Host defences against avian brood parasitism: an endocrine perspective

Abstract

Host defences against avian brood parasites are the outcome of well-documented coevolutionary arms races, yet important questions about variation in hosts' antiparasitic response traits remain poorly understood. Why are certain defences employed by some species or individuals and not by others? Here, we propose that understanding variability in and the evolution of host defences can be facilitated by the study of the underlying physiological mechanisms. Specifically, because antiparasitic strategies involve behaviours that have been shown to be hormonally regulated in other contexts, we hypothesize that host responses to brood parasites are likely to be mediated by related endocrine mechanisms. We outline the hallmarks of the endocrine bases of parasite defence-related avian behaviours, review the current understanding of antiparasitic host tactics and propose testable hypotheses about the hormonal mechanisms that may mediate host defences. We consider these mechanisms in a life-history framework and discuss how endocrine factors may shape variation in host defences. By providing a hypothesis-driven mechanistic framework for defences against parasitism, this perspective should stimulate the study of their endocrine bases to enhance our understanding of the intricate arms races in avian host–parasite systems.

1. Introduction

Obligate avian brood parasites, which lay their eggs into the nests of other species, comprise a little more than 100 species (approx. 1% of birds), but parasitize over 950 (nearly 10%) of host bird species worldwide [1,2]. By definition, brood parasitism has a negative, often severe, effect on host fitness and, as a consequence, hosts often become involved in a coevolutionary battle with their parasites [3,4]. A large proportion of host taxa have evolved a variety of defences to reduce the incidence of parasitism and/or minimize the negative fitness consequences of successful parasitism [5]. These defences are a diverse collection of behavioural, morphological and physiological traits that can be broadly categorized as front-line defences, which minimize the probability of the parasite laying an egg into the host's nest; egg-stage defences, which minimize the probability of hatching of the parasitic egg; and nestling- and fledgling-stage defences, which minimize the fitness consequences of a successfully hatched parasite (reviewed in [4]).

Among the most important evolutionary questions in brood parasite–host biology include why hosts employ some but not other antiparasitic defences; why there are differences in defences between and within host species; how and how fast these defences evolve; and why some species are more successful in preventing parasitism than others through certain defences [3,6,7]. These questions have traditionally been addressed with hypotheses based on the strength of selection, evolutionary time, behavioural repertoire, and morphology [2,8]. For example, the ‘strategy blocking’ hypothesis predicts that success of one defence strategy may weaken the selection for other types of defences [9,10]. On the other hand, the ‘strategy facilitation’ hypothesis suggests that certain host defences may favour evolution of additional antiparasitic behaviours, resulting in multi-component defences [5], while the ‘evolutionary lag’ hypothesis suggests that some defences might not have had time to evolve [11,12]. Egg rejection probability has also been linked to morphology, e.g. selection for small bill sizes may preclude some hosts from ejecting eggs [13]. While these hypotheses explain some of the interspecific variation in host defences, important aspects of the variation in the magnitude and type of host defences remain unexplained [14].

The majority of host defences across front-line, egg and nestling stages appear to employ behaviours and life-history strategies that have been shown to have a strong endocrine basis in other behavioural and life-history contexts that are distinct from host–parasite interactions [15–17]. In this article, we propose that avian host defences against brood parasites are also mediated or directly regulated by major vertebrate hormones (such as testosterone, corticosterone, and prolactin) and argue that studying the hormonal mechanisms may inform the evolution and expression of host defences, as well as individual, sex and species differences in defence type and efficacy.

Hormones have unique properties that make them especially informative in the evolution of behaviour and life histories: hormones are sensitive to the environment, they often regulate more than one phenotype (hormonal pleiotropy), and form interconnected hormonal networks, where secretion of one hormone often affects another hormone [18]. These properties of hormonal signalling may both constrain and facilitate evolution and expression of host defences, and may shape the types of defences we observe in nature. Furthermore, hormones may underlie variation in host defences at the within-individual (plasticity), between-individual and across-species levels, potentially facilitating evolution of plastic defences to fixed responses [19]. A small number of studies have begun to uncover the hormonal responses of avian hosts to brood parasites at the within-species [20–23] and across-species levels [24–26]. To better understand the role that hormones may play in host defence evolution and variation, however, we need an overarching hypotheses-driven framework that generates testable predictions.

