University of Groningen
Don’t underestimate father Lelono, Asmoro
DOI:
10.33612/diss.97045753
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Lelono, A. (2019). Don’t underestimate father: Effects of cryptic and non-cryptic paternal traits on maternal effect in a species without paternal care. Rijksuniversiteit Groningen.
https://doi.org/10.33612/diss.97045753
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Download date: 28-06-2021
Don’t underestimate father
Effects of cryptic and non-cryptic paternal traits on maternal effects in a species
without paternal care
Asmoro Lelono
Evolutionary Genetics, Behaviour and Development (EDGB) groups at the Groningen Institute for Evolutionary Life Sciences (GELIFES) according to the requirements of the Graduate school of Science (The Faculty of Science and Engineering, University of Groningen The Netherlands).
The research was supported by DIKTI scholarship from the Ministry of Research, Technology and Higher Education of the Republic of Indonesia awarded to Asmoro Lelono and Ton G G Groothuis and funds awarded to the latter.
Lay out : Legatron Electronic Publishing, Rotterdam, the Netherlands Figure : Asmoro Lelono
Photo : Marwedi Nurratyo Cover design : Asmoro Lelono
Printed by : IPSKAMP Printing, Enschede, the Netherlands ISBN: 978-94-034-1986-2
Don’t underestimate father
Effects of cryptic and non-cryptic paternal traits on maternal effects in a species
without paternal care
PhD thesis
to obtain the degree of PhD at the University of Groningen
on the authority of the Rector Magnificus Prof. C. Wijmenga
and in accordance with
the decision by the College of Deans.
This thesis will be defended in public on Friday 4 October 2019 at 12.45 hours
by
Asmoro Lelono born on 15 October 1968
in Wonogiri, Indonesia
Prof. A.G.G. Groothuis
Co-supervisor Dr. B.J. Riedstra
Assessment Committee Prof. J. Komdeur
Prof. M. Naguib Prof. W. Goymann
sesungguhnya Allah beserta orang-orang yang sabar.
O you who have believed, seek help through patience and prayer. Indeed, Allah is with the patient. QS, Al Baqarah: 153).
Chapter 1 General Introduction 9
Chapter 2 Female reproductive investment in relation to male 21 attractiveness in red junglefowl (Gallus gallus gallus)
Chapter 3 Ejaculate testosterone levels affect maternal investment in 37 red junglefowl (Gallus gallus gallus)
Chapter 4 The relationship between male social status, ejaculate and 67 circulating testosterone concentration and female yolk androgen transfer in red junglefowl (Gallus gallus gallus)
Chapter 5 Does paternal immunocompetence affect offspring 87 vulnerability to maternal androgens? A study in domestic
chickens
Chapter 6 Synthesis (General Discussion) 107
References 125
Summary 147
Samenvatting 151
Ringkasan 157
Acknowledgements 161
“Colorful”
“Colorful”
Chapter 1
General Introduction
Asmoro Lelono
1
Sexual selection, mate choice, and the study species
An individual’s reproductive success depends on how many surviving offspring’s it can produce. As males produce less costly sperm compared to the more costly eggs of females, males can increase their success by fertilizing many different females.
For females, producing eggs is much more limiting than producing sperm because of the size and the nutrient content of the eggs (Trivers, 1972), and that is why females have to be more selective in choosing a mate than males do (Giraudeau et al., 2011; Loyau et al., 2007). Females should, therefore, invest in carefully selecting the best male and supplying the offspring with adequate amounts of resources. Due to limited access to mates, females are not always able to select the best quality male for siring her offspring. There are several strategies by which females can deal with variation in mate quality (Johnsen and Zuk, 1996; Zuk et al., 1990b). They can compensate insufficient male quality by investing more in their offspring themselves, a compensatory strategy (Gowaty, 2008). Or they can decrease investment when fertilized by poor-quality males in order to reserve resources for the next reproductive attempt with another and better male (Gowaty and Hubbell, 2009; Horváthová et al., 2012). This is a core aspect of sexual selection theory and has given rise to numerous studies in the past decades (e.g. Olson et al., 2008; Parker, 2006).
From the above it is clear that one important aspect in female reproductive decisions is recognizing mate quality. This has been extensively studied in insects (Kotiaho et al., 2003; Pischedda et al., 2011; Wedell and Karlsson, 2003), fish (Evans et al., 2010), and birds (Bolund et al., 2009; Cunningham and Russell, 2000; Horváthová et al., 2012). For birds, the wild ancestor of the domestic chicken is a very suitable species as it is sexual dimorphic with the male being much more colourful and hardly contributing to parental care, having a harem of several females. As a consequence, the secondary sexual male traits reflect sexual selection and might reflect male genetic quality on the basis of which females may base their partner preference on.
Like in many species, male chickens have several traits, such as plumage coloration (Zuk et al., 1990b), tail length (Johnsen and Zuk, 1996), body mass (Parker and Garant, 2004), vocalisations (Wilson et al., 2008), fighting ability (Parker and Ligon, 2003), and courtship behaviour (Leonard and Zanette, 1998; Parker and Ligon, 2003), that may serve as an honest signal of quality. Especially the comb characteristics are important in this respect. In red junglefowl, the ‘ancestral’ chicken, increased comb size is associated with i.e. increased health (Zuk et al., 1990a) and dominance status
(Parker and Ligon, 2002; Zuk, and Johnsen, 2000; Zuk et al., 1990b). The dominance status is an important determinant for access to mattings. Moreover, comb size reflects male attractiveness (Collias and Collias, 1996; Johnsen and Zuk, 1996):
Females prefer large-comb males over males with small-comb (Parker and Ligon, 2003). Other comb characteristics are important too: e.g. females prefer to mate with brighter comb males (Parker and Ligon, 2003). The phenotypic characteristics of the male comb may therefore be treated as a proxy for male quality.
