How Sexual Conflict and Reproduction Drive the Sex Dependent Evolution of Life History Traits

How sexual conflict and reproduction drive the sex dependent evolution of life history traits

Teagan Joy Gale

A thesis in fulfilment of the requirements for the degree of Doctor of Philosophy

School of Biological, Earth and Environmental Sciences Evolution and Ecology Research Centre

UNSW

August 2018 i

THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet Surname or Family name: Gale First name: Teagan Other name/s: Joy Abbreviation for degree as given in the University calendar: PhD School: School of Biological, Earth and Environmental Sciences Faculty: Faculty of Science Title: How sexual conflict and reproduction drive the sex dependent evolution of life history traits Abstract: Sexual conflict is inevitable in all sexually reproducing species as males and females are selected to maximise their own fitness often at the expense of their mate. This conflict can intensify the already costly nature of reproduction for female mammals. To examine the effects of territory turnover of a dominant male to pregnant females and their resulting offspring, we exposed females to a novel male in late pregnancy. These females spent less time nursing their pups, who subsequently had a low weaning weight and suffered oxidative costs at adulthood. We believe the change in behaviour is a result of the females strategically altering their investment in relation to their perceived chances of offspring’s survival. To partition the costs of male presence and mating from those of reproduction we housed females with males of differing gonadal status and allowed half access to a refuge. We found no costs of male mating, presence or insemination, suggesting the costs associated with these behaviours may be subtler than currently predicted. However, females housed with castrated males exhibited more refuging behaviour and higher stress levels. To examine the cause of this response, we again housed females with males of differing gonadal status and supplemented each cage with the scent of an unfamiliar dominant, subordinate or castrated male. While we replicated the effect of castrated males on female stress levels, we found that this can be ameliorated after exposure to the scent of an intact male, either dominant or subordinate. This shows that the signal(s) eliciting this response in females can be transferred in soiled bedding, but that it is not due to subordination. Lastly, we investigated genetic sexual conflict to attempt to provide further evidence for Haig’s kinship theory of genomic imprinting. We used Igf2 knockout mice to explore whether paternally expressed genes in offspring reduce maternal reproductive success. We mated females to male Igf2 KO homozygotes and found that females did show reduced metabolic costs. This shows foetal/placental Igf2 upregulates maternal metabolism, imposing energetic demands on the mother in the manner predicted by Haig’s kinship hypothesis.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968.

I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research. FOR OFFICE USE ONLY Date of completion of requirements for Award:

Originality Statement

‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’

iii Authenticity Statement

‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’

‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

iv Date 16/02/2018 Summary of Collaborators’ Contributions

In all chapters and manuscripts submitted, and in preparation in which I was listed as first author, I was the primary investigator. This included being the main researcher in data collection, analysis and interpretation, and preparation of the manuscript. Professor Robert Brooks was my PhD supervisor and Dr Michael Garratt a close collaborator, who both also contributed to the design, analysis, interpretation and editing of the manuscripts. Dr Neil Youngson contributed to the designing of the Chapter 5 and provided training on dissection techniques. Professor Michael Constancia and Dr Ionel Sandovici provided genetic strains of mice necessary to conduct some of this research.

v Acknowledgements

In the words of Richard Feynman “Physics is like sex: sure, it may give some practical results, but that's not why we do it.” Which I feel is true for science in general. I am first and foremost grateful to have spent four years in a labour of love pursuing knowledge and chasing my curiosities. Although after this dedicated period of scholarly pursuit I have found that while I learned a great deal throughout this project, more so, I have learned just how much I don’t know. I hope that by reading this thesis I can impart the tiny fraction of knowledge I have contributed and simultaneously highlight the much larger fraction there is still to know in this field.

While I considered myself lucky to end up with Rob Brooks as my supervisor in my honours year, there was no luck involved when it came to my PhD. I knew I was choosing to work under an intellectual giant who is also somehow completely down to Earth. Rob, I would like to express my sincere gratitude for everything you have done for me during my honours and PhD and for giving me so many wonderful opportunities. Your continued support, patience and immense knowledge has been integral to the success of this PhD. I also remain indebted for your belief in me, which was the only thing that gave me the confidence to undertake this journey in the first place and continued to mitigate my imposter syndrome up until the very end.

My other honours supervisor, Mike Garratt, whom I also continued to work with throughout my PhD was also an essential source of assistance and guidance that I would not have been able to complete this project without. I am hugely appreciative for all your help from across the other side of the world. I would also like to take this opportunity to thank Neil Youngson, who not only responded to an email from a random PhD student from another faculty in need, but subsequently stepped in to help me with my genetics experiment without hesitation.

In my daily work I have been incredibly lucky to be surrounded with the ever supporting and banterful members of the Sex lab and EERC. Thank you Amany, vi

Hamish, Francesca, Dax, Jess and Justin, who helped keep my spirits high and made coming into lab everyday more enjoyable. A special thanks to Heather Try who not only helped me with my mouse work, but literally any question or request I had day or night. A massive thank you also must go to another fellow PhD student, and my now BFF, Emma Asbridge. Thank you for being one of my biggest cheer leaders, whether it was listening to conference talks numerous times or driving me to the start of my half marathon at 4:30am you were always there for me with untiring enthusiasm. You made the PhD highs higher and the lows shallower and I am truly grateful.

I also need to wholeheartedly thank my boyfriend, Laurie, who has supported me and been very understanding of the enormous time investment required for this project. Our numerous adventures provided me with a much needed escape and helped me to never let me lose perspective of the big picture. I have never been so happy. Thank you.

To my family, thank you for your omnipresent love, patience and support throughout this PhD and my life in general. Mum and Dad you guys are so amazing to just go along with every crazy pursuit I undertake and stand by my sometimes- erratic life decisions. Knowing you are there for me whenever I need you is a constant source of strength and comfort, I am forever grateful for everything you do for me. To my sister, Peita, who looks after me to the point she calls me her ‘dependent’, we have been through so much together and you are the one who have endured the entirety of this journey with me, consistently by my side. Words can’t even begin to articulate how grateful I am to you, but ultimately please just know you were invaluable in helping me see this project through to the end.

vii

Publications and Presentations

Publications

Gale, T. G., and Brooks, R. C. (in prep). Conflict between the sexes over pregnancy in mammals.

Gale, T. G., Garratt, M. and Brooks, R. C. (Submitted). Perceived threats of infanticide reduce maternal allocation during lactation and lead to elevated oxidative damage in offspring.

Gale, T. G., Garratt, M. and Brooks, R. C. (Submitted). No evidence that male mice impose mating or insemination costs on females, but females avoid castrated males.

Gale, T. G., Garratt, M. and Brooks, R. C. (in prep). Stress effects of castrated males on females is olfactory modulated and can be ameliorated by the scent of an intact male

Gale, T. G., Garratt, M., Constancia M., Sandovici I., Youngson, N., and Brooks, R. C. (in prep). Igf2 KO pups reduce metabolic costs to pregnant female mice.

Presentations

Gale, T. G. (2015, August). Perceived threats of infanticide reduce maternal allocation during lactation and lead to elevated oxidative damage in offspring. Paper presented at the Behaviour conference, Cairns, QLD, Australia.

Gale, T. G. (2016, August). No evidence that male mice impose mating or insemination costs on females, but females avoid castrated males. Paper presented at the International Society of Behavioural Ecology, Exeter, UK.

viii

Abstract

It has long been known that reproduction is costly and is of fundamental importance in life history evolution; however, research has only relatively recently begun to explore how sexual conflict influences sex specific life-histories and their individual costs during reproduction. While the field of sexual conflict abounds in theory in mammals, evidence still remains highly skewed towards invertebrates. Sexual conflict in mammals can be manifested in a variety of behavioural, anatomical and physiological traits linked to reproduction, such as mate choice and frequency, timing of fertilisation, cryptic choice, infanticide, genomic imprinting, number of offspring and parental investment (reviewed in Chapter 1). All of which can have varying consequences for the individuals, their mates and their future offspring. The aim of this PhD was to use an empirical approach to examine how conflict and reproduction drive sex-dependent evolution and attempt to quantify some of the costs and consequences of these in a mammalian organism (Mus musculus). We used a variety of strains of Mus musculus depending on which suited the focus and methodology of the experiment most appropriately.

In Chapter 2, we examined the life-history effects of territory turnover of a dominant male to pregnant females (post-implantation) and their resulting offspring. Females who were exposed to a novel male in late pregnancy were found to spend less time nursing their pups, which resulted in the pups having a low weaning weight. Following weaning these pups exhibited catch-up growth and there were no weight discrepancies at adulthood. While we found no effects to the offspring’s reproductive fitness, they were found to suffer increased oxidative challenges. As male mice are highly infanticidal, the change in maternal behaviour during lactation we believe to be a result of the females strategically altering their investment in relation to their perceived chances of offspring’s survival.

While in Chapter 3 we set out to partition the costs of male presence and mating from those of reproduction by housing females with males of differing gonadal status and allowing half of each treatment access to a refuge. We found no costs of male mating, presence or insemination, suggesting the costs associated with these

ix behaviours may be subtler than currently predicted. However, females housed with castrated males exhibited more refuging behaviour and had higher faecal glucocorticoid metabolites, indicating a stress response. While the cause of this response was unknown, we speculated that it may have been driven by male quality and subsequently examined this prediction in Chapter 4. In order to test whether the effects on females of being housed with castrated males are due to the absence of intact male scent, or the absence of the scent of a dominant male, we conducted a factorial experiment. Females were housed with either a vasectomised or castrated male or another female and each cage was supplemented with scent of an unfamiliar dominant, subordinate or castrated male. While I replicated the effect of castrated males on female stress levels, I found that this can be ameliorated after prolonged exposure (21 days) to the scent of an intact male, either dominant or subordinate. This study did confirm that the signal(s) eliciting this response in females can be transferred in soiled bedding, but as subordination did not mimic the effect this indicates that something other than male status is driving this response.

In Chapter 5 we investigated genetic sexual conflict between the sexes to attempt to provide further evidence for Haig’s kinship theory of genomic imprinting. Haig’s kinship theory hypothesises that genomic imprinting, a phenomenon where either the maternally- or the paternally-inherited allele is silenced, originated from this conflict as paternal genes act as foetal growth promoters and maternal genes as foetal growth inhibitors. The most well-known imprinted gene that behaves as predicted by the kinship hypothesis is Igf2, which produces a hormone (IGF2) that promotes gestational nutrient demand and thus offspring growth in mammals. We used Igf2 knockout mice to explore whether paternally expressed genes in offspring reduce maternal reproductive success, and maternally imprinted genes increase maternal reproductive success, as predicted by Haig's theory. To do this we mated females to male Igf2 KO homozygotes and predicted that they would pay smaller costs of reproduction (than females mated to wild-type controls) due to lower placental and foetal (Igf2 KO) demand for nutrients. We found that females did show reduced metabolic costs, however, we did not detect any differences in body composition. Our results that foetal/placental Igf2

x upregulates maternal metabolism, imposing energetic demands on the mother in the manner predicted by Haig’s kinship hypothesis.

This thesis shows that sexually antagonistic adaptations arising from sexual conflict between the sexes are intricate and varied, both behaviourally and genetically, while the costs of reproduction are much subtler than has been predicted. Disentangling the cause of these results and understanding how males and females influence each other’s life histories, will, in turn, also help us better understand mate choice strategies and the physiological costs of reproduction as well as their impact on life- history evolution.

Table of Contents

Thesis/Dissertation Sheet ...... ii Originality Statement ...... iii Authenticity Statement ...... iv Copyright Statement ...... iv Summary of Collaborators’ Contributions ...... v Acknowledgements ...... vi Publications and Presentations ...... viii Abstract ...... ix Chapter 1 ...... 1 Conflict between the sexes over pregnancy in mammals ...... 1 Sexual Conflict ...... 2 Sexual conflict in mammals ...... 6 Shared interests: Do we have a future? Will I know my children? Are we related? ...... 7 Kinship ...... 9 Mating and re-mating ...... 12 Mating ...... 12 Remating ...... 13 Fertilisation ...... 16 Implantation ...... 20 Infanticide Prevention ...... 20 Spontaneous abortion ...... 24 Gestation ...... 26 Placental Development ...... 26 Genomic Imprinting ...... 30 Pregnancy sickness, preeclampsia and gestational diabetes ...... 33 Pregnancy sickness ...... 33 Pre-eclampsia ...... 34 Gestational Diabetes ...... 36 Conclusion ...... 37 Chapter 2 ...... 39 Perceived threats of infanticide reduce maternal allocation during lactation and lead to elevated oxidative damage in offspring ...... 39 xii

Abstract ...... 39 Introduction ...... 40 Materials and methods...... 43 Animal Housing ...... 43 Experimental protocol ...... 44 Maternal Investment and offspring weights ...... 45 Maternal behaviour ...... 46 Glucocorticoid metabolites ...... 46 Oxidative Stress...... 47 Offspring reproductive fitness ...... 48 Scent-marking rates ...... 48 Major Urinary Protein Concentration ...... 48 Statistics ...... 49 Results ...... 49 Offspring weight ...... 49 Maternal behaviour ...... 52 Glucocorticoid metabolites in mothers ...... 54 Oxidative stress in offspring ...... 55 Discussion ...... 60 Chapter 3 ...... 65 No evidence that male mice impose mating or insemination costs on females, but females avoid castrated males ...... 65 Abstract ...... 65 Introduction ...... 66 Material and methods ...... 69 Animal Housing ...... 69 Experimental protocol ...... 70 Female refuge use ...... 71 Oxidative Stress...... 72 Glucocorticoid metabolites ...... 73 Statistics ...... 73 Results ...... 74 Female and offspring weight ...... 74 Radio Frequency Identification (RFID) of female movement ...... 75

xiii

Female movement ...... 76 Faecal Glucocorticoid Metabolites (FGM) ...... 78 Oxidative stress ...... 79 Discussion ...... 80 Chapter 4 ...... 86 Stress effects of castrated males on females is olfactory modulated and can be ameliorated by the scent of an intact male ...... 86 Abstract ...... 86 Introduction ...... 87 Material and methods ...... 90 Experimental design ...... 90 Animal Housing ...... 91 Experimental protocol ...... 92 Dominant/ subordinate odour donor trials ...... 92 Scent-marking trials ...... 92 Testosterone Levels ...... 93 Glucocorticoid metabolites ...... 93 Female behaviour ...... 94 Statistics ...... 95 Results ...... 95 Establishment of dominant/ subordinate hierarchy in male odour donors ...... 95 Faecal glucocorticoid metabolites in females ...... 96 Female Behaviour ...... 97 Castrated male treatment ...... 100 Vasectomised treatment ...... 101 Female treatment ...... 101 Discussion ...... 101 Chapter 5 ...... 106 Paternally expressed Igf2 KO pups reduce metabolic costs to pregnant female mice ...... 106 Abstract ...... 106 Introduction ...... 107 Materials and Methods ...... 109 Mice generation and breeding ...... 110 Genotyping ...... 110

xiv

Animal Housing ...... 111 Experimental Protocol ...... 111 Whole body metabolism ...... 113 Body Composition ...... 114 Statistics ...... 114 Results ...... 114 Physiological consequences of Igf2 during gestation day 18 ...... 114 Whole body metabolism (CLAMS) ...... 115 Body Composition (echoMRI) ...... 117 Discussion ...... 119 Chapter 6 ...... 124 Conclusions ...... 124 The costs of reproduction ...... 124 The costs of catch up growth ...... 127 Odour signals/ cues ...... 128 On the origin of genomic imprinting ...... 129 References ...... 131 Appendices ...... 182

Chapter 1

Conflict between the sexes over pregnancy in mammals

“Pregnancy is absolutely central to reproduction, and yet pregnancy doesn’t seem to work very well.” (David Haig in Zimmer 2006)

Worldwide, approximately 800 women die from pregnancy or childbirth-related complications daily (WHO and UNICEF 2012). There are various complications that may occur during and following pregnancy and childbirth which remain common despite a wealth of research. For example, pre-eclampsia and gestational diabetes each affect 3-5% of pregnancies globally (Ben-Haroush et al. 2004; Ananth et al. 2013) with pre-eclampsia/eclampsia being one of the three leading causes of maternal morbidity and mortality (Ghulmiyyah and Sibai). Less severe complications such as pregnancy sickness carry a much higher incidence with the meta-analytic average rate at 69.4% (Einarson et al. 2013). These complications, among many others, highlight the paradox David Haig noted that despite being ‘central to reproduction… it doesn’t work very well’. The aim of this review is to highlight the role of sexual conflict during pregnancy in all mammals (including humans) and we argue that the complications of pregnancy, beyond the usual ambit of clinical ‘pregnancy complications’ is a result of evolutionary conflicts of interests between the parties involved including the mother, father and offspring. These conflicting interests are not confined to humans but are pervasive throughout mammalian taxa occurring at every stage throughout pregnancy.

First, we summarise the history of the literature and define sexual conflict. We then go through the factors that can affect the degree to which sexual conflict is realised, where we further propagate the theory that kinship is the main mechanism underpinning the realisation of sexual conflict. Finally, we outline how and why

1 sexual conflict occurs at each stage of pregnancy including fertilsation, implantation and gestation (see Figure 1). In gestation we narrow in on pregnancy sickness, pre- eclampsia and gestational diabetes where we provide the evidence for, and argue the aetiology of these complications, lies in sexual conflict theory.

Figure 1: Flow chart showing the chronological stages of pregnancy concurrent with the potential realisations of sexual conflict at each stage.

Sexual Conflict

Initially biologists assumed that sexual reproduction was a purely cooperative union where males and females mutually came together to perpetuate the species and maximise the number of surviving young (e.g., Lack 1947, 1954). The idea that there may actually be some conflict in this union was first discussed by Williams (1966) in the 60’s. While Williams is credited as the first to use the term ‘sexual conflict’, he attributed the idea to Fisher (1930). Fisher and Williams’ ideas concerning sexual conflict didn’t really garner attention until Trivers (1972), and then Parker (1979)

2 both identified that males and females can experience selection to increase their own fitness at the expense of their mate’s fitness creating conflict.

Defining sexual conflict has been the source of some controversy that has plagued the theory, however, with discrepancies in the use of the term ‘sexual conflict’, contributing to non-trivial misunderstandings about the concept (Pizzari and Snook 2003; Arnqvist 2004). Many researchers have contributed to the body of research attempting to clarify the definition and shed light on the biological components (e.g. Chapman et al. 2003; Arnqvist and Rowe 2005; Parker 2006; Tregenza et al. 2006). Here, we use the definition of sexual conflict as; 1) “when each sex has differing optimal fitness strategies concerning reproduction which cannot be simultaneously achieved,” and 2) “subsequently one sex evolves a trait that decreases the fitness of the other sex”. Sexual conflict has also historically been confused with sexual selection, and in literature that does acknowledge them as separate entities, it has often presented them alternative explanations for trait evolution. Sexual selection, however, is a mode of selection arising from intra-sexual competition over access to mates (Darwin 1859; Andersson 1994), with the main mechanisms for sexual selection being mate competition and mate choice. Therefore, while sexual selection can generate sexual conflict, conflict extends to traits beyond sexual selection. For a review of the relationship between sexual conflict and sexual selection see Kokko and Jennions (2014), which argues that sexual conflict and sexual selection should not be presented as alternative explanations for trait evolution and that it is possible for one “win” the sexually antagonistic co-evolutionary arms race with the prediction that population fitness would be higher if females win.

Since Trivers (1972) and Parker’s (1979) seminal papers there has been an exponential increase of research on sexual conflict showing its far reaching consequences, not only for reproductive behaviour and morphology, but also for life histories (Bonduriansky et al. 2008), demography (Holland and Rice 1999), and even as a promoter of allopatric divergence and speciation (Rice 1998; Arnqvist et al. 2000).

Reviews of the topic tend to distinguish two modes of sexual conflict (Arnqvist and Rowe 2005); inter-locus and intra-locus sexual conflict. Intra-locus sexual conflict arises because, for most loci, half of the time the gene will find itself in a female and half the time in a male. The phenotype optimal for the fitness of one sex may not be optimal for the other. Therefore, most forms of allelic expression prevent either sex from reaching optimal trait values, leading to sexually antagonistic selection (Bonduriansky and Chenoweth 2009). In humans, for example, it is advantageous for females to have a larger hip width to accommodate childbirth, but larger hips are a disadvantage to males as it results in reduced agility and mobility (Rice and Chippindale 2001; Price and Hosken 2007).

Inter-locus sexual conflict (the focus of this review) arises due to differences in the evolutionary interests between males and females concerning reproduction. While a female and male coming together to reproduce both have a shared interest in the success of their current reproduction, this interest does not always extend to include their mate’s fitness or future reproduction. Selection often acts to maximise the fitness of the individual, even if this comes as a detriment to their mate (Arnqvist and Rowe 2005; Parker 2006). A large, experimentally rigorous body of research on invertebrates has demonstrated the many ways in which inter-locus sexual conflict can manifest, including the evolution of female-damaging penile spines (Crudgington and Siva-Jothy 2000), traumatic insemination (Carayon 1966), sexual cannibalism (Arnqvist and Rowe 2013), intimidating courtship (Chang and Jablonski 2010), and female life-span shortening seminal fluids (Chapman et al. 2003).

Distinguishing intra- and inter-locus sexual conflict provides a useful dichotomy only up to a point. This has been highlighted by the discovery of genomic imprinting, which is the expression of genes in a parent-of-origin specific manner. Typically, diploid organisms receive one set of genes from each parent which are both equally likely to be expressed. However, a small subset of genes subvert this biallelic expression and instead show monoallelic expression according to the parental origin of the allele. Genomic imprinting presents a non-Mendelian pattern of expression that ameliorates the intra-locus sexual conflict over sex specific 4 phenotypes. But true sex-specificity can arise according to models of genomic imprinting evolutionary origins (Haig 1993a, 2000; Day and Bonduriansky 2004) from conflicting interests of mates. Further support for inter-locus and intra-locus not being mutually exclusive comes from research showing inter-locus conflict can be mediated by alleles which are at the same locus (intra-locus) in males and females (Tregenza et al. 2006).

Inter-locus sexual conflict can result in sexually antagonistic co-evolution, where adaptations have evolved that benefit one sex but harm the other. In this way the confounding males and females can enter into a co-evolutionary arms race of offence and defence (Holland and Rice 1998; Arnqvist and Rowe 2005; Parker 2006). For example in the seed beetle (Callosobruchus maculatus) males evolved spiny genitalia to improve stability during mating and thereby greater reproductive success (Edvardsson and Tregenza 2005). In response to this, females evolved tougher copulatory ducts to resist the spines, and subsequently males evolved more spines (Rönn et al. 2007), a classic example of the progression of a sexually antagonistic co-evolutionary arms race.

While sexual conflict over mating results in females having a lower fitness than they would in the absence of such conflict (Arnqvist and Rowe 2005), this does not mean males are without any costs of mating. Reproduction is still energetically costly to males who invest in mate-searching, courtship, copulation, male–male combat and spermatogenesis, all of which have life history altering consequences and ultimately shortens lifespan (Cordts and Partridge 1996; Prowse and Partridge 1997; Gaskin et al. 2002; Hunt et al. 2004).

Although we focus on sexual conflict during gestation occurring between maternal and paternal genomes within individual embryos (and placentas), conflict can also occur between mothers and embryos as well as sibling embryos in the womb. Sexual reproduction and pregnancy are vulnerable to sexually antagonistic selection at several stages including; mate choice, courtship behaviours, gametogenesis, copulation, implantation, parturition, and parental care (Arnqvist and Rowe 2005).

Sexual conflict in mammals Eutherian mammals are predicted to be particularly vulnerable to sexually antagonistic selection due to the exaggerated asymmetry in mating optima and parental investment between the sexes, largely due to the costs of pregnancy and lactation. The quantitative genetic and experimental manipulations might not be as extensive as in invertebrates, but sexual conflict is definitely a notable force shaping mammalian physiology, behaviour and life histories. For a recent review of sexual conflict in mammals and the consequences for mating systems and life history see Aloise King et al. (2013), which highlights the incongruity that while theoretically there is a large potential for sexual conflict in mammalian mating systems, there a lack of empirical evidence of sexual conflict in mammalian species and the use of sexual conflict theory in models of mammalian life history theory.

Eutherian mammal mothers are well known for their prolonged placental gestation and the large investment of time, energy and other resources into pregnancy and subsequent lactation. The metabolic costs to female mammals are highlighted in species such as the seal (Arctocephalus gazella) where mothers transfer up to 16–18 % of the maternal body protein via milk (Oftedal 2000) and lose more than 50% of their stored body fat by the end of lactation (Bowen et al. 1992). This investment is sharply contrasted to the low levels of obligate investment required by mammalian males. For mothers the economic concept of diminishing returns is applicable where investment in current offspring procures benefits diminishing until a point at which it becomes beneficial for a mother to limit her investment and allocate the resources to somatic maintenance and future reproduction. Pregnancy also marks a time when viviparous female mammals are obligated to solely provide their offspring with essential investment of resources and therefore consequently making pregnancy a prime time for males to exploit female investment to their favour.

Almost all mammals are viviparous, meaning that development of the embryo occurs inside the body of the female who give birth to live young that require varying amounts of help from the parent(s)(Blackburn 2000). The exception to this is the monotremes, which are a group of mammals that lay eggs, however, the only extant monotremes include the platypus (Ornitorhyncus anatinus) and four species 6 of echidnas (Tachyglossidae) (Griffiths 2012). Viviparity creates a post-fertilisation arena for conflict that is absent in oviparous species and provide a conduit for male manipulation of female investment (Zeh and Zeh 1996, 1997). According to the viviparity-driven conflict hypothesis, divergent interests of mother, father and offspring lead to a rapid antagonistic co-evolution (Zeh and Zeh 2000). Internal fertilsation also creates paternity uncertainty, discussed in the Shared Interests section.

Humans, as mammals are no exception to the phenomenon of sexual conflict. The potential for sexual conflict is as pervasive in humans as in many plant and animal species (Chapman et al. 2003). It is surprising, however, that such an integral biological process central to reproduction continues to carry an array of complications that are still prevalent after many years of human evolution. It is theorised that the hazards that arise and continue to persist are a consequence of conflict, as like every other species that reproduces sexually, human pregnancy also involves the combining of two sets of genes with interests that are not exactly aligned (Haig 1993a).

Shared interests: Do we have a future? Will I know my children? Are we related?

Sexual conflict between mates arises because each individual has less interest in their mates’ future reproduction than their current. This conflict is modified by probability of mating again, paternity uncertainty and relatedness.

Probability of mating- monogamy to promiscuity Mating systems are never optimal for either sex (Lessells 2006; Bonduriansky et al. 2008), but the degree of conflict covaries with the mating system and population structure. The more likely that a breeding pair (or even a pair of potential mates) will reproduce together again in the future; the more aligned their interests and the weaker sexual conflict will likely be. If pairs breed together repeatedly, if one individual damages their partner they also subsequently damage their own long- term fitness (Lessells and Parker 1999). Under monogamous, or even largely 7 monogamous mating arrangements, both sexes will be selected to refrain from harming the other, because such harm diminishes not only their mate’s fitness, but also their own (Wigby and Chapman 2004). In contrast, sexual conflict is heightened in multi-male and multi-female mating arrangements, where mates’ interests are vastly different and where male-male competition can accelerate the evolution of adaptations that may be harmful to females and vice versa. When there is a high level of competition and males are trying to outcompete each other for fertilisations, there is also a higher chance male harassment, coercion and females are more likely to be inadvertently injured as collateral damage (e.g. Morrow et al. 2003; Parker 2006). This can vary not only with mating system but with the operational sex-ratio where female-biased mortality resulting from male mating behaviour is more common if there is a male-biased sex ratio, such has been found feral sheep (Ovis aries) where female mortality peaks in the main rutting peak (Réale et al. 1996).

This is especially pertinent to mammals, most of which are predominantly polygynous (Krebs and Davies 1993; Birkhead 2000; Eberle and Kappeler 2004) with approximately 90% adopting this mating system (Clutton-Brock 1989). A review by Wolff and Macdonald (2004), suggests that promiscuous mating in mammals is more common that currently accepted. Such uncertainty surrounding mammalian mating systems has prompted a need for a more thorough genetics- based investigation (McEachern et al. 2009). The benefits of multi-male mating to female mammals is also well supported which adds support to the idea that mammals may be more promiscuous than currently recognised. For example, saddle-backed tamarins (Saguinus fuscicollis) will mate with multiple males who will each help to provide parental care for her offspring (Terborgh and Goldizen 1985). Multiple male mating has also been proposed as a counter strategy for infanticide in various mammals such as rodents (e.g Labov et al. 1985), primates (e.g. O'Connell and Cowlishaw 1994; Agoramoorthy and Rudran 1995) and warthogs, Phacochoerus aethiopicus, (Somers et al. 1995). Therefore, conflict over mating is endemic to mammals and is expected to shape the evolution of female mammalian reproductive physiology and behaviour.

Paternity certainty Internal fertilsation coupled with gestation generally results in a lengthy time lag between the mating event and parturition meaning males cannot always be certain the mating event resulted in fertilisation or if any given offspring are his own. Some mammals also have concealed ovulation, where there are no visible signs of ovulation and males are unable to determine whether females are in oestrus or not. This is believed to have evolved as a female strategy to create further paternity uncertainty (Schoröder 1993; Hestermann et al. 2001). While females may benefit from paternity uncertainty by reducing aggression toward (and promoting protection of) infants, decreasing chances of infanticide and inciting sperm competition (Hrdy 1979; Van Schaik and Janson 2000b; Swedell and Saunders 2006), clearly uncertainty regarding offspring relatedness disadvantages the males. Paternity uncertainty in mammals can exacerbate conflict over pregnancy and essential anatomical structures during gestation that are not purely within the mothers control, such as the (mainly paternally derived) placenta, can make pregnancy even more risky for females. Paternity uncertainty has also been suggested to responsible for many phenomena in humans, such as sexual coercion, commitment scepticism and sexual overperception (Goetz and Shackelford 2009).

Trivers (1972) was first to suggest that males reduce parental investment for putative offspring as unlike females who have maternal certainty, males cannot as easily be 100% certain of genetic relatedness. This is supported by levels of parental investment seen in mammals where paternity uncertainty is also thought to be one of the main causes for the lack of parental care seen in majority of mammals (95-97%) where males provide little or no investment in their offspring (Platek and Shackelford 2006).

Kinship

Paternity uncertainty also has implications for genetic relatedness between parents and offspring, as well as between those offspring and the future offspring that a mother may yet bear in subsequent reproductive events. These kinship dynamics have been suggested to be the basis of the origins of genomic

9 imprinting. The kinship theory suggests that genomic imprinting evolves as a result of the misaligned interests of each parent in terms of the evolutionary fitness of their genes , as well as other types of interactions with asymmetric kin (Haig 1997, 2000). Both the maternally and paternally-derived alleles at any given locus share an interest in the reproductive success of the individual they reside in. Conflict can arise, however, when maternally and paternally-derived alleles are not equally likely to be shared by future offspring (Haig 1997, 2000). In short, paternally-derived genes are less likely than maternally-derived genes to be present in the offspring of a mother’s subsequent offspring, in a manner proportion to the probability that the male concerned also sires the subsequent offspring. According to Haig’s (1997) kinship theory, these asymmetric interests lead to the evolution of parent-of-origin genomic imprinting. Although genomic imprinting is rare, with only around one hundred genes in humans and mice constrained to maternally or paternally inherited allele (Barlow and Bartolomei 2014), imprinted genes can have significant effects which are largely seen during pregnancy, which we cover later in this review under the section Genomic Imprinting.

Relatedness Ultimately, sexual conflict arises from the fact that the two mating partners are genetically different (Lessells and Parker 1999). As such, high kinship due to relatedness between mates can help resolve sexual conflict by reducing the incentive to inflict harm. The net benefit males gain from harming females to gain extra matings, and higher offspring output are offset by the inclusive fitness costs of the conflict. Harmful adaptations inflicted by the male on the female, will not evolve if the benefit of the harm is less than the net benefit to the male (Rankin et al. 2011). This has recently been seen in the seed beetle (Callosobruchus maculatus) where males harmed females less when they were competing with familiar relatives (Lymbery and Simmons 2017), which suggests these males were plastically altering their harmfulness to maintain the balance of the net benefit of the harm in relation to the inclusive fitness costs of causing the harm. Higher relatedness among same sex members should also reduce competition for matings and hence lessen the benefit of conflict in the population (Cant and 10

Johnstone 2008). Therefore, it has been suggested that the realisation of sexual conflict is a function of both the probability of mates reproducing together in the future, and their level of relatedness (Bourke 2009).

While relatedness may align mates interests, inbreeding depression has been shown to be a widespread phenomenon in both experimental and wild populations (Charlesworth and Charlesworth 1999; Keller and Waller 2002) and inbreeding avoidance a common behavioural strategy for many animals (Szulkin et al. 2009; Brouwer et al. 2011; Lemaître et al. 2012). Despite the wealth of studies supporting the assumption that inbreeding depression drives selection for inbreeding avoidance, there have also been many studies that have failed to detect inbreeding avoidance in birds (Keller and Arcese 1998; Hansson et al. 2006; Holand et al. 2007; Szulkin et al. 2009), and even a few mammals including reindeer (Rangifer tarandus) (Holand et al. 2007), and bighorn sheep (Ovis canadensis) (Rioux-Paquette et al. 2010). Moreover, one study even reported active inbreeding in cichlid fish (Pelvicachromis taeniatus) (Thünken et al. 2007). Recently there has been an increased interest as to whether the inclusive benefits of inbreeding outweigh the costs of inbreeding depression and research is beginning to focus on when and if animals should or should not avoid it. Kokko and Ots (2006), developed a model that predicted that inbreeding tolerance should be common, and avoidance should only evolve if the costs associated with inbreeding were considerable. Inbreeding tolerance should also be higher in polygynous species than in monogamous species (Waser et al. 1986) if relatedness between mates increases the parent’s inclusive fitness (Bengtsson 1978). Which makes it is especially pertinent to mammals of which approximately 90% are polygynous (Clutton-Brock 1989). Some research by Pizzari et al. (2004) on red jungle fowl (Gallus gallus) found that while males initiated copulation faster with unrelated females, they were equally likely to copulate with related females, but the unrelated females retained more sperm.

Relatedness alone will not resolve sexual conflict. There are invertebrate species such as the spider (Stegodyphus lineatus) which displays high levels of inbreeding and sexual conflict (Maklakov and Lubin 2004; Bilde et al. 2005; 11

Maklakov et al. 2006; Maklakov and Lubin 2006). This is suggested to occur due to the benefits of relatedness in a population becoming offset by competition between relatives (West et al. 2002). Currently it is not clear whether animals should tolerate, avoid or actively inbreed and more research is needed on the evolution of inbreeding under sexual conflict (Szulkin et al. 2013).

Mating and re-mating

Sexual conflicts over mating, prevalent among many animal taxa including mammals, are relevant to pregnancy as they frequently can shape female reproductive behaviour and physiology. While our focus is on pregnancy and hence post mating conflict, for comprehensive reviews of sexual conflict over mating see Parker (2006) or Chapman et al. (2003).

Mating Conflict over mating occurs when it is advantageous for one sex to mate, but not the other. Differences may occur over not just whether to mate but also when and how frequently (Van Schaik et al. 2004; Arnqvist and Rowe 2005). Generally it is males who are the in the position in which mating is more favourable (unless majority parental investment is provided by the male) (Parker 2006). This asymmetry generally manifests as male-male competition over access to mates, resulting in males investing in sexually selected traits to attract more females, or evolving a range of antagonistic adaptations to overcome female resistance to mating, such as harmful reproductive morphology, male sexual coercion or sexual harassment (Smuts and Smuts 1993; Clutton-Brock and Parker 1995; Wells et al. 2014).