Here, we review the hormonal bases of behavioural traits that underlie major aspects of host defences (figure 1). We focus specifically on aggression, maternal behaviour and host offspring behaviour, which underlie the most widespread and well-studied behavioural defences at the front line [2], egg [27] and nestling stages [28]. We propose hypotheses about how hormones may regulate host-specific behaviours (table 1) and outline testable predictions for the patterns of covariation between host defences and hormones that we expect (electronic supplementary material, table S1). We conclude by conceptualizing hormone-mediated host defences in a more general life-history framework.

Figure 1.

Potential hormonal regulation of host defences against avian brood parasitism. Major vertebrate hormones have been shown to regulate specific aspects of behaviour and physiology that are also involved in host defence against brood parasites. CORT, corticosterone; T, testosterone; PROG, progesterone; PRL, prolactin; MT, mesotocin.

Table 1.

Mechanistic hypotheses for the hormonal basis of host defences. (T, testosterone; CORT, corticosterone; PRL, prolactin; MT, mesotocin; PROG, progesterone.)

defence stage	host defence	mechanistic hypothesis
front line	vigilance against brood parasites	CORT increases vigilance against brood parasites T decreases vigilance against brood parasites
		chronically high CORT in response to brood parasites has negative effect on host fitness
	aggression against brood parasites, blocking parasite access to nest	T and/or PROG regulate aggression towards brood parasites
		pleiotropic functions of T impose trade-offs between aggressive phenotype and parental care
egg and nestling stages	rejection of parasite eggs and nestlings, reduction of parental care towards parasitic nestlings	PRL, MT mediate response to egg, nestling and/or nest stimuli, and regulate the threshold for egg and nestling rejection and nest abandonment
		CORT and T downregulate host parental behaviour towards own and parasitic eggs and nestlings
egg and nestling stages	host nestlings outcompete parasitic nest-mates	maternally deposited yolk T and CORT increases host offspring ability to compete with a parasitic nestling
		chronically high maternal CORT elevates yolk CORT and decreases host offspring ability to compete with parasitic nestling
		T and CORT secreted by host nestlings in response to parasitic nestling increases host offspring ability to compete

2. Aggression towards brood parasites

One of the most noteable and widespread front-line defences against brood parasitism is the aggressive behaviour by the hosts towards the parasitic adults. Host species often aggressively mob, harass and physically attack adult parasites prior to parasitic egg laying [29,30] and, in some cases, can block parasites' access to the nest cup by sitting on the clutch and pecking at the intruder [31,32].

No study, to date, has investigated mechanisms underlying host aggression towards brood parasites. However, the hormonal control of aggression is well studied in other contexts. In birds, aggression has been most intensively studied in territorial interactions, where in many cases male and female aggression has been shown to be mediated by steroid hormones [15]. Testosterone (T) in peripheral circulation has been shown to have a permissive effect on the expression of aggression towards conspecifics [33], and it positively correlates with individual differences in aggression [34]. The effects of T on aggression may be sex-specific [35], e.g. in some species, aggression in females is instead mediated by progesterone (PROG) [36], which has also been shown to regulate non-territorial maternal aggression in rats [37]. In some cases, differences in aggression may be independent of circulating steroids and are instead explained by differential expression of steroid receptors in the brain [38], or by differences in the expression of aromatase, which converts T into oestradiol in the brain [39]. Finally, in other cases, conspecific aggression may be independent of steroid signalling and may be mediated more locally by serotonin and histamine signalling in the brain [40,41].

