave-
making ant Polyergus rufescens withstands initial vicious attack
from host Formica sanguinea workers, as she penetrates a host
colony, with an integument that is apparently especially
thickened for this purpose (Mori et al., 2000). The parasitic
queen then launches a brief and violent counter-attack from
which she emerges unscathed but which causes host workers
to lose antennae or legs (Mori, D’Ettorre & Le Moli, 1995).
The parasite’s success in counter-attacking is largely due
to a secretion from her Dufour’s gland, which acts as an
appeasement allomone and greatly reduces the incidence of
worker aggression. This further enables the parasitic queen to
move freely to attack the host queen (Mori et al., 2000) whom
she quickly kills with bites to the head, thorax and gaster (Mori
et al., 1995). Similar deployment of appeasement allomones
is observed in the congeneric slavemaker P. sumarai when
attacking host F. japonica colonies (Tsuneoka & Akino, 2009).
Avoidance of confrontation is taken to a greater level in
Sphecodes cuckoo bees and in the common cuckoo Cuculus
canorus. When parasitizing some hosts, for example, cuckoo
bees will only enter the host nest when the host female
is absent, and will sometimes even wait nearby until the
host has departed (Bogusch et al., 2006). Likewise, to avoid
mobbing by their hosts (Welbergen & Davies, 2009), common
cuckoos are exceptionally furtive around the host nest (Davies
& Brooke, 1988). Like the cuckoo bees, common cuckoo
females choose to visit the nest when the host is absent,
and they also lay their egg in the early afternoon rather
than the morning to avoid encountering the nest owner as
she herself lays an egg. In addition, the time spent by the
cuckoo at the host nest is very brief, because the act of
egg-laying is so rapid, and this too minimizes the likelihood
of a confrontation between the parasite and its host (Davies
& Brooke, 1988). Great-spotted cuckoos Clamator glandarius
are just as secretive around the nest because they risk serious
injury from attack if discovered by their larger corvid hosts.
In this brood parasite, males seemingly distract hosts away
from the nest with conspicuous calling behaviour as their
mate quietly glides to the host nest and quickly adds an egg
of her own (Davies, 2000).
While cuckoos are under selection to avoid coming face
to face with their hosts, there is some evidence that hosts are
counter-selected to increase the chance of a confrontation
with their parasite. For example, the loud host alarm calls
triggered by the presence of an avian brood parasite near
the nest attract the attention of nearby conspecifics and
even heterospecifics who join in mobbing the parasite until it
leaves the nest’s vicinity (Trivers, 1971; Welbergen & Davies,
2009). Mobbing behaviour appears to have counter-selected
for mimicry in adult common cuckoos, who now resemble
hawks with their barred chest plumage. Barring seemingly
induces fear in potential hosts, which limits the extent of
their mobbing, and thereby conceals the cuckoo’s presence
from at least some members of the host population (Davies
& Welbergen, 2008).
The entrance tubes that some Ploceus hosts of the diederik
cuckoo Chrysococcyx caprius weave on the front of their nests
may also function to render the parasite more apparent
to hosts. The tubes’ diameter is sufficiently narrow to
prevent the cuckoo from gaining rapid access to the nest
and there are anecdotal reports of hosts attacking diederik
cuckoos that become trapped as they attempt to sneak into
the host nest (Davies, 2000). Nevertheless, the entrance
tubes offer only a limited deterrent to parasites and Ploceus
species whose nests possess such structures are still frequent
Biological Reviews 86 (2011) 836–852 © 2011 The Authors. Biological Reviews © 2011 Cambridge Philosophical Society