Males in various taxa inflict costs on females through their reproductive morphology and mating frequency, which can negatively correlate with female lifespan. As the co-evolutionary arms race of sexual conflict proceeds, females will respond with adaptations to mitigate damage inflicted or costs associated with these reproductive behaviours and damaging morphologies. These may play a role in the evolution of the female reproductive tract or fertilsation which is pertinent to this review. For example male waterfowl have a phallus that everts from their body in a 12 counter-clockwise direction, and as a response to protect themselves from male coercion and forceful intromission, females developed protective vaginal structures and their vaginal spirals are coiled in the opposite direction, so copulation requires female cooperation (Brennan et al. 2007; Brennan et al. 2009). While the costs of mating to females are widely studied, the costs of female resistance strategies draw much less attention. Resistance strategies may be costly to the females and (as with all antagonistically selected traits) are only expected to evolve if the benefits of resistance exceed the costs (Mühlhäuser and Blanckenhorn 2002).

Behaviours such as sexual coercion, harassment and intimidation of females by males may arise and persist in a species if it increases the males’ reproductive success, as predicted by the sexual coercion hypothesis. There is much empirical support for this hypothesis. For example in the Kasekela Chimpanzee (Pan troglodytes schweinfurthii) where long-term patterns of intimidation to females by dominant males has shown to increase their reproductive success (Feldblum et al. 2014). However, females of various species pay substantial costs from these behaviours, including increased energy expenditure (Watson et al. 1998b) and predation risk in water striders (Gerris odontogaster) (Arnqvist 1989), decreased foraging returns in the solitary bee (Amphiphora plumipes) (Stone et al. 1995), or reduced reproductive fitness as in the bruchid beetle (Callosobruchus maculatus) (Gay et al. 2009). In extreme cases these behaviours can be fatal for females. Such is the case for elephant seals (Mirounga angustirostris) where although rare, majority of observed females’ deaths were caused by traumatic injuries inflicted by males during mating attempts (Le Boeuf and Mesnick 1991). Male harassment has also been observed to be fatal in the feral sheep (Ovis aries) (Réale et al. 1996).

Remating While multi male mating or subsequent remating with another male after copulation may offer benefits for the female, this is clearly disadvantageous from the perspective of the male who could not only lose paternity but also may experience sperm competition from the females’ other mates. When females mate with multiple males, any given individual males’ reproductive success can also be governed by cryptic female choice. Thus, these mechanisms are key drivers of sexual 13 selection, which inevitably favour the males who most frequently do not lose paternity (Parker 1970; Simmons and Siva-Jothy 1998). This has resulted in males evolving various behavioural, morphological and physiological adaptations to either delay or reduce the chances of the female remating (Eberhard 1996; Simmons 2001) or increase their chances of paternity. Whether or not the adaptations directly damage the female, if remating is in the interest of the female and the male prevents this, then it will carry a cost to the female by constraining their mate choice and ability to select the best mate possible. However, females also frequently suffer injuries as a result of collateral damage from manipulative male traits (Chapman et al. 2003). It is also important to note that many researchers argue that the costs to females of manipulative male traits are balanced by the benefits of having sons that are equally good manipulators. For reviews on sexual conflict featuring this argument see Cameron et al. (2003), Pizzari and Snook (2003) or Cordero and Eberhard (2003).

Despite Bateman’s seminal and highly controversial demonstration that female reproductive success does not increase with number of mating partners as it does for males (Bateman 1948) and the costs associated with mating that have been quantified in various species, females do accept and often actively pursue multiple partners (e.g., grey mouse lemurs (Microcebus murinus) (Radespiel et al. 2002)). While the reproductive success of the female may not increase, there are many other benefits females may yield from promiscuous mating. These benefits may include fertility insurance, replenishment of sperm, improved offspring genetic diversity (through reduced inbreeding), nuptial gifts, parental investment (Thornhill and Alcock 1983; Stockley et al. 1993; Lewis and Austad 1994; Yasui 1998; Jennions and Petrie 2000; Zeh and Zeh 2006; Gowaty et al. 2010) or potentially higher quality offspring through sperm competition and selection in the female genital tract (Simmons 2005).

Some behavioural adaptations that males may use to increase paternity and prevent remating include fending off rivals or physically guarding their mate for a period after mating. While this can be costly for the male as it requires extended proximity to the female post copulation and may involve physical altercations with rival males, 14 it also allows the male to continue to mate with the female, increasing the amount of sperm transferred and prevent sperm competition that may arise from any consecutive mates. There are various methods of mate guarding throughout the animal kingdom ranging from violence and the emotion of jealousy in humans (Buss 2017), males riding on the back of females for a period of up to two days after mating in water striders (Arnqvist 1988) to post-copulatory vocalisations in Columbian ground squirrels (Manno et al. 2007). Mate guarding may be associated with various costs to females including energetic costs (Watson et al. 1998a), increased risk of predation (Arnqvist 1989; Rowe 1994; Cothran 2004), or decreased survival (Jormalainen et al. 2001; Wedell et al. 2006). To combat the costs of male harassment, coercion and guarding behaviour, females may develop counterstrategies consistent with sexual antagonistic co-evolution. For example, in the water strider (Aquarius paludum), female abdominal spines function to increase female control over copulation by helping them to reject any unwanted matings by harassing males (Ronkainen et al. 2005).

In mammals, it is sometimes seen that alternatively females may tolerate mate guarding if the benefits outweigh the costs, which may arise if harassment from other males is a threat to females (Amano and Hayashi 1998; Watson et al. 1998a). For example, female Sumatran orangutans (Pongo pygmaeus abelii) can greatly reduce harassment levels by maintaining spatial association with adult males (Fox 2002) and wild female horses have been found to form strong bonds with males who provide protection from harassment (Rubenstein 1986; Kaseda et al. 1995; Linklater et al. 1999; Sundaresan et al. 2007).

Another rare but interesting tactic male ungulates have been found to employ in sexual conflicts over remating is deception (Bro-Jørgensen 2011). Male Topi antelopes (Damaliscus lunatus) alarm snort when they detect a predator (Estes 1991) to warn conspecifics. This same behaviour is also occasionally exhibited in the absence of danger, when a receptive female attempts to leave a male’s territory (Bro-Jørgensen and Pangle 2010). In response to the alarm snort the receptive female moves back into the male’s territory in precaution, allowing the male to

15 either mate with or remate/ mate guard the female (Bro-Jørgensen and Pangle 2010).

Fertilisation

If the female does remate, sexual conflict then moves into the female reproductive tract as the sperm from rival males compete for fertilsation. This competition opens up the possibility of post copulatory sexual selection, comprising both sperm competition and also cryptic female choice. While sperm competition is a key driver of sperm morphology, sperm number per ejaculate (Parker 1970) and testes size (Birkhead and Møller 1998), it is not just the sperm that males have utilised to skew fertilsation rates in their favour. Various animals produce ejaculates that contain a suite of additional substances as well as sperm that act to enhance male’s fertilsation rate and manipulate the female. Copulatory plugs, for example, are formed by seminal substances that coagulate and act to physically block the opening to the female reproductive tract. These plugs have evolved independently in many taxa including mammals (Hartung and Dewsbury 1978; Dixson 1998) and can prevent or delay female re-mating (Parker 1970). Sperm coagulation has also been found to positively correlate with primates that exhibit multi-male mating (Anderson and Dixson 2002) which suggests that male-male competition has enhanced selection for seminal coagulation and copulatory plug formation.

Ejaculates may also contain seminal substances that induce female refractory behaviour (Takami et al. 2008), decrease female attractiveness (Orr and Rutowski 1991; Polak et al. 2001) or neutralise previous sperm (Birkhead and Pizzari 2002). The most prominent example of this comes from invertebrates such as the fruit fly Drosophila whose seminal fluid contains over 80 different seminal peptides (known as accessory gland proteins) that have been found function in sperm competition defense in multiple ways (Wolfner 2002; Kubli 2003; Fiumera et al. 2006). These accessory gland proteins act increase their own chances of paternity by killing of sperm present from rival males (Fry and Wilkinson 2004; den Boer et al. 2010), components increase ovulation and egg laying, alter immunity, promote the storage of sperm (Neubaum and Wolfner 1999; Avila and Wolfner 2009) and also reduce the 16 chances of the remating by reducing female receptivity (Chapman et al. 2003; Liu and Kubli 2003). While these proteins may serve to benefit the male’s reproductive success, the female’s life span and reproductive success are reduced as a consequence of these male accessory proteins (Chapman et al. 1995; Wigby and Chapman 2005).

While the accessory gland proteins in Drosophila are the most noted example of manipulative male seminal fluid, this adaptation is not limited to invertebrates. Mammalian male seminal fluid has also been shown to have various effects on female reproductive physiology (McGraw et al. 2015; Schjenken and Robertson 2015). Recent studies have found that semen induces many molecular and cellular changes in the female reproductive tract that facilitate successful fertilsation and pregnancy. The first reported case of seminal fluid influencing female physiology in a mammalian organism occurred in 1952 where McDonald et al. (1952) observed an influx of leukocytes (white blood cells) in the female reproductive tract following semen exposure in rabbits. This immune response to seminal fluid has now been found in several mammals including pigs (Bischof et al. 1994), horses, (Kotilainen et al. 1994), sheep (Mattner 1968; Scott et al. 2006), cows (Mahajan and Menge 1967), and dogs (Schjenken and Robertson 2014). This ensures the effective removal of sperm and microorganisms, facilitates resolution of the endometrium to a receptive state (Schjenken and Robertson 2015) and aids immune tolerance to paternal antigens (Robertson 2007). Other examples of how seminal fluid can influence females comes from pigs where there is evidence that male seminal fluid increases the number of fertilised oocytes that attain the viable blastocyst stage and influences the timing of ovulation, corpus luteum development, and progesterone synthesis in females (Robertson 2007). The effects of seminal constituents are not limited to influencing the female. In mice, seminal plasma has been found to contain components essential for embryo development and the health of male offspring (Bromfield et al. 2014). Experiments where male rodents have had their seminal vesicles, prostate or coagulating glands removed each show that seminal plasma is integral to optimal reproductive outcomes (Queen et al. 1981; Peitz and Olds-Clarke 1986; Robertson 2005).

Human seminal fluid, too, is no exception with research of the potential role of the various constituents of an ejaculation gaining more attention. As sperm only makes up two to five percent of the average ejaculate total semen volume, the question still pertains of what roles do the rest of the ejaculate play? In humans currently approximately 2,000 seminal plasma proteins have been identified (Batruch et al. 2011) but our knowledge of their functions remains limited (Rodríguez‐Martínez et al. 2011). Interestingly some of the identified functions of proteins in human seminal fluid are not that far removed from those observed in invertebrate species such as Drosophila and they can have a range effects on conception and pregnancy. So far research has found that like the rabbits in McDonald et al. (1952) experiment, female humans also respond by leukocytosis in the reproductive tract following cervical deposition of spermatozoa (Pandya and Cohen 1985). This inflammatory response requires direct contact between the female reproductive tract and the male’s seminal fluid as the same response is not elicited from condom-protected intercourse.

Seminal proteins have also been found to play a role in influencing female tissues to promote the success of pregnancy (Wolfner 2002; Chapman and Davies 2004; De Jonge 2005; Robertson 2005; Rodriguez-Sosa et al. 2006). The benefits of seminal fluid contact for embryo implantation during IVF has been shown by various studies with rate of live births dramatically increasing if women engage in intercourse while undergoing IVF treatments (Bellinge et al. 1986; Tremellen et al. 2000; Crawford et al. 2014). One study by (Gordon Jr et al. 2002) surveyed 300 students and found that women who were having sex without condoms scored lower on a measure of depression. Sperm competition can potentially confer further benefits to females beyond nuptial gifts as post copulatory incited competition could bias success fertilsation to the higher quality or more compatible sperm (Jennions and Petrie 2000).

The female reproductive tract has evolved mechanisms to only allow a few spermatozoa to reach the oocytes. This may serve a dual function in both postcopulatory choice of the fertilising spermatozoon as well as preventing polyspermy, a condition where oocytes are simultaneously fertilised by multiple spermatozoa 18 and results in embryonic death, impeding female fertility (Gilbert et al. 1997). These counter-adaptions in females (such as ova defensiveness; Firman et al. (2014)) can create a delicate balance for females who walk an evolutionary tightrope of increasing resistance to only allow a few spermatozoa but not so much that they prevent fertilisation altogether (Arnqvist and Rowe 2005). Empirical evidence for this tightrope of resistance is limited but some correlational support is seen in Mus species where female ovum defenses covary with male sperm competitiveness (Martin-Coello et al. 2009).

Sperm competition may be mediated by females’ ability to ‘choose’ between the sperm of different males (Thornhill 1983; Immler et al. 2008; Fitzpatrick and Lüpold 2014). This ability is termed ‘cryptic female choice’ (Birkhead 1998) and has been researched and recognised in many species including mammals (Jennions and Petrie 2000; Qazi 2003; Hosken and Stockley 2004; Bussière et al. 2006; Briceño and Eberhard 2009). The word ‘cryptic’ is used as the choice over the spermatozoa takes place concealed within the female reproductive tract. The mechanistic basis of mammalian sperm selection in the female reproductive tract, however, is still unclear (for a recent review see Holt and Fazeli (2015)). This ‘hidden’ fertilisation occurring within the female reproductive tract in mammals further amplifies sexual conflict over fertilisation.

Reproductive delays During different stages of the reproductive process sexually antagonistic selection may elicit time lags to facilitate cryptic female choice in mammals. For example, reproductive delays may allow females to exert choice over her progeny by allowing time for multi mating and comparison of zygotes or selective re-absorption. The delays may also function to allow time to assess and avoid genetic incompatibility (Tregenza and Wedell 2000). Stockley (2003) presents comparative support for this idea, finding that females from promiscuous species have lower rates of reproductive failure than monogamous species. Stockley argued that these females have reduced the costs of reproductive failure as they have compared the fertilised eggs and avoided genetic incompatibility. These time lags may occur between mating and fertilisation, fertilisation and implantation, or implantation and

19 parturition, and are found in over 100 mammalian species (Hayssen 1993; Mead 1993) and are hence a common suite of reproductive tactics.

A rare but interesting form of reproductive delay occurs in the American mink (Neovison vison) and European badger (Meles meles) whereby females are capable of superfetation (conception during pregnancy; Shackelford (1952)) in combination with embryonic diapause (delayed implantation of embryos). Recently these have been suggested to have evolved as a result of sexual conflict in this species and may occur other Mustelid species (Yamaguchi et al. 2006). Superfetation has been recorded under medical pregnancy treatments in some domestic animals and humans (Hall 1987; Rottenstein 1989; Steck and Bussen 1997), however, in the wild superfetation has only been reported in a few small mammals including Ctenodactyles gundi (Gouat 1985), Lepus europaeus (Caillol et al. 1991), Proechimys semispinosus (Weir 1974). The combination of embryonic diapause, associated with superfetation may have evolved from sexual conflict with and longer delays may have evolved thereafter and may benefit females beyond scheduling parturition times by confusing paternity and preventing likelihood of infanticide (Nobuyuki Yamaguchi et al. 2006).

Implantation

Following fertilsation the embryo comes into physical and physiological contact with the uterus, implanting into the wall. Implantation is a feature of viviparous birth that allows mammals to protect and allocate essential investment to young during early development but also serves as a locus for sexual conflict that is mostly unique to mammals.

Infanticide Prevention One of the more extreme examples of sexual conflict that has been found to indirectly affect implantation success is infanticide. Over 100 species of mammals have been documented to commit infanticide (Hrdy and Hausfater 1984; Parmigiani and Vom Saal 1994; Ebensperger 1998b). Infanticide may be committed by either parent of the young or another conspecific. It can present an adaptive strategy for

20 the perpetrator to enhance their reproductive success, provide nutritional benefits, increase access to limited resources, increase reproductive opportunities, or it may prevent misdirecting parental care to unrelated offspring (Ebensperger and Blumstein 2007). Here, we focus on infanticide committed by a male conspecific as a manifestation of sexual competition for paternity and avoiding misdirecting parental care to unrelated offspring. This infanticide brings the male into sexual conflict with the mother.

As infanticide is costly to the reproductive success of mothers, females have evolved various counter-strategies such as maternal (or postpartum) aggression (reviewed in Ebensperger and Blumstein 2007; McGuire et al. 2007; Weber and Olsson 2008). While this aggression is usually greatest in early lactation, some species can begin in gestation. The golden hamster (Mesocricetus auratus), for example shows a peak in aggression about six days before parturition (reviewed in Svare 1981). The effectiveness of this tactic has been called into question and has been suggested to simply “delay,” rather than prevent, infanticide (Parmigiani et al. 1989), however this delay has been argued to result in maternal success under naturalistic conditions (Blumstein 2007). Evidence for maternal aggression occurring as a result of sexually antagonistic co-evolution comes from the positive correlation of maternal aggression with the magnitude of male infanticidal behaviour in genetic strains of mice (Parmigiani et al. 1999).

The evolution of female sociality (Van Schaik and Janson 2000b; Pradhan and van Schaik 2008; Teichroeb et al. 2012), sexual segregation (Wielgus and Bunnell 2000; Dahle and Swenson 2003; Crofoot 2007), or of changes in the group sex ratio (Van Schaik and Kappeler 1997; Van Schaik and Janson 2000a; Pradhan and van Schaik 2008; Teichroeb et al. 2012) may also be attributed to infanticide prevention as females form coalitions to deter and defend their offspring against males. Evidence for this has been found in various species including semi natural populations of house mice where females residing in communal nests experienced significantly lower rates of infanticide (by both sexes) then single-mother nests (Manning et al. 1995). Female lions living in groups of two or more have lower cub mortality rates (Pusey and Packer 1994) and female equids have higher foal birth rates and survival 21 when socially integrating with other unrelated females and reduces harassment by males (Cameron et al. 2009).

Additionally, many females may mate promiscuously to reduce the risk of infanticide by confusing paternity (Hrdy 1974; Hrdy 1979; Van Schaik and Janson 2000b), although this may be absent in species with pronounced sperm competition. It is also thought that mating systems may be largely shaped under selection for infanticide avoidance with it currently being the primary explanation for the underlying the evolution of polyandry in mammals (Van Noordwijk 2000; Wolff and Macdonald 2004). However, vulnerability to infanticide is only present in a few mammalian orders (Van Noordwijk 2000) and other explanations for the evolution of polyandry that do not include any form of sexual conflict have been suggested. These explanations include the moderate fitness benefits gained for females, potentially in the form of indirect genetic benefits (Huchard et al. 2012).

As well as counter-strategies, pre-partum females also exhibit loss-cutting strategies to infanticide. One such loss-cutting strategy is pregnancy termination, where females cut their losses in the current reproductive event to save wasting further investment in ‘doomed’ or even ‘low-quality’ offspring. Dawkins and Carlisle (1976) proposed that pregnancy termination is advantageous to females that were deserted by their original mated male, as they are able to quickly re-mate with a new male who can provide parental investment, while others suggest that females may terminate pregnancy as a form of post-copulatory mate choice (Huck et al. 1982; Coopersmith 1998). One interesting example of pregnancy termination brought about by threat of infanticide is the ‘Bruce effect’. The Bruce effect is a phenomenon whereby pregnant female mammal terminate their pregnancies following exposure to an unfamiliar male and return to oestrus (Bruce 1959; Parkes and Bruce 1961). This effect was first discovered by, and subsequently named after, Hilda Bruce in 1959. The leading adaptive hypothesis for the advantage of the Bruce effect is that it occurs as a counter-strategy for pregnant females whose offspring would be susceptible to infanticide by unfamiliar males post-partum (Hrdy 1979; Schwagmeyer 1979) and therefore prevents wasted reproductive investment (Labov 1981; Storey 1986). As reproductive investment is costly to females 22

(reviewed in Speakman 2008), by preventing the waste of this investment, the females reduce their costs of reproduction.

The Bruce effect it has been well studied in laboratory mice and confirmed in several other mammalian species including deer mice Peromyscus maniculatus; (Bronson and Eleftheriou 1963), meadow voles Microtus pennsylvanicus (Clulow and Langford 1971), prairie voles Microtus ochrogaster (Heske and Nelson 1984), collard lemmings (Dicrostonyx groenlandicus; Mallory and Brooks (1980)), Mongolian gerbils Meriones unguiculatus (Rohrbach 1982), rats (Rattus norvegicus; Marashi and Rülicke (2012)), domestic horses Equus caballus (Bartoš et al. 2011) and also more recently in primates (Theropithecus gelada; Roberts et al. (2012)). There is also some evidence for hamadryas baboons (Papio hamadryas; Colmenares and Gomendio (1988)), Hanuman langurs (Presbytis entellus; Agoramoorthy et al. (1988)) and yellow baboons (Papio cynocephalus; Pereira (1983)).

In Mus, the Bruce effect is triggered by olfactory signals present in the urine of an unfamiliar male that are detected by the vomeronasal organ of the females (Luo et al. 2003), activating specific neuroendocrine pathways that inhibit prolactin secretion (Rajendren and Dominic 1993). As prolactin is imperative for sustaining proper luteal function in rodents (Stormshak et al. 1987), deficiency of this hormone causes luteolysis and subsequently pregnancy termination in these females. While urine has long been identified as the primary source of pheromones, including those responsible for pregnancy block, recent evidence by Thomson et al. (2013), suggests bodily fluids such as saliva, blood serum or faecal extract, along with tissue extracts are capable of inducing the same pregnancy block, in a manner equivalent to female mice exposed to whole urine. The mechanism of the Bruce effect in other species has not been elucidated, but additionally or alternatively it could be triggered by physiological stress associated with new males or a threatening environment (Roberts et al. 2012; Catalano et al. 2016).

The period of sensitivity to the unfamiliar male’s signals that result in pregnancy termination range between species, limiting females capacity to terminate pregnancy. In Mus, the Bruce effect can only occur if females are exposed to the 23 unfamiliar males before embryonic implantation, which typically occurs after 4-5 days post-mating (Parkes and Bruce 1961). In some Microtine species pregnancy termination can result pre- and post-implantation, up to 17 days post-mating (Stehn and Jannett 1981) and in geladas pregnancy termination can occur up to their third trimester (gestation day 151) (Roberts et al. 2012). Pregnancy termination does not affect the success of the subsequent reproductive event, with the females typically re-mating with the new male upon return to oestrus (Bruce 1960). In Mus, females exposed to an unfamiliar male after this period of sensitivity (late gestation) have been shown to spend less time nursing their pups during lactation (Gale, Brooks and Garrett in press; see Chapter 2 of this thesis), and subsequently wean smaller offspring (Gale et al. 2013). This suggests females adjust post-partum investment in relation to pups’ perceived chances of survival, potentially benefitting maternal lifetime fitness at the expense of the current litter of offspring (Marshall and Uller 2007).

It has also been suggested that the Bruce effect may extend to humans. A study by Catalano et al. (2016) has proposed the possibility of a Bruce effect in humans after they examined the odds of twins among male and female infants in Norwegian birth cohorts following the Oslo Massacre of July 2011. Births of male infants in Norway fell below, while those among females rose above expected levels which they interpreted as the Bruce effect by way of operating to autonomically raise the standard of foetal fitness necessary to extend the gestation of twins. Even if vestigially, they suggest the response to environments that threaten offspring plausibly persists in humans.

Spontaneous abortion Akin to the Bruce effect, another potential loss-cutting strategy that occurs in all animals that experience pregnancy is spontaneous abortion. There are a variety of known risk factors in non-human mammals, of both infectious and non-infectious origin. For example, in the mare (Equus caballus), infections by Escherichia coli and Streptococcus zooepidemicus are the most common causes of foetal death, abortion and stillbirth (Acland 1993; Giles et al. 1993; Smith et al. 2003). Non-infectious causes of spontaneous abortion include malnutrition, ill health, stress and maternal 24 endocrine imbalance (Jonker 2004). For example, cows diagnosed with mastitis have significant higher rate of foetal loss (Risco et al. 1999). In humans, however, spontaneous abortion is much more common and usually occurs in the first trimester generally due to genetic or chromosomal anomalies in the embryo (Forbes 2002). While the Bruce effect is usually thought to be a loss cutting strategy induced by a novel male (Schwagmeyer 1979; Labov 1981), this maternal screening is a quality control measure to ensure that resources required to sustain a successful pregnancy is not wasted on low-quality progeny (see Forbes 1997). As many as 78% of all fertilised eggs are subject to implantation failure or are spontaneously aborted by the mother early in pregnancy (Nesse and Williams 1994), before the woman misses her first period and so might not even know she was pregnant (Haig 1993a). As with every type of maternal versus paternal investment over reproduction, sexual conflict results from their diverging interests.

It is thought to be to the mother's advantage to cut her losses early so that she can presumably preserve investment in gestation and lactation until she has an embryo that is more likely to thrive and reproduce (Forbes 2017). From the father’s evolutionary perspective, however, it is in his interest for the mother to invest in his progeny and not spontaneously abort it. Importantly, maternal stress has been shown to increase the likelihood of miscarriage (e.g., Nepomnaschy et al. 2006) thus, the conflict outcome appears to be partly regulated by the psychobiological state of the mother.

Males have evolved various adaptions to circumvent spontaneous abortion in humans and to manipulate the female into increasing her investment in the foetus. One adaptation that is thought to have evolved to increase the chance of embryo retention by women is human chorionic gonadotropin (hCG), a hormone the foetus secretes into the mother's bloodstream (McGregor et al. 1981). Pre-implantation the embryo produces high levels of this hormone which is essential for early pregnancy maintenance. hCG not only serves as the basis of pregnancy detection via urine test kits but prevents the corpus luteum from breaking down (preventing the mother from menstruating) and thus allows the foetus to remain implanted. Evidence for the role of hCG as a maternal quality control screen comes from various studies that 25 have found that low levels of hCGs are associated with increased risk of first- trimester miscarriage (Keay et al. 2004; Tong et al. 2006) and aneuploidy resulting in a high rate of miscarriage (Macklon et al. 2002). Furthermore, trisomy for chromosome 21 (resulting in Down syndrome), being an exception that is often maintained throughout pregnancy and is characterised by high levels of hCG (Saller and Canick 2008). Therefore, production of high levels of hCG appears to be an adaptation to subvert the mothers attempts to spontaneously abort the foetus as it is interpreted by the mother as an indicator of a viable and high-quality foetus worthy of retention.

Sexual conflict is implicated in this adaptation as hCGα and hCGβ are both produced at high levels in the male prostate and testis and subsequently transferred to the female in seminal fluid (Berger et al. 2007). Haig (1993c) also argues that chorionic gonadotropin gene is imprinted in humans, suggesting that higher expression of this gene coming from paternal chromosome. Under Haig’s conflict hypothesis this direction of imprinting would be expected, however, the role of genomic imprinting and whether or not it has evolved as a result of conflict is currently the focus of much research.

Gestation

The best evidence of sexual conflict during gestation comes from humans and this section of the review is human focused. This may be due in part to the challenges of human pregnancy, but also due to importance in human health. While increasing evidence is mounting to support the idea that the maternal and paternal interests diverge in utero and the embryonic genome is able to directly affect maternal physiology in several mammals (Haig 1993a; Crespi and Semeniuk 2004; Haig 2004), the pregnancy complications we review are human specific.

Placental Development Once the embryo is successfully implanted, another form of sexual conflict begins to unfold over the amount of foetal nutrition provided to the embryo. In Eutherian mammals, there is some cooperation between mother and foetus where the foetus

26 receives resources from the mother to enable growth in utero. The mother benefits from the investment in her offspring to the extent that investment of her personal fitness (Haig 1993a). But while it is in the interest of the females to allocate enough resources that sustains the current reproductive event but doesn’t compromise her future reproductive events, the embryo would clearly benefit from more maternal allocation than it is optimum for her to give. Generally, a larger, better-resourced embryo is more likely to survive and achieve a greater reproductive success in its lifetime. Here the kinship theory predicts that because the paternal genes will benefit from a greater share of maternal resources then it is optimal for the mother to give, therefore, selection will result in genomic imprinting at loci responsible for foetal nutrition- with paternally expressed genes acting as growth promoters and maternally expressed genes acting as growth inhibitors (Moore and Haig 1991; Van Cleve and Feldman 2007), which they frequently do.

As with every divergence in interests over reproductive investment, sexual conflict arises, and sexual antagonistic adaptations evolve to manipulate the opposite sex. Males are able to manipulate females’ investment in utero through the genes he passes onto the embryo currently implanted in her womb (Haig 1993a). There is a growing body of evidence that suggests that interactions between maternally and paternally expressed genes in the foetus can alter the amount of maternal investment allocated during individual pregnancies. These studies show that both the mother and foetus are able to influence the amount of investment and affect foetal growth and pregnancy outcome (Biensen et al. 1998; Kurz et al. 1999; Allen et al. 2002a; Allen et al. 2002b).

This conflict often plays out in the placenta- a temporary organ that regulates the physiological exchange between mother and embryo (Mossman 1987; Leighton et al. 1995; Haig 1996). The paternal genome has been suggested to have a major influence on placental development (Wang et al. 2013) and the action of paternally expressed imprinted genes in placental development has been supported by research in mouse models (Surani et al. 1984; McGrath and Solter 1986). This is also evident in humans when an abnormality occurs and two copies of the entire genome (the complete set of chromosomes) come from one parent. If two copies of the 27 paternal genome are present, generally a hydatiform mole results (Kajii and Ohama 1977; Wake et al. 1978). A hydatiform mole is a placenta that grows wildly with a poorly developed or absent embryo. It is through the physiological exchange over this maternal-foetal interface that the father manipulates the mother to influence maternal investment.

Despite having the same fundamental function in all mammals, the gross morphology of placentas vary greatly in shape, spatial arrangement (interdigitation), and the type and number of the layers that separate the maternal and embryonic bloods (invasiveness; Mossman 1987; Leiser and Kaufmann 1994; Wooding and Burton 2008). The antagonistic co-evolution between mother and developing embryo are thought to have driven rapid evolutionary change in the placenta (Haig 1993a; Wilkins and Haig 2003; Pollux et al. 2009). The implications for the variation seen in placentas, foetal growth, and gestation length are currently unknown, however, Haig (1993) suggests that placental invasiveness is related to conflict as a more invasive placenta allows the foetus greater access to (and control over) maternal nutrients. Garratt et al. (2013b) show that selection for a faster pace of life has intensified parent–offspring conflict and less-invasive placentas have redirected control over investment in utero back to the mother.

Invasiveness is the thickness of the separating tissue layers between mother and foetus. These maternal tissue layers have been greatly reduced in several lineages of placental mammals and this variation in invasiveness has been used to classify three placenta types; epitheliochorial placenta (e.g swine, horse), endotheliochorial placenta (e.g dog, cat) and hemochorial placenta (e.g human, rodents, primates). In the least invasive epitheliochorial placenta the maternal and foetal blood are separated by six tissue layers which create the placental barrier (Mossman 1987; Wooding and Burton 2008), the intermediary invasive endotheliochorial placenta where the maternal epithelial and connective tissue is lost but the maternal endothelium remains, and in the highly invasive hemochorial placenta the foetal tissues are directly bathed in maternal blood (Mossman 1987; Wooding and Burton 2008). This direct access to the mother’s blood in the more invasive placentas allows the foetus greater access to maternal resources and may enhance foetal 28 growth rate (Crespi and Semeniuk 2004; Elliot and Crespi 2009). However, many species with epitheliochorial placentas that have well developed alternative mechanisms of nutrient transfer which may negate the impact of invasiveness on nutrient transfer (Vogel 2005).

While a majority of research on placental evolution has focused on invasiveness of the placenta as a potential for influencing nutrient transfer via conflict, increasing evidence suggests that interdigitation is also important (Capellini et al. 2010). Interdigitation is the amount of contact of materno-foetal tissues at the sites of exchange, which greatly affects the surface area for exchange and shows great variation in mammals (Mossman 1987; Capellini 2012). As with invasiveness, variation in interdigitation allows varying levels of contact with maternal blood. In the least interdigitated villous placenta, such as those of most primates, foetal tissues form villi which can either be bathed in maternal blood or covered by maternal tissues. The most interdigitated placenta where villi are highly branched and fused to form a complex, ‘labyrinthine’ structure that provides a much larger surface area for exchange compared to the villous placenta (Benirschke et al.). Last, the trabecular placenta has an intermediary level interdigitation between the villous and labyrinthine (Mossman 1987) with less branched villi only partially connected with one another (Mossman 1987; Leiser and Kaufmann 1994) such as seen in the macaques Macaca sp. (Luckett 1970; Enders and Blankenship 1999).

Intra-species comparative studies have revealed that a larger surface boosts foetal growth rates. For example the horse (Equus caballus), mares with greater surface exchange area give birth to heavier foals (Allen et al. 2002b; Wooding and Burton 2008). However, it seems that interdigitation may be associated with reproductive trade-offs. Species with labyrinthine interdigitation do have higher foetal growth rates but also have shorter gestation time (Capellini 2012). Therefore, while species with villous or trabecular placentae give birth to neonates of similar size to those of species with labyrinthine placenta, the species with a labyrinthine placenta do it in less than half the time. The apparent trade-off between foetal growth and gestation length is thought to be of consequence of sexual conflict, with paternal genes favouring greater interdigitation and enhanced foetal growth rate, and maternal 29 genes responding with shorter gestation times (Capellini 2012). Villous interdigitation is also associated with slower pace of life (late reproduction, long lifespan), that selection for a faster pace of life (early reproduction, short lifespan) may intensify this conflict and the evolution of less invasive and interdigitated placentas are a counter adaption to reaffirm control over gestational nutrient transfer and alter their investment in reproduction across life (Garratt et al. 2013b).

Genomic Imprinting Genomic imprinting has been shown in a variety of taxa including flowering plants, but in animals it is mainly confined to viviparous mammals (Scott and Spielman 2006; Feil and Berger 2007; Suzuki et al. 2007; Smits et al. 2008). Further support for the view that placental development is a major site of sexual conflict comes from genomic imprinting with a large number of maternal genes that are expressed in placenta being silenced. In fact, it seems that the presence of genomic imprinting in marsupials and eutherian mammals, but not in monotremes, points towards imprinting evolving in concert with placentation, arising at various time points in mammalian evolution due to different selective pressures at different loci (Renfree et al. 2013).

Three main theories dominate the current literature on the origins of genomic imprinting; Haig and colleagues’ kinship theory (Haig and Westoby 1989; Haig 2000, 2004). Wolf and Hager (2006) maternal-offspring co-adaption theory (see also Wolf and Hager 2009; Wolf 2013) and sexual antagonism theory Day and Bonduriansky (2004).

The kinship hypothesis has the most extensive empirical support in mammals (Moore and Haig 1991; Haig 1997; Wilkins and Haig 2003). This hypothesis suggests a male–female conflict that arises from asymmetries in relatedness between individuals’ maternally and paternally derived alleles (Wilkins and Haig 2003; Haig 2004; Úbeda and Gardner 2012). Basically, if the male is not likely to father all of the female’s future offspring, frequently the case in mammals where monogamy is uncommon (Fuentes 1998), then paternal genes will benefit from securing a greater share of maternal resources then it is optimal for the mother to 30 give. The conflict here is not between the mothers and father’s genes but rather the maternally-derived and paternally-derived genes in the foetus and placenta, an evolutionarily significant distinction (Haig 1993a; Burt and Trivers 1998; Ubeda and Haig 2003; Haig 2014a). Therefore, like Patten et al. (2014), we use the terms ‘matrigenes’ and ‘patrigenes’ to denote this difference which encapsulates their different ploidies and expected relatedness to other individuals. Under this hypothesis because the mother expects to give birth to future offspring each of which has a 50% chance of carrying the same maternal gene, through relatedness this matrigenic gene has vested interest in mother’s future reproduction. However, the patrigenic gene typically does not have the same interest. Kinship theory predicts that paternally expressed genes will act as growth promoters increasing an offspring’s share of maternal resources, while, conversely, maternally expressed genes will act as growth inhibitors (Moore and Haig 1991; Van Cleve and Feldman 2007). This prediction has been empirically supported in both humans and in mice (Mochizuki et al. 1996; Haig 2004; Abu-Amero et al. 2006; Gregg et al. 2010), with the large majority of all imprinted genes being involved in placental and foetal growth (Barlow 1995; Tycko and Morison 2002; Coan et al. 2005; Constância et al. 2005).

