Social monogamy vs. polyandry: ecological factors associated with sex roles in two closely related birds within the same habitat

Why mainly males compete and females take a larger share in parental care remains an exciting question in evolutionary biology. Role‐reversed species are of particular interest, because such ‘exceptions’ help to test the rule. Using mating systems theory as a framework, we compared the reproductive ecology of the two most contrasting coucals with regard to sexual dimorphism and parental care: the black coucal with male‐only care and the biparental white‐browed coucal. Both species occur in the same lush habitat and face similar ecological conditions, but drastically differ in mating system and sexual dimorphism. Black coucals were migratory and occurred at high breeding densities. With females being obligatory polyandrous and almost twice as heavy as males, black coucals belong to the most extreme vertebrates with reversed sexual dimorphism. Higher variance in reproductive success in fiercely competing females suggests that sexual selection is stronger in females than in males. In contrast, resident white‐browed coucals bred at low densities and invariably in pairs. They were almost monomorphic and the variance in reproductive success was similar between the sexes. Black coucals were more likely to lose nests than white‐browed coucals, probably facilitating female emancipation of parental care in black coucals. We propose that a combination of high food abundance, high population density, high degree of nest loss and male bias in the adult sex ratio represent ecological conditions that facilitate role reversal and polyandry in coucals and terrestrial vertebrates in general.     

Competing Females and Caring Males. Polyandry and Sex-Role Reversal in African Black Coucals, Centropus grillii

Most species of birds show bi‐parental or female‐only care. However, a minority of species is polyandrous and expresses male‐only care. So far, such reversals in sex roles have been demonstrated only in precocial bird species, but there was suggestive evidence that such a mating system may occur in one altricial bird species, the black coucal, Centropus grillii. In a field study in Tanzania we investigated whether black coucals are sex‐role reversed and polyandrous. We found that males were mated to one female, rarely vocalized and provided all parental care from incubation of eggs to feeding of young. In contrast, female black coucals were about 69% heavier and 39% larger than males and polyandrous. They spent a large proportion of time calling from conspicuous perches, defended breeding territories, did not help in provisioning young and had a higher potential reproductive rate than males. We conclude that the black coucal currently represents the only altricial bird species with sole male parental care and a classical polyandrous mating system. High nest predation pressure and small territory sizes due to high food abundance may have been important factors in the evolution of sex‐role reversal and polyandry in this species.     

Evolution of reversed sex roles, sexual size dimorphism, and mating system in coucals (Centropodidae, Aves)

The evolution of greater male than female parental care remains poorly understood. In birds it is thought to be related to precocial chicks and small clutch size. This review shows, however, that such role reversal has also evolved in a family with altricial young and relatively large clutch size: coucals (Centropodidae, Cuculiformes). Males perform most nest building, incubation, and feeding of young. As predicted by sexual selection theory, coucals have also reversed sexual size dimorphism, females being larger than males in all 12 species for which size data are available. Most coucals that have been studied are monogamous, but the black coucal Centropus grillii appears to be polyandrous, and males perform almost all parental care, whereas females show more active advertisement behaviour. In this species, females are about 50% heavier than males. Polyandry in the black coucal seems to be associated with a shift to a habitat with seasonally rich food resources. Difficulties for female coucals of gathering enough resources for producing several clutches of relatively large eggs may favour mainly male parental care. Female sexual competition and resource storage, and male foraging economy, may explain why females are larger. Additional field studies are needed to test these hypotheses; the coucals are of great interest to sexual selection and mating systems theory.     