Maternal effects and hormones
Female choice may lead to a change in maternal investment that can in turn act as a pathway for a maternal effects. Mother preference could be stimulated by an exposure of different male quality which stimulate their reproductive investment. Maternal effects induce phenotypic changes in the offspring brought about by phenotypic aspects of the mother, which is currently much researched in the literature. Such maternal effects can be both postnatal (e.g. rearing, feeding and protecting the offspring) and prenatal (e.g. egg mass and exposure to hormones Groothuis and Schwabl, 2008). The latter is intriguing as it is usually much more cryptic than the former, is powerful because it can have long-lasting effects on the organization of brain, behaviour and physiology, and is often overlooked in genetic studies. It is also more difficult to study, especially in mammals where prenatal exposure to hormones may be variable in time and interactions between siblings may occur, while experimental manipulation of prenatal exposure to hormones inevitably also effect the mother, potentially leading to confounding effects. However, egg-laying species, such as birds, in which the embryo develops outside the mother individually in a concealed environment, facilitate studying prenatal maternal effects (Groothuis and Schwabl, 2008; Von Engelhardt and Groothuis, 2011).
One important component of prenatal maternal effects that has received a lot of attention in the past two decades, is that of embryonic exposure to maternal hormones (Groothuis et al., 2005b; Von Engelhardt and Groothuis, 2011). In many taxa, including mammals, embryos are exposed to maternal hormones such as testosterone. It is not surprising that most research on this phenomenon has been performed in birds, as their eggs contain substantial amounts of steroids (Schwabl,
1
1993) and the embryo is easily accessible for measuring and manipulating hormone exposure without interfering with the mother.
Several factors can influence maternal deposition of hormones to the egg, an intriguing one being partner attractiveness induced by the morphological traits or behaviour of the male (Gil et al., 2006, 2004, 1999; Loyau et al., 2007; Loyau and Lacroix, 2010; Von Engelhardt et al., 2001), reviewed in (Von Engelhardt and Groothuis, 2011). This is intriguing as it actually leads to a maternal effect that is induced by a cryptic paternal trait and this has hardly been studied.
One particular group of hormones that has received much attention in the field of hormone mediated maternal effects in birds is that of androgens. This group includes testosterone (T), androstenedione (A4) and 5 alpha-dihydrotestosterone (DHT) that have many functions in an organism and can exert strong effects during development. Mothers may use these androgens as a tool to influence the embryo’s development in order to adjust its phenotype to the prevailing or future environmental conditions (Groothuis et al., 2005b; Schwabl, 1996).
Increased levels of yolk androgens may induce a variety of effects on the chicks’
developmental time, growth, behavioural phenotype, immune functions and physiology, some being positive and others, such as on immune function and metabolic rate, detrimental (Eising et al., 2001; Eising and Groothuis, 2003; Gil et al., 2004; Müller et al., 2005, 2007; von Engelhardt et al., 2006; Von Engelhardt and Groothuis, 2011). However, mothers do not always allocate large amounts of androgens to (all of) her eggs. This suggests that not all effects of exposure to these hormones are beneficial in all circumstances and that perhaps the genetic quality of the offspring may determine their vulnerability to the negative effects of yolk androgens. Therefore, only the offspring of high-quality fathers may be able to cope with the costs associated with hormone exposure, which has been suggested to be chiefly in the domain of reduced immune function. As a consequence, females should adjust hormones levels according to the genetic quality of their mate (Von Engelhardt and Groothuis, 2011).
Paternal effects and the ejaculate
Males vary not only in traits externally visible, but also in more cryptic aspects such as sperm and ejaculate characteristics. Ejaculates need resources to produce and males
are therefore predicted to allocate resources preferentially to copulate with the most promising females (Cornwallis and O’Connor, 2009). females on the other hand may copulate with more than one male in a single breeding cycle, inducing selection for sperm competition. Physiological adaptations of the male to this sperm competition include relatively large testes, large sperm stores, and long spermatozoa. There are also behavioural adaptations such as increased mate guarding and increased copulation frequencies during the time frame that females are fertile (Cornwallis and Birkhead, 2006). Another important factor in sperm competition is sperm motility, determining the ability of the sperm cell to reach and fertilize the ovum (Birkhead et al., 1999), which depends partly on the chemical composition of the ejaculate.
Intra-sexual selection often leads to a skew in male reproductive success, due to male hierarchies in which the dominant males to a large extent monopolize copulations (Ligon et al., 1990; Olson et al., 2008). In mating systems where winners ‘take almost all’, there are only scarce opportunities for losers to propagate (Parker and Ligon, 2002; Young et al., 2007). This leads to selection in subordinate males where in order to increase their fertilization success, these males allocate more resources to sperm and ejaculate production to enhance sperm quality, in order to outcompete sperm of dominant males.
During reproduction, males contribute not only spermatozoa as genetic material but also hormones and other substances in the ejaculate supporting fertilization success (Anderson and Navara, 2011). The ejaculates fulfil their task to provide optimal conditions for fertilization and contain immunosuppressive substances that protect spermatozoa from damage in the reproductive tract (Pohanka et al., 2002).
In insects, the ejaculate substances can exert diverse behavioural and physiological effects in females, including altered longevity and reproductive output (South and Lewis, 2011). If such effects would also present in avian species, then that would provide the male, via manipulation of his ejaculate composition, a tool to influence female reproductive investment in her offspring.
In contrast to maternal effects, a cryptic paternal trait have been understudied.
Most of the focus has been on the females side since they usually play a larger role during reproduction, egg production and raising chicks (Horváthová et al., 2012;
Reed and Clark, 2011). However, males also have a vested interest in the quality of their offspring since they invest in reproduction via risky agonistic interactions that may increase their dominance status or by investing in ejaculates (Cornwallis et al., 2014; Cornwallis and Birkhead, 2007; Pizzari et al., 2007). As mentioned above,
1
male investment may influence female reproductive decisions, in interaction with maternal effects. Given that strategies of maternal resource allocation to the egg have been demonstrated to be dependent on male quality, and that sperm or ejaculate quality may differ between males of different quality (Birkhead et al., 1999;
Froman and Kirby, 2005; Pizzari et al., 2007), this provide us an excellent case to study the interactions between a cryptic paternal trait and maternal effects.