Maternally and paternally imprinted alleles usually reach a delicate balance, where if either side starts to ‘win’ the battle over resources, it actually results in severe health complications (even death) of one or both the mother and foetus. This balance, and the genetic antagonism between maternally and paternally derived alleles, becomes particularly apparent when there is a disruption in an imprinted gene. Usually when one gene gets damaged, a foetus has two sets of genes at each locus (a maternal and paternal genome) which means it has one normally functioning gene that can mitigate any symptoms that may arise. But in the case of imprinting where there is only one functioning gene at a locus, there is no counterpart to mitigate any symptoms should anything go wrong at that locus. Occasionally things do go wrong at an imprinted locus that result in rare disorders and severe health consequences (Reik and Walter 2001; Badcock and Crespi 2008; Rabinovitz et al. 2012) revealing the tension between paternally and maternally-derived genes within the growing foetus. 31

The first case of genetic disorders arising from disruption to imprinted genes were the reciprocally inherited in humans were Prader-Willi syndrome and Angelman syndrome. Prader-Willi and Angelman syndrome are both linked to the same imprinted region of chromosome 15. Some of the genes on chromosome 15 are maternally silenced and some are paternally silenced, so offspring who inherit a defect on chromosome 15 is missing different active genes, depending on which parent the chromosome was inherited. Prader-Willi syndrome is a neurological disorder associated primarily with loss of expression of paternally derived alleles, while Angelman syndrome is a complex genetic disorder that primarily affects the nervous system and is associated with loss of expression of the maternal allele (Nicholls and Knepper 2001).

Another human genetic disorder arising from disruption to imprinted genes is Beckwith-Wiedemann syndrome (BWS), which occurs when there are two copies of the paternal gene or when the maternal genes are silenced. This rare disorder also clearly illustrates the antagonism between maternally and paternally inherited genes. The hormone Insulin-like Growth Factor II (IGF2) stimulates growth during embryonic and foetal development by promoting placental invasion and the proliferation of cells in many tissues of the body, in accordance with paternal interests. As you might predict with the conflict hypothesis in mind, the gene that codes for this hormone, Igf2, is maternally imprinted. Meaning that the paternal copy of the Igf2 gene in the foetus manufactures IGF2, but the maternal copy is silenced with the effect of demanding less resources of the mother than if it had been expressed. The opposite is seen in a gene H19 which acts to regulate the action of IGF2, preventing excessive placental invasion and slowing foetal growth. In this case, the paternal copy of H19 is imprinted and the maternal copy expressed. If there is a disruption in either Igf2 or the corresponding H19 the embryo makes too much IGF2 and it will suffer from Beckwith-Wiedemann syndrome: a disease associated with have a variety of symptoms including an unusually large placenta, very large birth weight, a high susceptibility to tumours and cancer. Other imprinted genes that lend themselves as support for the kinship hypothesis include the paternally expressed genes Peg1 and Peg3 whose deletion results in smaller size of 32 the placenta, and conversely the maternally expressed genes H19, Igf2r, Cdkn1c, Phlda2 and Grb10 whose deletion results in overgrowth of the placenta (Reik et al. 2000; Tycko and Morison 2002; Coan et al. 2005).

Pregnancy sickness, preeclampsia and gestational diabetes

The predictions of Haig’s kinship hypothesis regarding the evolution of placental invasiveness (Haig 1993) and hCG as way for the foetus to subvert the mothers attempts to spontaneously abort it (Haig 1993c) can potentially extend to include many pregnancy complications as well as the origins of genomic imprinting. In light of this review focusing on sexual conflict during pregnancy, here we summarise the current evidence for conflict as evolutionary origin from which pregnancy sickness, preeclampsia and gestational diabetes have arisen and continue to persist.

Pregnancy sickness Pregnancy sickness, also known as morning sickness or nausea and vomiting of pregnancy (NVP), is an obstetric syndrome (Goodwin 2002) that accompanies the first trimester of two‐thirds of human pregnancies. For such a common phenomenon, afflicting women worldwide it is surprising that the physiological mechanisms and fitness consequences remain poorly understood. Pre sexual conflict theory explanations focused on the “maternal and embryonic protection,” (Profet 1988; Flaxman and Sherman 2000; Fessler et al. 2002; Flaxman and Sherman 2008), positing that the NVP evolutionary adaptation was to protect the embryo from potentially harmful foods and toxins.

Haig (1993a) by contrast suggests that the diverging interests of mothers and fathers over spontaneous abortion, in which fathers benefit from manipulating the mother into interpreting that his embryo is viable and high-quality worthy of retention and does so through imprinted genes. Haig implicates NVP as a by-product of this conflict. Forbes (2002), while supporting Haig’s conflict hypothesis, additionally suggests that hCG is the route through which males subvert spontaneous abortion, providing a probable proximate trigger for NVP. Here patrigenes benefit from averting miscarriage, despite any cost for the mother’s 33 future reproduction. Pregnancy sickness, then, may be a symptom of these paternal genes in the foetus signalling that the foetus is viable and healthy and should not be miscarried by the mother.

Indeed, there is a negative correlation between first-trimester spontaneous abortion and NVP, and NVP is associated with enlarged placental size and elevated levels of hCG (Zhou et al. 1999; Forbes 2002; Niebyl 2010). There is also very strong evidence to suggest that hCG is produced by imprinted genes. For a review of the current knowledge surrounding physiological mechanisms underlying NVP, (see works by Flaxman and Sherman 2000; Goodwin 2000; Furneaux et al. 2001).

Pre-eclampsia Pre-eclampsia is a human specific condition in females characterised by hypertension, proteinuria, and a poorly vascularised (invaded) placenta which can lead to perinatal morbidity and mortality (Kobayashi 2015). Two phases define pre- eclampsia (Winn et al. 2011). First, a shallow invasion of an abnormally developed placenta into the maternal tissues occurs early in gestation, causing the foetus and placenta to begin running short of oxygen. Second, the maternal response to a foetus struggling to obtain more oxygen from the mother’s blood leads to the clinical manifestations of pre-eclampsia. As the foetus nutritional needs are quite modest during the first two trimesters this makes little difference to either mother or foetus. However, during the third trimester when the foetus’ needs become much more substantial, problems start arising as the foetus isn’t receiving the required oxygen.

Again, conflict stands up as one of the most promising hypotheses behind the evolution and persistence of this condition. While is characteristic of the hemochorial placenta to be highly invasive into the maternal uterine lining, in human pregnancy it is especially deep and extensive compared to other primates with hemochorial placentation. The placenta modifies the maternal spiral arteries by destroying the arteriolar muscles that are responsible for regulating blood flow to the foetus, to preclude constriction (Haig 1993c; Haig 1996; Haig 1999; Barnett and Abbott 2003).

If the human placenta doesn’t sufficiently invade the maternal uterine lining the arteries are poorly remodelled and the capacity of the blood flow across the placenta is too small. Under Haig’s kinship theory it is thought that when the foetus needs greater nutrition, it will release substances into the mother's bloodstream that causes her arteries to constrict. Constriction of her arteries subsequently causes an elevation in her blood pressure, resulting in more blood flow to the foetus. More blood flow causes more oxygen and nutrients to the growing foetus. If the elevated blood pressure becomes too high, however, it risks damaging the mother’s tissues including the blood vessels that bathe the placenta in blood, as well as fine capillaries in kidney and liver damage lungs and brain, and potentially result in stroke; the symptoms of pre-eclampsia.

According to this scenario, pre-eclampsia occurs as a result of adaptations that benefit the foetus, even if it comes at a cost of inflicting damage of the mother. However, the consequences of pre-eclampsia for the foetus can be just as severe. In preeclamptic pregnancies the reduced blood flow and placental damage slows the foetus’s growth rate and results in a low birth weight. If this is in fact the mechanism by which pre-eclampsia occurs than it implies that paternal and foetal genes expressed in the placenta have evolved to increase maternal blood pressure to enhance foetal resource provisioning via the foeto-maternal interface (Naeye 1981; Haig 2008). While maternal genes have evolved to limit foetal growth (Fowden and Moore 2012). How deeply the placenta invades and the extent to which it remodels the mothers’ arteries are exactly the types of traits expected to be of conflicting optima between mother and foetus.

It is also important to note that it is of interest that this disease is confined to humans, other than a few isolated reports on pre-eclampsia-like symptoms in gorillas, (Baird Jr 1981; Thornton and Onwude 1992), as despite also having a hemochorial placenta, incomplete invasion the placentas of primates is the normal course of events in these species (Ramsey et al. 1976). There are also other mammals that exhibit the invasive hemochorial placentation but not pre-eclampsia,

35 including the majority of extant Euarchontogliran mammals (rodents, lagomorphs, treeshrews, colugos and primates) (Elliot and Crespi 2009). Through phylogenetic reconstruction it has been revealed that former members of this super order that used to have a hemochorial placenta have evolved towards less-invasive endotheliochorial or epitheliochorial placentation including tree shrews, strepsirrhine primates and heteromyid rodents (Elliot and Crespi 2015). Why the reduced invasion of maternal vessels in non-human (especially closely related) mammals is not only normal but doesn’t result in preeclampsia is currently unknown. The lack of non-human model species in has also contributed to the difficulty in elucidating the cause of, and a treatment for pre-eclampsia (McCarthy et al. 2011).

Gestational Diabetes Haig’s (1993) kinship hypothesis also offers a viable explanation for gestational diabetes, a metabolic disease characterized by high levels of circulating glucose and insulin in the maternal bloodstream with onset or first recognition during pregnancy. Usually when blood glucose levels rise (e.g. after a meal), the hormone insulin is secreted to increase glucose uptake by the liver and skeletal muscles which in turn lowers blood glucose levels. However, with gestational diabetes the mother’s tissues become temporarily resistant to insulin and therefore their blood glucose levels soar. The mechanism by which the mother’s insulin sensitivity is lost, is through a placental hormone known as human placental lactogen (hPL) – which plays an important role stimulating foetal growth. Again, under the conflict hypothesis it is thought that the foetus is releasing substances (mainly hPL) to purposely reduce maternal sensitivity to insulin, and subsequently elevate the level of blood glucose that it can draw upon for its own growth. Indeed, gestational diabetes is often associated with foetal overgrowth as the high levels of blood glucose provide more energy than the developing foetus requires, allowing it to store fat reserves. However, the foetus’s gain may be the mother’s loss, not only the short-term loss of energy but long-term damage to the pancreas.

While in modern times this condition can be easily managed, it may have been more relevant in our ancestors, for whom food was scarcer. In fact, when pregnancy 36 women fast the levels of hPL in their bloodstream elevate dramatically. This evolutionary hangover from our predecessors is also thought to follow the pattern of paternal genes as growth promoters and maternal genes as growth suppressors. The reciprocally imprinted duo H19 and Igf2 have shown to be a suspect in gestational diabetes as different forms of H19 gene have been found to vary in their ability to regulate insulin resistance. Moreover, preliminary evidence suggests that maternal glucose tolerance and offspring birth weights are associated with common polymorphic variation in the H19 gene (Petry et al. 2005). And more recent studies in mice have found evidence suggesting that variation in foetal Igf2 expression could affect risk for gestational diabetes (Petry et al. 2010).

In fact, most of the complications that plague human pregnancy such as infertility, pre-eclampsia, pregnancy sickness, gestational diabetes, intrauterine growth restriction, and polycystic ovary syndrome show evidence for a potential evolutionary origin in genomic conflicts. However, due to the lack of research in human models and also any other model entirely similar to humans that could be used, we have a severe lack of data and understanding of most of these complications.

Conclusion

In Eutherian mammals, the difference in reproductive optima between the sexes is exaggerated by the constraints imposed on females by viviparity and subsequent investment in lactation of altricial young which has lead to widespread patterns of sexual conflict. This review has outlined numerous circumstances in which the sexually antagonistic selection during pregnancy has been a key factor in shaping mammalian reproduction and life histories. While there is growing empirical evidence for sexual conflict in mammals, current literature is still highly skewed towards invertebrate species due to the difficulty in demonstrating that fitness benefits to one sex come at a cost to the other (Pizzari and Snook 2003) in larger, more complex organisms such as mammalian taxa. However, with growing research on mammals in this field by placing a greater emphasis on examining the underlying kin structure and relatedness of

37 sexual interactions, we can attain a much deeper understanding of the evolutionary origins and the conflicts they create (Bourke 2009).

The breadth of our knowledge of genomic imprinting is also lacking, and currently it is confined to only a few tissues in a small subset of mammalian and angiosperm taxa (Patten et al. 2014). While this allows some of the strongest support we have for the kinship hypothesis with imprinting occurring the placental mammals but not monotremes (Ferguson-Smith et al. 1991) and the placenta having more imprinted genes (Coan et al. 2005) related to growth (Morison et al. 2005). It doesn’t allow for clarity of the evolution of imprinting. Extensive phylogenetic trees are available for most groups of mammals and future research should look into systematically testing and comparing patterns of genome-wide data on the imprinting status of genes across a range of mammalian taxa.

For the human pregnancy complications, hopefully future research on genomic conflict will help elucidate the causes and facilitate the development of treatments. However, further interesting questions will be raised if genomic imprinting is behind pregnancy sickness and if we were to intervene and subvert it, would the female be more likely to show an increase in spontaneous abortions? If the embryo- protection theory is correct, would intervention prevent the protective mechanisms and end up harming the foetus?

While there remain many unanswered questions that this review calls attention to, what is clear is that mammalian life histories and each individual’s optimal partitioning of time energy into growth, survival and reproduction hinges largely upon their interactions with their mates and how the inevitable sexual conflicts are resolved. Gaining a better understanding of how kin selection affects sexual conflict is important in understanding the variation in which conflict is realised, how it can be resolved, and we believe will help explain the evolutionary source of pregnancy complications in mammals.

Chapter 2

Perceived threats of infanticide reduce maternal allocation during lactation and lead to elevated oxidative damage in offspring

Gale, T. G., Garratt, M. and Brooks, R. C. This chapter is currently in review.

Abstract

Maternal investment is costly to the mother but essential to offspring survival in altricial species. Infanticide by novel males results in loss of maternal investment, and maternal strategies have evolved to mitigate these losses. One such maternal strategy, the Bruce effect, involves spontaneous abortion by females of some mammal species when exposed to a novel male during pregnancy. In mice, the Bruce effect only occurs during early pregnancy, but we have previously found that female mice exposed to a novel male’s scent in late pregnancy weaned smaller offspring. Here we replicate that manipulation in order to resolve the cause of the reduced weaning weight and subsequent effects on offspring fitness. Females exposed to an unfamiliar male’s scent in late pregnancy spent significantly less time nursing their pups during lactation, suggesting that reduced maternal allocation contributes to slower offspring growth. The offspring with a reduced weaning weight exhibited catch-up growth and reached a normal weight at adulthood. These offspring, however, were found to bear oxidative damage in adulthood, revealing long-term effects on offspring condition. We conclude that female mice strategically alter their investment in lactation in relation to the likelihood of infanticide, but that this results in long term fitness costs to their offspring.

Key words: catch-up growth, infanticide, maternal effects, oxidative stress.

Introduction

In mammals, reproduction is characterised by large maternal investments in pregnancy, lactation and other behaviours critical for offspring survival, growth and development (Wolff and Sherman 2008). This investment such as nutrition, warmth, and protection can be costly to the mother (Bronson 1989; Clutton-Brock and Vincent 1991; Woodside 2007), and those costs are exacted in the currency of residual reproductive value (König et al. 1988). Generally, the relationship between mother and offspring is a cooperative interaction, with the mother benefitting from the success of her progeny. However, with current and future reproduction known to trade-off (Reznick 1985; Stearns 1992; Kirkwood and Austad 2000), a mother can also benefit from altering, and sometimes reducing, her investment in current offspring according to environmental conditions, in order to optimize her overall fitness.

For most mammals, lactation is the most energetically demanding episode of reproduction (Speakman 2008) and is a key time to modify investment in relation to contextual conditions. Therefore, when contextual conditions aren’t favourable to a positive reproductive outcome, females can benefit from reducing maternal investment in the current reproduction. This idea is supported by a recent meta- analysis by Berghänel et al. (2017) who found that in times of adversity prenatal stress causes mammalian mothers to reallocate energy away from offspring and towards themselves, resulting in an altered growth rate in the offspring. This benefits the mother’s future reproduction by reserving residual reproductive value. In mice it has been found that maternal investment in lactation is somewhat flexible (König 1985; König and Markl 1987), and offspring do not get more milk than corresponds to the mother’s optimal allocation during lactation, despite frequent suckling attempts (König and Markl 1987). Such flexibility may significantly enhance maternal fitness.

Female mice have also evolved other strategies to optimise their maternal investment when faced with the prospect of infanticide. Infanticide represents one

40 of the more extreme forms of sexual conflict that results in a significant loss (and waste) of maternal investment for females. Infanticide is very common in mice (Elwood and Ostermeyer 1984; Manning et al. 1995), and is thought to be an adaptive strategy for the perpetrator to enhance their reproductive success (Agrell et al. 1998). Social counter-tactics to infanticide females have been found to use include communal nesting for group defence (Agrell et al. 1998; Ebensperger 1998a), maternal aggression (Parmigiani et al. 1989) or multiple male mating to confuse paternity (Thonhauser et al. 2013).

The threat of infanticide has also been shown to have effects beyond the killing of the offspring. Parkes and Bruce (1961) reported that recently inseminated female mice (Mus musculus) terminate their pregnancies following exposure to the urinary scent of an unfamiliar male and return to oestrus. This phenomenon, now known as the ‘Bruce effect’, has been experimentally confirmed in numerous laboratory studies in house mice (Bruce 1960; Chipman and Fox 1966) and across a number of murine and microtine rodent species (Clulow and Clarke 1968; Clulow and Langford 1971; Mallory and Clulow 1977). It has also recently been observed in the wild in the primate Theropithecus gelada (Roberts et al. 2012).

The Bruce effect only occurs in mice if the female is exposed in early pregnancy, up to 4-5 days post mating (Bruce 1961), before embryo implantation. We have shown that if female mice are exposed to an unfamiliar male in late pregnancy (approximately day 14), offspring are of comparable birth weight, but growth was slower over lactation, resulting in lighter weaning weights (Gale et al. 2013) compared with controls. Following weaning, however, the experimental offspring grew faster and caught up to the control offspring by the onset of adulthood. Reduced weaning weight is possibly due to alteration of maternal behaviour during lactation. As unfamiliar males typically kill pups sired by another male (Vom Saal 1985), the females may strategically reduce their investment in lactation to prevent wasting it on ‘doomed’ offspring. If so, females are capable of utilising subtle tactics to optimise their investment in reproduction in relation to their perceived chances of offspring survival. In commensal habitats mice generally live in groups consisting of a number of breeding females, potentially some non-breeding females and 41 subordinate males, and a dominant male who aggressively defends his territory against unfamiliar males (Hurst 1990). As there is frequent territory turnover (Oakeshott 1974; Bronson 1979), females would be advantaged by these counterstrategies towards unfamiliar conspecifics.

However, whether reduced offspring growth was a consequence of a strategic alteration of investment by mothers in that study is unknown. Other possible causes of the offspring’s reduced weaning weight could include stress triggered by the unfamiliar male (stress has previously been found to result in litters with a reduced weaning weight following lactation (Barlow et al. 1978; Kinsley and Svare 1988; Meek et al. 2000)), an epigenetic modification in offspring gene expression or potentially reduced milk quality or volume.

Modifications in maternal investment have been found to have further effects other than just a reduced weaning weight. Mouse pups show great developmental plasticity, with various aspects of their early environment affecting their life-history trajectories and having lasting effects on adult phenotypes. Through the influence of maternal effects, mothers may alter the phenotype of their offspring and affect their physiological, sexual and behavioural responses as adults (Rossiter 1996; Sheldon and West 2004).

The aims of this investigation were to test if the degree of maternal investment during lactation is affected by the threat of infanticide, and whether any such effects 42 on maternal investment impose further costs to the offspring later in life. We predicted that: 1) When females are faced with the threat of infanticide they will strategically alter investment in their offspring by changing their behaviour during lactation. 2) Offspring that accelerate their growth rate following a reduced weaning weight will suffer oxidative costs as a consequence of the rapid growth. 3) Offspring that have a reduced weaning weight will suffer costs to components of reproductive effort such as scent-marking rates and composition for males and future reproduction for females.

Materials and methods

We followed our previous experimental methods (Gale et al. 2013), modified to include exposing females to either an unfamiliar male or the paternal male late in their pregnancy on three occasions over the last eight days of gestation. We added a third unmated control treatment, in which females were housed with another female instead of a male. The unmated treatment acted as a control for female weight and our marker of stress (glucocorticoid metabolites). We then conducted scan sampling during lactation where we examined female nursing behaviour, collected faecal samples to measure female stress, and measured offspring oxidative stress levels at adulthood to test for oxidative damage associated with compensatory growth.

Animal Housing Experimental mice were all sixth-generation captive-bred house mice (Mus musculus) originally derived from a population of wild mice acquired from a chicken farm in the Northwest of Sydney, Australia (as in Gale et al., 2013). Females were weaned at 28 days of age and were housed with their female siblings until the beginning of the experiment. Males were also weaned at 28 days of age but were housed individually. Mice were maintained on a 12:12 hour reversed light cycle. A dim red light was used for experimental procedures, which were all undertaken in the dark phase as mice are nocturnal. Each mouse was housed in a 315 x 180 x 125mm cage lined with corncob bedding and provided with tissues and shredded

43 newspaper for bedding. Vella Rat and Mouse Pellets and water were provided ad libitum.

Experimental protocol To investigate the consequences of a novel male’s presence to pregnant females (post-implantation of the embryo) we compared two controls (mated and unmated) with the experimental ‘novel male’ treatment. The unmated control was exposed only to the presence and odour of another unfamiliar female. The mated control or ‘familiar male’ treatment females were mated and then exposed to the scent and presence of the same mate. Our third treatment was the novel male treatment that was experimentally manipulated so that late gestation the mated females were exposed to the scent and presence of a different male who was unrelated to the mate. Because exposure to this new male was late in pregnancy, well beyond the time frame when a Bruce effect is thought to occur, females carried pregnancies through to parturition, therefore allowing us to examine whether females alter their reproductive investment in their offspring and whether this has any long-lasting effects on the offspring (as previously used in (Gale et al. 2013).

To begin each treatment replicate, a male or a female for the unmated control, unrelated to the female (all between 80-120 days of age) was placed in her cage for a two-week mating period. Males were subsequently removed, and the females were transferred to a clean cage. Female weights were recorded every three days, a weight gain of 3g indicated pregnancy (mean days ± s.e. from reaching the 3g weight threshold to giving birth for females exposed to the male they mated with: 6.9 ± 1.7; females mean days and exposed to an unfamiliar male: 7.1 ± 1.9). As the gestation period of Mus musculus is 21 days (Jones et al., 2009), the females were close to parturition and the embryos had undergone substantial development. When the 3g- weight gain threshold was reached, females were sequentially allocated into a treatment, to either exposure from the paternal male (n=25) or a novel male (n=24). The unmated control females had already been assigned to be exposed to another female (n=25). As the unmated control females never gained the weight concurrent with pregnancy, each day one of the females from the other two treatments hit the

44 weight gain threshold one of the unmated control females was randomly assigned to also begin their exposure until all of the control treatment had been exposed.

The mice (males or females) added to the females’ cages were separated by a metal divider with nine small holes (5 mm diameter) which only allowed limited contact. The exposures took place for three hours a day for three consecutive days. On each day a handful of the respective male or female’s used bedding (approximately 15- 20g) was placed into the female’s cage. Following exposure, the female (prior to parturition) was placed into a clean cage. There were no significant differences between the treatments in the number of days between the removal of the breeding male and the beginning of the male exposure (mean days ± s.e. for females exposed to the male they mated with: 2.16 ± 1.1; an unfamiliar male: 2.08 ± 1.2). There was also no difference in the time between the exposure and the subsequent birth (mean days ± s.e. for females exposed to the male they mated with: 3.2 ± 1.9; an unfamiliar male: 3.7 ± 1.5).

To test for further costs associated with reduced investment during lactation, offspring were tested for either effects to their reproductive fitness or for oxidative damage. At 13 weeks of age half of the male and female offspring from both treatments were randomly assigned to a reproductive fitness or oxidative stress group. The females in the reproductive fitness group were mated to an unrelated male and the size, weight and pup mortality was recorded. The males in the reproductive fitness group were tested for scent-marking rates and concentration of major urinary proteins.

Maternal Investment and offspring weights The body weights of the females were recorded every three days throughout the experiment. Offspring were weighed at birth and every three days during lactation (to the nearest 0.1g) as a quantification of maternal investment (Ortiz et al. 1984; Ross 1988; Pontier et al. 1989). The offspring were weighed collectively as a litter until weaning (4 weeks old), after which they were separated from their dam, and an individual weight was recorded. Another individual weight was collected for each of the offspring at adulthood (13 weeks old). 45

Maternal behaviour Maternal behaviour was evaluated by observing each of the females using a scan – sampling technique twice a day, every day for the duration of their lactation (four weeks). Observations were conducted during the dark phase using only a head torch with red light. The first observation occurred in the second or third hour after the change from the light to dark period and the second between four hours and one hour before the change from the dark to light period. Each of the females was randomly assigned an observation order at the beginning of each observation period and observations begun ten minutes after the red light was turned on to allow them to habituate.

A single observer sequentially recorded each female’s behaviour for a total of eight observations with a five minute gap between each of the eight recordings. As this was done morning and evening each day it totaled 16 scans a day. Behaviours recorded included; in nest, nursing, grooming, licking pups, eating/ drinking, nest building, resting or active (e.g see (Benus and Rondigs 1996; Koteja et al. 1999; Palanza et al. 2002). The unmated control females were not included in any of the scan sampling of maternal behaviour.

Glucocorticoid metabolites To determine whether the novel male elicits a stress response from the female, faecal samples were obtained from the females at three points through the experiment and tested for glucocorticoids metabolites (using methods as described by Palme and Möstl (1997b). Following a stressful event glucocorticoids are released into circulation (Sapolsky et al. 2000) and are hence used as an indicator of the stress response. The first sample was taken at the beginning of the experiment (three weeks prior to being paired for mating), the second on the second day of male exposure, and the third on day seven of lactation. The unmated females were randomly allocated for faecal sample collection in order to correspond with the timings of the females in the other treatments.

For collection of faecal samples, mice were placed in a large empty cage (565 x 387 x 203mm) made from H.D. polyethylene with a wire roof, for maximum of 45 minutes. The cages were placed topside down over another corresponding cage and faecal samples were collected from the bottom cage. Immediately after collection, faecal samples were frozen at -80 °C. Faecal samples were homogenized and an aliquot of typically 0.05g faeces (Palme et al. 2013) was extracted with 1ml of 80% methanol for 30min on a vortex. When there was insufficient sample the protocol was adjusted accordingly (e.g., 0.25 g faeces in 2.5 mL methanol). Samples were placed in a spinner overnight and then the supernatant was diluted (1:1000) with assay buffer (Trizma, pH 7.5). Samples were then analysed in a double-antibody 5a- pregnane- 3b,11b,21-triol-20-one enzyme immunoassay (EIA) which has been validated for use in mice to assess concentration of glucocorticoid metabolites as described by (Touma et al. 2003; Touma et al. 2004).

Oxidative Stress At 13 weeks of age, half of the female and male offspring from the novel male and familiar male treatment groups were culled humanely by cervical dislocation, and the liver, kidney, heart and gastrocnemius muscle were quickly removed, snap- frozen in liquid nitrogen and stored at −80°C. To assess oxidative stress in the mice two biomarkers of oxidation including protein thiol content and aconitase enzyme activity were analysed in each of the tissues (Gibson et al. 2015). Both of these biomarkers correlate negatively with oxidative stress.

Protein thiol content was measured by methods described by (Di Monte et al. 1984) but modified for use on a 96 well plate reader (Vasilaki et al. 2006b). Protein thiols are essential for stability of and optimum function of proteins, but are highly susceptible to oxidation (Halliwell and Gutteridge 1999), and therefore good markers of oxidative stress. Aconitase is an enzyme of the tricarboxylic acid cycle that is highly susceptible to deactivation by radical oxygen species (specifically superoxide) and therefore used as a marker to indicate levels of reactive oxygen species and concomitantly oxidative stress (Hausladen and Fridovich 1994; Gardner et al. 1995; Hausladen and Fridovich 1996b; Gardner 1997). As aconitase is located in part in the mitochondria (Wiegand and Remington 1986a; Gardner et al. 1995), 47 mitochondrial density was also assessed using citrate synthase activity. Citrate synthase is an enzyme commonly used as an indicator of the content of intact mitochondria (Holloszy et al. 1970) and was measured in homogenates according to (Pichaud et al. 2008).

Offspring reproductive fitness Dominant adults are known to scent mark more regularly than subordinates (Drickamer 1995) to communicate their territorial and sexual status (Bronson 1979; Hurst et al. 2001) and competitive ability (Rich and Hurst 1998). These chemical scent marks are of high importance to male fitness as they directly influence the attractiveness of a male to a female (Rich and Hurst 1998). The main involatile scent component of male mouse urine is major urinary proteins (MUPs) that bind volatile components of the urine and slowly release them from the scent marks (Hurst et al. 1998). Scent marks may not prevent intruders invading the territory, but they do allow males to use long-lasting signals of identity and dominance over a territory to alert competitors and potential mates (Hurst et al. 1998; Hurst and Beynon 2004). Therefore, both the rates and the composition of scent-marks can substantially influence male reproductive fitness.

Scent-marking rates Scent-marking rates of all of the individually housed male offspring were assessed at 13 weeks of age. Scent-marking rates were measured by placing the individual males into an empty (315 x 180 x 125mm) cage lined with Benchkote for one hour a day, for three consecutive days. The scent marks were measured by the number of spatially separate marks observed under UV light and the average number of marks for each male over the three trials was used for analysis.

Major Urinary Protein Concentration A urine sample was collected from each of the male offspring at 13 weeks of age. Males were confined in a large empty cage (565 x 387 x 203mm) made from H.D. polyethylene with a wire roof, for maximum of 180 minutes. The cages were placed topside down over another corresponding cage to allow the mouse urine to pool in

48 the bottom cage. Urine was then pipetted into an Eppendorf tube and frozen at -20 °C. The concentrations of major urinary proteins were established using Coomassie plus® protein assay reagent kit from Perbio Science UK Ltd (Cramlington, Northumberland, UK) as described by Cheetham et al. (2009). We also measured urinary creatinine (Beynon and Hurst 2004) using the method of Cheetham et al. (2009) to correct for the urinary dilution.

Statistics All statistical analyses were performed using SPSS software package version 2.1 (IBM Corp, Armonk, NY, USA). The analyses were done with dam ID and experimental block fitted as random factors to account for non-independence of individuals originating from the same litter and time differences of each group of experimental mice unless otherwise described. For the oxidative stress data we also fitted plate number as a random factor to control for between plate variability. Scent marking frequency was transformed to log (x+1) to account for measures of zero deposits and normalise the data. Significance was determined at p≤0.05.

Results

Offspring weight In the novel male treatment, the females gave birth to pups that were of similar weight and litter size as the familiar male treatment females (i.e. exposed to the paternal male). Out of 25 females mated in the familiar male treatment, 23 mothers gave birth within the time frame with a mean litter size of 4.217±0.77 and out of 24 mothers in the Novel Male treatment, 21 gave birth within the timeframe with a mean litter size of 4.14±0.89. The mean birth weight of novel male treatment pups (1.619± 0.432g) was not significantly different from that of the familiar male treatment (1.628 ± 0.433g; ANOVA: weight: F1, 182= 0.022, p= 0.882; litter size: F1,42= 0.070, p= 0.792). Two females in the control treatment and one female in the experimental treatment group destroyed their litters within three days of giving birth. Some pup mortality of unknown cause was observed over lactation; however, this is common in captive breeding mice (Weber et al. 2013). The pup mortality was not significantly different between the treatments (ANOVA: F1, 7= 0.778, p= 0.407)

49 with six mothers of the novel male treatment committing maternal infanticide (total 10 pups lost) and three mothers of the familiar male (total seven pups lost).

Although the novel male treatment females gave birth to pups of a similar weight and litter size as the familiar male treatment females, novel male treatment litters grew more slowly over lactation. To analyse the differences in offspring growth rate we used repeated measures ANOVA reporting within-subjects effects. Mauchly's Test of Sphericity indicated that the assumption of sphericity had been violated (χ2(2) = 34.210, p = <0.001), therefore degrees of freedom were corrected using Greenhouse-Geisser adjusted degrees of freedom. There was a difference in growth rate between the treatments but there was no effect of sex (treatment: F1.7,310=7.49, p= 0.001; interaction between treatment and time: F1.7,310=5668, p= <0.001; sex x treatment: F1.6,306=1.3, p= 0.270). This resulted in novel male treatment pups being significantly smaller (7.19±0.61g), at weaning (4 weeks old) (ANOVA: F1,182=27.11, p=<0.001) than the familiar male treatment pups (7.99±1.48g) (see Figure 2). This difference did not persist into adulthood however, as both males and females of the novel male treatment exhibited catch-up growth following weaning and there were no weight differences at adulthood (13 weeks old) (ANOVA: F1, 184=0.760, p=0.384) (see Figure 3). There were also no significant interactions of treatment and sex in weight at weaning (sex x treatment F1,180= 2.14, p=0.145) or at adulthood (sex x treatment F1, 180=0.044, p=0.868)).

4 0

N o v e l m a le

) 3 0

g (

F a m ilia r M a le

r 2 0

a e

M 1 0

0 1 3 6 9 1 2 1 5 1 8 2 1 2 4

A g e (d a y s o ld )

Figure 2: Average litter growth rate over lactation (4 weeks). Data are displayed as means ± 1 s.e.m.

N o ve l m a le 1 5

F a m ilia r M a le

)

(

g i

e 1 0

s f

f 5

e M

0 B ir th W e a n in g Ad u lth o o d Figure 3: Comparison of individual offspring birth, weaning and adult weight between the two treatments. Data are displayed as means ± 3 s.e.m.

Maternal behaviour There was no significant difference in the amount of time the mothers spent in nest, grooming, licking pups, eating/ drinking, nest building, resting or active (Table 1). The only behaviour showing significant differences between treatments was nursing, with the novel male treatment females spending nearly half the time that the familiar male females did (Table 1).