It remains unclear, however, if sex steroids regulate aggression in contexts other than those involving conspecific interactions, e.g. aggression against brood parasites or predators [42]. For example, in eastern bluebirds Sialia sialis, circulating T does not explain variation in aggression against heterospecific individuals [43]. In other taxa, however, T positively correlates with aggression against competitors as well as nest predators [44], implying that shared mechanisms may regulate aggression in different situations. Individual differences in aggressive behaviours across contexts are often positively correlated in many lineages, including birds [45,46]. This suggests that animals show a generalized aggression ‘behavioural syndrome’, and that aggression towards brood parasites may covary with territorial aggression. Behavioural syndromes are hypothesized to arise owing to shared mechanisms [42], therefore at least a partial overlap may exist between the mechanisms that control aggression towards brood parasites and aggression in conspecific contexts.

If T mediates aggression towards brood parasites, then individuals with higher T levels, or individuals with higher expression of aromatase and androgen or oestrogen receptors in the brain areas associated with aggression [47], should be more aggressive towards brood parasites. Alternatively, higher aggression towards brood parasites may be associated with lower PROG levels. Exposure to brood parasites is predicted to increase T in parasitized host individuals, and populations or species with higher brood parasitism levels should show higher T levels (or higher sensitivity to androgens, the brain areas associated with aggression). Supporting these predictions, Hahn et al. [22], in a first study of its kind, showed that eggs from parasitized nests had higher T levels relative to eggs from non-parasitized nests, suggesting that host mothers increase T in response to brood parasite sighting or recognizing parasitic eggs in their nests.

The regulation of aggression against brood parasites by sex steroids may have important consequences to host life histories. For example, individuals with higher T levels have been shown to invest less in parental care [48]. If aggression against brood parasites is mediated by circulating T, then this predicts that brood parasite hosts which are more aggressive should also invest less in parental care. This could lead to an interesting trade-off in mate choice, where females can choose between males who are good at defending nests against brood parasites, but are poor fathers, and males who are good fathers but are poor at nest defence.

Reduction in parental care may result in lower attachment to own eggs and offspring. If so, more aggressive individuals, or more territorial species, should be more likely to eject parasite eggs but perhaps also to mistakenly eject their own eggs compared with less aggressive individuals. Conversely, species or individuals with higher parental investment, which should have lower T [49], should be less aggressive towards brood parasites, and should more likely be an egg- and/or chick-acceptor species. Because aggression is probably regulated differently in males and females, these predictions are more likely to be observed within but not among sexes: whereas females have lower T, they are also often more aggressive than males against brood parasites [32].

These predictions assume that differences in aggression are regulated by circulating hormones, but it is important to note that that is not always the case [38]. However, some of these trade-offs may occur irrespective of the hormonal mechanisms, e.g. these behaviours may be traded-off in time. Furthermore, even if T does not regulate aggression against brood parasites, it may nevertheless affect host defences. For example, high T may lead to reduced vigilance against predators [50], and T-dependent behaviours, such as conspecific territorial aggression, may increase the probability of detection by brood parasites [51]. This predicts that high T individuals may actually experience higher levels of brood parasitism compared with low T individuals, contrary to the T-mediated aggression hypothesis that predicts the opposite pattern.

From a broader perspective, the behavioural syndrome hypothesis predicts that individuals from more highly territorial host species should be more aggressive towards brood parasites than group-living or colonial birds, and that species which are more aggressive against predators should be also more aggressive against brood parasites. Importantly, the former prediction is contradicted by studies which demonstrate that colonially and cooperatively breeding birds show greater aggression through more intense mobbing towards brood parasites [52,53] and by findings which show that more polygynous males are less aggressive against brood parasites [54]. This suggests that aggression against brood parasites and conspecifics may be disassociated, which may happen even if these behaviours are regulated by similar hormonal mechanisms (e.g. owing to changes in hormone receptor expression in the brain areas responsible for these behaviours [38]).

Context-general hormonally mediated aggression may also have important implications for the evolution of host aggression against brood parasites. For example, aggression against brood parasites may have evolved by co-opting the mechanisms of aggression against predators. Indeed, it is widely known that many brood parasites remove host eggs, resulting in partial predation of their clutches [3]. Host response to brood parasites may have evolved from a generalized response to a nest predator to a unique host defence against specific brood parasites, as shown in yellow warblers Setophaga petechia that show different defence strategies against brood parasites versus nest predators [32]. Unique behavioural responses to brood parasites according to this hypothesis may be more likely to arise in species that already possess the ability to differentiate between different predators or species where predator recognition is learned.