Outline of the main questions for this thesis
Based on the above there are several questions to address in relation to paternal and maternal reproductive strategies and the interaction between these two. Firstly, do differences in partner quality affect prenatal investment in reproductive traits such as the time to start a clutch, egg mass, clutch size, yolk hormone composition, and the quality of offspring? And if this is the case, is this than caused by the perception of the physical appearance of the male?, or perhaps by more cryptic components, such as the chemical composition of the males’ ejaculates, as suggested by Parker, (2003)? The ejaculate chemical composition is the inspiration for the second and third question: Is there a difference in ejaculate hormone levels between males of different social status (or quality), and if so, do these hormones affect female reproductive decisions directly?
The fourth question is related to sex-specific maternal effects. There are several studies reporting maternal effects on offspring in a sex-dependent way: in other words, the maternal effect is different between male offspring and female offspring (Tschirren, 2015). Moreover, there is even some evidence that this may come about by females allocating different amount of their hormones to male and female eggs (Badyaev et al., 2006; Müller et al., 2002). This opens the possibility that females differentially deposit hormones to her eggs depending on male attractiveness or quality, adding another layer or complexity. Although the underlying mechanism for this sex-specific hormone deposition is difficult to understand as the hormone is added to the yolk before fertilization, it would explain sex-dependent maternal effects and therefore urgently need replication. From this the fourth question emerged: Do females indeed expose their embryos to different amounts of hormones depending on the interactions between mate quality and offspring sex? For example, do females favour sons in hormone deposition when paired with attractive males, in line with
the sexy son hypothesis, or do they do this when paired with low-quality males to compensate for poorer genes?
Finally, the fifth and sixth question I address in this thesis are based on the following:
As the results of several studies show that mothers deposit more androgens in their eggs when exposed to attractive males, the question is raised why, if these hormones are so beneficial for the offspring, not all mothers deposit high amounts of such hormones in their eggs. One hypothesis is that maternal testosterone has also costs for the offspring, especially immune suppression (Groothuis et al., 2005a; Müller et al., 2005) and that only offspring sired by good-quality males can withstand those costs.
Therefore I tried to answer the following two questions: 5: Do females differentially allocate hormones to eggs when their mates differ in immunocompetence, and 6:
Are the sons and daughters of immunocompetent males better able to bear the costs of being exposed to hormones than offspring from ‘immuno-incompetent’
fathers?
The animal models
In order to answer these questions, I chose the chicken, both the ancestral red junglefowl (Gallus gallus gallus) and the domesticated white leghorn (Gallus gallus domesticus) as my animal models. I think that these animals are appropriate models for the following reasons: Firstly, they show clear sexual differences (dimorphism).
Males possess elaborate ornaments, colourful plumage, and fleshy comb and wattles on the head and neck that are very variable, whereas females are relatively drab and cryptic (Johnsen and Zuk, 1996; Collias & Collias 1996). This variability in male characteristics is related to male quality (Zuk et al., 1990; Parker and Ligon, 2002, 2007) and is therefore important since most of my research questions are based on the effects of variation in male quality on female reproductive investment. Secondly females show clear mate choice. Hens prefer to copulate with larger and brighter comb males (Ligon and Zwartjes, 1995; Zuk et al., 1995, 1990b). Thirdly, there is some evidence that variation in male quality in chickens induces differential maternal investment. Forkman and Corr, (1996) showed that female Leghorns invested more in egg mass when mated with more symmetrical wattles males, a trait that most likely is related to attractiveness, but there was a negative relationship between wattle size and the number of fertilized eggs produced. Furthermore, Müller et al., (2002) found
1
that the sex-specific differences in the yolk hormone allocation strongly depend on the social rank of the mother: Dominant female Leghorns, but not subordinate females, allocated significantly more testosterone to male eggs than to female eggs.
On the other hand, subordinate females increased the testosterone concentrations of female eggs. Fourthly, the procedure of collecting ejaculate samples through an abdominal massage method is relatively easy (Burrows W.H and Quinn J.P, 1939).
The ejaculate sampling is a common procedure for artificial insemination and also sperm hormonal analysis (Birkhead et al., 1999; Parker, 2003; Pizzari et al., 2007). This enables us to tease apart the effects of visual input of the female and the sperm- or ejaculate androgen concentration. Finally, both the wild species and the domestic form are available, including selection lines for immunity. The wild study species has the advantage that they can be used to test hypotheses about function and evolution of behaviour, as highly selected domestic strains may have lost certain aspects, such as laying a ‘normal’ clutch. However, the availability of selected strains, of which the red junglefowl is the original ancestor, gave us the possibility to test whether females mated with males of different immunocompetence invested differently in reproduction and whether this affected their offspring.
Overview of this thesis
This thesis contains the following chapters: In chapter 2, I tested the effect of variation in male comb size on maternal investment. Our hypothesis was that mothers are able to recognize the quality of their partner and differentially invest in reproduction according to mate quality. More specifically, Horváthová et al., (2012) showed that avian species invest more when mated to attractive, high-quality partners compared to when mated with less attractive partners. I therefore randomly mated females with a male with a large-comb or with a male with a small-comb and let them produce a clutch and raise the offspring. After the first clutch, females were then paired up with a male of the opposite comb size category and left to raise a second brood. I found that females mated with large-comb male initiated clutch production sooner than females mated with small-comb males. Moreover, I found differences in offspring growth that were affected by the interaction between offspring sex and male comb size. Sons of small-comb males grew faster than sons from large-comb males and daughters showed the opposite effect.
Parker, (2003) found that there was differential reproductive investment between females artificially inseminated with ejaculates of large-comb males or of small- comb males. He suggested that this was because of the differences in genetic quality between the different males, but he could not rule out that possible differences in the composition of the ejaculate was causing this. Anderson and Navara, (2011), showed that the seminal fluid (or ejaculates) of birds contain several steroid hormones. Steroid hormones are potent signal molecules that have diverse effects.
They were therefore of key interest to us, especially testosterone, since this hormone shows robust differences in blood level concentrations between roosters of different social status and sexual characters.