Cuckoos versus hosts in insects and birds 839
victims of the diederik cuckoo (Davies, 2000). Host front-line
defences can be effective means of deterring brood parasites
(e.g. Welbergen & Davies, 2009), sometimes even forever
(e.g. Mori et al., 1995; Ortolani & Cervo, 2010). Nevertheless,
empirical evidence to date suggests that defences on the front
lines are doomed to fail if the parasite responds by using
subterfuge rather than direct confrontation.
(2) Host recognition systems
Having bludgeoned its way past host front-line defences, or
circumvented them by more subtle means, the parasite
sets about commandeering host resources for its own
reproduction. At this point it encounters the elaborate
recognition systems used by hosts to protect the nursery’s
resources from marauders and these become the focus of
further co-evolution between the parasite and the host.
Theoretical genetic analyses show that parasites are placed
under selection to mimic their hosts, effectively forging their
host’s signature, which in turn selects hosts that escape
this mimicry through diversification and elaboration of their
signature (Gavrilets, 1997; Kopp & Gavrilets, 2006; Nuismer,
Doebeli & Browning, 2005; Takasu, 2003). Parasites thus
chase their hosts through signature space, sometimes in
circles and sometimes in branching linear trajectories,
depending on the particular assumptions of the theoretical
model, and polymorphisms in host signatures and parasite
forgeries are a common predicted outcome. As we shall
see, case studies from both the avian and insect social
parasites support these general predictions from theory and
also reveal co-evolutionary trajectories not yet imagined by
theoretical work. Importantly, the precise way in which the
parasite eludes the recognition system critically affects the
co-evolutionary arms race that ensues.
(a) Forgery of the host signature before parasitism: specialist parasites
This section considers evasion of recognition through forged
signatures in the parasite that are present before parasitism,
and that are probably inherited genetically. Among the
birds, eggshell colour and patterning are common signatures
of offspring identity. Although environmental conditions can
induce small variations in the precise colour and pattern
adorning an egg, most of the variation in egg appearance
is controlled genetically and individual females lay eggs of
a consistent phenotype throughout their lives, whether they
are hosts or parasites (reviewed by Kilner, 2006). Avian
egg signatures have been especially well characterized in
the many hosts of the common cuckoo. In general, hosts
discriminate against eggs that look odd by comparison with
their own and the greater the discrepancy in appearance,
the more likely they are to reject the egg (Brooke & Davies,
1988; Lahti, 2006; Moksnes, Røskaft & Braa, 1995). Egg
discrimination is a co-evolved response to parasitism because
species that have no evolutionary history of interaction with
the cuckoo lack this ability (Davies & Brooke, 1989). The
signatures themselves, the diversities of egg colouring and
the intricacies of egg patterning, are also an evolved response
to parasitism because former cuckoo hosts that are no longer
exposed to parasitism are less variably coloured and less
elaborately maculated (Lahti, 2005). Egg recognition and
rejection has in turn driven the evolution of cuckoo egg
mimicry: the more discriminating the host, the closer the
match between cuckoo and host eggs (Aviles et al., 2010;
Cassey et al., 2008; Spottiswoode & Stevens, 2010; Stoddard
& Stevens, 2010). The most recent work in this area takes
account of the fact that bird visual systems differ markedly
from our own, most notably by extending into the ultraviolet,
and uses techniques of avian visual modelling to quantify the
extent of mimicry in colour and pattern through the eyes
of a bird, the intended perceiver of the eggshell signature.
It has revealed finely tuned levels of mimicry in colour and
pattern that are essentially cryptic to the human eye (Aviles,
2008; Aviles et al., 2010; Cassey et al., 2008; Spottiswoode
& Stevens, 2010; Stoddard & Stevens, 2010). So there
is compelling evidence that discrimination and rejection
by hosts has driven the evolution of exquisite cuckoo egg
mimicry, at least in some instances (but see Moksnes et al.,
1995). One consequence has been that the common cuckoo
has split into genetically distinct host-specific lines, each
specializing on one host by laying an egg that resembles
their host’s clutch (Fossøy et al., 2011; Gibbs et al., 2000).
Mimetic cuckoo eggs have, in turn, caused hosts to diversify
their egg signatures. In general, there is more variation in
egg appearance among clutches of parasitized populations
than is seen in eggs laid by populations that have never been
exposed to brood parasitism (reviewed by Kilner, 2006). In
one host of the common cuckoo, the ashy-throated parrotbill
Paradoxornis alphonsianus from China, clutches have diversified
so much that hosts now possess an egg polymorphism (Yang
et al., 2010).
How have cuckoos responded to this increase in host
egg diversity? Genetic analyses show that there are several
mtDNA haplotypes within each host race, suggesting
that cuckoos routinely switch between hosts, perhaps
when temporarily defeated by increased diversification in
egg signatures produced by their former hosts, and the
corresponding increase in host discrimination that results
(Davies & Brooke, 1989; Gibbs et al., 2000; Marchetti, 2000).
If cuckoos can keep up with their hosts as they chase through
signature space, and cuckoo egg mimicry becomes more and
more refined, hosts are more likely to make discrimination
errors and mistakenly reject their own eggs instead of the
cuckoo’s, sometimes removing eggs from clutches that are
not even parasitized (Marchetti, 1992). At some point, when
rejection costs start to outweigh the benefits of discrimination,
hosts can gain greater fitness on average by accepting all eggs,
even the occasional cuckoo, especially at very low levels of
parasitism (Brooke, Davies & Noble, 1998; Davies, Brooke &
Kacelnik, 1996; Langmore & Kilner, 2009; Marchetti, 1992).
Consequently, hosts start to benefit by using phenotypically
plastic discrimination rules, only showing egg rejection at
high levels of parasitism when the benefits outweigh any
associated costs (e.g. Brooke et al., 1998; Hauber, Moskat
& Ban, 2006; Langmore et al., 2009a; Rodriguez-Girones &
Lotem, 1999). As we shall see in Section III.3, recognition
Biological Reviews 86 (2011) 836–852 © 2011 The Authors. Biological Reviews © 2011 Cambridge Philosophical Society