Front Cover

In most animals, competition for mating opportunities is higher among males, whereas females are more likely to provide parental care. In few species, though, these "conventional" sex roles are reversed such that females compete more strongly for matings and males provide most or all parental care. This "reversal" in sex roles is often combined with classical polyandry—a mating system in which a female forms a harem with several males. Here, we review the major hypotheses relating such role reversals to evolutionary and behavioural traits (anisogamy, phylogenetic history, sexy males, parental care, genetic paternity, trade‐off between mating and parenting, adult sex ratio) and to ecological factors (food supply, offspring predation). We evaluate each hypothesis in relation to coucals (Centropodinae), a group of nesting cuckoos of great interest for mating system and parental care theory. The black coucal (Centropus grillii) is the only known bird combining classical polyandry with altricial development of young, a costly trait with regard to parental care. Our long‐term study offers a unique possibility to compare the strongly polyandrous black coucal with a monogamous close relative breeding in the same area and habitat, the white‐browed coucal (C. superciliosus). We show that the evolution of sex roles in coucals and other animals has many different facets. Whereas phylogenetic constraints are important, confidence in genetic paternity is not. In combination with facilitating ecological conditions, adult sex ratios are key to understanding sex roles in coucals, shorebirds, and most likely also other animals. We plead for more studies including experimental tests to understand how biased adult sex ratios emerge and whether they drive sexual selection or vice versa. How do sex ratios and sexual selection interact and feedback on each other? Answers to these questions will be fundamental for understanding the evolution of sex roles in mating and parenting in coucals and other species.     

The reproductive pattern of male dusky salamanders (genus Desmognathus) is neither associated nor dissociated

Many seasonally breeding vertebrate species have an associated reproductive pattern: mating behavior, gonadal activity, and peak circulating androgen levels occur simultaneously. In these species, androgens influence the expression of male mating behavior. Other species have a dissociated reproductive pattern: mating behavior occurs at a different time than peak gonadal activity. In such species, it is hypothesized that mating behavior is not dependent on androgen levels [Crews, D., 1984. Gamete production, sex hormone secretion, and mating behavior uncoupled. Horm. Behav. 18, 22-28]. The salamander Desmognathus ochrophaeus mates in the spring and fall while spermatogenesis occurs during the summer, suggesting that it has a dissociated reproductive pattern and that androgens do not mediate mating behavior. To assess whether mating behavior is regulated by gonadal androgens, we castrated males to reduce endogenous androgens and implanted testosterone propionate (TP) to restore androgen levels. Castrated males mated significantly less than did control males. Castrated males given TP mated as much as control males. Compared to controls, circulating androgen levels (both testosterone (T) and dihydrotestosterone (DHT)) were reduced in castrated males and elevated in castrated males given TP implants. We also found that plasma corticosterone (CORT) levels were strongly and positively correlated with T levels. Together, these data indicate that, although spermatogenesis is dissociated in time from mating behavior, androgens are associated with the expression of mating. Thus, the associated-dissociated dichotomy does not adequately describe the reproductive pattern of D. ochrophaeus. We discuss the limitations of the associated-dissociated framework in clarifying hormone-behavior relationships in reptiles and amphibians.     

Polymorphic microsatellite loci in pheasant coucal (Centropus phasianinus)

Little is known about the effect of male parental care and behavioural sex‐role reversal on the mating system of birds because genetic markers for species with these characteristics are lacking. We developed primers for nine polymorphic microsatellite loci in pheasant coucals (Centropus phasianinus). Eight of the primers were also polymorphic in African black coucals (Centropus grillii). Pheasant coucals are of particular interest in the study of evolutionary and behavioural ecology, because their sex‐role reversal and extensive male parental care suggests low levels of extra‐pair fertilizations, yet they have large testes indicating sperm competition.     

Males paving the road to polyandry? Parental compensation in a monogamous nesting cuckoo and a classical polyandrous relative

Comparative studies have established the necessity for biparental care as an important factor for monogamy in freshwater fish and birds. However, whether two parents are really needed for offspring care remains an open question in many cases. I experimentally studied female and male contributions to offspring care in the white-browed coucal (Centropus superciliosus), a monogamous and biparental cuckoo with a balanced adult sex ratio, and contrasted it with the sympatric black coucal (C. grillii), a classically polyandrous species with a male-biased adult sex ratio and male-only care. To study the necessity for biparental care, I temporarily removed one partner for 2 days to see whether the remaining parent compensated for the absence of its partner. Both female and male white-browed coucals approximately doubled their feeding rates when their partner was absent, thus fully compensating the number of feeding visits to the nest. However, nestlings maintained their growth only, when males were present and females were removed. When males were removed and only females were present, nestling growth declined. Hence, only male white-browed coucals fully compensated for the temporary loss of the partner, suggesting that females could benefit most from nesting with additional males—if these should become available. Removing female black coucals had no consequence for nestling feeding rates of male black coucals. But male black coucals had to be returned to their territories within a few hours to avoid harming the brood because female black coucals typically would not commence feeding their offspring. In conclusion, the breeding system of white-browed coucals seems quite flexible and the relatively balanced adult sex ratio may stabilize monogamy in this species. Should ecological factors ever favour a stronger bias in the adult sex ratio towards males, female white-browed coucals may easily become polyandrous and relinquish parental care entirely to males.     