Another important hormone that may play a role in host nest defence against brood parasites is corticosterone (CORT), a hormone that is secreted in response to environmental stressors [17]. CORT has been shown to be positively correlated with vigilance behaviour [55,56], which is an important aspect of host defence against brood parasites [57,58]. Avian hosts that have been parasitized by a brood parasite show higher CORT levels [21]. Although it is as of yet unclear whether CORT increases in response to brood parasitism or an encounter with a brood parasite, increase in CORT in response the presence of brood parasites in the vicinity may cause host individuals be more vigilant against parasite attempts to access the nests, potentially leading to lower brood parasitism rates.

3. Maternal behaviour

If parasites succeed in laying an egg (or eggs) in a host's nest, many host species can recognize the foreign eggs and reject them [59,60]. In addition to rejecting eggs, hosts may also abandon a parasitized nest altogether, building a new nest elsewhere [11] or directly on top of the parasitized nest [60]. Some host species have also been shown to eject parasite nestlings upon hatching [61], or feed these nestlings less compared with their biological offspring [62,63].

Nest desertion and egg and nestling ejection are a stunning reversal from typical parental behaviour, specifically incubation and nestling provisioning, with the valence of typically associative stimuli (nests, eggs and nestlings) becoming negative. Maternal behaviour is strongly mediated by hormones [16], suggesting that hormones may also play a role in rejection of parasitic eggs or offspring. A limitation to this hormone-dependent maternal behaviour hypothesis is that systemic changes in hormone secretion are more likely to result in changes in behavioural and perceptual thresholds as opposed to directly causing an immediate response in response to specific stimuli. However, the time course of hosts' responses to parasitic eggs or nestlings can be as long as several days [64], suggesting that endocrine mechanisms may be involved in priming the neural mechanisms that control differential response of hosts to own and parasitic offspring.

The most likely hormones that may mediate the propensity of hosts to eject parasitic eggs or abandon their nests are those that regulate associative behaviours towards own offspring, such as incubation and maternal provisioning. In birds, prolactin (PRL) has been shown to regulate both of these behaviours and thus may mediate maternal responses to parasitic eggs or offspring. PRL plays an important role in the initiation and maintenance of incubation in many bird species [65,66]. In some, but not all, species, variation in PRL levels also correlates positively with parental behaviour during the nestling stage—individuals with higher PRL levels have higher offspring provisioning rates [67,68], and exposure to own young in birds increases PRL signalling in the hypothalamus [69]. An experimental increase in PRL can induce parental provisioning behaviour and reduce aggression towards nestlings, whereas disrupting PRL signalling has the opposite effect [70,71]. These results suggest that PRL may increase the sensitivity of individuals to associative egg or nestling stimuli, allowing the induction of parental behaviour towards offspring.

Another hormone that has been shown to play an important role in regulating maternal behaviour in birds is mesotocin (MT), an avian homologue to mammalian oxytocin. Inhibition of MT signalling in the brain results in reduction of brooding behaviour in chickens, and incubating chickens have a larger number of active MT neurons in the hypothalamus [72]. Exposure to young induces rapid activation of hypothalamic areas that contain MT-producing neurons in chickens [72].