840 Rebecca M. Kilner and Naomi E. Langmore
costs associated with egg rejection are crucial in determining
the outcome of co-evolution.
Recent work suggests that the equivalent recognition-
based co-evolutionary interactions of social insects and their
parasites are likely to follow remarkably similar trajectories to
those identified in cuckoos and their hosts. Here, recognition
centres on hydrocarbon signatures in the insect (or egg)
cuticle, in particular the alkenes, di- and trimethylalkanes
(Martin, Helanter
¨
a & Drijfhout, 2008a), which can occur
in a number of positional isomers and so can readily
encrypt information about colony or species identity (Lenoir
et al., 2001; Martin et al., 2010a, b; Martin & Drijfhout,
2009). Social insect parasites are versatile mimics of these
hydrocarbon signatures (e.g. Lenoir et al., 2001; Martin
et al., 2010a ), and in some cases mimicry is due to the
biosynthesis of host-specific signatures prior to parasitism
(e.g. Martin et al., 2010a, b). Among bumblebees (Bombus
spp.), for example, hosts discriminate against individuals
lacking the correct hydrocarbon signature, and the greater
the mismatch, the more violent their reaction (Dronnet
et al., 2005). To escape host recognition, different species of
Bombus (Psithyrus ) cuckoo bumblebees accurately reproduce
the different alkene isomer profiles of their particular Bombus
hosts, and these are apparently present in the cuckoo
bumblebee cuticle before it enters the host colony (Martin
et al., 2010a). So just as with the avian cuckoos, hosts
discriminate against individuals that are unlike their own
kind and this has selected parasites that can genetically forge
the host hydrocarbon signature and so evade recognition.
And, just like common cuckoo hosts, insect cuckoo hosts
appear to respond to mimicry by diversifying their signatures.
Populations of host Formica fusca ants that are exposed to
parasites possess more diverse hydrocarbon signatures, and
in particular more dimethylalkane isomers, than those that
are free from parasite pressure (Martin et al., 2010b). In
some cases, there is suggestive evidence that parasites may
have responded by switching hosts. For example, Martin
et al. (2010a) argue that British populations of the cuckoo
bumblebee Bombus (Ps.) sylvestris may recently have switched
to parasitizing B. pratorum. Nevertheless, parasites can track
even fine changes in the hydrocarbon signature and become
specialized on particular hosts (e.g. Bogusch et al., 2006). In
the extreme case of the hoverfly Microdon mutabilis parasite of
Formica lemani ant colonies, females (but not males) are host
specific at the colony level meaning that successful parasitism
involves reinfecting the same ant nest for generation after
generation (Sch
¨
onrogge et al., 2006). Although the precise
mechanisms underpinning recognition are still unknown,
host specificity that is confined to the female line is
reminiscent of some common cuckoo populations (Gibbs
et al., 2000), suggesting that the key recognition cues in this
system might also reside in the parasitic egg (Sch
¨
onrogge
et al., 2006 but see Fossøy et al., 2011).
(b) Forgery of the host signature after parasitism: generalist parasites
It is common for insect brood parasites to adopt a strategy
of chemical camouflage and acquire the colony-specific
hydrocarbon signature after parasitism. Parasites may