Effects of androgens on sex differentiation of the urodele Pleurodeles waltl

In nonmammalian vertebrates, steroids have been hypothesized to induce somatic sex differentiation, since manipulations of the steroidal environment of gonads have led to various degrees of sex reversal. Whereas the critical role of estrogens in ovarian differentiation is well documented, studies on androgens have produced a perplexing variety of results depending upon species variations and nature of androgens used. In this way, testosterone induces masculinization of females in some species but provokes paradoxical feminization of males in many other species such as the urodelan Pleurodeles waltl. In reptiles this phenomenon could be interpreted by conversion of exogenous testosterone to estradiol by aromatase. Treatments of Pleurodeles larvae with nonaromatizable androgens bring support to this hypothesis and suggest a role of androgens in sex differentiation. Dihydrotestosterone (DHT) could not induce the paradoxical feminization of ZZ larvae. In addition, DHT as well as 11beta-hydroxy-androstenedione could drive a functional male differentiation of ZW larvae. Moreover, other 5alpha reduced androgens also induced sex reversal of female larvae. Yet, the 5alpha reductase inhibitor CGP 53133 and antiandrogens such as flutamide or cyproterone acetate did not exert any effect on male sex differentiation of ZZ larvae. Though the precise role of androgens is still unknown, especially for 11-oxygenated androgens, our results suggest an implication in male sex differentiation. In this way, testosterone could play a pivotal role in being metabolized either into other androgens during testis differentiation or into estradiol during ovarian differentiation.     

Progesterone modulates aggression in sex-role reversed female African black coucals

Testosterone is assumed to be the key hormone related to resource-defence aggression. While this role has been confirmed mostly in the context of reproduction in male vertebrates, the effect of testosterone on the expression of resource-defence aggression in female vertebrates is not so well established. Furthermore, laboratory work suggests that progesterone inhibits aggressive behaviour in females. In this study, we investigated the hormonal changes underlying territorial aggression in free-living female African black coucals, Centropus grillii (Aves; Cuculidae). Females of this sex-role reversed polyandrous bird species should be particularly prone to be affected by testosterone because they aggressively defend territories similar to males of other species. We show, however, that territorial aggression in female black coucals is modulated by progesterone. After aggressive territorial challenges female black coucals expressed lower levels of progesterone than unchallenged territorial females and females without territories, suggesting that progesterone may suppress territorial aggression and is downregulated during aggressive encounters. Indeed, females treated with physiological concentrations of progesterone were less aggressive than females with placebo implants. This is one of the first demonstrations of a corresponding hormone-behaviour interaction under challenged and experimental conditions in free-living females. We anticipate that our observation in a sex-role reversed species may provide a more general mechanism, by which progesterone--in interaction with testosterone--may regulate resource-defence aggression in female vertebrates.     

Plasma steroid hormones in relation to behavioral sex role reversal in the spotted sandpiper, Actitis macularia

Plasma samples collected from spotted sandpipers during the reproductive season were analyzed for testosterone, dihydrotestosterone (DHT), estradiol-17 beta and progesterone. Prior to incubation, plasma testosterone and DHT levels were significantly greater in males than in females. Estradiol levels of paired females were significantly greater than those of paired males. Testosterone and DHT levels of unpaired resident and paired males were significantly greater than those of incubating and brooding males. A 25-fold decline in testosterone occurred in males from the 1- or 2-egg stage to the 3-egg stage, when incubation is initiated. In females, testosterone values were low in unpaired, brooding, and transient birds. Paired females had levels 7-fold greater than unpaired birds. In both sexes, there was a strong correlation between testosterone and DHT levels. Prolactin values were negatively correlated with testosterone and DHT in males. These results indicate that the high level of intrasexual competition for mates among female spotted sandpipers is not based upon a total reversal of the normal male/female levels of androgens and estradiol. Territoriality and intense competition for mates in females may be based upon enhanced receptivity of neural centers to moderate hormone levels. Relative changes in testosterone between unpaired and paired females indicates that this hormone may play a role in mate acquisition and territoriality of these sex role-reversed females.     