While PRL and MT are not likely to directly regulate parasitic egg or offspring ejection, they may mediate behavioural and perceptual thresholds for parasitic egg or offspring rejection, or changes in maternal behaviour affecting the whole clutch (e.g. overall reduction in maternal behaviour or nest abandonment). PRL levels are known to decrease in response to egg removal [73] and in response to stressors [74]; therefore, sighting of a brood parasite near the nest, or viewing a poorly mimetic parasitic egg or nestling in one's own nest, may result in a drop in PRL or MT levels. If PRL or MT mediates associative responses to eggs, this may result in nest abandonment or reduce attachment to eggs in the nest. A lower attachment to eggs may effectively reduce the rejection threshold of perceived differences between own and parasitic eggs (or chicks) [75], resulting in a higher probability of egg (or chick) rejection. The hypothesis predicts that individuals that have higher PRL of MT levels (or decrease PRL or MT less in response to brood parasites) are less likely to abandon nests or eject parasitic eggs and nestlings compared with individuals with lower PRL or MT levels. Another prediction is that nest abandonment or egg rejection is less likely to occur during periods when PRL or MT levels are highest. This may explain why egg rejection most often occurs during egg laying (e.g. [27]) when PRL levels are lower compared with incubation [16]. Individual- or species-level variation in nest abandonment or egg and nestling rejection may also be owing to differential expression of PRL or MT receptors—individuals and species that have higher expression of PRL or MT receptors in the preoptic area and other hypothalamic nuclei should be more likely to abandon nests or eject eggs and nestlings. PRL and MT may also mediate provisioning of nestlings by hosts: recognition of parasitic nestlings in the nest may result in lower PRL or MT levels, resulting in lower provisioning rates per capita.

Changes in host parental behaviour may also be mediated by hormones that typically inhibit parental behaviour, such as T and CORT. As outlined above, high T levels can have a negative effect on parental care. For example, T reduces nestling provisioning behaviour in males [76,77] and incubation behaviour in females [78]. CORT has also been shown to reduce nestling provisioning [79,80]. Host adults that have been naturally parasitized show higher CORT [21] and, probably, higher T levels [22], which may result in a reduction in an associative response to own nests, eggs, of offspring, potentially leading the nest abandonment, increase in egg or nestling rejection probability or/and overall reduction of offspring provisioning.

While we argue that hormones may mediate egg and nestling rejection, hormones, including PRL and MT, may also be the reason why these behaviours have not evolved in all avian brood parasite hosts. Egg and nestling rejection require a fundamental departure from hormonally regulated parental care and attachment to eggs and young, which may make it more difficult to evolve defences against brood parasitic eggs and offspring and explain why so many species do not eject parasitic eggs despite the severely negative fitness consequences of raising a parasitic offspring.

4. Nestling behaviour and growth rate

In evicting brood parasite species, such as the common cuckoo Cuculus canorus, nest-mates have little chance of successfully competing with the foreign chick because they are often eliminated as eggs or hatchlings [81]. In turn, in non-evicting brood parasite systems, such as brown-headed cowbirds Molothrus ater, the battle between the hosts and parasites is most intense at the nestling phase, when the parasitic nestling competes with host nestlings for parental care [82]. Many hosts have evolved parental solicitation strategies by nestlings to limit the fitness loss owing to feeding the parasitic chick. At the species level, host nestlings from species with higher brood parasitism rates develop on average faster [83] and beg louder [84,85] than nestlings from species with lower parasitism rates. At the within-species level, nestlings from parasitized nests beg louder than those from non-parasitized nests ([86], but see [28]). This shows that brood parasitism can select for changes in the average host nestling phenotype over evolutionary time as well as result in plastic responses to specific parasitism events.

An increasing body of evidence suggests that nestling behavioural and metabolic phenotype can be regulated by maternally deposited yolk hormones [87]. For instance, yolk hormones can influence nestling growth rate: higher levels of yolk T stimulate nestling growth [88], whereas higher levels of CORT can have the opposite effect ([89] but see [90]). Yolk hormones can also have organizational effects on behaviour during the development of the embryo: high yolk T can result in long-term changes in behaviours, such as aggression [91]. Higher T and CORT levels in yolk can result in higher begging rates and intensified begging postures [87]. Importantly, the amount of hormones deposited in yolk is affected by the stress environment, e.g. stressed females have higher yolk CORT [92].