In chapter 3, I therefore studied whether males of different social status (and comb size) also differed in the chemical composition of their ejaculates. My hypothesis was that males should adjust their allocation of the hormone to the ejaculate to their social status. Dominant males can secure mattings by monopolizing females while subordinate males have rare fertilization opportunities and should heavily invest in their ejaculate and sperm. I also explored whether the ejaculate T level mimicked circulation levels or whether there was a trade-off between the two. I tested T level in the circulation and ejaculate of dominant and subordinate males after a social challenge and found that dominant and subordinate males differed both in the ejaculate T level and circulating T level. Males seemed to trade-off these levels since dominant males had, as is frequently found in the literature, higher circulating T level, but also lower ejaculate T level. I decided to follow up these findings by collecting ejaculates of dominant males and experimentally elevate the T level to that of the subordinate males and artificially inseminated hens with either the T enriched ejaculates or control ejaculates. Females were then allowed to produce a clutch and I recorded female reproductive investment (clutch size, egg mass, and clutch size).
Directly after clutch completion the treatment was reversed and the same variables were recorded. I also raised the chicks of two different treatment conditions to study their growth and competitiveness. In line with my hypothesis, females invested differentially in her eggs: females produced heavier eggs when inseminated with enriched ejaculates. Moreover, I found a similar growth pattern for the offspring as in chapter 2. Sons of hens inseminated with control ejaculates grew slower then sons of hens inseminated with enriched ejaculates, whereas an opposite pattern was present in daughters. These differences in growth patterns between sons and daughters, in the interaction with male ‘quality’ could be explained by differential
1
exposure to androgens during embryonic development depending on the quality of the partner, as testosterone is known to enhance early growth (Casagrande et al., 2011; Müller et al., 2010).
In chapter 4, I tested whether females mated with different quality males indeed differentiated sons and daughters concerning yolk androgen deposition. I therefore designed and performed an experiment, using a similar design as in the previous experiments: I paired up hens with both large-comb (attractive) or small-comb (unattractive) roosters that had won or lost a staged dyadic agonistic encounter in full view of these hens, and subsequently left the hens to produce a clutch. After clutch completion I repeated the experiment but reversed partners (from winner to loser and vice versa). In both reproductive attempts I recorded egg mass, clutch size, clutch initiation time, and also measured yolk androgen deposition and embryo sex.
Here I reported that females indeed deposited androgens in the eggs in such a way that it could explain the differences in early growth found in chapter 2 and chapter 3.
Differential deposition depending on mate quality is the main question of chapter 5. Maternal androgens stimulate growth and competitiveness, why do then females not always provide her androgen in ample amounts for her offspring? This may be because there are also negative aspects (costs) of early exposure to testosterone. An old hypothesis, still frequently cited but never properly tested, proposes that early exposure to testosterone decreases immunocompetence and that only chicks sired by attractive high-quality fathers are able to bear these costs (e.g. (Gil et al., 1999)).
In chapter 5 I demonstrated an experiment in which I firstly paired white leghorn females (not selected on immunocompetence) with a rooster selected for high or low natural antibody (Nab) and measured androgen levels on freshly laid eggs and their egg production. Secondly, I manipulated yolk androgen levels in the eggs and tested whether the immunocompetence of chicks sired by fathers differentially selected for Nab was affected. Unfortunately, I did not find any supports for the hypothesis.
Finally, in chapter 6 (Synthesis/General Discussion) I synthesized all of the results and placed them in a broader context. I also delivered a new point of view to study how males in different social status allocated their investment and how the relationship between a cryptic paternal trait, maternal effect with the sex of the offspring in the evolutionary perspectives. In this chapter, I introduced a new approach to understand how paternal and offspring sex play an intriguing roles in the trend of evolution direction. This new approach will provide a better and complete story compared to the previous studies which focused mostly on the maternal side.
“Rooster”
Chapter 2
Female reproductive
investment in relation to male attractiveness in red junglefowl (Gallus gallus gallus)
Asmoro Lelono Bernd Riedstra Ton G. Groothuis
Submitted
Abstract
Mothers are predicted to invest differentially in offspring from partners of different quality in order to optimize their fitness. A recent meta-analysis showed that females in female-care only species do so by investing more in egg mass when mated with attractive males rather than in other reproductive traits. We found some discrepancies with this result after reviewing all the literature available. Here, we report on these discrepancies and on an experiment we performed on the relationship between male comb size, a signal for attractiveness, and maternal reproductive investment in clutch production and offspring development in red junglefowl. We randomly mated females with a male with a large- or a small-comb and let them produce a clutch and raise the brood. After this first clutch, we paired up females with a male of opposite comb size and allow them produce a second brood. We found that females paired with large-comb males produced eggs sooner after pairing than females mated with small-comb males, but we did not find differences in egg mass or clutch size. Male comb size also affected growth and body condition, but only in daughters.
Twenty four week old daughters of large-comb males were heavier and in better condition. These results support the positive differential allocation hypothesis (DA) for mate quality dependent reproductive investment rather than negative DA in which females would invest more in offspring sired from poor-quality males. This outcome adds to the discrepancies we found in the literature that warrants more study.
2
Introduction
If there is variability in mate quality, life history theory predicts that reproductive strategies will evolve that allow individuals to adjust their reproductive investment according to the expected fitness returns (Burley, 1986; Sheldon, 2000). Although in many species males invest heavily in reproduction through the acquisition and maintenance of territories and mates, females, in general, are limited by the number of eggs they can produce. Females invest initially more in the actual production of zygotes, embryos, and new-borns both through a more substantial investment in the ovum while males do in spermatozoids, as well as during incubation or gestation. It is therefore expected that the physiological and behavioural reproductive strategies to adjust reproductive investment to partner quality are especially pronounced in females, certainly in species where the male does not contribute much to embryonic development or postnatal care.
Reproductive strategies involve decision making about ‘when to reproduce’, ‘how many offspring to produce’, and ‘how much to invest in the offspring’ in order to optimize current and future reproductive success. If females of equal reproductive potential end up with mates of different quality, a dichotomy in reproductive strategies will emerge: either invest more in current offspring when mated with high-quality males because the offspring will be of high-quality too (differential allocation (DA) hypothesis), (Burley, 1986; Sheldon, 2000) or invest more in current offspring when mated with low-quality males to compensate the offspring for having a low-quality father (the compensation hypothesis), (Gowaty, 2008; Gowaty et al., 2007). The strategy chosen may depend on several aspects such as the species specific division of labour between the sexes in raising offspring (Horváthová et al., 2012). Currently, these two hypotheses are seen as one continuum and are now mostly referred to as positive and negative DA, where positive DA seems to be most common in birds (Haaland et al., 2017; Harris and Uller, 2009; Kindsvater and Alonzo, 2014).