Table 1: Maternal behaviour of Novel male treatment and Familiar male treatment females over lactation

Comparison of the number of times maternal behaviours were recorded during lactation

Mated Control Novel Male

F d.f P Mean SE Mean SE

In nest 3.77 1,41 0.059 114.34 4.15 103.47 3.54

Nursing 11.56 1.41 0.002 15.52 1.34 8.4 1.53

Licking Pups 1.79 1,41 0.188 1.82 0.63 0.80 0.34

Nest Building 0.12 1,41 0.730 0.65 0.27 0.80 0.31

Eating/ 0.14 1,41 0.707 5.82 0.62 5.52 0.52 drinking

Grooming 0.14 1,41 0.705 9.73 0.68 10.14 0.82

Active 0.78 1,41 0.382 131.34 6.21 139.76 7.45

Resting 1.15 1,41 0.288 168.73 5.00 179 7.96

Glucocorticoid metabolites in mothers Three faecal samples were taken from each females: The first sample was taken before the experiment began to determine the normal level of faecal glucocorticoid metabolites (FGMs) for each female. The second collection was taken on day two of the exposures to a male and the third coincided with mid lactation in the mated treatments. Some of the females did not provide a sample within the timeframe they were in the collection cages (45min) so the sample sizes for the glucocorticoid metabolites are novel male n=20, familiar male n= 21 and Unmated control n= 18. Using a repeated measures ANOVA we found that FGM’s did change significantly over time (effect of time: F2, 112 = 131, p= <0.001; interaction between time and housing companion: F4,112= 29.5, p<0.001). In the novel male and familiar male treatments the FGM’s were higher in the second collection and dropped back down closer to their normal levels of FGM’s at the third collection. The was no difference in the FGMs in the females before the experiment began (ANOVA: F2,56=0.094, p=

0.911), but there was following exposure to their partners (ANOVA: F2,56=43.72, p=<0.001) and at the end of the experiment which coincided with mid lactation for the reproducing treatments (F2,56=4.6, p=0.014)(See Figure 4). After removing the unmated control from the analysis we found that the effect of time was still significant (F2, 78= 143, p=<0.001) but the interaction between time and housing companion was not (F1,78=1.89, p=0.158). While there was no difference between the FGMs between the novel male treatment and the familiar male treatment, there was a difference between the mating treatments and females who were not mated. Females from both mated treatments experienced a rise in FGM’s levels in the second collection following exposures to a male, however, by the third collection at mid lactation their FGM’s had returned closer to their normal level (collection one). The unmated control females FGM levels did not show any pronounced variability over the three collections.

1 2 0 0 0

M id la c ta tio n s

e 9 0 0 0 t

i F o llo w in g e x p o su re

)

s b

e P re -E x p e rim e n t

e a

f 6 0 0 0

(

n a

e 3 0 0 0 M

0 N o v e l m a le F a m ilia r M a le U n m a te d c o n tro l

T r e a tm e n t

Figure 4: Concentration of faecal glucocorticoid metabolites in females taken before the experiment started, on the second after exposure to a male/female and then mid lactation in the mated treatments. Females were either exposed to a novel male (novel male treatment) their mated male (familiar male) or another female (unmated control). Data are displayed as means ± s.e.m.

Oxidative stress in offspring Both markers of oxidative stress showed consistent evidence of oxidative damage with the novel male treatment (n=47) in two (liver and heart) out of four organs tested having lower protein thiol concentrations and aconitase enzyme activity (indicating higher levels of oxidative damage) than the familiar male treatment (n=38). There was an effect of sex with novel male treatment offspring males having lower protein thiol concentrations than those of the familiar male treatment males in the heart but there were no other sex differences in any of the analyses for oxidative damage (see Table 2 and Figure 5 and 6).

The same trend was observed when testing the aconitase enzyme activity to citrate synthase activity (mitochondrial density) ratio. The liver and heart in the novel male treatment were also found to have a significantly lower aconitase enzyme activity to

55 citrate synthase activity ratio indicating higher levels of reactive oxygen species (ROS) production in these organs. This decrease in aconitase enzyme activity was not observed in either the kidneys or the gastrocnemius and there were no effects of sex (See Table 2 and Figure 5 and 6).

Table 2: Markers of oxidative stress in offspring of both treatments at adulthood. Showing protein thiol concentration (μmol g-1 protein) and aconitase enzyme activity/citrate synthase (units/mg protein). Models were also fitted with plate number and block as a random factor to control for variation between assay plates.

Oxidative stress results for protein thiol concentration and aconitase enzyme activity

Liver Kidney Protein thiol Aconitase enzyme activity Protein thiol Aconitase enzyme concentration concentration activity F d.f P F d.f P F d.f P F d.f P Treatment 60.65 1,80 <0.001 14.1 1,81 <0.001 1.44 1,80 0.234 1.47 1,80 0.228 Sex 0.130 1,80 0.719 0.20 1,81 0.656 0.01 1,80 0.997 0.32 1,80 0.569 Sexx treatment 1.43 1,80 0.234 1.59 1,81 0.210 0.61 1,80 0.436 1.85 1,80 0.177 Heart Gastrocnemius Protein thiol Aconitase enzyme activity Protein thiol Aconitase enzyme concentration concentration activity F d.f P F d.f P F d.f P F d.f P Treatment 10.34 1,80 <0.001 20.42 1,81 <0.001 0.17 1,80 0.681 1.76 1,80 0.188 Sex 8.49 1,80 0.005 0.98 1,81 0.755 2.88 1,80 0.093 0.78 1,80 0.379 Sexx treatment 1.80 1,80 0.183 0.41 1,81 0.523 0.001 1,80 0.971 2.11 1,80 0.150

2 5 0

n N o ve l m a le

a g

r F a m ilia r m a le

e 2 0 0

)

t t

n 1 5 0

l o

o 1 0 0

(

e t

o 5 0

a e

M 0 L iv e r K id n e y H e a r t G a s tr o c n e m iu s

O r g a n

Figure 5: Mean protein thiol concentrations. Data is presented as estimated marginal means ± 1 s.e.m. for each measure from general linear mixed models for each tissue sample.

0 .0 8

N o ve l m a le

n a

g F a m ilia r m a le

r o

0 .0 6

)

g 0 .0 4

(

c a

0 .0 2

e M

0 .0 0 L iv e r K id n e y H e a r t G a s tr o c n e m iu s

O r g a n

Figure 6: Ratio of mean aconitase enzyme activity to citrate synthase. Data is presented as estimated marginal means ± 1 s.e.m. for each measure from general linear mixed models for each tissue sample.

Male and female offspring fitness components For the male offspring fitness components sample sizes were both n=17 for the treatments. There was no difference in the concentration of major urinary proteins among the male offspring produced by the novel male treatment females compared to the control offspring (ANOVA: F1, 30= 0.76, p= 0.543). The mean protein (mg/mg creatinine) ± s.e for the novel male treatment was 18.44± 3.60, and the familiar male treatment males were 20.73±2.74. Experimental block also did not have an effect

(ANOVA: F1, 30= 5.8, p= 0.249). The mean ± s.e scent-marking frequency for the novel male treatment offspring was 28.28±1778.56 and the familiar male treatment males was 36.96±1873. A repeated measures ANOVA found that there was no difference in the frequency of scent-marks between the treatments (F2, 64= 3.00, p= 0.57). For the female offspring produced that were bred at adulthood there was no difference in the pup weights that the female offspring gave birth to (ANOVA: F1,40=0.35, p=

0.556) or of their pups weaning weights at the end of lactation (F1,40=1.002,

59 p=0.323) and a repeated measures ANOVA showed that there was no difference in growth rate between the treatments (Sex x Treatment F1,41=1.29 p=0.262). The mean birth weights between the treatments were 1.26± 0.04g and 1.355± 0.13g for the novel male and familiar male treatment females respectively and the mean weaning weights were 12.75± 2.09g and 12.27± 0.13g respectively. Experimental block also had no effect (ANOVA: F1,40= 0.69, p= 0.837).

Discussion

Our results support the prediction that the ‘Late Bruce Effect’, in which female mice exposed to a novel male late in pregnancy weaned at lower weights (Gale et al. 2013), may be due to a strategic reduction in maternal investment in lactation. Notably, females exposed to a novel male’s scent spent less time nursing pups than females exposed to the scent of their mate. Our results support the prediction that females adjust post-partum investment in relation to pups’ perceived chances of survival, potentially benefitting maternal lifetime fitness at the expense of the current litter of offspring (Marshall and Uller 2007).

This finding suggests that patterns of post-partum investment can be adjusted by mothers in relation to the risk of infanticide, in ways that mirror the strategic, spontaneous abortion of pups under the Bruce effect (Bruce 1960, 1961; Hrdy 1979; Schwagmeyer 1979; Storey 1986), as well as evidence that a mechanism to abort foetuses strategically at the later stages of pregnancy has evolved in other species including the gelada (Roberts et al. 2012) and prairie vole (Clulow and Clarke 1968; Clulow and Langford 1971). As mice are only able to abort in early pregnancy, altering their investment post-partum provides an opportunity to respond to the threat of infanticide. Akin to the Bruce effect, is it not currently known if the response is driven by the costs of sexually antagonistic selection or a benefit to females during social instability which could be the focus of future work.

Other studies have shown that female mice strategically moderate investment during lactation and that they may alter investment relative to the reproductive value of their offspring. Mashoodh et al., (2012) found that females increased their 60 investment during lactation when mated with a male housed in lifelong socially enriched conditions compared with females mated with males housed in impoverished conditions. This suggests that females may invest more in offspring on the basis of paternal condition. Konig and Markl (1987) showed that despite frequent suckling attempts, offspring do not get more milk than corresponds to the maternal optimum during lactation, suggesting mothers have control over reproductive allocation during lactation.

It is important to note, however, that there is little evidence that time spent nursing indexes milk transferred from mother to offspring, especially if measured through scan sampling. Mendl and Paul (1989) found that scan sampling of suckling did not measure milk intake in mice. While we base our prediction that there was a reduction in milk transfer from mother to offspring on the fact that not only did we observe that females exposed to novel males spent less time nursing, their pups also had a reduced weaning weight. Future studies could examine nutritional allocation by mothers by measuring milk transfer to determine if this is the cause of the offspring’s slower growth rate.

Another potential cause of the reduced weaning weight could be a stress response. In rodents, maternal stress has been previously found to cause depressed growth in offspring (Barlow et al. 1978; Kinsley and Svare 1988; Meek et al. 2000). Previous research on the effects of prenatal stress on offspring is frequently contradictory. The nature, timing and length of the stress inflicted varies from study to study, as do the results. One study, for example that used crowding as a stressor found no difference in time spent nursing between stressed and non-stressed dams (Moore and Power 1986), while another study that used novelty stress found that stressed dams spent significantly more time nursing compared to non-stressed dams using (Muir et al. 1985). It is also important to note there is also evidence that maternal stress leads to faster (Dantzer et al. 2013) and increased growth of offspring at birth and weaning (Szuran et al. 1991).

We found while females all had similar stress levels before the experiment began, females that were mated and then exposed to either a novel or their mated male had 61 much more FGMs than the females that were not mated in the experiment. Females from the mated treatments both displayed a rise in their FGMs following the exposures, suggesting that the females were equally stressed by encountering the paternal male as they were a novel male. This may suggest that separation of the female from the paternal male in the mated control treatment may have be sufficient to increase the risk of infanticide by the paternal male as the separation could result in less paternity certainty. Higher levels of FGM could also result from exposure to a male, as our experiment used another female as an unmated control we are unable to distinguish this result properly. Only one other study we could find used a novel male conspecific as a source of stress. Lerch et al. (2016) stressed pregnant or lactating females using unfamiliar male faeces and examined maternal and offspring behaviour to investigate whether early adverse experiences elevate the risk of developing psychiatric disorders. However, unlike our study that used a familiar male as a control, they compared pregnant or lactating females exposed to unfamiliar male faeces with a control group that didn’t receive any faeces. The scarcity of research that uses another conspecific as a stressor is surprising as this could be expected to be a more environmentally relevant challenge that other manipulations like forced immobilization challenges, for example. While the over production of maternal glucocorticoids can be harmful (Korgun et al. 2012) levels are known to increase during pregnancy in mice (Barlow et al. 1974; Dalle et al. 1978) as they are essential for foetal development (Korgun et al. 2012). Post - partum, glucocorticoids also have important roles in milk secretion and lactogenesis (Chida et al. 2011). The higher FGM levels that we found in females from both of the mated, compared to the unmated treatment, may, therefore, be just a normal consequence of pregnancy.

As in our previous study (Gale et al. 2013), the offspring of the females that were exposed to the novel male were smaller at the end of lactation, but they caught up in size by maturity. The catch-up growth exhibited by the offspring coincides with the time after lactation when offspring begin to feed themselves on solid food. Offspring can acquire the resources to accelerate their growth rate themselves. For compensation or catch-up growth to occur, the benefits must outweigh the costs of not accelerating growth. While there may be a positive association between size and 62 fitness (Roff, 1992) and accelerated growth may increase overall reproductive success, many studies have found that compensatory growth inflicts various costs over different time-scales (reviewed in Metcalfe and Monaghan (2001)). Costs that have been documented in rats (Rattus norvegicus) include deficiencies in protein maintenance (Samuels and Baracos 1995), telomere abrasion rate (Jennings et al. 1999) insulin regulation (Ozanne and Hales 1999), adult obesity (Waterland and Garza 1999), and, perhaps most importantly, reduced lifespan (Jennings et al. 1999; Rollo 2002). While compensatory growth may reduce lifespan it could still be adaptive if it increases overall reproductive success (Metcalfe and Monaghan 2001) or short-term survival chances (Arendt 1997). In mice, reproductive allocation in adulthood is influenced by size, and so compensatory growth may allow individuals to attain a normal reproductive rate in adulthood, at least early in life. Female offspring that exhibited catch-up growth produced litters of the same size and weight as those female offspring from the control group; male offspring showed equivalent scent marking abilities in our assays. Thus, catch up growth seems to allow offspring to attain a similar reproductive output early in life that is comparable with the steadier growth in the familiar male group, although costs may be paid for this in terms of late life reproduction or lifespan.

The existence of compensatory and catch up growth shows that organisms do not grow at their maximal rate, but rather at a rate influenced by, and potentially optimized to, their circumstances. Hector and Nakagawa (2012) distinguish these two terms by defining compensatory growth as a faster than usual growth rate and catch-up growth an attainment of control size. Mangel & Munch (2005) propose that growth leads to an accumulation of damage at the cellular level that is expressed at the level of the organism and is an important cost of compensatory growth. We tested for damage on a cellular level in form of oxidative damage and oxidative stress. We found evidence for oxidative damage in the livers and hearts of offspring from the novel male exposure treatment, highlighting that an olfactory change in a pregnant mother’s environment can elicit a variety of maternal and offspring responses, ultimately influencing offspring physiological condition in adulthood. This may be a consequence of catch-growth, but could also be a consequence of odour exposure itself, or changes in maternal allocation in response to this. In future 63 studies it may be of interest to limit an offspring’s ability to show compensatory growth (through a nutritional or genetic manipulation) and test whether oxidative damage in offspring still occurs in adulthood.

Wild mice are highly territorial and turnover of the dominant male is a common occurrence in wild populations (Oakeshott 1974; Bronson 1979). For the females in the territory that means that they will be exposed to novel males which present the threat of infanticide (Ebensperger 1998a). Our experiment was designed with the rationale to mimic this turnover by exposing the females to a novel male to see the effects it would have on reproductive allocation and offspring fitness. Our results suggest that the very prospect of male territorial turnover can have physiological consequences for the offspring and potentially alter their life history. We also suggest that females are capable of strategically modulating their investment relative to their current contextual conditions, which may offer significant fitness benefits in the wild where offspring survival is much more variable and infanticide is a common threat.

Chapter 3

No evidence that male mice impose mating or insemination costs on females, but females avoid castrated males

Gale, T. G., Garratt, M. and Brooks, R. C. This chapter is currently in review

Abstract

Sexual conflict is inevitable in all sexually reproducing species with members of each sex selected to maximize their own fitness often at the expense of the other. This conflict, and the physical necessity of intromission, can intensify the already costly nature of reproduction for female mammals. To identify and partition the costs males may inflict on females during mating and reproduction in mice, we paired females with either vasectomised, castrated or intact males, or with other females, and gave half of each of these groups access to a refuge from their partner. We found that females did use the refuges but that this was dependent on the gonadal status of their partner. Females paired with vasectomised and castrated males spent the most time in their refuge, and the females housed with castrated males also had increased glucocorticoid levels, indicating an increased stress response. This suggests that females actively avoid males with low levels of testosterone, and that housing with such males is sufficient to generate a physiological stress response. There was no effect of refuge or housing partner on female weight or oxidative stress levels.

Key words: refuge, stress, costs of reproduction, castration

Introduction

Males and females can increase the costs of mating and reproduction experienced by their mates (Trivers 1972; Arnqvist and Rowe 2005; Parker 2006; Wedell et al. 2006). Each of the sexes can have different optima for whether, when or how often to mate, how many different individuals to mate with, how many offspring to have, and how much parental care to invest in those offspring. Sometimes the result of this is sexually antagonistic evolution, which has led to adaptations that benefit one sex and harm the other (Rice 1992; Holland and Rice 1998; Parker 2006; Arnqvist and Rowe 2013). Sexually antagonistic coevolution can embroil males and females in a co-evolutionary arms race of offence and defence (Holland and Rice 1998; Arnqvist and Rowe 2002; Parker 2006).

In mammals, females bear greater facultative physiological costs of reproduction due to prolonged gestation and subsequent lactation, which contrast sharply with the less onerous, typically facultative, investments of mammalian fathers (Clutton- Brock and Parker 1992; Tregenza and Wedell 2000; Scantlebury et al. 2002; Clutton-Brock et al. 2004; Aloise King et al. 2013). Males can impose further costs of reproduction on females through courtship, harassment or coercion of females into mating (Clutton-Brock and Parker 1995), infanticide (Lukas and Huchard 2014) and genetic conflict during pregnancy (Moore and Haig 1991; Haig 1993a; Foerster et al. 2007; Mainguy et al. 2009). Partitioning the costs that arise from male presence and mating behaviour from those that arise from gestation and lactation can be difficult because, at least in the absence of in-vitro fertilization, pregnancy ensues only after a period of colocation with a male.

Glucocorticoid (GC) stress hormones represent one physiological route by which mates might impose costs on one another via, inter alia, mating behaviour, harassment, or insemination. Male mating behaviour has been demonstrated to increase female GC levels in primates through sexual coercion and aggression (Muller et al. 2007; Thompson et al. 2010) and changes in male rank (Beehner et al. 2005; Carnegie et al. 2011). Non-preferred mates may also induce stress in females. This has been demonstrated in the socially monogamous Gouldian finch where

66 females paired with poor-quality mates had higher levels of circulating corticosterone (Griffith et al. 2011). While it is well established that GC stress hormones play crucial adaptive roles in foetal development (Baxter and Rousseau 1979), prolonged high secretion can also have suppressive effects on female reproductive physiology and behaviour (Tilbrook et al. 2000; Wingfield and Sapolsky 2003). Increased GC’s may even have further costs related to reproduction as they have been found to have a significant effect on oxidative stress in long-term experiments (Costantini et al. 2011), and prenatal maternal stress has long-term consequences for offspring (Lemaire et al. 2000; Maccari et al. 2003; Yang et al. 2006; Tamashiro et al. 2009).

Oxidative stress has been suggested to be an important cost of reproduction (Harshman and Zera 2007; Costantini 2008; Metcalfe and Alonso-Alvarez 2010). Previous studies that have assessed oxidative stress as a cost of reproduction have found varying results. Avian studies have found a positive correlation between hatching success and oxidative damage (Metcalfe and Alonso-Alvarez 2010) and increased brood size caused a significant decrease in antioxidant defences (Alonso- Alvarez et al. 2004) although antioxidant decline does not necessarily indicate oxidative stress (Costantini 2008; Monaghan et al. 2009). In mammals, Nussey et al., (2009) found no significant relationship between reproductive effort and oxidative damage in wild soay sheep (Ovis aries) and our own research group Garratt et al., (2010) and Garratt et al. (2013a) found no consistent effects of pregnancy, or post- partum pregnancy while lactating in captive wiled-derived house mice (Mus musculus). Indeed, increasing reproductive effort of mother actually increased antioxidant defences in (Garratt et al. 2013c). Bergeron et al., (2011) however, did find a positive relationship between oxidative stress and litter size in eastern chipmunks (Tamias striatus).

If males impose costs on females over and above the costs of reproduction, then females may be able to avoid costly mating’s or male harassment by hiding from males. Refuging behaviour in female sharks (Scyliorhinus canicula), for example, has been proposed as an avoidance strategy to reduce the occurrence of aggressive and potentially costly mating events during egg-laying and gestation periods (Sims 67

2003; Sims et al. 2005). Avoidance of males to elude unsolicited mating attempts and harassment has also been suggested to explain sexual segregation in other marine mammals including humpback whales (Megaptera novaeangliae) (Smultea 1994; Clapham 2000), and Galapagos (Zalophus californianus wollebaek) and southern sea lions (Otaria flavescens) (Connor 2002; Wolf et al. 2005). When sexual harassment levels are high, female guppies (Poecilia reticulata) may even actively select areas of high predation risk if male presence is low in those areas (Darden and Croft 2008) as an avoidance strategy. This segregating behaviour in response sexual harassment has also been found in a number of invertebrate species (Robbins et al. 1987; Krupa et al. 1990; Eldakar et al. 2009) and hypothesized to be a key driver for the sexual segregation seen in ungulates (Bon and Campan 1996). Harassment, infanticide and other coercive male reproductive behaviours can not only encourage sexual segregation but can also be key determinants in the development of some mating systems (Linklater et al. 1999; Van Schaik et al. 2004). In horses, for example, it has been shown that mares benefit from long-term relations with a single stallion as they avoid intraspecific aggression (Kaseda et al. 1995), and this avoidance of stallion aggression is suggested to be the reason for the polygynous band-structure of horses (Linklater et al. 1999).

It is also important to note that being alone may also carry costs that could potentially surpass those incurred by residing with a male during a low-cost scenario (i.e. no immediate risk of infanticide, low harassment levels etc). Females can also preferentially mate with males that are able to provide protection, paternal care and/ or access to resources (Clutton-Brock 1988; Clutton-Brock et al. 1989). Furthermore, in some species mated males become more paternal rather than infanticidal towards their own offspring and it has been found that in some species of mice paternal male removal results in lower offspring survival (Gubernick and Teferi 2000).

Here we seek to partition the costs of reproduction from those of male presence, mating and insemination in female house mice (Mus musculus) in a 4 by 2 factorial experiment. We paired female house mice with either a sham-vasectomised (intact) male, a vasectomised male, a castrated male or a control female. Castrated males don’t 68 show most sexual behaviours (Hull and Dominguez 2007b) such as mounts, intromissions and ejaculations (Larsson, 1979) (Clemens et al. 1988; Hull and Dominguez 2007b), vasectomised males will still mate with females, but without impregnating them, and sham-vasectomised males can additionally impregnate females. This suite of treatments thus allows us to partition the physiological and life history costs of male presence (female control vs other groups), copulation (castrated vs vasectomized males), and insemination, pregnancy and reproduction (sham vasectomized vs vasectomized males). In addition, we sought to study female refuge- seeking responses to these costs by offering half the females in each treatment a refuge where they could escape from their allocated partner.

The aims of this investigation were to test if females would choose to avoid males when they have control over their time spent with them, and to split potential costs of male presence from mating and reproduction. We predict that: 1) The females that are housed with other females or castrated males will spend the least amount of time in their refuges as there will be no harassment or mating behaviour. 2) Females without refuges from their male partners will have higher glucocorticoid levels. 3) Females that are housed with sham-vasectomised and vasectomised males will have higher oxidative stress levels.

Material and methods

This experiment was approved by the University of New South Wales Animal Care and Ethics Committee (approval: 14/75B).

Animal Housing The female experimental mice were all sixth-generation captive-bred house mice (Mus musculus) originally derived from a population of wild mice acquired from a chicken farm in the Northwest of Sydney, Australia (Garratt et al. 2013a). Wild females were weaned at 28 days of age and were housed with their female siblings until the beginning of the experiment. The partner mice were of the C57BL/6 strain

69 and were purchased for the Australian BioResource Center (ABR, Mossvale, NSW, Australia). Male partners were operated on (castrated vasectomised or sham vasectomised) when aged between 7-8 weeks old prior to being shipped to University of New South Wales (UNSW). Upon arrival, they were given 4 weeks in individual cages to habituate to the new environment. The partner females were weaned at 21 days and housed with their female siblings until the beginning of the experiment. Mice were maintained on a 12:12 hour reversed light cycle with lights off at 6pm. A dim red light was used for experimental procedures, which were all undertaken in the dark phase. Prior to the experiment each mouse was housed in a 315 x 180 x 125mm cage lined with corncob bedding and provided with tissues and shredded newspaper for bedding. Vella Rat and Mouse Pellets and water were provided ad libitum.

Experimental protocol The experimental protocol consisted of two treatments in a 2 x 4 factorial design. The first treatment factor was the presence or absence of a refuge for the females from their assigned partner. The second factor is the type of partner the females were paired with; either a male that had undergone a vasectomy (n=20), a sham vasectomy (i.e. intact) (n=20), a full castration (n=20) or another female (n=10). We replicated this experiment to increase our sample size and completed two blocks of n=20 for each treatment group, resulting in our final sample size of n=40 for each treatment. For the second block we reused the same partner mice (vasectomised, castrated or another female C57BL/6 strain lab mouse) but used different female mice. Females that were housed with the sham-vasectomised male but did not give birth were excluded from the analysis as this treatment aimed to investigate the effect of reproduction on the measures tested.

A Passive Integrated Transponder (PIT) tag (Trovan ID100, 2 by 12 mm) was inserted subcutaneously behind the neck of each female using a Trovan manufacturer-supplied implanter, and she was given a week to adjust and recover. She was then placed into a cage (315 x 180 x 125mm) that was joined to another identical cage via a tube. Vella Rat and Mouse Pellets and water were provided ad libitum. on both sides of the connected cages. In the “refuge present” treatment, the 70 tube diameter was so small (15mm) that only the wild strain female could fit through. This gave the female a refuge from her larger C57BL/6 strain partner. In the “refuge absent” treatment the tube was much larger (70mm) so both the C57BL/6 strain partners and the wild strain females could move freely between the cages. Females were weighed at the beginning (prior to any experimental work) and again at the end. Oestrus was induced in each participating female by exposing her to the scent of her assigned mate two days before she entered the experiment. Females with access to a refuge could control the time they spent with their paired partner. This allowed us to measure the time each female would choose to spend with a male when given control, and to determine if female refuge use could be used to reduce costly male interactions. The movements of females in and out of their refuges was recorded for the duration of the experiment and, for the non-refuge treatment, the location of both the male and the female were recorded.

Female refuge use Antennae were placed at either end of the tubes connecting the adjoining cage, and attached to a reading device (Trovan 665), which registered the date and time when the female passed underneath from their passive integrated transponders. Two antennae were necessary to determine the direction of the movement of the females. Every four days the reading device was removed, and the data was downloaded with Trovan LID650/665 software version 603 up until 12 days into the experiment. While the females in the no-refuge treatment were tagged, their movements were not logged as both the partner and the wild females were able to move freely between the cages. We only measured the first twelve days of the experiment as this part of the method was very labour intensive and we only wanted to gain an insight as to whether the focal animal sampling data was sufficient to pick up the movements of the mice.

To further monitor the females’ movements, scan sampling was conducted twice a day every day for the duration of the experiment. The scan sampling took place within the dark phase 1-3 hrs after the change from light to dark and then 1-3 hrs before the change back from dark to light. A red light was turned on and the mice were given 10mins to habituate. The females were then randomly assigned an order 71 and sequentially scanned to record their placement in the cages; either in their side, in the tube or in the male’s side. In the refuge absent treatment the location of both the male and the female were recorded. Each of the females were scanned eight times with five minutes between each scan.

Oxidative Stress Females paired with sham-vasectomised males were left to reproduce until mid- lactation when housing partners were removed, and females were culled to measure oxidative stress in the liver, kidney, heart and gastrocnemius. As there was only one reproductive treatment (sham-vasectomised), each time a female from this treatment reached day 14 of lactation, one female from each of the other treatments was randomly chosen (control, vasectomised and castrated) to be culled. To assess oxidative stress two biomarkers of oxidation including protein thiol content and aconitase enzyme activity were analysed in each of the tissues. Females were humanely culled by cervical dislocation at 14 days post-partum and the liver, kidney, heart and gastrocnemius muscle were quickly removed, snap-frozen in liquid nitrogen and stored at −80°C.

Protein thiol content was measured by methods described by Di Monte et al., (1984) but modified for use on a 96 well plate reader (Vasilaki et al. 2006a). Protein thiols are essential for stability of and optimum function of proteins, but are highly susceptible to oxidation (Halliwell 1999), and therefore good markers of oxidative stress. Aconitase is an enzyme of the TCA cycle that is highly susceptible to deactivation by radical oxygen species (specifically superoxide) and therefore used as a marker to indicate levels of reactive oxygen species and concomitantly oxidative stress (Hausladen and Fridovich 1994; Gardner et al. 1995; Hausladen and Fridovich 1996a; Gardner 1997). As aconitase is located predominately in the mitochondria (Wiegand and Remington 1986b; Gardner et al. 1995), mitochondrial density was also assessed using citrate synthase activity. Citrate synthase is an enzyme commonly used as an indicator of the content of intact mitochondria (Holloszy et al. 1970) and was measured in homogenates according to (Pichaud et al. 2010).

Glucocorticoid metabolites Faecal samples were obtained from the females after they had been housed with their partners for 13 days and tested for glucocorticoid metabolites (using methods as described by Palme & Möstl, 1997) to determine their stress levels. To collect the faecal samples the females were placed into a large empty cage (565 x 387 x 203mm) made from H.D. polyethylene with a wire roof. This cage was then placed topside down over another identical cage and the faecal samples were collected from the bottom cage. Females were held within these cages for a maximum of 45 minutes. Immediately after collection, faecal samples were frozen at -80 °C.

Faecal samples were then assayed after the second block of the experiment, so all samples were analysed at the same time. Samples were first homogenized and an aliquot of 0.05g faeces was extracted with 1ml of 80% methanol for 30min on a vortex (Palme et al. 2013). In cases where there wasn’t sufficient sample the protocol was adjusted accordingly (e.g., 0.25 g faeces in 2.5 mL methanol). Samples were placed in a spinner overnight and then the supernatant was diluted (1:1000) with assay buffer (Trizma, pH 7.5). Samples were then analysed in a double- antibody 5a-pregnane- 3b,11b,21-triol-20-one enzyme immunoassay (EIA) which has been validated for use in mice to assess concentration of glucocorticoid metabolites as described by Touma et al., 2003, 2004.

Statistics All statistical analyses were performed using SPSS software package version 2.1 (IBM Corp, Armonk, NY, USA). The analyses were done with dam ID and experimental block fitted as random factors to account for non-independence of litter and block time unless otherwise described. For the oxidative stress and glucocorticoid data, we also fitted plate number as a covariate to control for between plate variability. Scent marking frequency was transformed to log (x+1) to account for measures of zero deposits and normalise the data. Significance was determined at p≤0.05.

Results

Female and offspring weight A univariate analysis of variance showed that there were no differences between weights of the females from the various treatments at the beginning of the experiment (start weight x refuge: F1,62=0.017, P= 0.896; start weight x partner:

F3,62=0.692, P= 0.560). To analyse differences in female weight gain from the beginning to the end of the experiment we used repeated measures ANOVA reporting within-subjects effects with sphericity assumed. We saw no differences in weight gain between blocks (F1,61= 0.008, p= 0.930), refuge treatments (F1,61= 0.66, p=0.798), or partner type x refuge treatment (F3,61=0.426, p= 0.735), but there were significant differences in weight gain between partner types (F3,61= 20.14, p= <0.001) (See Figure 7). This difference in weight was due to a straightforward effect of pregnancy, with the sham-vasectomised treatment females gaining more weight than all other treatments. When this treatment was removed, there was no longer any difference between the weights of the females in the other treatments (repeated measures effect of partner F2,44= 0.145, p= 0.866).

In the sham-vasectomised treatment (the only reproducing treatment), we found no difference in the time to parturition or in the litter size between the refuge present and refuge absent females (ANOVA: Time to parturition: F1,15= 0.55, p= 0.818; difference between litter size: (F1,15= 646, p= 0.435). The mean days to parturition were 23.3±0.33 and 23.16±0.47 for the refuge and no refuge treatments respectively and the mean litter sizes were 5.4±0.54, and 4.66±0.76 for the refuge and no refuge treatments respectively. However, we did find most pregnancies were from refuge treatment with 10 out of the 16 females that gave birth having had access to a refuge

2 (X [1] = 2.0, p= 0.157) and that all females in the refuge treatment gave birth on the side of the apparatus where the male was located, a chi squared tests revealed this

2 effect was significant (X [1] = 5.0, p= 0.0433).

2 0

S ta rt W e ig h t 1 8

E n d W e ig h t

) g

( 1 6

g i

e 1 4 W

1 2

1 0 F e m a le V a s e c to m is e d C a s tr a te d In ta c t

T r e a tm e n t

Figure 7: Female weight at the beginning and end of the experiment for each treatment. Data is presented as estimated marginal means ± 1 s.e.m

Radio Frequency Identification (RFID) of female movement The RFID data was only collected on females with a refuge. A repeated measures ANOVA revealed that over the first twelve days of the experiment time spent in the refuge declined in favour of spending more time with their assigned partners (time:

F11,297= 16.50, p= <0.001), and that the rate of this decline differed between treatments (day x partner F33,297= 7.201, p= <0.001). There was no effect of block

(day x block F11,297= 1.52, p=0.120). The females spent the least time with the castrated males (i.e. more time in refuge), and this was the only treatment in which the average time spent in the refuge was greater than in the partner’s enclosure after the 4th day of the trials (See Figure 8). A post hoc Tukey’s test revealed that this treatment was significantly different from all other treatments (p= <0.001 for all three treatment comparisons), while none of the other treatments significantly differed from each other.

To explore the day x partner effect we split the data up and looked at the first and last day as these time points represent periods where females responded differently to housing partners (see Figure 8). On day one of the experiment females spent 75 significantly more time in their refuge if they were housed with a sham vasectomised males than if they were with any of the other partner types. A univariate analysis of day one showed that partner was significant (F3, 31= 7.85, p= <0.001) and a post hoc LSD test that the sham vasectomised treatment was significantly different from the female treatment (p= <0.001) and the vasectomised treatment (p= 0.014) but not the castrated treatment (p= 0.646). However, by day 12 of the experiment females spent the most time in their refuges if they were with castrated males (castrated vs all other treatments; p= <0.001). Those females with vasectomised males also spent significantly more time in their refuges than those with intact males (vasectomized vs intact; p= 0.009).

Figure 8: Time the females spent in their refuge over the initial twelve days of the experiment. Data is displayed as second order polynomial trend lines.

Female movement The scan-sampling showed that the difference in the time the females spent with their partners seen in the RFID data persisted over the whole duration of the experiment. We used a repeated measures ANOVA and as Mauchly's Test of Sphericity indicated that the assumption of sphericity had been violated (χ2(594)

=940.28, p = <0.001), degrees of freedom were corrected using Greenhouse-Geisser estimates. There was a significant effect of day x partner (F51.3,1045= 4.260, p=

<0.001) as seen in Figure 8. There was also an effect of day x refuge (F17.1,1045= 2.02, p= <0.001) and day x refuge x partner (F51.3,1045=1.65, p=0.003). Block did not have an effect (F17.1, 1045= 0.60, p=0.889). A post hoc Tukey’s test revealed that the castrated male treatment was significantly different from each other over time (all p= <0.000), but none of the other treatments significantly differed. Data is graphed as time spent with their partner as this measurement included the ‘no refuge’ treatment where the partners also had access to both sides of the adjoined cages. If both the female and the partner were in the same side of the adjoined cages (regardless of which side) they were marked as ’same side’.