Maternally controlled differences in yolk hormones may underlie differences in the mean behaviour or growth rate of host nestlings across species, or changes in behaviour or growth rate in response to specific parasitism events [22]. At the between-species level, females from frequently parasitized species are predicted to deposit higher levels of T in their eggs compared with less frequently parasitized species, resulting in higher offspring growth rates and higher begging rates; indeed, small cowbird host species that pay a greater cost of competition with the parasitic nestling [82] have higher yolk T levels than do larger host species [25]. In addition to being exposed to higher T levels, embryos of host species may express higher levels of androgen receptors: increased sensitivity to T may lead to increased begging behaviour or growth rates. Yet another, thus far untested possibility is that species with higher parasitism rates may achieve higher begging rates and faster growth by changing the CORT levels in the yolk or the expression or glucocorticoid or mineralocorticoid receptors in the embryonic tissues. Because CORT can have different effects on nestling phenotype in different species, it is possible to predict both positive and negative relationships between CORT and nestling begging rate or growth.

Maternally derived yolk hormones may also mediate more immediate, plastic response to specific parasitic events or to high parasite density in the environment. Hahn et al. [22] showed that host eggs in nests parasitized by cowbirds have higher yolk T content than eggs from non-parasitized nests. They suggest that host females may increase yolk T either in response to observing a brood parasite that is prospecting its nest, or by recognizing a parasitic egg.

In addition to T, females may also be able to rapidly modify CORT levels in yolk in response to perceiving a brood parasite, increasing a future offspring begging rate. If, on the other hand, CORT negatively affects offspring growth, then such a response may result in a maladaptive reduction in host nesting growth rate or begging intensity. Females may also manipulate offspring phenotype via thyroid hormone (TH), which can regulate the growth of embryos or nestlings [93]. We are awaiting studies that experimentally investigate the role of TH on nestling phenotype, especially in the context of competitive growth and begging in broods of frequently parasitized hosts of non-evictor parasites.

In addition to maternally controlled phenotypes, host nestlings in parasitized nests may also respond to the parasitic nestling directly. European magpie Pica pica nestlings have higher CORT levels in nests parasitized by the great spotted cuckoo Clamator glandarius [20]. The authors suggest that increase in CORT might be either owing to the hunger of nestlings caused by increased competition of food, or may be in response to the increased begging displays by the parasitic nestlings. Higher CORT is linked to low body condition in other species [94] and could either increase glucose metabolism [17] or increase begging behaviour in nestlings [87,95], allowing them to become more competitive against parasitic nest-mates. A higher begging rate may also be achieved if nestlings elevated T in response to their parasitic nest-mates [87].

5. Parasites as stressors

While CORT may mediate adaptive responses to brood parasites, chronic elevation of CORT may result in negative fitness consequences [96]. CORT is a hormone that fulfils multiple functions, which may change depending on the levels of CORT in circulation [97] and the expression of its receptors in target tissues [98]. As suggested by Mark & Rubenstein [21], if brood parasites cause long-term elevation in CORT, this may result in negative fitness consequences to host fitness beyond the direct negative effects of brood parasitism. High CORT levels can have negative effects on body condition [99] and suppress immunity [100,101], and suppress growth [17]. If hosts mount a repeated stress response in response to brood parasitism or seeing a brood parasite near their nests, hosts may develop a lower body condition or suppressed immunity, potentially leading to increased ability to escape predators, provision offspring or resist infections. This may result in a negative effect on population growth rate independent of direct parasitism effects.

As shown by Hahn et al. [22], host hormonal responses to brood parasites may depend on the host species morphology or ecology—small hosts that suffer more severe negative fitness consequences of brood parasitism may mount a stronger stress response in response to brood parasites than larger host species, and as a result potentially suffer more negative fitness consequences. Along the same line, hosts that reject parasite eggs may mount a lower stress response than those that do not reject parasite eggs.