The females’ non-genetic contributions that influence offspring development, collectively known as maternal effects (Mousseau and Fox, 1998), have been shown to be related to paternal attractiveness in several avian species but not always in a consistent manner (for a review see (Horváthová et al., 2012; Von Engelhardt and Groothuis, 2011). This maternal effect study is of special interest as it consists of a quantitative meta-analysis, which showed that, in favour of the positive DA over
negative DA, avian females invest more in current reproduction when paired with attractive high-quality males. From this analysis, it also emerged, but with small to moderate effect sizes, that females in species with bi-parental care invest in producing more eggs, whereas females of species with female only care invest in larger eggs when paired with high-quality attractive males. There was no effect on
‘when to produce’ except an advance in clutch initiation of first year female mallards (Anas platyrhynchos) when mated with more attractive drakes (Sheppard et al., 2013). The investment in more cryptic components such as the embryonic exposure to immune stimulants and exposure to androgens, has no consistent effect. This latter finding may in itself not be evidence against the positive DA hypothesis since it is unknown whether adding hormones or immuno-enhancing molecules are costly to females both in the present or in future reproductive events (Groothuis and Schwabl, 2008; Sheldon, 2000). However, the results of this meta-analysis with respect to bi-parental versus uniparental care should be taken with some cautions as the meta-analysis contained only 8 studies on 5 species that show female only care: the chinese quail Coturnix chinensis, (Uller et al., 2005), the grey partridge Perdix perdix, (Garcia-Fernandez et al., 2010), the houbara bustard Chlamydotis undulata, (Loyau and Lacroix, 2010), the mallard Anas platyrhynchos, (Bluhm and Gowaty, 2004; Cunningham and Russell, 2000; Giraudeau et al., 2011), and the peafowl Pavo cristatus, (Loyau et al., 2007).
We found 7 additional studies, 4 older than at the time of publication of this meta- analysis (Petri and Williams 1993 (peafowl), Forkman and Corr 1996 (domestic chicken – Gallus gallus domesticus), Rintamaki et al. 2000 (black grouse - Tetrao tetrix), Parker 2003 (red junglefowl – Gallus gallus gallus) and 3 of similar date or newer (Cucco et al. 2011, Alonso-Alvarez et al. 2012, both in the red legged partridge – Alectoris rufa, Sheppard et al. 2013 and mallard), that relate female reproductive investment to male quality in female care only species. In contrast to the results of the meta-analyses (Horváthová et al., 2012), four of these studies report on egg size (mass/volume) and all report that there is no relationship between attractiveness and egg size (Petri & Williams 1993, Parker 2003, Cucco et al. 2011, and Alonso- Alvarez et al. 2012) with the exception of Forkman and Corr (1996) that did find a positive relationship (but see below). Also, four studies report on the relationship with laying date, 2 of these (Petri and Williams 1993, Cucco et al. 2011) report no relationship and two (Alonso-Alvarez et al. 2012, and Sheppard et al. 2013) indicate an advanced laying date when mated with more attractive males. One of the studies
2
(Forkman and Corr, 1996) showed, contrary to expectation, that although female leghorns (Gallus gallus domesticus) invested more in egg mass when mated with males with more symmetrical wattles, a trait that may be related to attractiveness, there was a negative relationship between wattle size and the number of fertilized eggs produced. Moreover, in four of these studies, the number of eggs laid increased with male attractiveness (Petri and Williams 1993, Rintamaki et al. 2000, Parker 2003, and Alonso-Alvarez et al. 2012). These extra results make the generalization of the conclusion that females mate with attractive males invests more in egg mass rather than the number of eggs in female only care species, less convincing.
Given the inconsistent results presented above, the current study aimed was to add new data to answer the question whether in species where only females perform offspring care there is the differential investment in current reproduction that is related to mate quality. Many traits may signal quality of males; these may be behavioural traits, such as food provision and vigilance (Pizzari 2003, Wilson et al.
2008) or morphological traits such as the size of the comb. Comb size in junglefowl positively affects female mate choice (Collias and Collias, 1996; Johnsen and Zuk, 1995; Parker and Ligon, 2003; Zuk et al., 1995). Moreover, comb size is heritable (Parker, 2003) and is related to social dominance (Parker et al., 2002; Parker and Ligon, 2003). Furthermore, increased comb size is related to decreased immuno- competence (Zuk et al. 1995), which indicates that only high-quality males can bear the cost of a decreased immunocompetence (Zuk 1992). These relationships indicate a correlation between phenotypic quality and attractiveness where males with large-comb are of higher quality because they are more dominant and have fewer circulating lymphocytes (Folstad and Karter, 1992; Verhulst et al., 1999; Zuk et al., 1995) and are therefore more attractive to females than small-comb males.
We therefore firstly paired up hens with roosters with a large-comb or a small-comb (selected at random) and allowed them to produce a clutch and raise a brood to see whether rooster quality influenced female reproductive investment in the clutch.
Importantly, we also analysed whether offspring development was positively affected when females were mated with high-quality males, as was the case in the mallard (Bluhm and Gowaty, 2004), the houbara bustard (Loyau and Lacroix, 2010), and the peafowl (Loyau et al., 2007; Petrie and Williams, 1993). Studies often do not report on offspring development even though this is an important outcome of reproductive investments. Subsequently, we switched mating partners: Hens those were first paired up with a large-comb (high-quality) rooster were now paired up
with a small-comb (low-quality) rooster and vice versa. Hens were then allowed to produce a second clutch and brood. Here we report on the relationship between male quality (as reflected by comb size) and ‘when to reproduce’ (clutch initiation time), ‘how many offspring to produce’ (clutch size), ‘how much to invest in offspring’
(egg mass), and offspring growth into early adulthood.