biosynthesize the appropriate hydrocarbon signature
themselves, through altered gene expression, or it may be
acquired by mechanical transfer soon after entering the
host nest (Lenoir et al., 2001). For example, the caterpillar
of the cuckoo butterfly Maculinea rebeli exploits several
different Myrmica antspecies,eachwiththeirownsignature
(Elmes et al., 2002), which the parasitic caterpillars acquire
after adoption (Akino et al., 1999). Similarly, the generalist
social parasite paper wasp Polistes atrimandibularis changes
its hydrocarbon signature to mimic its host, but only after
taking over the host colony (Bagneres et al., 1996). Just one
equivalent example is known so far from the avian brood
parasites (Langmore et al., 2008). The generalist Horsfield’s
bronze-cuckoo Chalcites basalis exploits diverse hosts whose
nestlings differ in their begging calls (Langmore et al., 2008).
Hosts abandon chicks with odd-sounding begging calls
(Langmore, Hunt & Kilner, 2003) and the cuckoo nestling
flexibly adjusts the structure of its call after hatching to mimic
the calls of the particular host’s own young. Remarkably,
there are no models from whom the cuckoo chick can learn
because it evicts host young from the nest soon after hatching.
Instead, host parents must somehow train the young parasite
to make the appropriate-sounding begging call (Langmore
et al., 2008), and so are inadvertently complicit in their own
deception.
The principal co-evolutionary consequence of forging the
host’s signature after parasitism, rather than expressing it
beforehand, is that parasites can be individual generalists,
capable of flexibly adapting to exploit any of their hosts.
Consequently there is no segregation into genetically distinct
host-specific lineages within species (Als et al., 2004; Fanelli
et al., 2005; Langmore et al., 2008). Otherwise, co-evolution
proceeds in a broadly similar way to the cases where the
forged signature is expressed before parasitism. Insect hosts
place parasites under selection to refine their mimicry of
the host hydrocarbon signature which, in some cases, gives
rise to parasites that become more and more chemically
invisible themselves, effectively presenting a blank slate
to be daubed with their hosts’ particular hydrocarbons
(Brandt et al., 2005a; D’Ettore & Errrard, 1998; Lenoir
et al., 2001). Parasites place hosts under selection to escape
mimicry, although presumably with this mode of forgery,
signature diversification alone is not sufficient to prevent
the parasite from acquiring the signature upon entering the
host nest. Instead, hosts may diversify their signatures in a
very particular way, by specifically incorporating particular
hydrocarbons that parasites find hard to absorb onto their
cuticles. There is evidence of this from Leptothorax hosts of the
slave-making ant Harpagoxenus sublaevis (Bauer et al., 2010).
Whereas closely related ant species usually have similar
hydrocarbon profiles, hosts L. muscorum and L. acervorum are
unusually distinct, suggesting that they have diversified under
selection from their slave-making parasite. In particular, the
L. acervorum
signature lacks the short-chained hydrocarbons
that dominate the signature of L. muscorum, and that are
more easily transferred to the parasite. Perhaps it is no
Biological Reviews 86 (2011) 836–852 © 2011 The Authors. Biological Reviews © 2011 Cambridge Philosophical Society