6. An integrative perspective

The multifaceted roles of hormones in behaviour and reproductive physiology as well as other organismal and life-history functions (such as self-maintenance, immunity and energetics) provide opportunities for both constraint and facilitation in the evolution and expression of host defences. As mentioned above, mechanisms that regulate maternal behaviour may constrain the evolution of host defences at the egg and nestling stages. On the other hand, the pleiotropic actions of CORT may allow the hosts to develop coordinated and complex defence responses, which may include increased vigilance against parasites, an increase in egg rejection propensity or reduction of feeding behaviour in parasitized nests, and increase in offspring begging rate. However, high CORT levels may also result in adverse effects to host fitness or offspring growth rate via yolk hormones. The effect of hormonal responses to brood parasitic adults or young may extend beyond host defences: high CORT levels may suppress reproduction, the immune system and may impose other somatic costs, reducing the probability of breeding in the future [21,101,102]. Host defences against brood parasites may thus need to be viewed in a more general framework of allostatic load [103].

T can likewise may both facilitate defence strategies as well as impose life-history trade-offs [76]. For example, high yolk T levels may facilitate the ability of the host offspring to compete for food against a brood parasitic nestling, as well as result in higher aggression against brood parasites in adulthood. On the other hand, high T phenotype that is advantageous in response to brood parasitism may be maladaptive in terms of survivorship, immunity and parental care [76].

Furthermore, CORT and T, and PRL, are sensitive to other environmental and social stimuli, such as stress [104] and competition [105]. Changes in hormone levels owing to factors that are not directly related to brood parasitism may nevertheless affect host defences that are mediated by these hormones (figure 1). For example, animals experiencing stressors that elevate CORT levels may show reduced attachment to eggs and nests, resulting in higher nest abandonment or higher propensity to reject eggs in response to parasitism. Understanding variation in host defences may therefore necessitate understanding the physiological state of the animal in response to the overall environment.

7. Future directions

The field of avian brood parasitism is ripe for mechanistic studies that test the endocrine mechanisms underlying host defences. The hypotheses outlined in this paper can be tested by adopting well developed and affordable [106] pharmacological techniques that have been used to experimentally manipulate hormone levels in wild animals in other systems (e.g. conspecific aggression [107]) and combining it with equally fine-tuned experimental techniques for studying host responses to parasites (reviewed in [4]). To understand the endocrine basis of host defences, it is imperative to focus both on hormones as well as their receptors, as variation in both can cause differences in behaviour [38,97,98]: studies on host endocrinology to date have focused mainly on variation in hormone levels, but receptor expression for these hormones in the target tissues may yield equally illuminating findings.

We emphasize that in addition to steroid hormones and PRL, other hormones, such as TH [93], catecholamines [17] and galanin [108], mediate the behavioural and reproductive phenotypes that we have discussed here and thus should be considered in future studies. Furthermore, hormones can act in a paracrine and autocrine fashion [109], and these effects may affect the relationships between peripheral hormones and behaviour. Finally, correlations between hormones, their receptors and behaviours may come about because of hormonal regulation of that behaviour, induction of hormonal changes by the behaviour, or unrelated third factors [110]. We therefore advocate for entertaining alternative hypotheses when interpreting relationships between physiology and host defences.

Lastly, many of the studies on host endocrinology so far have documented species differences in their responses to brood parasites [20,22–25]. Future studies should therefore embrace a comparative approach when investigating hormonal bases of host defences, as the costs and benefits of hormonally mediated phenotypes may differ between species [22].

8. Summary

The study of the arms-race between avian brood parasites and their hosts is a productive model system in behavioural ecology and evolutionary biology. Here, we argue that this system is also ripe for mechanistic studies on the proximal basis of host and parasite behaviours. Specifically, we suggest that understanding the hormonal basis of host defences has the potential to inform major questions in avian brood parasitic systems, such as why hosts employ some but not other defences against brood parasites, why individuals vary in their defence behaviours and how these defences evolve (table 1 and the electronic supplementary material, table S1).

Data accessibility

This article has no additional data.

Authors' contributions

M.A.-A. and M.E.H. drafted the manuscript. Both authors gave final approval for publication.

Competing interests

We declare we have no competing interests.

Funding

M.E.H. was funded by the US National Science Foundation (IOS-1456524) and the University of Illinois’ Harvey Jones Van Cleave Professorship.