Methods
Experimental animals
In this experiment, we used 14 pairs of sexually naive captive bred red junglefowl (Gallus gallus gallus) from our aviary at the University of Groningen The Netherlands.
All hens and 10 roosters were two to three years old; the other 4 roosters were 1 year old. The experiment was initiated in late spring of 2013 and lasted until autumn of 2013. Prior to the experiment, hens and roosters were housed in two single sex groups in separate outdoor aviaries.
Experimental design
Two types of biometrical measurements of all parental birds were taken at the beginning of pairing the birds with each other in order to asses male quality and balance hens over the two experimental groups. We measured body mass to the nearest gram and comb surface area (comb size) was determined as follows:
Firstly, we attached a removable circular sticker (diameter 0.8 cm) to the comb of the males and then photographed all male comb using a digital camera (Canon SX 500 IS: diaphragm 4,3 – 129,0 mm). Pictures were made only of the left side of the rooster heads. The images were imported in GIMP 4.8, and the comb were manually outlined, and the number of pixels in the area was determined. Comb size was then determined by contrasting this number of pixels against the number of pixels of the sticker (from the same photograph) with a known surface area determined. Similarly, comb length was determined by measuring the longest straight line from the upper beak (front comb) to the end of the comb. Furthermore, we determined comb colour using a spectrophotometer and procedure identical to that was described in Riedstra et al. (2013). Roosters were then ranked by of comb size and the 7 birds with the largest comb were qualified as ‘large-comb’ roosters, the remaining 7 with
2
the smallest comb size as ‘small-comb’ roosters. Male phenotypic characteristics are described in Table 1.
We then paired up females with either a large-comb male or a small-comb male in such a way that the average body mass of the two female groups did not differ (see Table 2). These pairs were each housed in one of 14 identical aviaries (1.5 × 3 × 2.5 m (l × w × h)) with ad-libitum access to water and food (standard chicken pellets), a dustbathing area, a perch, and a nest site. All pairs were checked daily for egg production. When eggs were produced, they were weighed and marked with a non- poisonous felt-tipped pen for identification purposes and placed back in the nest and left for hatching.
Chick housing and growth
On the day of hatching, each chick was weighed, and individually colour marked using flexible rubber leg bands. After that, they remained in the home cage condition with both parents present for five weeks. We choose for this setup because the males’ phenotypic characteristics could reinforce the hens’ reproductive decisions during the total period of incubation and chick rearing, even though females (both red junglefowl and feral chickens) living in natural conditions are generally classified as solitary when having small dependent chicks (Collias & Collias 1996, McBride et al. 1969). They may however frequently encounter males or other broody females during foraging because their home ranges and territories overlap and it has even been observed that males tidbit to chicks (McBride et al. 1969), which we also found on only a few occasions. Casual observations indicated that the males in our experiment remained at a distance from the broody hen and her chicks and hardly interacted with them.
After three weeks the leg bands were replaced by a numbered metal wing tag for permanent identification. At the age of five weeks, all chicks of the first reproductive attempt were relocated to a single large aviary. The chicks from the second clutch were also relocated at the age of 5 weeks, to a different but identical aviary. At the age of three months, the roosters and hens of the first clutch were separated and divided into two aviaries. Hens of the second clutch were housed with the hens of the first clutch when they were three months old, whereas the roosters of the second clutch were housed in a separate aviary to avoid escalation of aggression.
At the age of 24 weeks (about 5 and a half months), we determined body mass and tarsus length (using sliding callipers) of all offspring.
Reversal of treatment
After the first brood was raised and removed, the treatment was reversed after at least one week of recuperation: females those were first paired up with a large-comb male were now paired up with a small-comb male and vice versa. At this point, biometrical measurements of the males were taken again, and the next procedures were done as mentions above (experimental design). Since females first mated with large-comb males, initiated clutch laying a week earlier than females paired with small-comb males (see results section), they had on average a week longer to recuperate from the first reproductive attempt.
Statistical analysis
Based on the distribution of the raw data and the residuals of models we used parametric tests.
We first analysed whether our divisions between large- and small-comb males resulted in two groups that differed in comb size (surface area), and length whether other phenotypic characteristics (comb colour and body mass) differed between the two groups using independent t-tests (see Table 1). Similarly, we then analysed whether female body mass differed between the two treatments at both the onset of the first and the second reproductive attempt (see Table 1).
To analyse female reproductive performance (clutch size, average egg mass and clutch size) we conducted a multivariate test. The test analysed on the difference between clutch initiation time, average egg mass and clutch size between the reproductive attempt with a large-comb male and the attempt with a small-comb male, using a general linear model. Furthermore, we analysed the difference in hatching success separately (because our main predictions were on clutch initiation time, clutch size, and average egg mass) using a one sample t-test. Two females, those firstly paired with a small-comb and secondly with a large-comb male, did not reproduce at all within the given time frame of the experiment and were removed from all analyses on reproductive investment because no other data then clutch size (0) was available. Furthermore, two hens produced a clutch in both conditions, but none of their eggs showed embryonic development and did not hatch. These clutches were omitted from the analyses on hatching success.
The data of chick body mass at hatching and 24 weeks old was analysed using generalized linear mixed models on body mass averaged per mother over same- sex siblings, mother as a random factor and male quality, offspring sex and the
2
interaction between male quality and offspring sex as fixed variables. Here we also performed 2 post-hoc tests within daughters and 2 within sons because there was a significant interaction effect on growth (body mass and condition), of offspring sex and male quality. All statistical analyses were performed in SPSS 24.
Results
Phenotypic differences between males and between females
Our selection of comb size resulted in significant differences in comb size and comb length between the two groups of males (see Table 1). Large-comb males had comb surface areas about twice the size of the small-comb males, and comb length was about 1.3× longer. Furthermore, large-comb males were heavier than small-comb males in the first reproductive attempt, but this was not so in the second. Overall there were no differences in comb colour characteristics between the groups (see Table 1). There was also no effect of treatment, reproductive attempt (first or second clutch) on female body mass (Table 1, first row). Therefore, our objective was reached namely that the treatment groups only differed in male comb size.