Frequently asked questions

Q1. What have the authors contributed in "Cuckoos versus hosts in insects and birds: adaptations, counteradaptations and outcomes" ?

Despite their considerable taxonomic disparity, here the authors show striking parallels in the way that co-evolution between brood parasites and their hosts proceeds in insects and birds. Nevertheless hosts may win when the co-evolutionary trajectory follows arms race a, c or d. Next, the authors show that there are four common outcomes of the co-evolutionary arms race for hosts. The authors propose an adaptive explanation, which centres on the relative strength of two opposing processes: strategy-facilitation, in which one line of host defence promotes the evolution of another form of resistance, and strategy-blocking, in which one line of defence may relax selection on another so completely that it causes it to decay. Empirical work suggests that hosts are doomed to lose arms races b and e to the parasite, in the sense that parasites typically evade host defences and successfully parasitize the nest. The authors suggest that when strategy-facilitation outweighs strategy-blocking, hosts will possess complex defence portfolios and they identify selective conditions in which this is likely to be the case. 

Q2. What are the future works mentioned in the paper "Cuckoos versus hosts in insects and birds: adaptations, counteradaptations and outcomes" ?

The authors suggest an adaptive explanation for this variation, which centres on the relative strength of two opposing processes: strategy-facilitation, in which each line of host defence promotes the evolution of another form of resistance, and strategy-blocking, in which one line of defence may relax selection on another so completely that the former causes the latter to decay. In future work, it would be interesting to use comparative analyses to evaluate the extent to which co-evolutionary outcomes are determined by hosts or their brood parasites. ( 4 ) Empirical work suggests that hosts are doomed to lose arms races b and e to the parasite, in the sense that parasites typically evade any host defences and successfully parasitise the nest. Nevertheless hosts may beat parasites completely when the co-evolutionary trajectory follows arms race a, c or d. ( 5 ) Each of five types of co-evolutionary arms race has one of four potential outcomes for the host. 

Q3. What is the main consequence of forging the host signature after parasitism?

The principal co-evolutionary consequence of forging the host’s signature after parasitism, rather than expressing it beforehand, is that parasites can be individual generalists, capable of flexibly adapting to exploit any of their hosts. 

Q4. What is the main reason why some species have relatively thin defence portfolios?

Evolutionary lag is classically invoked to explain why dunnocks Prunella modularis fail to recognize odd-looking common cuckoo eggs in their nest (Brooke & Davies, 1988) and might also be invoked to explain why some species have relatively thin defence portfolios, comprising few lines of defence. 

Q5. What is the likely outcome of tolerance?

When traits in the parasite inflate the cost of host resistance, by retaliation or retribution for example, tolerance is the more likely outcome than acceptance. 

Q6. What is the main effect of parasite shell thickening for hosts?

One possibility is that they may thicken their own eggshells, so that host eggs are no longer collaterally damaged during puncture rejection of the parasitic egg. 

Q7. What is the third way in which parasites can evade the host recognition system?

(c) Escaping host recognition through crypsisA third way in which parasites can evade the host recognition system is effectively to become invisible. 

Q8. What is the common strategy for acquiring the colony-specific signature after parasitism?

It is common for insect brood parasites to adopt a strategy of chemical camouflage and acquire the colony-specifichydrocarbon signature after parasitism. 

Q9. What is the significance of the chemical insignificance of ant social parasites?

Chemical insignificance is of particular importance for ant social parasites, especially the queen-tolerant and queen-intolerant inquilines as well as the Polyergus spp. and Myrmoxenus spp. slavemakers, who rely on their invisibility to slip unnoticed into host colonies and completely lack fighting adaptations with which to battle their way in or to defend themselves if they are spotted by hosts (Brandt et al., 2005a). 

Q10. Why is it adaptive for hosts to accept a parasite when resistance carries high costs?

In general, it is adaptive for hosts to accept a parasite when resistance carries high costs, perhaps because host recognition systems are relatively unsophisticated or because the host clutch is inadequately protected from the physical damage associated with the rejection of foreign eggs. 

Q11. What is the evidence that avian host recognition systems can defeat cuckoos?

Further indirect evidence that avian host recognition systems can defeat cuckoos is provided by the error-free rejection of oddly marked experimental eggs by hosts that are currently not exploited by brood parasites. 

Q12. What is the definition of a parasite’s first task in appropriating this resource?

(1) Front-line parasite attack and host defenceBird and social insect nurseries (see Mock & Parker, 1997 for a definition of ‘nursery’) are extremely well defended by their owners, so the parasite’s first task in appropriating this resource commonly involves breaching the various physical lines of defence that protect the nest. 

Q13. Why are nave individuals more likely to accept parasites than experienced breeders?

It also explains why naïve individuals are more likely to accept parasites than experienced breeders, all else being equal, because they have yet to develop the error-free recognition systems that are honed by breeding experience (e.g. Langmore et al., 2009a; Lotem, Nakamura & Zahavi, 1992; Lotem et al., 1995). 

Q14. What are the main types of co-evolutionary arms race?

(3) Evidence from the better studied brood parasites and their hosts suggests that several types of co-evolutionary arms race, each focused on different modes of host defence, can play out concurrently between a single host and its brood parasite. 

Q15. Why do parasites place hosts under selection to escape mimicry?

Parasites place hosts under selection to escape mimicry, although presumably with this mode of forgery, signature diversification alone is not sufficient to prevent the parasite from acquiring the signature upon entering the host nest.