To further explore the 3 way interaction we then analysed the no refuge and refuge treatments separately in a repeated measures ANOVA over the first 18 days of the experiment and then over the last 17 days. For the no refuge treatment looking at the first 18 days of the experiment Mauchly's Test of Sphericity indicated that the assumption of sphericity had been violated (χ2(152) =374.37, p = <0.001), degrees of freedom were corrected using Greenhouse-Geisser estimates. There was still a significant effect of day x partner (F22,206= 2.66, p=<0.001). When looking at the last 17 days of the experiment the day x partner interaction was no longer significant

(F22,213= 0.89, p= 0.612). A post-hoc Tukey’s test showed that the castrated male treatment was only significantly different from the sham vasectomised treatment.

For the refuge treatment Mauchly's Test of Sphericity indicated that the assumption of sphericity had been violated (χ2(152) =211.77, p = 0.002), and again the degrees of freedom were corrected using Greenhouse-Geisser estimates. The day x partner effect was significant over the first 18 days of the refuge treatment (F29,338= 4.70, p=<0.001). When looking at the last 18 days of the refuge treatment the day x partner effect was again no longer significant (F23,270= 1.25, p=0.195). A post-hoc Tukey’s test showed that all treatments significantly differed from each other. This result highlights that the 3 way interaction is due to the first 18 days of the experiment mimicking the findings in the RFID data which was only conducted over the start of the experiment. After this period (days 16-35) there is only an effect of 77 partner x refuge (F3, 62= 4.83, p= 0.004) and day is no longer significant (treatment x day, partner x day and treatment x partner x day all p= >0.01), see Figure 9. Females spent the most amount of time with the intact males and spent less time with the castrated males if they had access to a refuge.

2 0

1 5 R e fu g e

N o R e fu g e

s n

a 1 0

c S

0 F e m a le V a s e c to m is e d C a s tr a te d In ta c t

T r e a tm e n t

Figure 9: Mean time spent with partner from days 16- 35. Data is presented as estimated marginal means ± 1 s.e.m for the mean of the 16 observational scans done per day

Faecal Glucocorticoid Metabolites (FGM) We used a univariate analysis to examine the faecal glucocorticoid metabolites of the females after they had been housed with their partner for 13 days. This revealed that there was an effect of partner (F3, 62= 7.11, p= <0.001), refuge (F1, 62= 4.54, p=

0.037) and a non-significant interaction of refuge x partner (F3, 62, =2.08, p= 0.086).

Adding in block as a covariate showed that there was no effect of block (F1, 62= 0.028, p= 0.868).

When looking at just the no refuge treatment the effect of partner was significant

(F3, 31= 9.36, p = <0.001). A post hoc LSD test revealed that this was completely

78 driven by the castrated male treatment which was significantly different from all other groups (all p=<0.002), see Figure 10. However, when looking at just the refuge treatment the effect of partner was no longer significant (F3, 31= 1.21, p= 0.320), since providing females housed with castrated males with a refuge reduced their FGM levels (see Figure 9).

1 5 0 0 0

R e fu g e

s N o R e fu g e

l )

o 1 0 0 0 0

n (

n 5 0 0 0

e M

0 F e m a le C a s tr a te d In ta c t V a s e c to m is e d

T r e a tm e n t

Figure 10: Mean faecal glucocorticoid levels of females at day 13. Data is presented as estimated marginal means ± 1 s.e.m.

Oxidative stress We found no evidence of differences in oxidative stress between treatments relating to either housing partner or whether a refuge was present or absent (Table 3). Block didn’t have an effect when analyzing housing partner (F3,61= 0.02, p= 0.996) or whether a refuge was present/ absent (F1,61= 0.052, p= 0.821). There also wasn’t any interaction between housing partner and whether the refuge was present or absent (F3, 61= 2.1, p= 0.279)

Table 3: Markers of oxidative stress in females of all treatments. Showing protein thiol concentration (μmol g-1 protein) and aconitase enzyme activity/citrate synthase (units/mg protein). General linear multivariate model was also fitted with plate number and block as a covariate.

Marker Housing Partner Refuge present/absent df F P df F P Liver Thiols 3,61 0.20 0.895 1,61 0.25 0.616 Kidney Thiols 3,61 1.41 0.248 1,61 2.60 0.112 Heart Thiols 3,61 0.75 0.524 1,61 0.11 0.745 Gastroc Thiols 3,61 0.44 0.724 1,61 0.94 0.334 Liver Aconitase 3,61 0.82 0.487 1,61 1.00 0.320 Kidney Aconitase 3,61 1.28 0.288 1,61 0.02 0.967 Heart Aconitase 3,61 0.43 0.729 1,61 0.22 0.636 Gastroc Aconitase 3,61 0.85 0.469 1,61 1.37 0.245

Discussion

We set out to partition the costs imposed by males on females, and to test whether the availability of a refuge mitigated or amplified any such costs. We found that female refuge usage was dependent on which treatment partner the female was housed with. While all females used the refuge in the first days of the trial, females paired with castrated males spent similar amount of time in the refuge in the second half of the trial, whereas females in other treatments spent much less time in the refuge as the trials progressed. Females housed with castrated males and did not have access to a refuge also had higher glucocorticoid levels, indicating greater stress, than the females paired with a vasectomised male, intact male or another female. We found no effect of housing partner or refuge use on oxidative stress levels.

Females housed with intact males showed a significant increase in body weight, as expected for the only reproducing treatment; with weight gain attributable to pregnancy and lactation. We also expected to see an increase in body weight in the females housed with the vasectomised males as it has been shown that mating with these males can induce pseudo-pregnancy, a temporary state that increases female body mass by approximately 5–10% (Dewar 1957; Dewar 1959), returning to normal 20 days after mating (Dewar 1957). We observed no pseudo-pregnancy effects on weight. This may be due to a difference in methods, as the vasectomised males in these previous studies were removed from the cages after mating (e.g. for 24–48 hours), whereas males were left in the cages for the duration of our experiment. Excluding the females that underwent reproduction we found no effect of housing partner or refuge use on female weight. This conflicts with recent research by Garratt et al (2016) that found housing C57BL/6 females with males of the same strain increases female weight irrespective of male gonadal status. This again may be attributed to a difference in strains, as it is important to note that we used wild-derived females. We were unable to find any other studies that have housed or mated wild-derived mice with C57BL/6 strain mice and therefore any effects resulting from the crossing of these two strains are not currently known.

While we speculated that females housed with intact males might use their refuge more frequently to avoid paying higher costs of male presence and mating, these females actually spent the least amount of time in their refuges. As females from this treatment gained weight during pregnancy, they would have gradually been unable to fit through the tube into their refuge and all of these females either chose to have their pups on the male’s side of the double cages or were caught on that side once they were too large to fit through the tunnel by chance. Doing so rather than giving birth in the seclusion of a refuge might protect the pups from infanticide. Male mice are highly infanticidal to unrelated pups (McCarthy and vom Saal 1985), and females may choose to remain close to the male so that he recognizes the pups as his own once they are born. This is important as an ejaculation induced inhibitory mechanism that reduces male aggression (and infanticidal behaviour) towards

81 infants has been reported (Perrigo et al. 1990; Perrigo et al. 1991) which is significantly affected by mating and post mating social experience (Elwood 1985; Kennedy and Elwood 1988). However, it still may be that the females did use their refuge to avoid paying higher costs of mating as we found that females spent more time in their refuge when housed with vasectomised males at the end of the experiment when compared to intact males or other females. This may be because once the females with intact males were pregnant, the male will stop mating with her, but as the females were not getting pregnant the vasectomised males will have continually mated with the females throughout the experiment. This continual mating and courtship behaviours may have caused these females to refuge more frequently. As sexual conflict theory states (Trivers 1972; Parker 2006), there is frequently conflicting mating optima between sexes, and the refuge use may have allowed the females more control over the frequency.

Females housed with castrated males displayed the most refuging behaviour. This was unexpected, as castrated males are less aggressive than intact males (Lofgren et al. 2012) and are known to not display most sexual behaviours (Hull and Dominguez 2007a) including mounts, intromissions and ejaculation (Hull & Dominguez, 2007; Wee, Weaver, & Clemens, 1988; Larsson, 1979). It may be that females housed with castrated males recognise no infanticide risk and no chance of mating, and thus choose to assort randomly with respect to their castrated partners, or that they actively avoid those castrated males.

Females housed with castrated males and did not have access to a refuge also had increased faecal glucocorticoid metabolites, which supports previous findings in mice (Garratt et al. 2016) and indicates a stress response. Increased stress among these females may explain why these females spent the most amount of time in a refuge, away from their partner. The reasons females avoid castrated males, and experience greater stress when with those males, remain unknown. One possibility is that it is the chemical signals in the urine of the castrated males driving the response from the females and potentially even signaling they are housed with a low genetic or phenotypic quality male. Mice use chemical signals in scent-marks to communicate information regarding social and dominance status (Jones and Nowell 82

1973, 1974; Hurst et al. 1993), territory ownership (Hurst 1987, 1989), sex, individual and kin recognition (Bowers and Alexander 1967; Hurst and Beynon 2004; Arakawa et al. 2008), reproductive and health status (Barnard and Fitzsimons 1988; Schellinck et al. 1995; Mossman and Drickamer 1996; Kavaliers et al. 2005). As castration has been shown to alter male mouse urine by reducing urinary volatile compounds, androgens, the production of lactones (potential mediators of chemical communication) (Soini et al. 2009) and major urinary proteins (Johnston et al. 2012; Guo et al. 2015), their urine and subsequent chemo signals are very different from an intact male and may cause the female to perceive the castrated male as low condition and/ or quality. Their urine will also be signaling to females that they are subordinate, as castration prevents testosterone production in male mice (Bartke et al. 1973), and testosterone relates positively to aggression and/or dominance in house mice (Compaan et al. 1993; Zielinski and Vandenbergh 1993; Matochik et al. 1994; Bronson 1996). Hence, females prefer males with stronger testosterone- controlled signals (Xiao et al. 2004) and the castrated males urine will be unattractive to them. Females may additionally avoid subordinate males as these are more likely to be usurped by a dominant male and the females will be at greater risk of losing offspring to infanticide. Low testosterone may also be signalling to the females that the castrated male may be sick or is sick or parasitized. Many parasitic diseases reduce circulating testosterone levels in male rodents (Hublart et al. 1990; Tavares et al. 1994; Hillgarth and Wingfield 1997; Willis and Poulin 2000). It could also be that the urine of castrated males may be ‘alien’ to females, failing to communicate all the attributes of either a normal male or a female, and while there is no previous research on whether this affects females, we suggest that it is possible that it may contribute to the refuging behaviour and physiological stress response we observed.

Despite the pervasive nature of sexual conflict that motivated our study we didn’t find the intact males to impose any stress, oxidative or mating costs on females. While our experimental design set up to partition the physiological costs of male presence (female control vs other groups), copulation (castrated vs vasectomized males), and insemination, pregnancy and reproduction (sham vasectomized vs vasectomized males) we found no differences between the intact and female control. 83

This therefore suggests that any costs of presence, copulation, insemination or reproduction are not observable in the measures we tested and on the scale of this experiment. Only the experimentally manipulated males (vasectomised and castrated) had an effect on the females perhaps denoting that while a normal healthy male doesn’t impose any of the costs we tested for, males that depart from this, such as those with abnormal behaviour that enhances conflict over optimal mating rates or that has low gonadal hormones representing a low-quality male, do impact females.

Our results also provide further evidence suggesting oxidative stress does not occur as a consequence of investment in reproduction (see also Garratt et al., 2011; Nussey, Pemberton, Pilkington, & Blount, 2009; Speakman & Garratt, 2014). There was no difference in oxidative damage between any of the treatments even though the females in the intact male treatment were mid lactation, the most energetically demanding time for them. The presence or absence of refuges did not have any significant effect on oxidative damage, and nor did the gonadal status of the housing partners, consistent with previous research of the same nature (Garratt et al. 2016). While it has been found that elevated glucocorticoids have a significant effect on oxidative stress (Costantini et al. 2011), we found no link between the increased FGMs in the castrated male treatment and oxidative stress. This may be because these effects tend to be stronger in longer-term experiments and perhaps if the females housed with castrated males were monitored over a longer term there may be a detectable difference in the oxidative damage in the females housed with them.

While it was necessary in our design to have reproducing and non-reproducing treatments to partition the costs of reproduction from those of male presence and mating, this also presents a possible limitation of our study. In using pregnant and non-pregnant females the differences in the measures we collected may have been driven by very different factors, as the costs and benefits of being with males vary with the female’s and male’s reproductive state. The costs to the female’s that got pregnant relate to infanticide (which may be low when housed with the paternal male) whereas the female’s housed with vasectomised males are assessing costs related to mating behaviour. 84

In conclusion, our results suggest that being housed with a castrated male has behavioural and stress effects on female mice, but no oxidative stress effects. Further studies are required to investigate the cause of this response. Potential studies could prevent contact but allow olfactory exposure to the castrated male and also have treatments of differing male quality and condition with contact to establish if these may play a role in female stress, or determine whether mating in fact reduces female stress. Our results also highlight that females would not, given the choice, seek refuge from intact males or vasectomised males to reduce costs of male presence or mating. This is consistent with other studies in Drosophila melanogaster where the absence of a spatial refuge didn’t increase male-induced harm to females (i.e. sexual conflict), even though they did find females with a refuge mated less often (Byrne et al. 2008). These results may additionally suggest that the costs or reprouction are much more subtle than has been predicted. Disentangling the cause of these results and understanding how each of the sexes may (or may not) influence each other’s life history strategies and trajectories will in turn also help us better understand mate choice strategies and the physiological costs of reproduction as well as their impact on life-history evolution.

Chapter 4

Stress effects of castrated males on females is olfactory modulated and can be ameliorated by the scent of an intact male

Gale T., Garratt M., & Brooks., RC. This chapter is currently in prep.

Abstract

Females have been found to be stressed by, and less attracted to, low quality and unhealthy males. Previously we found that females housed with castrated males had higher faecal glucocorticoid metabolites, indicating a stress response. One possibility is that the removal of gonadal hormones in these males cues females that they are housed with an unhealthy or low genetic or phenotypic quality male. In order to test whether the effects on females of being housed with castrated males are due to the absence of intact male scent, or the absence of the scent of a dominant male, we conducted a factorial experiment, housing females with either a vasectomised or castrated male or another female and supplementing each cage with scent of an unfamiliar dominant, subordinate or castrated male. We replicate the effect of castrated males on female stress levels but find that this can be ameliorated after prolonged exposure (21 days) to the scent of an intact male, either dominant or subordinate. Therefore, we conclude that the stress of being housed with a castrated male is not, alone, due to the lack of signals of dominance, but rather due to some other element missing from all intact males.

Key words: male quality, condition, stress

Introduction

In mammals, females are often burdened with the majority of the facultative physiological costs of reproduction due to the asymmetry of essential investment required during gestation and lactation by mammalian mothers compared to fathers (Clutton-Brock and Parker 1992; Tregenza and Wedell 2000; Scantlebury et al. 2002; Clutton-Brock et al. 2004; Aloise King et al. 2013). However, males can impose further costs of reproduction on females through courtship, harassment or coercion of females into mating (Clutton-Brock and Parker 1995), infanticide (Lukas and Huchard 2014) and genetic conflict during pregnancy (Moore and Haig 1991; Haig 1993; Foerster et al. 2007; Mainguy et al. 2009). These effects can have varying consequences for female behaviour, physiology and life history trajectories and highlight the importance of females exerting mate choice preferences where possible.

My collaborators and I have, in two recent studies, found that female physiology can be influenced by castrated males with female mice having elevated faecal glucocorticoid metabolites after being housed with castrated males (Chapter 3 of this thesis; Garratt, Kee, Palme, & Brooks, 2016), indicating a stress response. These females also took refuge in areas not accessible to males, thus avoiding their castrated partner (Chapter 3). This was surprising, as castrated males are less aggressive than intact males (Lofgren et al. 2012) and are known to not display most sexual behaviours (Hull and Dominguez 2007a) including scent marks, mounts, intromissions and ejaculation (Larsson 1979; Kimura and Hagiwara 1985; Wee et al. 1988; Hull and Dominguez 2007a). This is in part due to their low testosterone levels as castration prevents testosterone production in male mice (Bartke et al. 1973). The reasons females avoid castrated males, and experience greater stress when housed with those males, remain unknown, although males with low testosterone in the wild are likely to be those that are ill (Spindler 1988; Hublart et al. 1990; Lin et al. 1990; Tavares et al. 1994; Larralde et al. 1995), subordinate (McKinney and Desjardins 1973) or juvenile. Castration has also been shown to alter the molecular make up of mouse urine by reducing urinary volatile compounds, androgens, the production of lactones (potential mediators of chemical 87 communication) (Soini et al. 2009) and major urinary proteins (Johnston et al. 2012; Guo et al. 2015). This significantly alters their odour profile making their urine and subsequent chemo-signals different from an intact male, which may allow females to identify these individuals from smell alone.

Like castrated males, subordinate males also exhibit lower testosterone levels and are less aggressive than dominant males, as testosterone relates positively to aggression and dominance in house mice (Compaan et al. 1993; Zielinski and Vandenbergh 1993; Matochik et al. 1994; Bronson 1996). Furthermore, females have been found to prefer the scent of dominant males (e.g. Mossman & Drickamer 1996; Kruczek 1997; Gosling & Roberts 2001) and males with stronger testosterone-controlled signals (Xiao et al. 2004). Females may additionally avoid subordinate males as these are more likely to be usurped by a dominant male and the females will be at greater risk of losing offspring to infanticide. One possibility is that the response from the females is olfactory mediated and the castrated males’ urine could be interpreted by females as a cue that the male is of low genetic or phenotypic quality.

To assess male quality, females interpret male signals (Hurst and Beynon 2004). Many male animals signal their condition, identity and social status to conspecifics (Smith and Harper 2003). The expression of these signals are largely testosterone- dependent and honesty is regulated by the costs associated with of producing/maintaining them. It has also been suggested by Folstad and Karter (1992), that testosterone dependent traits are an honest form of male quality as testosterone impairs immune function, and only individuals of high quality can endure the costs immunosuppressive costs associated with high testosterone levels. However, support for this idea, known as the Immunocompetence Handicap Hypothesis (IHH), is currently equivocal (Roberts et al. 2004).

Many animals gain information about and assess the quality of potential mates from their odours (reviewed by Brown and Macdonald (1985)). In mice, chemical signals in male scent-marks are detected by the female’s olfactory system and communicate

88 a suite of information regarding a male’s identity and quality (Gosling and Roberts 2001). This information includes territory ownership (Hurst 1987, 1989), sex, individual and kin recognition (Bowers and Alexander 1967; Hurst and Beynon 2004; Arakawa et al. 2008), reproductive and health status (Barnard and Fitzsimons 1988; Schellinck et al. 1995; Mossman and Drickamer 1996; Kavaliers et al. 2005) and social and dominance status (Jones and Nowell 1973, 1974; Hurst et al. 1993). Dominant males are known to scent-mark more frequently (Drickamer 1995) and females prefer dominant male scent (Jones and Nowell 1974; Hayashi 1990). As the frequency of scent-marking and the preputial gland size are heritable (Horne and Ylönen 1998), and sons of high-quality males are preferred by females (Drickamer 1992), females can gain additional benefits from mating with dominant males in the form of sexy sons. However, these urinary scent-marks are costly for the males to invest in (Gosling et al. 2000; Garratt et al. 2014) which suggests that high frequency scent-marking is an 'honest' signal of quality (Zahavi 1975).

This research aimed to resolve why females paired with castrated males experience elevated stress levels by testing if this response is mediated by scent and determining if the effect on surrounding females is mimicked when they are housed with subordinate males. To do this we housed females with either a vasectomised or castrated male or another female and then supplemented these females with additional scent. The scent either came from a dominant male or subordinate male with no scent used as a control. The females housed with another female also had a fourth scent treatment that received castrated male scent. We predicted that:

1. As previously found (Gale et al. 2018), females housed with castrated males will spend less time with their partner (huddling) and will have elevated glucocorticoid metabolites. 2. Females housed with a vasectomised male or another female housing partner, but are supplied with subordinate male scent, will also spend less time with their partner (huddling) and will have elevated glucocorticoid metabolites. This would suggest that subordinate males are able to affect female physiology and behaviour similar to our previous findings with castrated males.

3. Females housed with other females but provided with castrated male scent will spend less time with their partner (huddling) and will have elevated glucocorticoid metabolites. This would indicate that the stress response of castrated males is olfactory mediated. 4. Females housed with a castrated male but supplied with dominant male scent will spend more time with their partner (huddling) and will not have elevated glucocorticoid metabolites when compared to females housed with castrated males that do not receive any supplementary odour. This would further indicate that male status influences female behaviour and physiology and that this effect is olfactory mediated.

Material and methods

This experiment was approved by the University of New South Wales Animal Care and Ethics Committee (approval: 17/58B).

Experimental design This experiment consisted of two treatments in a 2 x 4 factorial design. The first treatment factor was the type of partner females were paired with, either: a male that has undergone a vasectomy (n=20), a full castration (n=20) or another female (n=20). The second treatment factor is whether or not the females are provided with additional odour cues from another male. To examine whether the castrated males induce a stress response in the females due to the male being of low quality (the female is effectively enclosed with a male she would not normally have chosen), we supplemented one third of the females paired with each of the treatment males (vasectomised or castrated) with soiled bedding (urine) from an intact dominant male, or an intact subordinate male. The third odour treatment was a control of ‘no odour’, where the bedding inside the cage already was moved around the cage. This provides the females with the odour cues that signal a high-quality male even though physically they are paired with a low-quality male. The female control treatment had fourth odour treatment of castrated male soiled bedding. Additional bedding was added to the cages every third day. We replicated this experiment to

90 increase our sample size and completed two blocks of n=20 for each treatment group, resulting in our final sample size of n=40 for each treatment. For the second block we reused the same partner mice (vasectomised, castrated or another female C57BL/6 strain lab mouse) but used different female mice.

Animal Housing The female experimental mice were all seventh-generation captive-bred house mice (Mus musculus) originally derived from a population of wild mice trapped at a chicken farm in the Northwest of Sydney, Australia (Garratt et al. 2013a). Female house mice were weaned at 28 days of age and were housed with their female siblings until the beginning of the experiment. While we used female wild-derived captive house mice, for the male partners we used C57BL/6 strain lab mice as this study was also following on from a previous study (Gale, Brooks and Garratt, in press) where the experimental protocol required a large difference in size between males and females which these two strains provided. Therefore, in order to replicate the effect of castrated males on females from the previous study we kept the design as similar as possible. Using C57BL/6 strain lab male mice also allowed us to use genetically identical males that show the same expression levels of proteins that mediate sexual attractiveness in scent as there is limited variation in the major urinary proteins of laboratory mice (Cheetham et al 2009).

The partner mice were of the C57BL/6 strain and were purchased through the Australian BioResource Center (ABR, Mossvale, NSW, Australia). Male partners were either castrated or vasectomised when aged between 7-8 weeks old prior to being shipped to University of New South Wales (UNSW). Upon arrival they were given 2 weeks in individual cages to habituate to the new environment. The partner females were weaned at 21 days and housed with their female siblings until the beginning of the experiment. Mice were maintained on a 12:12 hour reversed light cycle with lights off at 6pm. A dim red light was used for experimental procedures, which were all undertaken in the dark phase. Mice were housed in a 315 x 180 x 125mm cage lined with corncob bedding and provided with tissues and shredded newspaper for

91 bedding. Vella Rat and Mouse Pellets and water were provided ad libitum on both sides of the connected cages.

Experimental protocol

Dominant/ subordinate odour donor trials We used 10 dyads of intact C57/B6 male mice to supply the supplementary soiled bedding. To establish a dominant and subordinate status between the dyads were allowed to interact with a female for 5mins and then were randomly paired in large cages and allowed to interact freely until one member of each dyad assumed submissive postures (Grant and Mackintosh 1963) upon each of three consecutive attacks. A metal divider containing nine small holes (5mm diameter) was then put into each cage to separate the males. This divider was removed at regular 2-day intervals, and the mice allowed to fight for a short period (maximum 10 mins) to ensure the dominant/ subordinate relationship was maintained for the rest of the experiment (see Jones and Nowell, 1974; Jones and Nowell, 1989). Two out of the ten dyads did not establish a dominant/ subordinate status and were excluded from the experiment.

Scent-marking trials Dominant adults are known to scent mark more regularly than subordinates (Drickamer 1995) to communicate their territorial and sexual status (Bronson 1979; Hurst et al. 2001) and competitive ability (Rich and Hurst 1998). After the first 10 days each of the male’s scent marking rates were assessed as an additional check to ensure dominant/ subordinate status. The scent-marking rates were assessed by placing the individual males into an empty cage lined with Benchkote for one hour a day over three consecutive days. The scent marks were measured by the average number spatially separate marks observed under UV light and the average marks over the three consecutive days was used for comparison.

Testosterone Levels To further validate the status of the male mice we also used a physiological measure of dominance by assessing the male’s testosterone levels. While mice typically use aggression to establish dominance, the relationship is not simple or direct (Klomberg et al. 2002). Many researchers, however, have found testosterone to positively correlate with dominance in mice (Compaan et al. 1993; Zielinski and Vandenbergh 1993; Matochik et al. 1994; Bronson 1996). Following the last day of the scent-marking trials a blood sample was collected from each of the male odour donor pairs. A blood sample was also collected from all of the castrated and vasectomised experimental mice to examine testosterone variation within each of the treatments.

Blood was collected in vacutainer tubes and kept on ice until centrifugation (15 min at 1000 r.p.m). Supernatant was aspirated at room temperature and plasma was stored at –80°C degrees until ELISA was performed. We used the commercial kit for testosterone (ALPCO 55-TESMS-E01) and instructions were followed as indicated by the supplier. The assays were read in an automated microplate reader (which one?) with Gen5 software (BioTek, Winooski, VT, version 2.04.11). The samples were all run on the same day with intra-assay variation 7.8-9.6%, and inter-assay variation 8.4-9.1%.

Glucocorticoid metabolites Faecal samples were collected from the females at three points throughout the experiment and tested for glucocorticoid metabolites (using methods as described by , a)Palme & Möstl, 1997) to determine their stress levels. The first-time point consisted of three samples taken from the females over three consecutive days prior to being placed into their assigned partners cage (day zero). These three samples were used to get an average of the female’s baseline glucocorticoid metabolites. The second sample was taken three days after the females had been placed into their partners cages (day three) and the third sample was taken three weeks after being placed into their partners cages (day 21). The final faecal sample marked the end of the experiment. To collect the faecal samples the females were placed into a large empty cage (565 x 387 x 203mm) made from H.D. polyethylene with a wire roof. 93

This cage was then placed topside down over another identical cage and the faecal samples were collected from the bottom cage. Females were held within these cages for a maximum of 45 minutes. Immediately after collection, faecal samples were frozen at -80 °C.

Female behaviour Behaviour was evaluated by observing each of the females using a scan –sampling technique twice a day, every day for the duration of the experiment (three weeks). Observations were conducted during the dark phase using only a head torch with red light. The first observation occurred in the second or third hour after the change from the light to dark period and the second between four hours and one hour before the change from the dark to light period. Each of the females was randomly assigned an observation order at the beginning of each observation period and observations begun ten minutes after the red light was turned on to allow them to habituate.

A single observer sequentially recorded each female’s behaviour for a total of eight observations with a five-minute gap between each of the eight recordings. As this was done morning and evening each day it totalled 16 scans a day. Behaviours recorded included; body contact (any contact including mounting, pursuing or sniffing each other) walking, huddling, sniffing substrate, immobile, grooming, and

94 eating/ drinking. These behaviours were taken from the social interaction test methods of (An et al. 2011), which we adapted for scan sampling.

Statistics All statistical analyses were performed using SPSS software package version 2.1 (IBM Corp, Armonk, NY, USA). The analyses were done with dam ID and experimental block fitted as random factors to account for non-independence of individuals originating from the same litter and time differences of each group of experimental mice unless otherwise described. For the testosterone and glucocorticoid data we also fitted plate number as a random factor to control for between plate variability. Significance was determined at p≤0.05.

Results

Establishment of dominant/ subordinate hierarchy in male odour donors

Scent-marking The mean ± s.e scent-marking frequency for the dominant odour donor males was 122.04 ± 9.14 and the subordinate odour donor males was 7.41±1.29. Repeated measures ANOVA (sphericity assumed, Mauchly's test χ2[2] = 0.95, p =0.750) showed that while there was no day x treatment effect (F1,14= 2.3, p=0.151) there was a significant difference in the frequency of scent-marks between the treatments

(F1,14= 108.0, p= <0.001) with the dominant male’s scent-marking more frequently.

Testosterone For the odour donor males (n=16) plasma testosterone levels ranged from 9.25 to 23.14 ng/ml with a mean ± s.e for the dominant males of 16.74±1.82 and the subordinate males ranged from 4-17.5 ng/ml with a mean ± s.e of 9.91±1.59. The dominant males’ testosterone levels were significantly higher than those of the subordinate (ANOVA: F1,15= 7.93, p=0.014).

Partner male’s testosterone levels The testosterone levels for the vasectomised (n=20) and castrated experimental males (n=20) ranged from 0.67 to 21.95 with a mean ± s.e for the vasectomised males of 7.41±1.7 and for the castrated males 0.38±0.03. The vasectomised males had significantly higher testosterone levels (ANOVA: F1,39= 16.90, p=<0.001).

Faecal glucocorticoid metabolites in females The faecal sample was taken from each of the females at day 21 after being placed into their partners cage. The sample sizes for the glucocorticoid metabolites are vasectomised n=40, castrated n= 40 and female control n= 40.

A two-way analysis of variance showed significant effects of housing partner (F2, 110=

5.41, p= 0.006) and partner x odour (F4, 110= 2.47, p= 0.048), but no significant main effect of odour (F3, 110= 0.90, p= 0.442). Due to the significant partner by odour treatment interaction, we explored the effects of odour within each of the partner treatments in separate one-way ANOVAS with LSD post-hoc tests. In the castrated male housing partner treatment there was an effect of odour (F2, 37= 5.01, p= 0.012) (see Figure 11), which a post hoc LSD test found was driven by differences between the odour supplemented treatments and the no odour treatment (no odour vs dom; p= 0.007, no odour vs sub; p= 0.011). This was driven by the odour supplemented groups having significantly lower levels of FGMs than those that were not exposed to odours. In the female housing partner treatment there was also an effect of odour

(F3, 36= 4.54, p= 0.008) which was driven purely by the castrated male odour (cast male odour vs all other odour treatments; all p= <0.009). Within the vasectomised male treatment there was no effect of odour (F2,40= 1.94, p= 0.157).

O d o u r

C a s tra te d m a le 2 0 0 0 0 D o m in a n t m a le

S u b o rd in a te m a le

s e

t 1 5 0 0 0

i N o o d o u r

)

a f

m 1 0 0 0 0

(

n a

e 5 0 0 0 M

0 C a s tr a te d m a le F e m a le V a s e c to m is e d m a le

P a r tn e r

Figure 11: Mean faecal glucocorticoid levels of females taken 21 days after being placed into their partners cage. Data is presented as estimated marginal means ± 1 s.e.m.

Female Behaviour Experimental manipulation had an effect on sniffing behaviour and we found that every third day when supplementary scent was applied the sniffing substrate behaviour was much more frequent. Therefore, we removed the data from the days the substrate was added (every third day) and analysed this behaviour separately. A repeated measures analysis showed that there was no effect of partner (F2, 110= 1.570, p=0.213), or partner x odour effect (F4, 110= 2.3, p=0.063) but there was an effect of odour (F3, 110= 7.5, p<0.001) with the castrated male urine treatment having significantly lower levels of sniffing behaviour compared to the other treatments (see Figure 17 in appendices). We analysed the frequency of all other behavioral variables including body contact, walking, huddling, grooming, eating/ drinking, over time in a Double Repeated Multivariate Analysis of Variance (DRMANOVA). Mauchly’s test of sphericity indicated that all behavioural variables departed from the assumption of sphericity (all p = <0.001), therefore degrees of freedom were corrected using Greenhouse-Geisser adjusted degrees of freedom.

Multivariate tests showed significant effects of day x partner, day x odour and day x partner x odour treatment interactions (Table 4). Univariate tests showed that day had a significant effect on all behaviours examined (Table 4).

Table 4: Multivariate repeated-measures analysis of variance of the effects of day, partner and odour on female behaviour.

Within-subjects effects Day Day x Partner Day x Odour Day x Partner x Odour

Wilk’s Lambda F dfn dfd P F dfn dfd P F dfn dfd P F dfn dfd P MANOVA 14.65 100 10620.3 <0.001 6.50 200 10817.4 <0.001 1.31 300 10855.2 <0.001 1.28 400 10868.5 <0.001 Body Contact 17.9 9.26 1009.6 <0.001 6.48 18.52 1009.6 <0.001 1.48 27.78 1009.6 0.053 1.65 37.05 1009.6 0.009 Walking 11.89 11.92 1300.1 <0.001 16.73 23.85 1300.1 <0.001 1.93 35.78 1300.1 0.053 1.73 47.71 1300.1 0.002 Huddling 55.43 11.46 1249.4 <0.001 14.60 22.92 1249.4 <0.001 1.45 34.38 1249.4 0.043 1.52 45.85 1249.4 0.015 Grooming 2.6 14.51 1582.5 0.001 0.93 29.03 1582.5 0.570 0.91 43.55 1582.5 0.629 0.81 58.07 1582.5 0.843 Eating/ 3.65 10.28 1121 <0.001 1.02 20.57 1121 0.435 0.94 30.85 1121 0.557 0.84 41.14 1121 0.747 Drinking Between-subjects effects Partner Odour Partner x Odour

Wilk’s Lambda F dfn dfd P F dfn dfd P F dfn dfd P MANOVA 18.01 10 210 <0.001 5.69 15 290.2 <0.001 4.73 20 349.1 <0.001 Body Contact 1.78 2 109 0.172 9.68 3 109 <0.001 7.26 4 109 <0.001 Walking 10.58 2 109 <0.001 12.3 3 109 <0.001 7.73 4 109 <0.001 1 Huddling 90.26 2 109 <0.001 4.56 3 109 0.005 4.26 4 109 0.003 Grooming 9.81 2 109 <0.001 4.58 3 109 0.005 4.93 4 109 0.001 Eating/ 0.001 2 109 0.999 0.21 3 109 0.889 0.51 4 109 0.724 Drinking * Greenhouse-Geisser adjusted df

Castrated male treatment In the DRMANOVA there was a day x partner x odour effect in huddling because females increased time spent huddling over time in the female control and vasectomised male treatments but there was no change over time in the castrated male treatment (Table 4). Using a GLMM to look at just the castrated male housing partner treatment we found an effect of odour on huddling (F2,37= 10.43, p= <0.001), and grooming (F2,37=8.82, p=0.001). A post-hoc Tukey’s test revealed that difference in huddling was between the no odour and the other two treatments (no odour x dom odour p=0.001; no odour x sub odour p=0.001) (see Figure 12). Females housed with castrated males but not supplemented with any additional scent spent less time huddling with their partners. The grooming was driven by significant differences between the subordinate male odour treatment and the other two treatments (sub x dom p= 0.001; sub x no odour p=0.010). Females housed with castrated males but supplemented with subordinate male scent spent more time grooming.

8 0

g 6 0

n 4 0

f 2 0 o

0 C a s tr a te d D o m in a n t N o o d o u r S u b o r d in a te m a le m a le m a le

O d o u r

Figure 12: Mean number of total scans (± 1 s.e.m) of huddling behaviour recorded for each odour treatment throughout the experiment (days 1-21).