Table 1 | Phenotypic characteristics (mean ±SE in brackets) of large – (N = 7) and small-comb roosters (N = 7): Body mass is expressed in gram, comb size in cm2, comb length in mm, brightness as the percentage reflectance, chroma is dimensionless, and hue is expressed in nm
first clutch second clutch
male comb size large small T p large small T p
male body mass 1232.3 (68.5) 1043.3 (42.0) 5.06 0.046 1131.0 (61.6) 1005.2 (27.3) 2.78 0.121 comb size 18.9 (0.98) 10.5 (1.09) 32.0 < 0.001 19.0 (0.89) 10.0 (1.20) 37.4 < 0.001 comb length 78.3 (2.38) 59.8 (2.11) 32.7 < 0.001 78.0 (2.13) 60.2 (2.31) 32.8 < 0.001 brightness 9.4 (0.71) 9.3 (0.93) 0.02 0.885 9.3 (0.52) 11.2 (1.61) 1.45 0.252
chroma 1.29 (0.14) 1.41 (0.16) 0.32 0.580 1.32 (0.08) 1.36 (0.21) 0.03 0.859
hue 624.6 (3.41) 611.9 (8.83) 2.01 0.183 612.6 (5.93) 597.7 (14.02) 1.11 0.311 female body mass 874.1(29.2) 828.8(36.9) 0.96 0.350 833.4 (19.5) 859.3 (20.0) 0.85 0.370
Female reproductive investment
There was no effect of male comb size on either clutch size (multivariate GLM:
F(1,11) = 0.02, p = 0.905) or egg mass (multivariate GLM: F(1,11) = 0.152, p = 0.705).
However, hens started laying egg approximately one week sooner when paired up with a large-comb rooster than a small-comb rooster (Figure 1 multivariate GLM:
F(1,11) = 5.05, p = 0.048). There were no differences in hatching success (one sample
t-test T = 0.583, P = 0.576), see Table 2 for reproductive variables.
Figure 1 | Mean of clutch initiation time (in days ±SE) of hens that were first paired up with a large- comb male and then with a small-comb male are represented by the closed dots. The reverse treatment is represented by the open dots
Table 2 | Reproductive performance (mean ±SE., and sample size in brackets) of females paired up firstly with large- and secondly small-comb roosters or vice versa
first clutch second clutch
male comb size large small small large
clutch initiation time (days) 10.1 ± 2.2 (7) 17.6 ±1.6 (5) 15.7 ± 3.5 (7) 8.2± 2.01 (4) clutch size 6.0 ± 0.6 (7) 6.6 ± 1.2 (5) 5.7 ± 0.7 (7) 6.0 ± 0.4 (4) average egg mass 30.8 ± 1.2 (7) 31.8 ± 0.8 (5) 32.4 ± 1.2 (7) 33.5 ± 0.6 (4) hatching success
(percentage) 63.6 ± 10.9 (5) 71.1 ± 8.5 (5) 64.7 ± 10.3 (5) 50.0 ± 9.5 (4) hatchling mass 23.3 ± 1.2 (22) 24.0 ± 0.8 (22) 21.0 ± 1.1 (16) 22.8 ± 0.5 (12)
2
Offspring growth
There was no effect of male comb size, offspring sex or the interaction between male comb size and offspring sex on body mass at hatching (GLMM: F-values < 1.44, p-values > 0.706). Twenty-four weeks after hatching, sons had outgrown daughters (GLM, F(1,16) = 37.3, p < 0.001), and there was no direct effect of male comb size
(F(1,15.1) = 0.29, p = 0.599), but there was an interaction between male comb size and
offspring sex on body mass (F(1,15.1) = 5.29, p = 0.036). Sons of large-comb males (N = 13) did not differ from sons from small-comb males (N = 19; figure 2; GLMM;
F(1,2.4) = 6.66, P = 0.101), but daughters (N = 13) from large-comb males were heavier
than daughters (N = 14) of small-comb males (F(1,11) = 11.82, p = 0.006, Figure 2).
Furthermore, there was an interaction between male comb size and offspring sex on body condition (body mass / tarsus length; N = 59, F(1,15.4) = 7.98, p < 0.013): daughters of small-comb fathers had a lower condition score than daughters of large-comb fathers (12.3 ± 0.19 vs. 14.1 ± 0.47; N = 27, F(1,11) = 13.82, p = 0.003), whereas there was no difference in sons (13.9 ± 0.78 vs 13.3 ± 0.47; N = 32, F(1,3.6) = 0.881, p = 0.406).
Figure 2 | Mean of chicks body mass (in gram ±SE.) of sons and daughters sired by large-comb roosters (dark bars) or small-comb roosters (white bars) 24 weeks after hatching
Offspring growth
There was no effect of male comb size, offspring sex or the interaction between male comb size and offspring sex on body mass at hatching (GLMM: F-values < 1.44, p-values > 0.706). Twenty-four weeks after hatching, sons had outgrown daughters (GLM, F(1,16) = 37.3, p < 0.001), and there was no direct effect of male comb size
(F(1,15.1) = 0.29, p = 0.599), but there was an interaction between male comb size and
offspring sex on body mass (F(1,15.1) = 5.29, p = 0.036). Sons of large-comb males (N = 13) did not differ from sons from small-comb males (N = 19; figure 2; GLMM;
F(1,2.4) = 6.66, P = 0.101), but daughters (N = 13) from large-comb males were heavier
than daughters (N = 14) of small-comb males (F(1,11) = 11.82, p = 0.006, Figure 2).
Furthermore, there was an interaction between male comb size and offspring sex on body condition (body mass / tarsus length; N = 59, F(1,15.4) = 7.98, p < 0.013): daughters of small-comb fathers had a lower condition score than daughters of large-comb fathers (12.3 ± 0.19 vs. 14.1 ± 0.47; N = 27, F(1,11) = 13.82, p = 0.003), whereas there was no difference in sons (13.9 ± 0.78 vs 13.3 ± 0.47; N = 32, F(1,3.6) = 0.881, p = 0.406).