Are Androgens Related to Aggression in House Wrens?

Elevated circulating testosterone levels are hypothesized to allow male animals to direct resources into territorial and mating behaviors at the expense of reducing paternal care of offspring. For this hypothesis to apply, testosterone must facilitate territorial/mating behaviors and have antagonistic effects on paternal care, but this pattern has only been supported in some, not all, species. I tested whether androgens correlate with aggressive behaviors in male house wrens ( Troglodytes aedon), a double‐brooded species where paternal and aggressive behaviors overlap temporally. House wrens may therefore benefit from having a hormonal mechanism that allows males to rapidly change behavioral states. However, I found no evidence that androgens (testosterone and 5α‐dihydrotestosterone) relate to aggression in house wrens: Androgens did not increase in response to playback, and endogenous‐circulating androgens were not correlated with how aggressively males responded to those playbacks. Moreover, androgen levels were low during the pre‐breeding stage of the second brood, when many males establish new territories and attract new mates. This study adds to a growing body of the literature suggesting that the relationship between circulating androgens and aggressive behavior is more complex than originally thought.     

Shield characteristics are testosterone-dependent in both male and female moorhens

The immunocompetence handicap hypothesis proposes that the expression of secondary sexual characteristics is positively related to testosterone levels, but that elevated testosterone levels also impose costs from immune suppression. Hence, testosterone-dependent characteristics should accurately reflect male quality because only high-quality males are able to invest in large sexual characteristics without detrimental effects upon their own immune system. Most studies to date have focused on the role of testosterone in the expression of male ornaments and on the possible immunosuppressant effects of androgens in males. In the moorhen (Gallinula chloropus), a sexually monomorphic monogamous bird species showing a partial sex-role reversal, both sexes have a prominent frontal shield. We implanted both sexes with testosterone-filled implants to examine the effects of testosterone on shield characteristics and immune function. Shield size, thickness, and color were significantly increased by an experimental increase in testosterone concentrations in both males and females. Likewise, removal of the implants led to a rapid decrease in shield size and thickness in both males and females, suggesting that both sexes responded quickly to an increase or a decrease in testosterone. Moorhens implanted with testosterone had higher intensities of ectoparasite infestations than control birds, but other indirect measures of immunocompetence did not differ significantly between the two categories of birds.     

Effects of prenatal treatment with antiandrogens on luteinizing hormone secretion and sex steroid concentrations in adult spotted hyenas,Crocuta crocuta

Prenatal androgen treatment can alter LH secretion in female offspring, often with adverse effects on ovulatory function. However, female spotted hyenas (Crocuta crocuta), renowned for their highly masculinized genitalia, are naturally exposed to high androgen levels in utero. To determine whether LH secretion in spotted hyenas is affected by prenatal androgens, we treated pregnant hyenas with antiandrogens (flutamide and finasteride). Later, adult offspring of the antiandrogen-treated (AA) mothers underwent a GnRH challenge to identify sex differences in the LH response and to assess the effects of prenatal antiandrogen treatment. We further considered the effects of blocking prenatal androgens on plasma sex steroid concentrations. To account for potential differences in the reproductive state of females, we suppressed endogenous hormone levels with a long-acting GnRH agonist (GnRHa) and then measured plasma androgens after an hCG challenge. Plasma concentrations of LH were sexually dimorphic in spotted hyenas, with females displaying higher levels than males. Prenatal antiandrogen treatment also significantly altered the LH response to GnRH. Plasma estradiol concentration was higher in AA-females, whereas testosterone and androstenedione levels tended to be lower. This trend toward lower androgen levels disappeared after GnRHa suppression and hCG challenge. In males, prenatal antiandrogen treatment had long-lasting effects on circulating androgens: AA-males had lower T levels than control males. The sex differences and effects of prenatal antiandrogens on LH secretion suggest that the anterior pituitary gland of the female spotted hyena is partially masculinized by the high androgen levels that normally occur during development, without adverse effects on ovulatory function.     