100

Vasectomised treatment When analysed by treatment in an GLMM we found that there was an effect of odour on the huddling (F2,37=8.38, p= 0.001). A post-hoc Tukey’s test revealed that difference in huddling was between the dominant odour and subordinate odour treatments (p=0.001). Females housed with vasectomised males but supplemented with dominant odour huddled with their partners more frequently. There was also an effect of odour on the grooming behaviour (F2,37=4.28, p= 0.021) which a post- hoc Tukey’s test revealed was driven by a difference between the dominant and no odour treatment (p=0.019). Females not supplemented with any odour spent more time grooming than females supplemented with dominant odour. Finally, there was also an effect on body contact (F2,37=47.9, p=<0.001) behaviour of the females. For the body contact there was a significant difference between all three treatments (dom vs sub p=0.003; dom vs no odour p=<0.001; no odour vs sub p=<0.001). Females housed with vasectomised males but provided with dominant or subordinate male scent lower scores for body contact. Females supplemented with the dominant male scent had the lowest body contact scores.

Female treatment For the females housed with other females there was an effect of odour on walking

(F3,36=20.90, p=<0.001). A post-hoc Tukey’s test revealed that they were nearly all different from each other. Females supplemented with either dominant or subordinate odour were recorded to walk less than the females provide with cast male or no odour.

Discussion

Females housed with castrated males did spend less time huddling with their partner and had elevated glucocorticoid metabolites, supporting previous findings in mice (Chapter 3, (Garratt et al. 2016). Further evidence for the effect of castrated males on females came from the females who were housed with other females but provided with castrated male scent that also had higher FGM levels after 21 days. However, while we predicted that subordinate male odour would have the same

101 effect as castrated male odour our results did not support this prediction. Females housed with subordinate males did not exhibit high glucocorticoid metabolites and also contrary to our predictions both the subordinate male odour and dominant male odour mediated the effect of castrated males equally as well after a 21-day period. As the effect of castrated male odour on female stress levels can be ameliorated by the addition of scent from an intact male, this suggests that it is driven by olfactory signals. However, while the females housed with castrated males but provided with subordinate male scent had lower FGMs than the no odour treatment, they still had nearly double the FGMS when compared to females housed with a vasectomised male but provided with subordinate male odour. This adds further evidence against our prediction that subordinate males will elicit a stress response in the females.

The reasons females huddle less with castrated males, and experience greater stress when with those males, remain unknown. Although, increased stress among these females may explain why these females spent the least time huddling with (and actively avoiding) these males. Castration reduces androgens such as testosterone in males which results in them being less aggressive than intact males (Simon and Lu 2006; Lofgren et al. 2012), as well as decreasing their vocalisations (Kerchner 2004) and most sexual behaviours (Hull and Dominguez 2007a). Castrated males’ urine also does not induce the same physiological (priming) responses in females such as the Vandenbergh effect (puberty acceleration in young females) (Vandenbergh 1967), the Whitten effect (oestrus synchronisation) (Whitten 1966) and reduces males capacity to induce the Bruce effect (pregnancy block) (Bruce 1960). Therefore, the responses observed in females housed with castrated males are surprising.

One possibility is that as castrated male urine is low in testosterone and as these males also don’t show most sexual behaviours, females may be interpreting this information as the castrated male is sick or parasitized. Many parasitic diseases reduce circulating testosterone levels in male rodents (Hublart et al. 1990; Tavares et al. 1994; Hillgarth and Wingfield 1997; Willis and Poulin 2000), and as testosterone plays a key role in stimulating sexual behaviours, infected males often 102 show a lack of these behaviours such as scent marking (Zala et al. 2004), mounting, intromission, and ejaculation (Morales et al. 1996; Barthelemy et al. 2004). While we could not find any rodent studies that examined whether being housed with an infected male raised female glucocorticoid levels, there is research evidence linking higher circulating corticosterone to being paired with a poor-quality mate in the Gouldian finch (Erythrura gouldiae) (Griffith et al. 2011). Moreover, numerous studies suggest that females preferentially mate with healthy, disease-resistant males (reviewed in Beltran-Bech and Richard (2014)). The castrated males in our study could represent a poor-quality, disease and/ or a non-preferred mating partner, which would imply that the social environment and health of the surrounding males can impact female physiology. It may also be that castrated males are ‘alien’ to females, not communicating all the attributes of either a normal male or a female, or that they might be that their low testosterone scents may be communicating that the male is of low genetic quality.

While we predicted that females housed with a vasectomised male but supplied with subordinate male scent would have elevated FGM, we found that both subordinate and dominant male scent elevated FGM levels after 21 days in this treatment. Females housed with vasectomised males but provided with dominant or subordinate male scent also had lower body contact scores. Body contact included mounting, pursing or sniffing, which meant that the addition of an intact novel male’s scent reduced the frequency of these behaviours. The elevated FGMs and reduced body contact in both of these treatments that received supplementary odour may be caused by the threat of infanticide that a novel male presents. Many species have been documented to commit infanticide, which is thought to occur so the perpetrator can increase their reproductive success (Agrell et al. 1998), reproductive opportunities, access to resources, provide nutritional benefits, or ensure parental care is directed to the perpetrator’s own offspring (Ebensperger and Blumstein 2007). As in our study, just the urinary scent of an unfamiliar male presenting the threat of infanticide has been shown to have effects beyond the killing of the offspring in mice. Recently inseminated female mice (Mus musculus) exposed to the urinary scent of an unfamiliar male terminate their pregnancies following exposure to the urinary scent of an unfamiliar male, and return to oestrus 103

(Parkes and Bruce 1961). This phenomenon, now known as the ‘Bruce effect’, has been experimentally confirmed in numerous laboratory studies in house mice (Bruce 1960; Chipman and Fox 1966). Both subordinate and dominant males have the ability to induce the Bruce effect equally (Labov 1981), which could explain why females housed with vasectomised males provided with dominant and subordinate male scent both experienced elevated FGMs. We also found that females housed with vasectomised males but supplemented with dominant odour huddled with their partners more frequently, this might be a behavioural response to the infanticide threat.

We also found differences in grooming behaviour between treatments. Females housed with castrated males but supplemented with subordinate male scent spent more time grooming. As did the females housed with vasectomised males in the no odour treatment when compared to female in dominant odour treatment. Self- grooming is an innate behaviour that serves multiple purposes in mice such as hygiene maintenance, thermoregulation, social communication, de arousal and stress reduction (Fentress 1968; Spruijt et al. 1988; Spruijt et al. 1992; Denmark et al. 2010; Kalueff et al. 2015). It may also be triggered by sexual behaviour, pain, exposure to predators or novelty reviewed in (Kalueff and Tuohimaa 2004; Kalueff et al. 2007). High levels of self-grooming has been used as a measure of stress and anxiety (Barros et al. 1994; Kalueff and Tuohimaa 2005a; Leppänen et al. 2006; Estanislau 2012). Our study, however, supports research that indicates that higher stress levels doesn’t necessarily translate into an increase in grooming activity (Kalueff et al. 2004; Kalueff and Tuohimaa 2005b; Boccalon et al. 2006; Bouwknecht et al. 2007). We found no correlation between stress level and grooming behaviour, especially with the females housed with vasectomised males in the no odour treatment having one of the lowest levels of FGMs but highest levels of grooming. The lack of correlation between stress and grooming, however, may also indicate that the grooming is working to reduce stress levels.

In conclusion, our results support previous findings that being housed with a castrated male has behavioural and physiological effects on female mice. We also find that subordination does not mimic the effects of castrated males and therefore 104 indicates that something other than male status is driving this response. Further studies are required to investigate the cause of this response which we believe to be driven by olfactory signals. Our study does confirm that the signal(s) eliciting this response in females can be transferred in soiled bedding. Potential studies could examine whether infected male odour has the same stress response in females as castrated male odour to see if the health status of the male affects the glucocorticoids of the females.

105

Chapter 5

Paternally expressed Igf2 KO pups reduce metabolic costs to pregnant female mice

Gale T., Garratt M., Sandovici I., Constancia M., Youngson N., & Brooks., RC. This chapter is still ongoing with further data collection.

Abstract

Mammalian pregnancy involves conflict between the mother and the father over the amount of resources the mother should provision to the developing foetus(es). Haig’s kinship theory hypothesizes that genomic imprinting, a phenomenon where either the maternally- or the paternally-inherited allele is silenced (i.e. imprinted), originated from this conflict as some paternally-expressed genes act as foetal growth promoters and maternal genes as foetal growth inhibitors. The most well- known imprinted gene that behaves as predicted by the kinship hypothesis is Igf2, which produces a hormone (IGF2) that promotes gestational nutrient demand and thus offspring growth in mammals. We present the first-ever attempt to test the prediction of Haig’s theory that maternally imprinted growth-promoting genes like Igf2 will impose costs on future maternal reproduction. We studied the effects on female mice bearing litters with either 100% normal paternally-derived Igf2 expression (i.e. control), or with 50% of pups with the paternally-derived Igf2 allele knocked out. Females bearing Igf2 knockout pups did show reduced metabolic costs, but we did not detect any differences in body composition. Our results suggest that foetal/placental Igf2 upregulates maternal metabolism, imposing energetic demands on the mother in the manner predicted by Haig’s kinship hypothesis.

Key words: genomic imprinting, Igf2, costs of reproduction, sexual conflict

106

Introduction

Sexually antagonistic selection favours sex-specific traits that maximise the fitness of the individual even if this comes as a detriment to their mate. Adaptations that benefit one sex but harm the other often spur counter-adaptation in the other sex to defend against or mitigate the harm, potentially inciting a co-evolutionary arms races between the sexes (Holland and Rice 1998; Arnqvist and Rowe 2005; Parker 2006). As well as morphological and behavioural antagonistic adaptations, genetic conflict can also result from the diverging interests of the sexes (Holland and Rice 1998).

At some genetic loci, either the maternally- or the paternally-inherited allele is silenced; a phenomenon known as parent-of-origin genomic imprinting (DeChiara et al. 1991; Wilkins and Haig 2003). One of the strongest evolutionary hypotheses for the origins of imprinting in mammals is Haig’s kinship hypothesis (Haig 1993a, 2000, 2004), which invokes sexual conflict between maternal and paternal evolutionary interests, and specifically, between the maternally- and paternally- derived alleles in the foetus and the placenta. Few mammalian females mate with only one male in a lifetime, and, as a consequence, paternally-inherited alleles in a developing embryo or neonate have lower average kinship with the mother’s future offspring than maternally-inherited alleles do (Haig 1993a, 2014b). Therefore, alleles that enhance offspring fitness by enabling it to extract more resources from the mother than is optimal for her residual reproductive value should be imprinted (i.e. not expressed) when inherited maternally but expressed when inherited paternally. By contrast, alleles that dampen current demand on maternal investment in favour of future maternal reproduction should be paternally imprinted (i.e. paternal copy not expressed) as their expression favours the mother’s genetic interests at the expense of the father’s (Moore and Haig 1991; Haig 1993a).

Since Haig’s seminal paper proposing the kinship theory of genomic imprinting (Haig 1993a) its predictions concerning offspring growth have been upheld in a number of studies (Vrana et al. 1998; Tycko and Morison 2002; Constância et al.

107

2005). Consistent with Haig’s Kinship hypothesis, most known imprinted genes are confined to viviparous mammals (Suzuki et al. 2007) and many are expressed in the placenta, the primary site of physiological exchange between mother and foetus (Barlow 1995; Tycko and Morison 2002; Reik et al. 2003; Fowden et al. 2006). During embryonic development, maternally inherited alleles (i.e. at loci where paternal alleles are silenced) are often involved in growth suppression whereas paternally inherited alleles are involved in growth enhancement (with maternally inherited alleles silenced) (Cattanach et al. 2004; Cattanach et al. 2006). Likewise, imprinted genes are known to influence the size of the placenta, with maternally- derived alleles at loci that enhance placental growth (e.g. Peg1 and Peg3) and paternally-derived alleles that restrict placental growth (e.g. Cdkn1c, Phlda2 and Grb10) silenced by imprinting (Tycko and Morison 2002; Reik et al. 2003; Coan et al. 2005).

The most notable and widely cited example of such an imprinted gene involved in placental and foetal growth is Igf2, which is responsible for the production of Insulin-like Growth Factor II (IGF2). IGF2 plays a key role as one of the major growth factors regulating foeto-placental development (DeChiara et al. 1991; Ferguson-Smith et al. 1991; Murrell et al. 2001; Constância et al. 2002). The Igf2 gene is maternally imprinted, hence only the paternally-inherited copy is expressed in the foetus and placenta (DeChiara et al. 1991; Ferguson-Smith et al. 1991). Supporting Haig’s hypothesis of paternal expressed (and maternally imprinted) genes maximising maternal allocation and maternally expressed (paternally imprinted) genes restricting allocation, deletion of Igf2 leads to severe growth retardation of both the placenta and foetus (DeChiara et al. 1991), while over expression, either by transcriptional or post-translational mechanism, leads to placental and foetal overgrowth (Eggenschwiler et al. 1997). Furthermore, complete ablation of the Igf2 gene results in more severe growth retardation when compared to a placental specific Igf2P0+/− mutant (Constância et al. 2005) and during maternal under nutrition Igf2 KO mutants are unable to adapt the placenta to enhance amino acid supply like WT controls could (Sferruzzi-Perri et al. 2011).

108

On the same chromosome as Igf2 in mice is the H19 gene, which regulates the expression of Igf2 (Forné et al. 1997; Wilkin et al. 2000) by suppressing the receptor through which IGF2 signals (Furutachi et al. 2013). This suppression prevents excessive placental invasion and restricts the foetus from growing too rapidly. Consistent with the kinship hypothesis’ predictions, H19 is imprinted in the opposite way to Igf2 (Bartolomei et al. 1991; Zemel et al. 1992; Thorvaldsen et al. 1998), with the paternally-derived copy in the foetus and placenta silenced, and the maternally-derived copy expressed.

Despite a wealth of evidence that imprinted alleles affect placentation and offspring growth as predicted by Haig’s kinship hypothesis, the consequences of parent-of- origin imprinting for maternal residual reproductive value (i.e. future survival and reproduction) remain unexamined. Such consequences are often assumed to exist because the costs of reproductive investment provide the cornerstone of life-history theory (Stearns 1992; Nilsson and Svensson 1996; Harshman and Zera 2007). And yet the costs of reproduction, including pregnancy and lactation, are often far smaller and more difficult to detect than expected from a simple reading of life history theory. A reduction in maternal residual reproductive value as a result of paternal gene expression is an explicit prediction of Haig’s kinship hypothesis, explaining why imprinting of the maternal allele of genes like Igf2 has evolved.

Here we test the prediction that female house mice bearing some pups in which the paternally-derived copy of Igf2 has been knocked out will pay smaller costs of reproduction than females not carrying knockout pups due to lower placental and foetal demand for nutrients in the knockout pups (see Figure 11).

Materials and Methods

This experiment was approved by the University of New South Wales Animal Care and Ethics Committee (approval: 15/128B) and was performed in accordance with relevant guidelines and regulations at UNSW Australia.

109

Mice generation and breeding This study exploits the Cre/loxP knockout system. Cre (cyclization recombination) recombinase is an enzyme that recombines a pair of short target sequences called the loxP sites (locus of crossing over of bacteriophage P1). Cre-mediated recombination between two loxP sites results in deletion of the DNA and doesn’t require the insertion of any extra supporting proteins or sequences.

Igf2 knock out mice were generated by the Cre/loxP recombination system to delete the Igf2 gene. Flanking loxP sites were introduced, followed by recombination between the loxP sites induced by transient expression of Cre. Meox2tm1(cre)Sor (referred to as Meox2Cre) expresses Cre recombinase in epiblast-derived tissues as early as embryonic day 5 (Tallquist and Soriano 2000).

Heterozygous Igf2 floxed (Igf2 fl/+) and heterozygous Meox2Cre (Meox2Cre/+) mice were imported from Prof Miguel Constancia at Cambridge University. The generation of these mice has been previously described (Murrell et al. 2001; Ferrón et al. 2015). The lines had to be rederived upon entry to Australia via In-vitro fertilization with sperm harvested from a male. Both lines were kept on C57BL6 backgrounds as we built up numbers to commence the experiment. The dams in our study were crossed to create a 129S6/SvEv x C57BL/6 hybrid background as Igf2 knockout mice on a C57BL6 background are not viable and die before birth (see Han et al. (2010). Igf2 KO have a high survival frequency in inbred strain 129/SvEv (DeChiara et al. 1990; DeChiara et al. 1991).

Genotyping Igf2 and Meox2Cre mutant mice were sent to the Garvan Institute of Medical Research to be genotyped by PCR analysis of DNA extracted from mouse from tail tips with the following primers; Igf2 floxed strain was genotyped using three primers to identify WT allele (II-delF: 5’-TTACAGTTCAAAGCCACCACG-3’), the targeted allele (II-delRW: 5’-GCCAAAGAGATGAGAAGCACC-3’) and the deleted allele (II-delRD: 5’- GCCAAACACAGTAAAAAGAAATGC-3’). For the Meox2Cre strain the three primers used for genotyping are: Meox2-F: 5’-GGACCACCTTCTTTTGGCTTC-3’ Meox2-R: 5’-AAGATGTGGAGAGTACGGGGTAG-3’ 110

Meox2-Cre: 5’-CAGATCCTCCTCAGAAATCAGC-3’

Animal Housing The mice were kept under controlled conditions housed in open top cages (48 × 11.5 × 12 cm) in the same room with a 12-h light/dark cycle (0700/1900 h) with ad libitum access food and water throughout. Temperature (22-26ºC) and humidity (40-70%) were regulated and diet was standardised chow purchased from Gordon’s Specialty Stockfeeds Pty Ltd, NSW, AU.

Experimental Protocol Our experimental treatment was a homozygote Igf2fl/fl sire mated to a Meox2Cre/+ dam on a 129S6/SvEv x C57BL/6 hybrid background. This cross knocks out the paternally-inherited Igf2 allele in 50% of pups, and those KO offspring grow slowly (DeChiara et al. 1990). We then had three control treatments. Control 1 crossed a C57BL/6 WT sire mated to a Cre-strain hybrid +/- dam. The pup phenotype for Control 1 is Igf2 wild-type (i.e. no knockout, paternally-inherited Igf2 expressed in all pups). Control 2 paired a vasectomised male with a WT (129S6/SvEv x C57BL/6) hybrid dam and Control 3 was a WT (C57BL/6) female/ WT (129S6/SvEv x C57BL/6) hybrid female treatment both of which did not produce pups (see Table 5). Control 2 was used to partition the costs associated with male presence and insemination from those of reproduction in the measurements collected and Control 3 was used as a measure of the standard female.

The experiment was completed in a series of blocks, each block containing mice from all of the treatments. Female mice were first weighed and split into individual cages, this marked day 0. On day 3 females were then placed in the echoMRI to assess body composition and subsequently the CLAMS (Comprehensive lab animal monitoring system) to assess whole body metabolism for 18hrs. On day 7 the assigned sires were then added to the cages. Females were checked for mating plugs daily, gestational day 1 (G1) was marked as soon as a mating plug was sighted. On gestational day 15 females were weighed and placed back into the CLAMS. Since we wanted to avoid any infanticide that male mice might cause to newborn young in the

111

mated treatments, we removed the males when those females housed with intact males were heavily pregnant (G15).

On gestational day 18 females were weighed again and placed into the echoMRI and then dissected. The embryos were weighed and genotyped, and the placenta, liver and retroperitoneal fat pads weighed.

Table 5: Treatment groups with genotype and background strain of individuals

Treatment Sire Dam Offspring Name Experimental Igf2 fl/fl Sire Meox2Cre/+ i.e. knockout of paternally- Treatment (C57BL/6) (129S6/SvEv x inherited Igf2 in 50% of pups, C57BL/6) offspring should grow slowly. Normal growth in other 50% of pups. Control 1 Igf2 +/+ Sire Meox2Cre/+ No knockout, wild type (C57BL/6) (129S6/SvEv x phenotype expressed C57BL/6) Control 2 Vasectomised male WT +/+ Non-reproductive control WT (C57BL/6) (129S6/SvEv x C57BL/6) Control 3 Female WT +/+ WT +/+ Non-reproductive control (C57BL/6) (129S6/SvEv x C57BL/6)

112

2 5 0

2 0 0

c u

1 5 0 d

1 0 0 f

s o

5 0 C

0 C o n tr o l 1 E x p e r im e n ta l C o n tr o l 2 C o n tr o l 3 T r e a tm e n t

T r e a tm e n t

Figure 13: This shows the predictions of our experiment. Igf2 knockout is expected to reduce the physiological costs of pregnancy relative to a full pregnancy with intact Igf2 (i.e. Control 1). Controls 2 and 3 allow us to partition the costs of reproduction producing Igf2 knockout pups from the background costs of male presence and insemination and a non-reproducing/ unmated female.

Whole body metabolism Whole body metabolism was assessed using Columbus Laboratory Animal Monitoring System (CLAMS/Oxymax). Mice were housed individually during this procedure with ad libitum access to food and water. Measurements began at 09:00 AM and continued for 18 h. The first 7 h of the measurement period was an acclimatisation period and the last 11 h were used for analysis, similarly to previously published methods (Garratt et al. 2014).

The calorimeter was calibrated before the experiment with a standard span gas (0.504% CO2, 20.43% O2 balanced with N2) and cross-calibrated with room air. Data were collected every 13 min over this 12-h period. The metabolic

measurements included the volume of carbon dioxide produced (VCO2), the volume

113 of oxygen consumed (VO2), and RER (VCO2/VO2). The data are represented as the mean values over each 5-h period.

Body Composition Body composition was analysed using the EchoMRI-900 Body Composition Analyzer (EchoMRI, Biological Resources Imaging Laboratory, Mark Wainwright Analytical Centre, UNSW, Sydney, Australia) following manufacturer’s instructions, at days 0, G18 and L14. The mass each mouse was related to the reported mass components (Fat, Lean, Free Water, Total Water) with the percentage body fat calculated as ([body fat mass]/ [body fat mass + lean mass]) × 100.

Statistics All statistical analyses were performed using SPSS software package version 2.1 (IBM Corp, Armonk, NY, USA). The analyses were done with experimental block fitted as random factors to account for non-independence of litter and block time unless otherwise described. Significance was determined at p≤0.05.

Results

Physiological consequences of Igf2 during gestation day 18

Mothers A univariate analysis of females culled at gestation day 18 showed significant differences in liver weight (F3, 12= 15.39, p= <0.001) and retroperitoneal fat pad mass (F3, 12= 3.55, p= 0.048) (see Table 6). A post hoc LSD test on both of these measures revealed that these differences occurred between the mated treatments (Experimental treatment and Control 1) compared to the unmated treatments (Control 2 and Control 3) (all p= <0.034) showing both of these effects were a result of pregnancy not treatment. There also was no significant difference in litter size (F1,

9= 2.29, p= 0.164) or the number of reabsorbed foetuses F1,9= 0.43, p= 0.527) at G18. There was only one miscarriage observed in the Experimental treatment.

114

Embryos A univariate analysis on the embryo and placenta weight found no difference in placental weight between the Experimental treatment and Control 1 (F1, 76= 0.25, p

=0.614) or of foetal sex (F1, 76= 0.08, p= 0.770) but did find a treatment effect on embryo weight (F1, 54= 16.33, p= <0.001). Embryos from the Control 1 weighed significantly more (1.25 g ± 0.01) than the Experimental treatment (0.94 g ± 0.04) (see Figure 14, Table 6). Genotype also had a significant effect on embryo weight in the Experimental treatment (F1, 76= 10.61, p= <0.001). Igf2 KO embryos weighed less than WT embryos. This was expected as Igf2 knockout pups are known to weigh less at birth than wild-type pups (DeChiara et al. 1991).

1 .3

1 .2

)

g (

1 .1

g i

e 1 .0 W

0 .9

0 .8 E x p e r im e n ta l C o n tr o l 1 T r e a tm e n t

T r e a tm e n t

Figure 14: Weights of dissected embryos at gestation day 18. Data is presented as estimated marginal means ± 1 s.e.m.

Whole body metabolism (CLAMS) Normally we would use a Repeated Measures ANOVA to infer changes within- subjects and between-subjects effects. Due to several equipment breakdowns, however, we were lost some measures for some individuals, and it would have

115 surrendered much of the power of our study to eliminate all individuals with one or more missing measure. We therefore present separate analyses for each time interval.

Pre-experiment A univariate analysis of the female’s metabolism before the experiment began showed that there was no difference in the total oxygen consumed (F3,52=0.47, p=

0.703), respiratory exchange ratio (F3, 52= 0.39, p= 0.756), amount of food consumed

(F3, 52= 0.41, p= 0.741) or their weights (F3, 52= 0.10, p= 0.959).

Gestation

Light Period We split the data up into the light and dark periods over which time the females were in the CLAMS. The light period included was from 7am-9am at which time the females should be less active, while the dark period was from 9pm to 5am during which the females should have been most active, as mice are nocturnal. First, we used a univariate analysis to see if there were any differences in weight between the treatments at gestation. There was also an effect of weight (F3, 26= 10.11, p=<0.001) due to the difference between the reproducing and non-reproducing treatments. This effect was expected as the females gained weight associated with pregnancy. A post-hoc LSD test showed that there was no difference in weight during gestation between the two reproduction treatments (p= 0.671).

During gestation there was an effect of treatment on the total oxygen consumed (F3,

26= 6.31, p= 0.002) which a post hoc LSD test showed was due to the Control 1 consuming more oxygen than the three other treatments (all p=<0.006) (see Figure 13). To further examine this result, we ran an ANCOVA to control for body weight. Total oxygen consumption differed between the groups after correcting for body weight and a planned contrast showed that Igf2 KO and reproducing females still differed in oxygen consumption after correcting for body weight (F3,23= 3.51, p=

0.031). There was no effect of treatment on respiratory exchange ratio (F3, 26= 0.86, p= 0.474) or the amount of food consumed (F3, 26= 1.17, p= 0.340). 116

Dark period

During gestation there was no effect of treatment on respiratory exchange ratio (F3,

26= 1.18, p= 0.336) or the amount of food consumed (F3, 26= 2.27, p= 0.104) but, as for the light period there was an effect of total oxygen consumed (F3, 26= 5.35, p= 0.005) (see Figure 15). A post hoc LSD test revealed this effect was driven by differences between the two non-reproductive controls and the mated control (both p=<0.005) and by a difference between the experimental treatment and Control 3 (p=0.047). I.e. there was no difference between the experimental treatment and

Control 1. There was also a difference in weight (F3, 26= 10.11, p= <0.001) again between the reproducing and non-reproducing treatments again due to pregnancy both (p=<0.008).

0 .4 5

0 .4 0 L ig h t P e rio d

) n

i D a rk P e rio d m

/ 0 .3 5

(

2 0 .3 0

O V

0 .2 5

0 .2 0 C o n tr o l 1 E x p e r im e n ta l C o n tr o l 2 C o n tr o l 3 T r e a tm e n t

T r e a tm e n t

Figure 15: Total amount of oxygen consumed during the light phase (7am-9am) and dark phase (9pm-5am) during gestation day 15. Data is presented as estimated marginal means ± 1 s.e.m. This is the time when the females had been in the CLAMS the longest and therefore more accurate as the females have had time to acclimatise.

Body Composition (echoMRI) Pre-experiment, using a univariate analysis of variance, we found no difference in female body fat percentage (F3,51=2.17, p= 0.102). At gestational day 18, however, 117 we did observe a difference between treatments in body fat percentage (F3,23= 3.74, p= 0.025) (see Figure 16). A post-hoc LSD test revealed this effect was driven by differences between the mated treatments and the female-female control (treatments; Experimental treatment x Control 3 p= 0.003, Control 1 x Control 3 p= 0.021), with the mated treatments having more body fat than the unmated female control (see Table 6).

2 5

2 0

t 1 5

d 1 0

o B

0 C o n tr o l 1 E x p e r im e n ta l C o n tr o l 2 C o n tr o l 3 T r e a tm e n t

T r e a tm e n t

Figure 16: Body fat percentage of females at gestation day 18. Body fat % was calculated as ([body fat mass]/ [body fat mass + lean mass]) × 100. Data is presented as estimated marginal means ± 1 s.e.m.

118

Table 6. Body composition of females at gestation day 18. Data is displayed as marginal means ± 1 s.e.m.

Treatment Control 1 Experimental Control 2 Control 3 Treatment Body fat % 19.59 ± 1.85 21.27 ± 1.75 11.91 ± 1.31 17.86 ± 2.4 Fat mass 6.14 ± 0.73 6.64 ± 0.56 2.74 ± 0.35 4.81 ± 0.91 Lean mass 25.25 ± 2.05 24.76 ± 1.48 20.12 ± 0.67 20.95 ± 0.69 Total water 23.83 ± 2.54 23.69 ± 1.91 17.56 ± 0.67 18.42 ± 0.74 Total RP fat pad 0.12 ± 0.01 0.13 ± 0.02 0.02 ± 0.01 0.04 ± 0.01 mass Liver 1.43 ± 0.03 1.53 ± 0.07 0.92 ± 0.03 1.05 ± 0.12 Weight 34.60 ± 2.63 34.90 ± 2.04 24.32 ± 0.99 26.98 ± 1.52 Litter size 4.66 ± 1.66 8 ± 1.76 - -

Discussion

This is the first attempt, as far as we are aware, to test the prediction, derived from Haig’s (2000) kinship hypothesis for genomic imprinting, that the expression of paternally-derived alleles for which the maternally-derived counterpart is imprinted will impose costs on the mother. Only one other paper has examined mothers carrying paternally expressed Igf2 KO embryos, in which they aimed to investigated inter-brood competition and in utero growth. Charalambous et al. (2003) found that females mated to Igf2 KO males invested 20% less into their first brood, and females that invested little in their first brood also invested little into their second despite being subsequently mated to a WT male. Despite a modest sample size, our results provide promising evidence of physiological costs of paternal Igf2 expression and suggest that attempts to measure longer-term fitness costs are worthwhile. Our results show that mothers carrying Igf2 KO offspring have reduced metabolic costs during gestation. This lends support for Haig’s kinship theory of genomic imprinting and may provide further support for the role of Igf2 in pre-eclampsia pathogenesis.

119

Female mammals undergo major physiological and metabolic adaptations during pregnancy to provide the essential needs for the developing embryo and prepare for lactation. The metabolic adaptations of pregnancy are principally mediated by hormones produced by the placenta and maternal pituitary gland which are drastically transformed during gestation (Freemark 2006). The elevation of metabolic rate during pregnancy is a significant energetic cost of reproduction and is considered a result of increased oxygen consumption due to the increased demand for maternal circulation, respiration, tissue mass and renal function (Hytten 1980). Our results show that females who carry some embryos with the paternally derived Igf2 allele knocked out have lower metabolic costs compared to females not carrying knockout pups. This demonstration that Igf2 expression elevates, either directly or indirectly, the physiological and metabolic adaptations that occur during pregnancy serves as a first confirmation of Haig’s kinship hypothesis’ prediction that maternal costs are an important locus of the conflict that drives the evolution of genomic imprinting. In the kinship theory, it is predicted that paternally expressed genes will act as growth promoters increasing an offspring’s share of maternal resources at a cost to maternal residual reproductive value, (Haig 1993a, 2000; Van Cleve and Feldman 2007). In our study the paternally expressed Igf2 is potentially working to elevate the females metabolic rate so the foetus can extract more resources from the mother. In humans, foetal weight has been found to correlate with cumulative increase in BMR during the second half of pregnancy but not during the first half (Forsum et al. 1988) and larger foetal weight tends to result in an increase in basal metabolic rate and lower rates of maternal energy storage (King 2000).

An increase in metabolic rate during gestation may also reduce the mother’s future reproductive success and/or lifespan by increasing the costs of reproduction. A fundamental concept in our understanding of life history evolution is that trade-offs exist between the various life-history traits (Fisher 1930; Stearns 1992). Trade-offs are thought to occur due to constraints in the allocation of an organism’s finite resources (Kirkwood and Holliday 1979) where an increase in the investment to one life history trait is linked to a decrease in another life history trait. For example, investment in reproduction could mean less resource allocation to somatic 120 maintenance and future reproductive success (Reznick 1985; Stearns 1992; Kirkwood and Austad 2000).

The reduced total oxygen consumption we observed in our Igf2 KO treatment may also support recent research in humans which have found evidence that Igf2 may be involved in the placental supply of maternal nutrients and may participate in pre- eclampsia pathogenesis (He et al. 2013; Guo et al. 2017). Pre-eclampsia is a human specific condition in females characterised by hypertension, proteinuria, and a poorly vascularised (invaded) placenta which can lead to perinatal morbidity and mortality (Kobayashi 2015). Two phases define pre-eclampsia (Winn et al. 2011). First, a shallow invasion of an abnormally developed placenta into the maternal tissues occurs early in gestation, causing the foetus and placenta to begin running short of oxygen. Second, the maternal response to a foetus struggling to obtain more oxygen from the mother’s blood leads to the clinical manifestations of pre- eclampsia. The Igf2 KO females in our study had significantly less total oxygen consumption, which would result in the foetus having less oxygen available- i.e the later stage of pre-eclampsia in humans. In murine models, Lai et al. (2011), found that a preeclampsia-like disease in mice was triggered by oxygen levels to the foetus falling under a certain threshold, and Rueda-Clausen et al. (2014) found that exposure foetal oxygen deficiency during gestation increased the both the severity of preeclampic-like phenotpye and growth restriction. Furthermore, pre-eclampsia is strongly associated with intrauterine growth restriction and placental abnormalities in humans (Silasi et al. 2010; Tallarek and Stepan 2012; Kwiatkowski et al. 2017), which are also found in Igf2 KO mouse pups.

Despite much research, the exact pathophysiology of pre-eclampsia is yet to be elucidated but it is thought to be due to placental malperfusion resulting from poorly remodelled maternal arteries (Staff et al. 2013; Chaiworapongsa et al. 2014). Currently, Haig’s kinship theory is one of the most promising hypotheses behind the evolution and persistence of this condition (Hollegaard et al. 2013; Christians et al. 2017). Evidence for this hypothesis comes from the fact that it is the paternal complement of genes that is primarily responsible for constructing the placenta, and at least three maternally inherited genes expressed in the foetus have been 121 identified that alter the chances of pre-eclampsia. The association between pre- eclampsia and incomplete placental invasion as well as research showing that females who have elevated blood pressure during pregnancy have lower rates of spontaneous abortion (Haig 1993a) also point towards a conflict causality. Although highly speculative, as our results suggest that paternally expressed imprinted Igf2 plays a role in elevating female metabolism and oxygen consumption during gestation in mice it may also add indirect evidence for conflict surrounding the cause of pre-eclampsia. While the Igf2 KO’s in our study showed significant decrease in foetal weight compared to the WT controls, mimicking previous results (DeChiara et al. 1990; Constância et al. 2002; Fowden et al. 2006), we didn’t find any decrease in placental weight, in contrast to recent research (Sibley et al. 2004; Guo et al. 2008). It is important to note that our results could be affected by the genetic back ground of the mice, which is known to influence phenotypes of genetically manipulated mice (Doetschman 2009), the lack of power in our study may have prevented us from detecting this effect.