Discussion
In this study, we assessed whether the comb size of red junglefowl roosters affected female reproductive investment and offspring development. Comb size is a proxy for male quality, since this feature is a proven indicator for male attractiveness, is related to social dominance, reduces immunocompetence (which only good- quality individuals may be able to bear the costs off) and is heritable (Collias and Collias, 1996; Johnsen and Zuk, 1995; Johnsen and Zuk, 1996; Parker, 2003; Parker and Ligon, 2003). Roosters in our population with large-comb (both in maximum length and size) were heavier but did not differ in comb colour from small-comb (low-quality) roosters. Based on the assumption that in general, positive DA is most common in birds (Harris and Uller, 2009; Ratikainen and Kokko, 2010) and that females of female care only species invest in egg size or mass when paired with attractive males (Horváthová et al., 2012), one may expect that females paired with large-comb males to produce heavier eggs. However, as suggested by additional literature (see introduction), we did not find this effect on egg mass. Neither did we find an effect on clutch size. However, we did find that hens paired with large-comb
2
males, initiated egg laying sooner. In general, early breeding is best in seasonal environments (Rowe 1994, Verhulst et al. 1995).
Our results are therefore in line with a) the parental quality hypothesis where the earliest breeders (Rowe, 1994) may be those of highest quality (Low et al., 2015;
Verhulst and Nilsson, 2008; Wardrop and Ydenberg, 2003), and b) with the positive DA hypothesis where the attractiveness of the partner stimulates reproductive investment (Cunningham and Russell, 2000; Harris and Uller, 2009). Disregarding possible differences in behaviour between the males with different comb sizes (Pizzari 2003, Wilson et al. 2008), there were no apparent phenotypic differences between the females in the two treatment groups, and they did not invest differentially in egg mass or clutch size. Both Wilson et al. (2008) and Pizzari (2003) did not find strong positive correlations between behaviours that predict male reproductive success and comb size, and we may, therefore, assume that these behaviours were randomly divided over our treatments, possibly having confounding effects on our results, but nevertheless the perceived male quality (comb size) must, therefore, have been the main factor affecting clutch initiation time.
Furthermore, hens mated with large-comb males produced heavier daughters with a higher condition score, but no such effects were found in sons. The absence of effects in sons was unexpected since Parker (2003) already found a positive effect of paternal comb size on offspring condition. Parker (2003) also showed that paternal genetics may underlie differences in development but could not rule out non- genetic paternal effects. Parker (2003) suggested that such non-genetic paternal effects could be transmitted via the chemical compounds in the roosters’ ejaculates.
A novel idea that has not received much attention.
The ejaculate of cocks is known to contain several steroid hormones (Anderson and Navara 2011). Hens have steroid receptors in the oviduct (Takeda et al. 1990, Yoshimura et al. 2000, Walters et al. 2010, Chang et al. 2013) and also deposit androgens in their eggs (Riedstra et al. 2013). We recently showed that testosterone concentrations in the ejaculate differ between large- and small-comb males (Lelono et al. 2019). This opens the possibility that the earlier initiation of egg laying may have been caused by differences in ejaculate composition between large- and small-comb males, which open for further testing. It is also conceivable that the testosterone in the ejaculate enters the egg during fertilization, affecting chick growth, although this then should have a sex-specific effect which is not unusual for these maternal hormones (Groothuis et al., 2019).
Contrary to Parker (2003) our set up was not designed to separate maternal responses from paternal effects. Therefore, maternal effects such as sex-specific yolk testosterone deposition may explain in our results, since a) this steroid is known to enhance early growth (Groothuis et al. 2005b), b) differential hormone deposition in relation to paternal quality is present in avian species (Gil et al. 1999, 2004, Rutstein et al. 2005, von Engelhardt et al. 2006, Williamson et al. 2006, Garcia-Fernandez et al. 2010, Loyau and Lacroix 2010), although some other studies did not find this (reviewed in Kingma et al. 2009, Cucco et al. 2011, Horvathova et al. 2012), and c) there is evidence that avian mothers can deposit testosterone in the yolk depending on the sex of the embryo (Muller et al. 2002, Badyaev et al. 2005, Rutstein et al.
2005, Gilbert et al. 2005, Pariser et al. 2012). From this, the intriguing hypothesis emerges that maternal (sex-specific) hormone deposition may be mediated by a cryptic paternal trait, namely the ejaculate hormone composition of fathers (Lelono et al., 2019a). Heavier daughters sired by large-comb males support the positive DA hypothesis. Nevertheless, based on the current results we did not find evidence in the development of sons that supports either the positive or negative DA hypothesis.
In conclusion: In line with previous studies, we found that paternal attractiveness affects female reproductive investment and offspring growth in red junglefowl.
Whether this effect is caused purely by visual perception of the male comb by females or perhaps via mechanisms involving the chemical substances in the ejaculates of males, as suggested as a possible path way by Parker (2003), is currently under investigation. However, although our results support the positive DA hypothesis, because females mated with large-comb males produce eggs earlier than females mated with small-comb males, we found no evidence that in the red junglefowl, females invest more in egg mass as was found by Horváthová et al. (2012). These results together with the studies presented in the introduction, therefore, weaken the generalized conclusion that females in female care only species invest more in egg mass when mated with attractive/good-quality males.
2
Ethical note
The procedures followed the relevant guidelines and regulations of the animal welfare committee of the University of Groningen and were approved by the committee under DEC license 6710A, 2013.
All handling and treatment of animals were carried out by experienced scientist with a licence (Certificate number 685412, DG VGZ/VVP (Stcrt.135), 25 January 2013), and animal caretakers, to perform animal experiments. The welfare of all birds was assessed on a daily basis.
Funding
This work is supported through a Direktorat Pendidikan Tinggi (DIKTI) Scholarship, The Ministry of Research, Technology and Higher Education, The Republic of Indonesia and the University of Groningen The Netherlands to A.L. and T.G.G.
Author contributions
A.L., B.R., and T.G.G. designed the experiment. A.L. performed the experiments. A.L., B.R., and T.G.G. analysed the data. A.L. wrote the first draft of the manuscripts and B.R., and T.G.G. wrote with A.L. the final version. All authors read and approved the final manuscript.
Acknowledgments
Special thanks to Saskia Helder for her assistance in animal caretaking.
“Semar”