Endocrine correlates of alternative phenotypes in the white-throated sparrow (Zonotrichia albicollis)

Many vertebrate species exhibit alternative phenotypes (or morphs), in which one sex displays phenotypic variation equal to or greater than the variation between the sexes. Males in such species typically display differences in reproductive strategies and morphology. Steroid hormones such as testosterone are known modulators of reproductive behavior and morphology and therefore are obvious candidates for the mediation of phenotypic differences between morphs. We conducted a year-round study in the white-throated sparrow (Zonotrichia albicollis) that exhibits alternative phenotypes in plumage coloration and behavior in both sexes: during the breeding season, white-striped males and females are more aggressive and have higher song rates than tan-striped individuals. At the beginning of the breeding season, free-living white-striped males had higher plasma testosterone concentrations than tan-striped males. However, this finding might have been due to different social experiences because captive male morphs sampled at similar times of year did not differ in testosterone concentrations. Captive white-striped males had larger testis and cloacal protuberance sizes than tan-striped males, which might be related to the divergent mating strategies of the morphs. Male morphs showed similar increases in luteinizing hormone following injections of gonadotropin-releasing hormone, but white-striped males showed larger increases in testosterone, indicating differences between morphs in gonadal testosterone production. Females had low concentrations of testosterone, and morphs did not differ. Plasma dehydroepiandrosterone (DHEA) concentrations were elevated in both sexes and morphs during the breeding and non-breeding seasons. These data do not support the hypothesis that testosterone activates behavioral differences between alternative phenotypes in the white-throated sparrow. Alternative testable hypotheses include hormonal effects during early development and direct genetic effects.     

Tight hormonal phenotypic integration ensures honesty of the electric signal of male and female Brachyhypopomus gauderio

Hormones mediate sexually selected traits including advertisement signals. Hormonal co-regulation links the signal to other hormonally-mediated traits such that the tighter the integration, the more reliable the signal is as a predictor of those other traits. Androgen administration increases the duration of the communication signal pulse in both sexes of the electric fish Brachyhypopomus gauderio. To determine whether the duration of the signal pulse could function as an honest indicator of androgen levels and other androgen-mediated traits, we measured the variation in sex steroids, signal pulse duration, and sexual development throughout the breeding season of B. gauderio in marshes in Uruguay. Although the sexes had different hormone titres and signal characteristics, in both sexes circulating levels of the androgens testosterone (T) and 11-ketotestosterone (11-KT) were strongly related to signal pulse duration. Consequently, signal pulse duration can serve as an honest indicator of circulating androgens in males and females alike. Additionally, through phenotypic integration, signal pulse duration also predicts other sexual traits directly related to androgen production: gonad size in males and estradiol (E2) levels in females. Our findings show that tight hormonal phenotypic integration between advertisement signal and other sex steroid-mediated traits renders the advertisement signal an honest indicator of a suite of reproductive traits.     

Evaluation of the pituitary-gonadal response to GnRH, and adrenal status, in the leopard (Panthera pardus japonensis) and tiger (Panthera tigris)

Frequent blood samples were collected to study hormonal responses to GnRH in male and female leopards and tigers. Animals were anaesthetized with ketamine-HCl and blood samples were collected every 5 min for 15 min before and 160 min after i.v. administration of GnRH (1 micrograms/kg body weight) or saline. No differences in serum cortisol concentrations were observed between sexes within species, but mean cortisol was 2-fold greater in leopards than tigers. GnRH induced a rapid rise in LH in all animals (18.3 +/- 0.9 min to peak). Net LH peak height above pretreatment levels was 3-fold greater in males than conspecific females and was also greater in tigers than leopards. Serum FSH increased after GnRH, although the magnitude of response was less than that observed for LH. Basal LH and FSH and GnRH-stimulated FSH concentrations were not influenced by sex or species. Serum testosterone increased within 30-40 min after GnRH in 3/3 leopard and 1/3 tiger males. Basal testosterone was 3-fold greater in tiger than leopard males. LH pulses (1-2 pulses/3 h) were detected in 60% of saline-treated animals, suggesting pulsatile gonadotrophin secretion; however, in males concomitant testosterone pulses were not observed. These results indicate that there are marked sex and species differences in basal and GnRH-stimulated hormonal responses between felids of the genus Panthera which may be related to differences in adrenal activity.     