Pregnancy is well documented to result in changes in body composition including increased weight gain, fat mass and fat free mass and basal metabolic rate which intensify in late gestation (Barnett and Widdowson 1971; Forsum et al. 1988; Goldberg et al. 1993; Lof et al. 2005; Bhardwaj et al. 2013). Our results concur with the current literature as we observed a collective increase in weight, body fat, liver weight, and metabolic rate in the pregnant treatments, compared to the non- pregnant treatments which are all due to necessary adaptations to sustain successful pregnancy and lactation. Interestingly, we also found females housed with vasectomised males had higher body fat percentages compared to females that were housed with other females. This fat gain may be attributed to the induction of pseudopregnancy, since it is known that females mated with vasectomised males can become pseudo-pregnant resulting in a temporary increase in body weight (approximately 5–10%)(Dewar 1957; Dewar 1959) . While early research showed that this effect is transient and generally females return to their original weight about 20 days after mating (Dewar 1957), other studies have found this weight gain to persist long term (Garratt et al. 2016). Here our study not only shows that females mated to vasectomised males exhibit weight gain, but that this gain is due to 122 an increase in fat deposits, which has not been shown before. However, rodent studies have found pseudopregnancy causes hyperphagia (Augustine and Grattan 2008) which would likely result in fat gain.

In conclusion, we do find support for our prediction that females mated to male Igf2 KO homozygotes would pay smaller costs of reproduction (than females mated to wild-type controls) due to lower placental and foetal (Igf2 KO) demand for nutrients. Females carrying Igf2 KO had reduced metabolic costs, however, we did not detect any differences in body composition between the mated treatments. Therefore, we believe that this study provides promising evidence that Igf2 is involved in regulating female metabolism during gestation and may play a role in human pregnancy related diseases such as preeclampsia. These results provide further evidence supporting Haig’s hypothesis of paternal expressed genes maximising maternal allocation and maternally expressed genes restricting allocation. Future studies could also look at the reciprocally imprinted maternally expressed gene H19 to determine whether the opposite effect is induced and test for any effects to the females life history as a consequence of the increased reproductive costs during gestation, such as future reproductive success/lifespan.

123

Chapter 6

Conclusions

Sexual conflict occurs when the evolutionary interests of the two sexes diverge and each sex is selected to maximise their own fitness, even if this comes at the expense of their mate. This conflict, and the physical necessity of close interaction required for mating, can intensify the already costly nature of reproduction for female mammals (reviewed in Chapter 1). Despite the costs of reproduction being arguably more intense in complex, longer-lived species such as most mammals, most of the experimental research on sexual conflict has focused on invertebrate organisms. The easiest model organism we can use to investigate sexual conflict in mammals is the mouse. Mice have relatively short lifespans, mature early in life, reproduce quickly and have been extensively researched making much of their genetics, physiology and behaviour already well documented.

During this research we have utilised both laboratory and wild-derived mice, which has allowed me to take advantage of genetic modification technology and manipulate treatments by varying the gonadal status of males with routine surgeries. In this thesis we show that sexually antagonistic adaptations arising from sexual conflict between the sexes are intricate and varied, both behaviourally and genetically, while the costs of reproduction are much subtler than has been predicted.

The costs of reproduction

The pivotal trade-off around which life-histories are thought to evolve is the costs associated with reproduction (Stearns 1992; Nilsson and Svensson 1996; Harshman and Zera 2007). While investment in reproduction generates fitness benefits (viable

124 offspring), it also entails costs as well. Yet while the costs of reproduction are probably the most significant component underlying life-history trade-offs, our understanding of them and their underlying proximate causes is still very undeveloped (reviewed in Speakman 2008) and a greater understanding of life- history evolution will only transpire once these costs are better understood (Speakman 2008).

In mice, as in most mammals, lactation is the most energetically demanding episode of reproduction (Speakman 2008) and is a key time to modify investment in relation to contextual conditions. Females have evolved various strategies to optimise their maternal investment and reduce the costs of reproduction when faced with the prospect of infanticide. Infanticide represents one of the more extreme forms of sexual conflict that results in a significant loss (and waste) of maternal investment for females. Previously, we found that exposing female mice to the scent of an unfamiliar male in late pregnancy resulted in smaller offspring at weaning (Gale et al. 2013), while the cause was unknown, which we speculated could have been due to the females strategically altering their investment in relation to their perceived chances of offspring’s survival.

In Chapter 2, we replicated this manipulation but additionally examined the females’ behaviour during lactation and measured her stress levels to determine the cause of the reduced weaning weight. The females exposed to the unfamiliar male’s scent spent less time nursing their offspring throughout lactation, suggesting that the reduced weaning weight observed in the offspring was due to a change in maternal behaviour throughout weaning. As male mice are highly infanticidal, these results support the prediction that females adjust post-partum investment in relation to pups’ perceived chances of survival, potentially benefitting maternal lifetime fitness at the expense of the current litter of offspring (Marshall and Uller 2007).

Post weaning (once the offspring could feed themselves) the offspring exhibited catch-up growth and reached a normal weight by adulthood. While we did not find any costs to the male offspring’s sexual fitness or female offspring’s subsequent 125 reproduction, however, they were found to bear oxidative damage in adulthood, revealing long-term effects on offspring condition. Here we concluded that when females are unable to exhibit early loss-cutting counter-strategies to infanticide (i.e the Bruce effect), female mice strategically alter their investment in lactation in relation to the likelihood of infanticide, to prevent wasting investment in ‘doomed offspring’. Although this results in long -term fitness costs to their offspring which we currently quantify as oxidative stress.

In Chapter 3, we aimed to further investigate the costs of reproduction by using an experimental design that partitioned the physiological costs of male presence (female control vs other groups), copulation (castrated vs vasectomized males), and insemination, pregnancy and reproduction (sham vasectomized vs vasectomized males). In addition, we sought to study female refuge-seeking responses to these costs by offering half the females in each treatment a refuge where they could escape from their allocated partner. Despite the pervasive nature of sexual conflict that motivated our study we did not find that intact males imposed any stress, oxidative or mating costs on females. This, therefore suggests that any costs of presence, copulation, insemination or reproduction are not observable in the measures we tested and on the scale of this experiment, or that they may be much subtler than originally predicted. Only the experimentally manipulated males (vasectomised and castrated) influenced the females perhaps denoting that while a normal healthy male doesn’t impose any of the costs we tested for, males that depart from this, such as those with abnormal behaviour that enhances conflict over optimal mating rates or that has low gonadal hormones representing a low-quality male, do impact females. Our results also highlight that females would not, given the choice, seek refuge from intact males or vasectomised males to reduce costs of male presence or mating. This is consistent with other studies in Drosophila melanogaster where the absence of a spatial refuge didn’t increase male-induced harm to females (i.e. sexual conflict), even though they did find females with a refuge mated less often (Byrne et al. 2008).

As both of my projects on this ‘late Bruce’ phenomenon were short-term, future research could use the same methodology to partition the costs of reproduction 126 from those of male presence, insemination and mating but instead conduct a long- term study to examine any effects to female lifespan or life time reproductive success. A fundamental understanding in the evolution of life-history strategies is the concept that trade-offs exist between the various life-history components (Fisher 1930). The trade-offs are thought to occur due to constraints in the allocation of an organism’s finite resources (Kirkwood and Holliday 1979) where an increase in the investment to one life history trait is tied to a decrease in another life history trait. Therefore, investment in reproduction means less resource allocation to other life-history traits including survival and future reproductive success (Reznick 1985; Stearns 1992; Kirkwood and Austad 2000).

The costs of catch up growth

Accelerating growth requires an increase in metabolic activity that can damage the organism (Morgan et al. 2000). One such cost of accelerated growth, oxidative stress, has been documented in zebra finches (Alonso‐Alvarez et al. 2007) and damsel flies (De Block and Stoks 2008a). Oxidative stress results from an imbalance between the production of harmful reactive oxygen species (ROS) and an organism’s ability to mitigate and detoxify the damaging effects (Monaghan et al. 2009). Failure to moderate this balance can result in oxidative damage to key biological molecules such as DNA, proteins and lipids (Monaghan et al. 2009) and can limit investment in other life history stages (Costantini 2008). Mangel and Munch (2005) propose that growth leads to an accumulation of damage at the cellular level that is expressed at the level of the organism and is an important cost of compensatory growth. We tested for damage on a cellular level in form of oxidative damage and oxidative stress. We found evidence for oxidative damage in the livers and hearts of offspring from the novel male exposure treatment, highlighting that an olfactory change in a pregnant mother’s environment can elicit a variety of maternal and offspring responses, ultimately influencing offspring physiological condition in adulthood. This may be a consequence of catch-growth but could also be a consequence of odour exposure itself, or changes in maternal allocation in response to this. In future studies it may be of interest to limit an offspring’s ability to show compensatory

127 growth (through a nutritional or genetic manipulation) and test whether oxidative damage in offspring still occurs in adulthood.

Odour signals/ cues

While we did not find any significant costs of reproduction in Chapter 3, surprisingly we did find that the females housed with castrated males were found to not only spend significantly more time in their refuge, but had elevated faecal glucocorticoid metabolites, indicating a stress response. This was surprising, as castrated males are less aggressive than intact males (Lofgren et al. 2012) and are known to not display most sexual behaviours (Hull and Dominguez 2007b) including scent marks, mounts, intromissions and ejaculation (Larsson 1979; Kimura and Hagiwara 1985; Wee et al. 1988; Hull and Dominguez 2007b). This is in part due to their low testosterone levels, as castration prevents testosterone production in male mice (Bartke et al. 1973). The reasons females avoid castrated males, and experience greater stress when housed with those males, remain unknown, although males with low testosterone in the wild are likely to be those that are ill (Spindler 1988; Hublart et al. 1990; Lin et al. 1990; Tavares et al. 1994; Larralde et al. 1995), subordinate (McKinney and Desjardins 1973) or juvenile.

Castration has also been shown to alter the molecular make up of mouse urine by reducing urinary volatile compounds, androgens, the production of lactones (potential mediators of chemical communication) (Soini et al. 2009) and major urinary proteins (Johnston et al. 2012; Guo et al. 2015). As females have been found to be stressed by and are less attracted to low quality and unhealthy males, we speculated that the removal of gonadal hormones in these males may cue females to interpret that they are housed with a low genetic or phenotypic quality male through their altered urine profile. In order to test whether the observed response from females of being housed with castrated males was olfactory mediated and related to male quality, we housed females with either a vasectomised male, castrated male or another female, and supplemented each cage with scent of an unfamiliar dominant, subordinate or castrated male (Chapter 4). As previously found, females housed with castrated males spent less time in contact with their 128 partner (observed through huddling behaviour) and had elevated glucocorticoid metabolites. However, this effect can be ameliorated after prolonged exposure (21 days) to the scent of an intact male, either dominant or subordinate. This shows that subordinate male odour does not induce the response in females as castrated male odour and therefore indicates that something other than male status is driving this response. Therefore, we concluded that the stress of being housed with a castrated male is not, alone, due to the lack of signals of dominance, but rather due to some other missing element that is present in all intact males. And that the signal(s) eliciting this response in females can be transferred in soiled bedding.

Further studies are required to investigate the cause of this response which we believe to be driven by olfactory signals. Potential studies could examine whether infected male odour induces the same stress response in females as castrated male odour does, in order to see if the health status of the male affects the glucocorticoids of the females. Our results also suggest that the threat of infanticide presented by just the scent of an unfamiliar male can elevate glucocorticoid metabolites in females.

On the origin of genomic imprinting

While there are many examples of physical and behavioural antagonistic adaptations, genetic conflict can also result from the diverging interests of the sexes (Holland and Rice 1998). This genetic conflict can result in monoallelic expression of genes at some loci, a phenomenon known as parent-of- origin genomic imprinting (DeChiara et al. 1991; Wilkins and Haig 2003). One of the strongest evolutionary hypotheses for the origins of imprinting Haig’s kinship hypothesis (Haig 1993b, 2000, 2004), invokes sexual conflict between maternal and paternal evolutionary interests within the developing foetus and placenta. In Chapter 5 we used genetic knockouts of the most notable and widely cited example of an imprinted gene that behaves as predicted by the kinship hypothesis, Igf2. We compared these knockouts to control foetuses in order to test the prediction that paternally expressed genes such as Igf1 reduce maternal reproductive success, Igf2 is one of the major growth factors regulating feto-placental development in mammals (DeChiara et al. 1991; 129

Ferguson-Smith et al. 1991; Murrell et al. 2001; Constância et al. 2002). As IGF2 is paternally expressed (maternally silenced), we mated females to male Igf2KO homozygotes and predicted that they would pay smaller costs of reproduction (than females mated to wild-type controls) due to lower placental and foetal (Igf2 KO) demand for nutrients. There females were found to have reduced metabolic costs during gestation, however, they did not detect any differences in body composition compared to the WT controls. This study provides promising evidence that Igf2 is involved in regulating female metabolism during gestation and provides further support for Haig’s kinship hypothesis. Future studies will also look at the reciprocally imprinted maternally expressed gene H19 to determine whether the opposite effect (Females suffering higher metabolic costs) is induced by this, and also examine the females not only in gestation but lactation which is the most energetically demanding stage of reproduction for Mus musculus.

In conclusion, disentangling the cause of these results and understanding how males and females influence each other’s life histories, will, in turn, also help us better understand mate choice strategies and the physiological costs of reproduction as well as their impact on life-history evolution. Further, examining the evolutionary theory behind the cause of pregnancy complications may lead to improved understanding of the aetiology and development of treatments for these conditions.

130

References

Abu-Amero, S., D. Monk, S. Apostolidou, P. Stanier, and G. Moore. 2006. Imprinted genes and their role in human fetal growth. Cytogenetic and Genome Research 113:262-270. Acland, H. 1993. Abortion in mares. Equine reproduction:554-562. Agoramoorthy, G., S. Mohnot, V. Sommer, and A. Srivastava. 1988. Abortions in free ranging Hanuman langurs (Presbytis entellus)—a male induced strategy? Human Evolution 3:297-308. Agoramoorthy, G. and R. Rudran. 1995. Infanticide by Adult and Subadult Males in Free‐ranging Red Howler Monkeys, Alouatta seniculus, in Venezuela. Ethology 99:75-88. Agrell, J., J. O. Wolff, and H. Ylönen. 1998. Counter-strategies to infanticide in mammals: costs and consequences. Oikos:507-517. Allen, W., S. Wilsher, F. Stewart, J. Ousey, and A. Fowden. 2002a. The influence of maternal size on placental, fetal and postnatal growth in the horse. II. Endocrinology of pregnancy. Journal of endocrinology 172:237-246. Allen, W., S. Wilsher, C. Turnbull, F. Stewart, J. Ousey, P. Rossdale, and A. Fowden. 2002b. Influence of maternal size on placental, fetal and postnatal growth in the horse. I. Development in utero. Reproduction 123:445-453. Aloise King, E. D., P. B. Banks, and R. C. Brooks. 2013. Sexual conflict in mammals: consequences for mating systems and life history. Mammal Review 43:47-58. Alonso-Alvarez, C., S. Bertrand, G. Devevey, J. Prost, B. Faivre, and G. Sorci. 2004. Increased susceptibility to oxidative stress as a proximate cost of reproduction. Ecology Letters 7:363-368. Alonso‐Alvarez, C., S. Bertrand, B. Faivre, and G. Sorci. 2007. Increased susceptibility to oxidative damage as a cost of accelerated somatic growth in zebra finches. Functional Ecology 21:873-879. Amano, H. and K. Hayashi. 1998. Costs and benefits for water strider (Aquarius paludum) females of carrying guarding, reproductive males. Ecological Research 13:263-272.

131

An, X.-L., J.-X. Zou, R.-Y. Wu, Y. Yang, F.-D. Tai, S.-Y. Zeng, R. Jia, X. Zhang, E.-Q. Liu, and H. Broders. 2011. Strain and sex differences in anxiety-like and social behaviors in C57BL/6J and BALB/cJ mice. Experimental Animals 60:111-123. Ananth, C. V., K. M. Keyes, and R. J. Wapner. 2013. Pre-eclampsia rates in the United States, 1980-2010: age-period-cohort analysis. Bmj 347:f6564. Anderson, M. J. and A. F. Dixson. 2002. Motility and the midpiece in primates. Nature 416:496. Andersson, M. B. 1994. Sexual selection. Princeton University Press. Arakawa, H., D. C. Blanchard, K. Arakawa, C. Dunlap, and R. J. Blanchard. 2008. Scent marking behavior as an odorant communication in mice. Neuroscience & Biobehavioral Reviews 32:1236-1248. Arendt, J. D. 1997. Adaptive intrinsic growth rates: an integration across taxa. The quarterly review of biology 72:149-177. Arnqvist, G. 1988. Mate guarding and sperm displacement in the water strider Gerris lateralis Schumm. (Heteroptera: Gerridae). Freshwater Biology 19:269-274. Arnqvist, G. 1989. Multiple mating in a water strider: mutual benefits or intersexual conflict? Animal Behaviour 38:749-756. Arnqvist, G. 2004. Sexual conflict and sexual selection: lost in the chase. Evolution 58:1383-1388. Arnqvist, G., M. Edvardsson, U. Friberg, and T. Nilsson. 2000. Sexual conflict promotes speciation in insects. Proceedings of the National Academy of Sciences 97:10460-10464. Arnqvist, G. and L. Rowe. 2002. Antagonistic coevolution between the sexes in a group of insects. Nature 415:787-789. Arnqvist, G. and L. Rowe. 2005. Sexual conﬂict. Prince—ton University Press, Princeton, NI. Arnqvist, G. and L. Rowe. 2013. Sexual conflict. Princeton University Press. Augustine, R. A. and D. R. Grattan. 2008. Induction of Central Leptin Resistance in Hyperphagic Pseudopregnant Rats by Chronic Prolactin Infusion. Endocrinology 149:1049-1055.

132

Avila, F. W. and M. F. Wolfner. 2009. Acp36DE is required for uterine conformational changes in mated Drosophila females. Proceedings of the National Academy of Sciences 106:15796-15800. Badcock, C. and B. Crespi. 2008. Battle of the sexes may set the brain. Nature 454:1054. Baird Jr, J. N. 1981. Eclampsia in a lowland gorilla. American journal of obstetrics and gynecology 141:345-346. Barlow, D. P. 1995. Gametic imprinting in mammals. Science-AAAS-Weekly Paper Edition 270:1610-1624. Barlow, D. P. and M. S. Bartolomei. 2014. Genomic imprinting in mammals. Cold Spring Harbor perspectives in biology 6:a018382. Barlow, S., A. Knight, and F. Sullivan. 1978. Delay in postnatal growth and development of offspring produced by maternal restraint stress during pregnancy in the rat. Teratology 18:211-218. Barlow, S. M., P. Morrison, and F. Sullivan. 1974. Plasma corticosterone levels during pregnancy in the mouse: the relative contributions of the adrenal glands and foeto-placental units. Journal of Endocrinology 60:473-483. Barnard, C. and J. Fitzsimons. 1988. Kin recognition and mate choice in mice: the effects of kinship, familiarity and social interference on intersexual interaction. Animal Behaviour 36:1078-1090. Barnett, D. K. and D. H. Abbott. 2003. Reproductive adaptations to a large‐brained fetus open a vulnerability to anovulation similar to polycystic ovary syndrome. American journal of human biology 15:296-319. Barnett, S. and E. Widdowson. 1971. Organ weights and body composition of parturient and lactating mice, and their young, at 21° C and− 3° C. Journal of reproduction and fertility 26:39-57. Barros, H., S. Tannhauser, M. Tannhauser, and M. Tannhauser. 1994. The effects of GABAergic drugs on grooming behaviour in the open field. Basic & Clinical Pharmacology & Toxicology 74:339-344. Barthelemy, M., C. Gabrion, and G. Petit. 2004. Reduction in testosterone concentration and its effect on the reproductive output of chronic malaria- infected male mice. Parasitology research 93:475-481.

133

Bartke, A., R. Steele, N. Musto, and B. Caldwell. 1973. Fluctuations in plasma testosterone levels in adult male rats and mice. Endocrinology 92:1223-1228. Bartolomei, M. S., S. Zemel, and S. M. Tilghman. 1991. Parental imprinting of the mouse H19 gene. Nature 351:153. Bartoš, L., J. Bartošová, J. Pluháček, and J. Šindelářová. 2011. Promiscuous behaviour disrupts pregnancy block in domestic horse mares. Behavioral ecology and sociobiology 65:1567-1572. Bateman, A. J. 1948. Intra-sexual selection in Drosophila. Heredity 2:349-368. Batruch, I., I. Lecker, D. Kagedan, C. R. Smith, B. J. Mullen, E. Grober, K. C. Lo, E. P. Diamandis, and K. A. Jarvi. 2011. Proteomic analysis of seminal plasma from normal volunteers and post-vasectomy patients identifies over 2000 proteins and candidate biomarkers of the urogenital system. Journal of proteome research 10:941-953. Baxter, J. and G. Rousseau. 1979. Glucocorticoid hormone action: an overview. Monographs on endocrinology 12:1-24. Beehner, J., T. Bergman, D. Cheney, R. Seyfarth, and P. Whitten. 2005. The effect of new alpha males on female stress in free-ranging baboons. Animal Behaviour 69:1211-1221. Bellinge, B. S., C. M. Copeland, T. D. Thomas, R. E. Mazzucchelli, G. O’Neil, and M. J. Cohen. 1986. The influence of patient insemination on the implantation rate in an in vitro fertilization and embryo transfer program. Fertility and sterility 46:252-256. Beltran-Bech, S. and F.-J. Richard. 2014. Impact of infection on mate choice. Animal Behaviour 90:159-170. Ben-Haroush, A., Y. Yogev, and M. Hod. 2004. Epidemiology of gestational diabetes mellitus and its association with Type 2 diabetes. Diabetic Medicine 21:103- 113. Bengtsson, B. 1978. Avoiding inbreeding: at what cost? Journal of Theoretical Biology 73:439-444. Benirschke, K., P. Kaufmann, and R. Baergen. Pathology of the human placenta.-- 2000. New York: Springer.–1070 p. Benus, R. and M. Rondigs. 1996. Patterns of maternal effort in mouse lines bidirectionally selected for aggression. Animal Behaviour 51:67-75. 134

Berger, P., M. Gruschwitz, G. Spoettl, S. Dirnhofer, S. Madersbacher, R. Gerth, W. E. Merz, E. Plas, and N. Sampson. 2007. Human chorionic gonadotropin (hCG) in the male reproductive tract. Molecular and Cellular Endocrinology 260- 262:190-196. Berghänel, A., M. Heistermann, O. Schülke, and J. Ostner. 2017. Prenatal stress accelerates offspring growth to compensate for reduced maternal investment across mammals. Proceedings of the National Academy of Sciences. Beynon, R. J. and J. L. Hurst. 2004. Urinary proteins and the modulation of chemical scents in mice and rats. Peptides 25:1553-1563. Bhardwaj, S., D. Verma, and S. Kapoor. 2013. Body composition and basal metabolic rate in pregnant women. Anthropological Review 76:163-171. Biensen, N. J., M. E. Wilson, and S. P. Ford. 1998. The impact of either a Meishan or Yorkshire uterus on Meishan or Yorkshire fetal and placental development to days 70, 90, and 110 of gestation. Journal of animal science 76:2169-2176. Bilde, T., Y. Lubin, D. Smith, J. M. Schneider, and A. A. Maklakov. 2005. The transition to social inbred mating systems in spiders: role of inbreeding tolerance in a subsocial predecessor. Evolution 59:160-174. Birkhead, T. 1998. Cryptic female choice: criteria for establishing female sperm choice. Evolution 52:1212-1218. Birkhead, T. 2000. Promiscuity: an evolutionary history of sperm competition. Harvard University Press. Birkhead, T. R. and A. P. Møller. 1998. Sperm competition and sexual selection. Academic Press. Birkhead, T. R. and T. Pizzari. 2002. Evolution of sex: postcopulatory sexual selection. Nature Reviews Genetics 3:262. Bischof, R. J., C. S. Lee, M. R. Brandon, and E. Meeusen. 1994. Inflammatory response in the pig uterus induced by seminal plasma. Journal of Reproductive Immunology 26:131-146. Blackburn, D. G. 2000. Classification of the reproductive patterns of amniotes. Herpetological Monographs:371-377. Blumstein, D. T. 2007. The evolution, function, and meaning of marmot alarm communication. Advances in the Study of Behavior 37:371-401.

135

Boccalon, S., B. Scaggiante, and L. Perissin. 2006. Anxiety stress and nociceptive responses in mice. Life sciences 78:1225-1230. Bon, R. and R. Campan. 1996. Unexplained sexual segregation in polygamous ungulates: a defense of an ontogenetic approach. Behavioural Processes 38:131-154. Bonduriansky, R. and S. F. Chenoweth. 2009. Intralocus sexual conflict. Trends in ecology & evolution 24:280-288. Bonduriansky, R., A. Maklakov, F. Zajitschek, and R. Brooks. 2008. Sexual selection, sexual conflict and the evolution of ageing and life span. Functional ecology 22:443-453. Bourke, A. F. G. 2009. The kin structure of sexual interactions. Biology Letters 5:689- 692. Bouwknecht, J. A., F. Spiga, D. R. Staub, M. W. Hale, A. Shekhar, and C. A. Lowry. 2007. Differential effects of exposure to low-light or high-light open-field on anxiety-related behaviors: relationship to c-Fos expression in serotonergic and non-serotonergic neurons in the dorsal raphe nucleus. Brain research bulletin 72:32-43. Bowen, W. D., O. T. Oftedal, and D. J. Boness. 1992. Mass and energy transfer during lactation in a small phocid, the harbor seal (Phoca vitulina). Physiological Zoology 65:844-866. Bowers, J. M. and B. K. Alexander. 1967. Mice: individual recognition by olfactory cues. Science 158:1208-1210. Brennan, P. L., C. J. Clark, and R. O. Prum. 2009. Explosive eversion and functional morphology of the duck penis supports sexual conflict in waterfowl genitalia. Proceedings of the Royal Society of London B: Biological Sciences:rspb20092139. Brennan, P. L., R. O. Prum, K. G. McCracken, M. D. Sorenson, R. E. Wilson, and T. R. Birkhead. 2007. Coevolution of male and female genital morphology in waterfowl. PLoS one 2:e418. Briceño, R. D. and W. G. Eberhard. 2009. Experimental demonstration of possible cryptic female choice on male tsetse fly genitalia. Journal of insect physiology 55:989-996.

136

Bro-Jørgensen, J. 2011. Intra- and Intersexual Conflicts and Cooperation in the Evolution of Mating Strategies: Lessons Learnt From Ungulates. Evolutionary Biology 38:28-41. Bromfield, J. J., J. E. Schjenken, P. Y. Chin, A. S. Care, M. J. Jasper, and S. A. Robertson. 2014. Maternal tract factors contribute to paternal seminal fluid impact on metabolic phenotype in offspring. Proceedings of the National Academy of Sciences 111:2200-2205. Bronson, F. 1979. The reproductive ecology of the house mouse. The Quarterly review of biology 54:265-299. Bronson, F. 1996. Effects of prolonged exposure to anabolic steroids on the behavior of male and female mice. Pharmacology Biochemistry and Behavior 53:329- 334. Bronson, F. and B. Eleftheriou. 1963. Influence of strange males on implantation in the deermouse. General and comparative endocrinology 3:515-518. Bronson, F. H. 1989. Mammalian reproductive biology. University of Chicago Press. Brouwer, L., M. Van De POL, E. Atema, and A. Cockburn. 2011. Strategic promiscuity helps avoid inbreeding at multiple levels in a cooperative breeder where both sexes are philopatric. Molecular Ecology 20:4796-4807. Brown, R. E. and D. W. Macdonald. 1985. Social odours in mammals. Clarendon Press. Bruce, H. M. 1959. An exteroceptive block to pregnancy in the mouse. Nature 184:105-105. Bruce, H. M. 1960. A block to pregnancy in the mouse caused by proximity of strange males. Journal of Reproduction and Fertility 1:96-103. Bruce, H. M. 1961. Time relations in the pregnancy-block induced in mice by strange males. Journal of Reproduction and Fertility 2:138-142. Burt, A. and R. Trivers. 1998. Genetic conflicts in genomic imprinting. Proceedings of the Royal Society of London B: Biological Sciences 265:2393-2397. Buss, D. M. 2017. Sexual Conflict in Human Mating. Current Directions in Psychological Science 26:307-313. Bussière, L. F., J. Hunt, M. D. Jennions, and R. Brooks. 2006. Sexual conflict and cryptic female choice in the black field cricket, Teleogryllus commodus. Evolution 60:792-800. 137

Byrne, P. G., G. R. Rice, and W. R. Rice. 2008. Effect of a refuge from persistent male courtship in the Drosophila laboratory environment. Integrative and Comparative Biology 48:e1-e1. Caillol, M., M. Mondain-Monval, and B. Rossano. 1991. Gonadotrophins and sex steroids during pregnancy and natural superfoetation in captive brown hares (Lepus europaeus). Journal of reproduction and fertility 92:299-306. Cameron, E., T. Day, and L. Rowe. 2003. Sexual conflict and indirect benefits. Journal of Evolutionary Biology 16:1055-1060. Cameron, E. Z., T. H. Setsaas, and W. L. Linklater. 2009. Social bonds between unrelated females increase reproductive success in feral horses. Proceedings of the National Academy of Sciences 106:13850-13853. Cant, M. A. and R. A. Johnstone. 2008. Reproductive conflict and the separation of reproductive generations in humans. Proceedings of the National Academy of Sciences 105:5332-5336. Capellini, I. 2012. The evolutionary significance of placental interdigitation in mammalian reproduction: contributions from comparative studies. Placenta 33:763-768. Capellini, I., C. Venditti, and R. A. Barton. 2010. Placentation and maternal investment in mammals. The American Naturalist 177:86-98. Carayon, J. 1966. Traumatic insemination and the paragenital system. Monograph of Cimicidae (Hemiptera, Heteroptera). College ParkMd: Entomological Society of America:81-166. Carnegie, S. D., L. M. Fedigan, and T. E. Ziegler. 2011. Social and environmental factors affecting fecal glucocorticoids in wild, female white‐faced capuchins (Cebus capucinus). American Journal of Primatology 73:861-869. Catalano, R. A., K. B. Saxton, A. Gemmill, and T. Hartig. 2016. Twinning in Norway following the Oslo massacre: evidence of a ‘Bruce Effect’in humans. Twin research and human genetics 19:485-491. Cattanach, B., C. Beechey, and J. Peters. 2006. Interactions between imprinting effects: summary and review. Cytogenetic and genome research 113:17-23. Cattanach, B. M., C. V. Beechey, and J. Peters. 2004. Interactions between imprinting effects in the mouse. Genetics 168:397-413.

138

Chaiworapongsa, T., P. Chaemsaithong, L. Yeo, and R. Romero. 2014. Pre-eclampsia part 1: current understanding of its pathophysiology. Nature Reviews Nephrology 10:466. Chang, H. and P. G. Jablonski. 2010. Male water striders attract predators to intimidate females into copulation. Nature communications 1:52. Chapman, T., G. Arnqvist, J. Bangham, and L. Rowe. 2003. Sexual conflict. Trends in Ecology & Evolution 18:41-47. Chapman, T. and S. J. Davies. 2004. Functions and analysis of the seminal fluid proteins of male Drosophila melanogaster fruit flies. Peptides 25:1477-1490. Chapman, T., L. F. Liddle, J. M. Kalb, M. F. Wolfner, and L. Partridge. 1995. Cost of mating in Drosophila melanogaster females is mediated by male accessory gland products. Nature 373:241. Charalambous, M., A. Ward, and L. D. Hurst. 2003. Evidence for a priming effect on maternal resource allocation: implications for interbrood competition. Proceedings of the Royal Society of London. Series B: Biological Sciences 270:S100-S103. Charlesworth, B. and D. Charlesworth. 1999. The genetic basis of inbreeding depression. Genetics Research 74:329-340. Cheetham, S. A., A. L. Smith, S. D. Armstrong, R. J. Beynon, and J. L. Hurst. 2009. Limited variation in the major urinary proteins of laboratory mice. Physiology & Behavior 96:253-261. Chida, D., K. Miyoshi, T. Sato, T. Yoda, T. Kikusui, and Y. Iwakura. 2011. The Role of Glucocorticoids in Pregnancy, Parturition, Lactation, and Nurturing in Melanocortin Receptor 2-Deficient Mice. Endocrinology 152:1652-1660. Chipman, R. and K. Fox. 1966. Oestrous synchronization and pregnancy blocking in wild house mice (Mus musculus). Journal of reproduction and fertility 12:233-236. Christians, J. K., K. Leavey, and B. J. Cox. 2017. Associations between imprinted gene expression in the placenta, human fetal growth and preeclampsia. Biology Letters 13. Clapham, P. J. 2000. Cetacean societies—field studies of dolphins and whales:173- 196.

139

Clemens, L. G., B. E. Wee, D. R. Weaver, E. J. Roy, B. D. Goldman, and B. Rakerd. 1988. Retention of masculine sexual behavior following castration in male B6D2F1 mice. Physiology & behavior 42:69-76. Clulow, F. and J. Clarke. 1968. Pregnancy-block in Microtus agrestis and induced ovulator. Clulow, F. and P. Langford. 1971. Pregnancy-block in the meadow vole, Microtus pennsylvanicus. Journal of reproduction and fertility 24:275-277. Clutton-Brock and G. A. Parker. 1992. Potential reproductive rates and the operation of sexual selection. The Quarterly Review of Biology 67:437-456. Clutton-Brock and G. A. Parker. 1995. Sexual coercion in animal societies. Animal Behaviour 49:1345-1365. Clutton-Brock, A. Russell, and L. Sharpe. 2004. Behavioural tactics of breeders in cooperative meerkats. Animal Behaviour 68:1029-1040. Clutton-Brock, T. and A. C. Vincent. 1991. Sexual selection and the potential reproductive rates of males and females. Nature 351:58-60. Clutton-Brock, T. H. 1989. Review lecture: mammalian mating systems. Pp. 339-372. Proc. R. Soc. Lond. B. The Royal Society. Coan, P., G. Burton, and A. Ferguson-Smith. 2005. Imprinted genes in the placenta–a review. Placenta 26:S10-S20. Colmenares, F. and M. Gomendio. 1988. Changes in female reproductive condition following male take-overs in a colony of hamadryas and hybrid baboons. Folia Primatologica 50:157-174. Compaan, J. C., G. Van Wattum, A. J. de Ruiter, G. A. Van Oortmerssen, J. M. Koolhaas, and B. Bohus. 1993. Genetic differences in female house mice in aggressive response to sex steroid hormone treatment. Physiology & behavior 54:899- 902. Connor, R. C. 2002. Ecology of group living and social behaviour. Marine mammal biology: An evolutionary approach 2:353-360. Constância, M., E. Angiolini, I. Sandovici, P. Smith, R. Smith, G. Kelsey, W. Dean, A. Ferguson-Smith, C. P. Sibley, and W. Reik. 2005. Adaptation of nutrient supply to fetal demand in the mouse involves interaction between the Igf2 gene and placental transporter systems. Proc. Natl. Acad. Sci. U. S. A. 102:19219-19224. 140

Constância, M., M. Hemberger, J. Hughes, W. Dean, A. Ferguson-Smith, R. Fundele, F. Stewart, G. Kelsey, A. Fowden, and C. Sibley. 2002. Placental-specific IGF-II is a major modulator of placental and fetal growth. Nature 417:945-948. Coopersmith, C. B. 1998. Pregnancy block in house mice (Mus domesticus) as a function of t-complex genotype: examination of the mate choice and male infanticide hypotheses. J. Comp. Psychol. 112:82. Cordero, C. and W. G. Eberhard. 2003. Female choice of sexually antagonistic male adaptations: a critical review of some current research. Journal of evolutionary biology 16:1-6. Cordts, R. and L. Partridge. 1996. Courtship reduces longevity of maleDrosophila melanogaster. Animal Behaviour 52:269-278. Costantini, D. 2008. Oxidative stress in ecology and evolution: lessons from avian studies. Ecology letters 11:1238-1251. Costantini, D., V. Marasco, and A. P. Møller. 2011. A meta-analysis of glucocorticoids as modulators of oxidative stress in vertebrates. Journal of Comparative Physiology B 181:447-456. Cothran, R. D. 2004. Precopulatory mate guarding affects predation risk in two freshwater amphipod species. Animal behaviour 68:1133-1138. Crawford, G., A. Ray, A. Gudi, A. Shah, and R. Homburg. 2014. The role of seminal plasma for improved outcomes during in vitro fertilization treatment: review of the literature and meta-analysis. Human reproduction update 21:275-284. Crespi, B. and C. Semeniuk. 2004. Parent-offspring conflict in the evolution of vertebrate reproductive mode. The American Naturalist 163:635-653. Crofoot, M. C. 2007. Mating and feeding competition in white-faced capuchins (Cebus capucinus): the importance of short-and long-term strategies. Behaviour 144:1473-1495. Crudgington, H. S. and M. T. Siva-Jothy. 2000. Genital damage, kicking and early death. Nature 407:855. Dahle, B. and J. E. Swenson. 2003. Home ranges in adult Scandinavian brown bears (Ursus arctos): effect of mass, sex, reproductive category, population density and habitat type. Journal of Zoology 260:329-335.