Seasonal variation in circulating testosterone and oestrogens of wild-caught California ground squirrels (Spermophilus beecheyi).

Blood samples were obtained each month for 18 months in 1984-85 from wild-caught California ground squirrels. Circulating testosterone and total oestrogens were extracted from the plasma, measured by radioimmunoassay and compared with concurrent changes in plasma progesterone and prolactin in the same individuals. Male plasma testosterone concentrations peaked in January, shortly before mating, whereas female testosterone concentrations were low throughout the year. Female plasma oestrogen concentrations peaked in February, during the mating period. Juvenile males exhibited a transient increase in circulating testosterone in September, followed by testis growth, preputial separation, and the appearance of epididymal spermatozoa. Juvenile females exhibited a transient increase in circulating oestrogens in November. By the start of the first mating season after their births, neither juvenile males nor females differed significantly from same-sex adults with respect to plasma concentrations of oestrogen or testosterone. Plasma concentrations of progesterone and prolactin in killed individuals were similar to those obtained in previous studies of free-living S. beecheyi. Mean plasma concentrations of all measured hormones varied significantly with reproductive condition in adults of both sexes.     

Profiling of Androgen Response in Rainbow Trout Pubertal Testis: Relevance to Male Gonad Development and Spermatogenesis

The capacity of testicular somatic cells to promote and sustain germ cell differentiation is largely regulated by sexual steroids and notably androgens. In fish species the importance of androgens is emphasized by their ability to induce sex reversal of the developing fries and to trigger spermatogenesis. Here we studied the influence of androgens on testicular gene expression in trout testis using microarrays. Following treatment of immature males with physiological doses of testosterone or 11-ketotestosterone, 418 genes that exhibit changes in expression were identified. Interestingly, the activity of testosterone appeared stronger than that of 11-ketotestosterone. Expression profiles of responsive genes throughout testis development and in isolated germ cells confirmed androgens to mainly affect gene expression in somatic cells. Furthermore, specific clusters of genes that exhibit regulation coincidently with changes in the natural circulating levels of androgens during the reproductive cycle were highlighted, reinforcing the physiological significance of these data. Among somatic genes, a phylogenetic footprinting study identified putative androgen response elements within the proximal promoter regions of 42 potential direct androgen target genes. Finally, androgens were also found to alter the germ line towards meiotic expression profiles, supporting the hypothesis of a role for the somatic responsive genes in driving germ cell fate. This study significantly increases our understanding of molecular pathways regulated by androgens in vertebrates. The highly cyclic testicular development in trout together with functions associated with regulated genes reveal potential mechanisms for androgen actions in tubule formation, steroid production, germ cell development and sperm secretion.     

Development of intraspecific size variation in black coucals,white-browed coucals and ruffs from hatching to fledging

Most studies on sexual size dimorphism address proximate and functional questions related to adults, but sexual size dimorphism usually develops during ontogeny and developmental trajectories of sexual size dimorphism are poorly understood. We studied three bird species with variation in adult sexual size dimorphism: black coucals (females 69% heavier than males), white-browed coucals (females 13% heavier than males) and ruffs (males 70% heavier than females). Using a flexible Bayesian generalized additive model framework (GAMM), we examined when and how sexual size dimorphism developed in body mass, tarsus length and bill length from hatching until fledging. In ruffs, we additionally examined the development of intrasexual size variation among three morphs (Independents, Satellites and Faeders), which creates another level of variation in adult size of males and females. We found that 27–100% of the adult inter- and intrasexual size variation developed until fledging although none of the species completed growth during the observational period. In general, the larger sex/morph grew more quickly and reached its maximal absolute growth rate later than the smaller sex/morph. However, when the daily increase in body mass was modelled as a proportion, growth patterns were synchronized between and within sexes. Growth broadly followed sigmoidal asymptotic models, however only with the flexible GAMM approach, residual distributions were homogeneous over the entire observation periods. These results provide a platform for future studies to relate variation in growth to selective pressures and proximate mechanisms in these three species, and they highlight the advantage of using a flexible model approach for examining growth variation during ontogeny.