141

Dalle, M., J. Giry, M. Gay, and P. Delost. 1978. Perinatal changes in plasma and adrenal corticosterone and aldosterone concentrations in the mouse. Journal of Endocrinology 76:303-309. Dantzer, B., A. E. Newman, R. Boonstra, R. Palme, S. Boutin, M. M. Humphries, and A. G. McAdam. 2013. Density triggers maternal hormones that increase adaptive offspring growth in a wild mammal. Science 340:1215-1217. Darden, S. K. and D. P. Croft. 2008. Male harassment drives females to alter habitat use and leads to segregation of the sexes. Biology letters 4:449-451. Darwin, C. 1859. On the origin of the species by natural selection. Dawkins, R. and T. R. Carlisle. 1976. Parental investment, mate desertion and a fallacy. Nature 262:131. Day, T. and R. Bonduriansky. 2004. Intralocus sexual conflict can drive the evolution of genomic imprinting. Genetics 167:1537-1546. De Block, M. and R. Stoks. 2008a. Compensatory growth and oxidative stress in a damselfly. Proceedings of the Royal Society of London B: Biological Sciences 275:781-785. De Block, M. and R. Stoks. 2008b. Compensatory growth and oxidative stress in a damselfly. Proceedings of the Royal Society B: Biological Sciences 275:781- 785. De Jonge, C. 2005. Biological basis for human capacitation. Human Reproduction Update 11:205-214. DeChiara, T. M., A. Efstratiadis, and E. J. Robertsen. 1990. A growth-deficiency phenotype in heterozygous mice carrying an insulin-like growth factor II gene disrupted by targeting. Nature 345:78. DeChiara, T. M., E. J. Robertson, and A. Efstratiadis. 1991. Parental imprinting of the mouse insulin-like growth factor II gene. Cell 64:849-859. den Boer, S. P., B. Baer, and J. J. Boomsma. 2010. Seminal fluid mediates ejaculate competition in social insects. Science 327:1506-1509. Denmark, A., D. Tien, K. Wong, A. Chung, J. Cachat, J. Goodspeed, C. Grimes, M. Elegante, C. Suciu, S. Elkhayat, B. Bartels, A. Jackson, M. Rosenberg, K. M. Chung, H. Badani, F. Kadri, S. Roy, J. Tan, S. Gaikwad, A. Stewart, I. Zapolsky, T. Gilder, and A. V. Kalueff. 2010. The effects of chronic social defeat stress on

142

mouse self-grooming behavior and its patterning. Behavioural Brain Research 208:553-559. Dewar, A. 1957. Body weight changes in the mouse during the oestrous cycle and pseudopregnancy. Journal of Endocrinology 15:230-233. Dewar, A. D. 1959. Observations on pseudopregnancy in the nouse. Journal of Endocrinology 18:186-190. Di Monte, D., D. Ross, G. Bellomo, L. Eklöw, and S. Orrenius. 1984. Alterations in intracellular thiol homeostasis during the metabolism of menadione by isolated rat hepatocytes. Archives of Biochemistry and Biophysics 235:334- 342. Dixson, A. 1998. Primate sexuality. Wiley Online Library. Doetschman, T. 2009. Influence of genetic background on genetically engineered mouse phenotypes. Pp. 423-433. Gene Knockout Protocols. Springer. Drickamer, L. C. 1992. Oestrous female house mice discriminate dominant from subordinate males and sons of dominant from sons of subordinate males by odour cues. Animal Behaviour 43:868-870. Drickamer, L. C. 1995. Rates of urine excretion by house mouse (Mus domesticus): differences by age, sex, social status, and reproductive condition. Journal of chemical ecology 21:1481-1493. Ebensperger, L. A. 1998a. Do female rodents use promiscuity to prevent male infanticide? Ethology Ecology & Evolution 10:129-141. Ebensperger, L. A. 1998b. Strategies and counterstrategies to infanticide in mammals. Biol. Rev. 73:321-346. Ebensperger, L. A. and D. T. Blumstein. 2007. Nonparental infanticide. Rodent societies: an ecological and evolutionary perspective:267-279. Eberhard, W. 1996. Female control: sexual selection by cryptic female choice. Princeton University Press. Eberle, M. and P. M. Kappeler. 2004. Selected polyandry: female choice and intersexual conflict in a small nocturnal solitary primate (Microcebus murinus). Behavioral Ecology and Sociobiology 57:91-100. Edvardsson, M. and T. Tregenza. 2005. Why do male Callosobruchus maculatus harm their mates? Behavioral Ecology 16:788-793.

143

Eggenschwiler, J., T. Ludwig, P. Fisher, P. A. Leighton, S. M. Tilghman, and A. Efstratiadis. 1997. Mouse mutant embryos overexpressing IGF-II exhibit phenotypic features of the Beckwith–Wiedemann and Simpson–Golabi– Behmel syndromes. Genes & development 11:3128-3142. Einarson, T. R., C. Piwko, and G. Koren. 2013. Quantifying the global rates of nausea and vomiting of pregnancy: a meta analysis. J Popul Ther Clin Pharmacol 20:e171-183. Eldakar, O. T., M. J. Dlugos, J. W. Pepper, and D. S. Wilson. 2009. Population structure mediates sexual conflict in water striders. Science 326:816-816. Elliot, M. G. and B. J. Crespi. 2009. Phylogenetic Evidence for Early Hemochorial Placentation in Eutheria. Placenta 30:949-967. Elliot, M. G. and B. J. Crespi. 2015. Genetic recapitulation of human pre-eclampsia risk during convergent evolution of reduced placental invasiveness in eutherian mammals. Philosophical Transactions of the Royal Society B: Biological Sciences 370:20140069. Elwood, R. W. 1985. Inhibition of infanticide and onset of paternal care in male mice (Mus musculus). J. Comp. Psychol. 99:457. Elwood, R. W. and M. C. Ostermeyer. 1984. Does copulation inhibit infanticide in male rodents? Animal Behaviour 32:293-294. Enders, A. C. and T. N. Blankenship. 1999. Comparative placental structure. Advanced Drug Delivery Reviews 38:3-15. Estanislau, C. 2012. Cues to the usefulness of grooming behavior in the evaluation of anxiety in the elevated plus-maze. Psychology & Neuroscience 5:105. Feil, R. and F. Berger. 2007. Convergent evolution of genomic imprinting in plants and mammals. Trends in Genetics 23:192-199. Feldblum, Joseph T., Emily E. Wroblewski, Rebecca S. Rudicell, Beatrice H. Hahn, T. Paiva, M. Cetinkaya-Rundel, Anne E. Pusey, and Ian C. Gilby. 2014. Sexually Coercive Male Chimpanzees Sire More Offspring. Current Biology 24:2855- 2860. Fentress, J. C. 1968. Interrupted ongoing behaviour in two species of vole (Microtus agrestis and Clethrionomys britannicus). II. Extended analysis of motivational variables underlying fleeing and grooming behaviour. Animal Behaviour 16:154-167. 144

Ferguson-Smith, A. C., B. Cattanach, S. Barton, C. Beechey, and M. Surani. 1991. Embryological and molecular investigations of parental imprinting on mouse chromosome 7. Nature 351:667. Ferrón, S., E. Radford, A. Domingo-Muelas, I. Kleine, A. Ramme, D. Gray, I. Sandovici, M. Constancia, A. Ward, and T. Menheniott. 2015. Differential genomic imprinting regulates paracrine and autocrine roles of IGF2 in mouse adult neurogenesis. Nature communications 6:8265. Fessler, D. T., T. Bayley, L. Dye, J. Brown, S. Flaxman, B. Leeners, R. Nappi, K. OConnor, D. Sellen, and P. Sherman. 2002. Reproductive immunosuppression and diet: An evolutionary perspective on pregnancy sickness and meat consumption. Current anthropology 43:19-61. Firman, R. C., M. Gomendio, E. R. Roldan, and L. W. Simmons. 2014. The coevolution of ova defensiveness with sperm competitiveness in house mice. The American Naturalist 183:565-572. Fisher, R. 1930. The Genetical Theory of Natural Selection Clarendon Press Oxford Google Scholar. Fitzpatrick, J. L. and S. Lüpold. 2014. Sexual selection and the evolution of sperm quality. Molecular Human Reproduction 20:1180-1189. Fiumera, A. C., B. L. Dumont, and A. G. Clark. 2006. Natural variation in male-induced ‘cost-of-mating’and allele-specific association with male reproductive genes in Drosophila melanogaster. Philosophical Transactions of the Royal Society B: Biological Sciences 361:355-361. Flaxman, S. M. and P. W. Sherman. 2000. Morning sickness: a mechanism for protecting mother and embryo. The Quarterly review of biology 75:113-148. Flaxman, S. M. and P. W. Sherman. 2008. Morning sickness: adaptive cause or nonadaptive consequence of embryo viability? The American Naturalist 172:54-62. Foerster, K., T. Coulson, B. C. Sheldon, J. M. Pemberton, T. H. Clutton-Brock, and L. E. Kruuk. 2007. Sexually antagonistic genetic variation for fitness in red deer. Nature 447:1107-1110. Folstad, I. and A. J. Karter. 1992. Parasites, bright males, and the immunocompetence handicap. The American Naturalist 139:603-622.

145

Forbes, L. S. 1997. The evolutionary biology of spontaneous abortion in humans. Trends in Ecology & Evolution 12:446-450. Forbes, S. 2002. Pregnancy sickness and embryo quality. Trends in Ecology and Evolution 17:115-120. Forbes, S. 2017. Embryo quality: the missing link between pregnancy sickness and pregnancy outcome. Evolution and Human Behavior 38:265-278. Forné, T., J. Oswald, W. Dean, J. R. Saam, B. Bailleul, L. Dandolo, S. M. Tilghman, J. Walter, and W. Reik. 1997. Loss of the maternal H19 gene induces changes in Igf2 methylation in both cis and trans. Proceedings of the National Academy of Sciences 94:10243-10248. Forsum, E., A. Sadurskis, and J. Wager. 1988. Resting metabolic rate and body composition of healthy Swedish women during pregnancy. The American journal of clinical nutrition 47:942-947. Fowden, A. and T. Moore. 2012. Maternal-fetal resource allocation: co-operation and conflict. Placenta 33:e11-e15. Fowden, A., C. Sibley, W. Reik, and M. Constancia. 2006. Imprinted genes, placental development and fetal growth. Hormone Research in Paediatrics 65:50-58. Fox, E. A. 2002. Female tactics to reduce sexual harassment in the Sumatran orangutan (Pongo pygmaeus abelii). Behavioral Ecology and Sociobiology 52:93-101. Freemark, M. 2006. Regulation of maternal metabolism by pituitary and placental hormones: roles in fetal development and metabolic programming. Hormone Research in Paediatrics 65:41-49. Fry, C. L. and G. S. Wilkinson. 2004. Sperm survival in female stalk-eyed flies depends on seminal fluid and meiotic drive. Evolution 58:1622-1626. Fuentes, A. 1998. Re‐evaluating primate monogamy. American Anthropologist 100:890-907. Furneaux, E. C., A. J. Langley-Evans, and S. C. Langley-Evans. 2001. Nausea and vomiting of pregnancy: endocrine basis and contribution to pregnancy outcome. Obstetrical & gynecological survey 56:775-782. Furutachi, S., A. Matsumoto, K. I. Nakayama, and Y. Gotoh. 2013. p57 controls adult neural stem cell quiescence and modulates the pace of lifelong neurogenesis. The EMBO journal 32:970-981. 146

Gale, T., M. Garratt, and R. Brooks. 2018. No evidence that male mice impose mating or insemination costs on females, but females avoid castrated males (Unpublished manuscript). University of New South Wales, Australia. . Gale, T., A. B. Gibson, R. C. Brooks, and M. Garratt. 2013. Exposure to a novel male during late pregnancy influences subsequent growth of offspring during lactation. Journal of Evolutionary Biology 26:2057-2062. Gardner, P. R. 1997. Superoxide-driven aconitase FE-S center cycling. Bioscience reports 17:33-42. Gardner, P. R., I. Raineri, L. B. Epstein, and C. W. White. 1995. Superoxide radical and iron modulate aconitase activity in mammalian cells. Journal of Biological Chemistry 270:13399-13405. Garratt, N. Pichaud, E. A. King, and R. Brooks. 2013a. Physiological adaptations to reproduction. I. Experimentally increasing litter size enhances aspects of antioxidant defence but does not cause oxidative damage in mice. Journal of Experimental Biology 216:2879-2888. Garratt, M., J.-M. Gaillard, R. C. Brooks, and J.-F. Lemaître. 2013b. Diversification of the eutherian placenta is associated with changes in the pace of life. Proc. Natl. Acad. Sci. U. S. A. 110:7760-7765. Garratt, M., A. J. Kee, R. Palme, and R. C. Brooks. 2016. Male Presence can Increase Body Mass and Induce a Stress-Response in Female Mice Independent of Costs of Offspring Production. Scientific Reports 6:23538. Garratt, M., N. Pichaud, E. N. Glaros, A. J. Kee, and R. C. Brooks. 2014. Superoxide dismutase deficiency impairs olfactory sexual signaling and alters bioenergetic function in mice. Proceedings of the National Academy of Sciences 111:8119-8124. Garratt, M., N. Pichaud, E. D. A. King, and R. C. Brooks. 2013c. Physiological adaptations to reproduction. I. Experimentally increasing litter size enhances aspects of antioxidant defence but does not cause oxidative damage in mice. The Journal of Experimental Biology 216:2879-2888. Garratt, M., A. Vasilaki, P. Stockley, F. McArdle, M. Jackson, and J. L. Hurst. 2011. Is oxidative stress a physiological cost of reproduction? An experimental test in house mice. Proceedings of the Royal Society of London B: Biological Sciences 278:1098-1106. 147

Gaskin, T., P. Futerman, and T. Chapman. 2002. Increased density and male–male interactions reduce male longevity in the medfly, Ceratitis capitata. Animal Behaviour 63:121-129. Gay, L., P. Eady, R. Vasudev, D. Hosken, and T. Tregenza. 2009. Costly sexual harassment in a beetle. Physiological Entomology 34:86-92. Ghulmiyyah, L. and B. Sibai. Maternal Mortality From Preeclampsia/Eclampsia. Seminars in Perinatology 36:56-59. Gibson, A. B., M. Garratt, and R. C. Brooks. 2015. Experimental evidence that litter size imposes an oxidative challenge to offspring. The Journal of Experimental Biology 218:3911-3918. Gilbert, S., P. Timoney, W. McCollum, and D. Deregt. 1997. Detection of equine arteritis virus in the semen of carrier stallions by using a sensitive nested PCR assay. Journal of clinical microbiology 35:2181-2183. Giles, R., J. Donahue, C. Hong, P. Tuttle, M. Petrites-Murphy, K. Poonacha, A. Roberts, R. Tramontin, B. Smith, and T. Swerczek. 1993. Causes of abortion, stillbirth, and perinatal death in horses: 3,527 cases (1986-1991). Journal of the American Veterinary Medical Association 203:1170-1175. Goetz, A. and T. Shackelford. 2009. Sexual conflict in humans: evolutionary consequences of asymmetric parental investment and paternity uncertainty. Animal Biology 59:449-456. Goldberg, G., W. Prentice, W. Coward, H. Davies, P. Murgatroyd, C. Wensing, A. Black, M. Harding, and M. Sawyer. 1993. Longitudinal assessment of energy expenditure in pregnancy by the doubly labeled water method. The American Journal of Clinical Nutrition 57:494-505. Goodwin, T. M. 2000. Human chorionic gonadotropin and hyperemesis gravidarum. Nausea and vomiting of pregnancy: state of the art:15-22. Goodwin, T. M. 2002. Nausea and vomiting of pregnancy: an obstetric syndrome. American Journal of Obstetrics & Gynecology 186:S184-S189. Gordon Jr, G., R. L. Burch, and S. M. Platek. 2002. Does semen have antidepressant properties? Archives of Sexual Behavior 31:289-293. Gosling, L. and S. Roberts. 2001. Scent-marking by male mammals: cheat-proof signals to competitors and mates. Advances in the Study of Behavior 30:169- 217. 148

Gosling, L., S. Roberts, E. Thornton, and M. Andrew. 2000. Life history costs of olfactory status signalling in mice. Behavioral Ecology and Sociobiology 48:328-332. Gouat, J. 1985. Notes sur la reproduction de Ctenodactylus gundi rongeur Ctenodactylidae. Zeitschrift für Säugetierkunde 50:285-293. Gowaty, P. A., Y.-K. Kim, J. Rawlings, and W. Anderson. 2010. Polyandry increases offspring viability and mother productivity but does not decrease mother survival in Drosophila pseudoobscura. Proceedings of the National Academy of Sciences 107:13771-13776. Grant, E. and J. Mackintosh. 1963. A comparison of the social postures of some common laboratory rodents. Behaviour 21:246-259. Gregg, C., J. Zhang, B. Weissbourd, S. Luo, G. P. Schroth, D. Haig, and C. Dulac. 2010. High-Resolution Analysis of Parent-of-Origin Allelic Expression in the Mouse Brain. Science 329:643-648. Griffith, S. C., S. R. Pryke, and W. A. Buttemer. 2011. Constrained mate choice in social monogamy and the stress of having an unattractive partner. Proceedings of the Royal Society of London B: Biological Sciences:rspb20102672. Griffiths, M. 2012. The biology of the monotremes. Elsevier. Gubernick, D. J. and T. Teferi. 2000. Adaptive significance of male parental care in a monogamous mammal. Proceedings of the Royal Society of London B: Biological Sciences 267:147-150. Guo, H., Q. Fang, Y. Huo, Y. Zhang, and J. Zhang. 2015. Social dominance‐related major urinary proteins and the regulatory mechanism in mice. Integrative zoology 10:543-554. Guo, L., S. Choufani, J. Ferreira, A. Smith, D. Chitayat, C. Shuman, R. Uxa, S. Keating, J. Kingdom, and R. Weksberg. 2008. Altered gene expression and methylation of the human chromosome 11 imprinted region in small for gestational age (SGA) placentae. Developmental Biology 320:79-91. Guo, X., M. Zhong, Y. Luo, L. Xiao, and X. Chen. 2017. The expression of imprinted genes IGF2 and PHLDA2 in mid-pregnancy have predictive values for the development of pre-eclampsia. Int J Clin Exp Med 10:8934-8941.

149

Haig, D. 1993a. Genetic conflicts in human pregnancy. Quarterly Review of Biology:495-532. Haig, D. 1993b. Genetic conflicts in human pregnancy. The Quarterly review of biology 68:68:495-532

Haig, D. 1993c. Genomic imprinting, human chorionic gonadotropin, and triploidy. Prenatal Diagnosis 13:151-151. Haig, D. 1996. Placental hormones, genomic imprinting, and maternal—fetal communication. Journal of Evolutionary Biology 9:357-380. Haig, D. 1997. Parental antagonism, relatedness asymmetries, and genomic imprinting. Proceedings of the Royal Society of London. Series B: Biological Sciences 264:1657-1662. Haig, D. 1999. Genetic conflicts of pregnancy and childhood. Evolution in health and disease:77-90. Haig, D. 2000. The kinship theory of genomic imprinting. Annual review of ecology and systematics 31:9-32. Haig, D. 2004. Genomic imprinting and kinship: how good is the evidence? Annu. Rev. Genet. 38:553-585. Haig, D. 2008. Placental growth hormone-related proteins and prolactin-related proteins. Placenta 29:36-41. Haig, D. 2014a. Coadaptation and conflict, misconception and muddle, in the evolution of genomic imprinting. Heredity 113:96-103. Haig, D. 2014b. Troubled sleepNight waking, breastfeeding and parent–offspring conflict. Evolution, Medicine, and Public Health 2014:32-39. Haig, D. and M. Westoby. 1989. Parent-specific gene expression and the triploid endosperm. The American Naturalist 134:147-155. Hall, W. 1987. Bovine superfetation by natural conception secondary to an embryo transfer pregnancy. The Cornell veterinarian 77:282-283. Halliwell, B. 1999. Detection of free radicals and other reactive species: trapping and fingerprinting; Assays of total antioxidant activity. Free Radicals in Biology and Medicine:422-424.

150

Halliwell, B. and J. Gutteridge. 1999. Detection of free radicals and other reactive species: trapping and fingerprinting. In: Free Radicals in Biology and Medicine, edited by Halliwell B and Gutteridge JMC Pp. 351-425. Oxford University Press Oxford. Han, L., P. E. Szabó, and J. R. Mann. 2010. Postnatal Survival of Mice with Maternal Duplication of Distal Chromosome 7 Induced by a Igf2/H19 Imprinting Control Region Lacking Insulator Function. PLoS Genetics 6:e1000803. Hansson, B., L. Jack, J. K. Christians, J. M. Pemberton, M. Åkesson, H. Westerdahl, S. Bensch, and D. Hasselquist. 2006. No evidence for inbreeding avoidance in a great reed warbler population. Behavioral Ecology 18:157-164. Harshman, L. G. and A. J. Zera. 2007. The cost of reproduction: the devil in the details. Trends in Ecology & Evolution 22:80-86. Hartung, T. G. and D. A. Dewsbury. 1978. A comparative analysis of copulatory plugs in muroid rodents and their relationship to copulatory behavior. Journal of Mammalogy 59:717-723. Hausladen, A. and I. Fridovich. 1994. Superoxide and peroxynitrite inactivate aconitases, but nitric oxide does not. Journal of Biological Chemistry 269:29405-29408. Hausladen, A. and I. Fridovich. 1996a. [4] Measuring nitric oxide and superoxide: Rate constants for aconitase reactivity. Methods in enzymology 269:37-41. Hausladen, A. and I. Fridovich. 1996b. [4] Measuring nitric oxide and superoxide: Rate constants for aconitase reactivity. Pp. 37-41. Methods in Enzymology. Academic Press. Hayashi, S. 1990. Social condition influences sexual attractiveness of dominant male mice. Zoological science 7:889-894. Hayssen, V. 1993. Empirical and theoretical constraints on the evolution of lactation. Journal of Dairy Science 76:3213-3233. He, J., A. Zhang, M. Fang, R. Fang, J. Ge, Y. Jiang, H. Zhang, C. Han, X. Ye, D. Yu, H. Huang, Y. Liu, and M. Dong. 2013. Methylation levels at IGF2 and GNAS DMRs in infants born to preeclamptic pregnancies. BMC Genomics 14:472-472. Hector, K. L. and S. Nakagawa. 2012. Quantitative analysis of compensatory and catch‐up growth in diverse taxa. Journal of Animal Ecology 81:583-593.

151

Heske, E. J. and R. J. Nelson. 1984. Pregnancy interruption in Microtus ochrogaster: laboratory artifact or field phenomenon? Biol. Reprod. 31:97-103. Hestermann, M., T. Ziegler, C. P. van Schaik, K. Launhardt, P. Winkler, and J. K. Hodges. 2001. Loss of oestrus, concealed ovulation and paternity confusion in free-ranging Hanuman langurs. Proceedings of the Royal Society of London. Series B: Biological Sciences 268:2445-2451. Hillgarth, N. and J. C. Wingfield. 1997. Testosterone and immunosuppression in vertebrates: implications for parasite-mediated sexual selection. Pp. 143- 155. Parasites and pathogens. Springer. Holand, Ø., K. R. Askim, K. H. Røed, R. B. Weladji, H. Gjøstein, and M. Nieminen. 2007. No evidence of inbreeding avoidance in a polygynous ungulate: the reindeer (Rangifer tarandus). Biology Letters 3:36-39. Holland, B. and W. R. Rice. 1998. Perspective: chase-away sexual selection: antagonistic seduction versus resistance. Evolution:1-7. Holland, B. and W. R. Rice. 1999. Experimental removal of sexual selection reverses intersexual antagonistic coevolution and removes a reproductive load. Proceedings of the National Academy of Sciences 96:5083-5088. Hollegaard, B., S. G. Byars, J. Lykke, and J. J. Boomsma. 2013. Parent-Offspring Conflict and the Persistence of Pregnancy-Induced Hypertension in Modern Humans. PLOS ONE 8:e56821. Holloszy, J., L. Oscai, I. Don, and P. Mole. 1970. Mitochondrial citric acid cycle and related enzymes: adaptive response to exercise. Biochemical and biophysical research communications 40:1368-1373. Holt, W. V. and A. Fazeli. 2015. Do sperm possess a molecular passport? Mechanistic insights into sperm selection in the female reproductive tract. MHR: Basic science of reproductive medicine 21:491-501. Horne, T. J. and H. Ylönen. 1998. Heritabilities of dominance‐related traits in male bank voles (Clethrionomys glareolus). Evolution 52:894-899. Hosken, D. J. and P. Stockley. 2004. Sexual selection and genital evolution. Trends in Ecology & Evolution 19:87-93. Hrdy, S. B. 1974. Male-male competition and infanticide among the langurs (Presbytis entellus) of Abu, Rajasthan. Folia primatologica 22:19-58.

152

Hrdy, S. B. 1979. Infanticide among animals: a review, classification, and examination of the implications for the reproductive strategies of females. Ethology and Sociobiology 1:13-40. Hrdy, S. B. and G. Hausfater. 1984. Comparative and evolutionary perspectives on infanticide: introduction and overview. Infanticide: Comparative and evolutionary perspectives:xiii-xxxv. Hublart, M., D. Tetaert, D. Croix, F. Boutignon, P. Degand, and A. Boersma. 1990. Gonadotropic dysfunction produced by Trypanosoma brucei brucei in the rat. Acta tropica 47:177-184. Huchard, E., C. I. Canale, C. Le Gros, M. Perret, P.-Y. Henry, and P. M. Kappeler. 2012. Convenience polyandry or convenience polygyny? Costly sex under female control in a promiscuous primate. Proceedings of the Royal Society B: Biological Sciences 279:1371-1379. Huck, U. W., R. L. Soltis, and C. B. Coopersmith. 1982. Infanticide in male laboratory mice: effects of social status, prior sexual experience, and basis for discrimination between related and unrelated young. Animal Behaviour 30:1158-1165. Hull, E. and J. Dominguez. 2007a. Sexual behavior in male rodents. Hormones and behavior 52:45-55. Hull, E. M. and J. M. Dominguez. 2007b. Sexual behavior in male rodents. Hormones and behavior 52:45-55. Hunt, J., R. Brooks, M. Jennions, M. Smith, C. Bentsen, and L. Bussière. 2004. High- quality male field crickets invest heavily in sexual display but die young. Nature 432:1024. Hurst, J., C. Payne, C. Nevison, A. Marie, R. Humphries, D. Robertson, A. Cavaggioni, and R. Beynon. 2001. Individual recognition in mice mediated by major urinary proteins. Nature 414:631-634. Hurst, J. L. 1987. The functions of urine marking in a free-living population of house mice, Mus domesticus Rutty. Animal Behaviour 35:1433-1442. Hurst, J. L. 1989. The complex network of olfactory communication in populations of wild house mice Mus domesticus Rutty: urine marking and investigation within family groups. Animal Behaviour 37:705-725.

153

Hurst, J. L. 1990. Urine marking in populations of wild house mice Mus domesticus rutty. I. Communication between males. Animal Behaviour 40:209-222. Hurst, J. L. and R. J. Beynon. 2004. Scent wars: the chemobiology of competitive signalling in mice. Bioessays 26:1288-1298. Hurst, J. L., J. Fang, and C. Barnard. 1993. The role of substrate odours in maintaining social tolerance between male house mice, Mus musculus domesticus. Animal behaviour 45:997-1006. Hurst, J. L., D. H. Robertson, U. Tolladay, and R. J. Beynon. 1998. Proteins in urine scent marks of male house mice extend the longevity of olfactory signals. Animal behaviour 55:1289-1297. Hytten, F. 1980. Clinical physiology in obstetrics. Immler, S., S. Calhim, and T. R. Birkhead. 2008. Increased postcopulatory sexual selection reduces the intramale variation in sperm design. Evolution 62:1538-1543. Jennings, B., S. Ozanne, M. Dorling, and C. Hales. 1999. Early growth determines longevity in male rats and may be related to telomere shortening in the kidney. FEBS letters 448:4-8. Jennions, M. D. and M. Petrie. 2000. Why do females mate multiply? A review of the genetic benefits. Biol. Rev. 75:21-64. Johnston, R. E., D. Müller-Schwarze, and P. W. Sorenson. 2012. Advances in chemical signals in vertebrates. Springer Science & Business Media. Jones, R. and N. Nowell. 1973. Aversive and aggression-promoting properties of urine from dominant and subordinate male mice. Learning & Behavior 1:207- 210. Jones, R. and N. Nowell. 1974. A comparison of the aversive and female attractant properties of urine from dominant and subordinate male mice. Learning & Behavior 2:141-144. Jonker, F. H. 2004. Fetal death: comparative aspects in large domestic animals. Animal Reproduction Science 82-83:415-430. Jormalainen, V., S. Merilaita, and J. Riihimäki. 2001. Costs of intersexual conflict in the isopod Idotea baltica. Journal of Evolutionary Biology 14:763-772. Kajii, T. and K. Ohama. 1977. Androgenetic origin of hydatidiform mole. Nature 268:633. 154

Kalueff, A., Y.-R. Lou, I. Laaksi, and P. Tuohimaa. 2004. Increased grooming behavior in mice lacking vitamin D receptors. Physiology & behavior 82:405-409. Kalueff, A. V., J. W. Aldridge, J. L. LaPorte, D. L. Murphy, and P. Tuohimaa. 2007. Analyzing grooming microstructure in neurobehavioral experiments. Nature protocols 2:2538-2544. Kalueff, A. V., A. M. Stewart, C. Song, K. C. Berridge, A. M. Graybiel, and J. C. Fentress. 2015. Neurobiology of rodent self-grooming and its value for translational neuroscience. Nature Reviews Neuroscience 17:45. Kalueff, A. V. and P. Tuohimaa. 2004. Grooming analysis algorithm for neurobehavioural stress research. Brain Research Protocols 13:151-158. Kalueff, A. V. and P. Tuohimaa. 2005a. The grooming analysis algorithm discriminates between different levels of anxiety in rats: potential utility for neurobehavioural stress research. Journal of Neuroscience Methods 143:169- 177. Kalueff, A. V. and P. Tuohimaa. 2005b. Mouse grooming microstructure is a reliable anxiety marker bidirectionally sensitive to GABAergic drugs. European journal of pharmacology 508:147-153. Kaseda, Y., A. Khalil, and H. Ogawa. 1995. Harem stability and reproductive success of Misaki feral mares. Equine Veterinary Journal 27:368-372. Kavaliers, M., E. Choleris, and D. W. Pfaff. 2005. Recognition and avoidance of the odors of parasitized conspecifics and predators: differential genomic correlates. Neuroscience & Biobehavioral Reviews 29:1347-1359. Keay, S. D., M. Vatish, E. Karteris, E. W. Hillhouse, and H. S. Randeva. 2004. Review: The role of hCG in reproductive medicine. BJOG: An International Journal of Obstetrics & Gynaecology 111:1218-1228. Keller, L. F. and P. Arcese. 1998. No evidence for inbreeding avoidance in a natural population of song sparrows (Melospiza melodia). The American Naturalist 152:380-392. Keller, L. F. and D. M. Waller. 2002. Inbreeding effects in wild populations. Trends in Ecology & Evolution 17:230-241. Kennedy, H. F. and R. W. Elwood. 1988. Strain differences in the inhibition of infanticide in male mice (Mus musculus). Behavioral & Neural Biology.

155

Kerchner, M. 2004. Ultrasonic courtship vocalizations of adult male mice: a laboratory exercise illustrating comparable activation by either estradiol or testosterone. Journal of Undergraduate Neuroscience Education 2:A50. Kimura, T. and Y. Hagiwara. 1985. Regulation of urine marking in male and female mice: Effects of sex steroids. Hormones and Behavior 19:64-70. King, J. C. 2000. Physiology of pregnancy and nutrient metabolism–. The American journal of clinical nutrition 71:1218S-1225S. Kinsley, C. and B. Svare. 1988. Prenatal stress alters maternal aggression in mice. Physiology & behavior 42:7-13. Kirkwood, T. B. and S. N. Austad. 2000. Why do we age? Nature 408:233-238. Kirkwood, T. B. and R. Holliday. 1979. The evolution of ageing and longevity. Proceedings of the Royal Society of London B: Biological Sciences 205:531- 546. Klomberg, K. F., T. Garland, J. G. Swallow, and P. A. Carter. 2002. Dominance, plasma testosterone levels, and testis size in house mice artificially selected for high activity levels. Physiology & behavior 77:27-38. Kobayashi, H. 2015. The Impact of Maternal-Fetal Genetic Conflict Situations on the Pathogenesis of Preeclampsia. Biochemical Genetics 53:223-234. Kokko, H. and M. D. Jennions. 2014. The relationship between sexual selection and sexual conflict. Cold Spring Harbor perspectives in biology 6:a017517. Kokko, H. and I. Ots. 2006. When not to avoid inbreeding. Evolution 60:467-475. König, B. 1985. Die Brutpflegestrategie der Mäusemutter (Mus musculus): Eltern- Nachkommen-Konflikt, Toleranz gegen fremde Junge und Reaktion auf Nahrungsmittelmangel. König, B. and H. Markl. 1987. Maternal care in house mice. Behavioral Ecology and Sociobiology 20:1-9. König, B., J. Riester, and H. Markl. 1988. Maternal care in house mice (Mus musculus): II. The energy cost of lactation as a function of litter size. Journal of Zoology 216:195-210. Korgun, E. T., A. Ozmen, G. Unek, and I. Mendilcioglu. 2012. The effects of glucocorticoids on fetal and placental development. development 21:26.

156

Koteja, P., T. Garland, J. K. Sax, J. G. Swallow, and P. A. Carter. 1999. Behaviour of house mice artificially selected for high levels of voluntary wheel running. Animal behaviour 58:1307-1318. Kotilainen, T., M. Huhtinen, and T. Katila. 1994. Sperm-induced leukocytosis in the equine uterus. Theriogenology 41:629-636. Krebs, J. and N. Davies. 1993. Parental care and mating systems. An Introduction to Behavioural Ecology. London: Blackwell:208-243. Krupa, J. J., W. R. Leopold, and A. Sih. 1990. Avoidance of male giant water striders by females. Behaviour 115:247-253. Kubli, E. 2003. Sex-peptides: seminal peptides of the Drosophila male. Cellular and Molecular Life Sciences CMLS 60:1689-1704. Kurz, H., U. Zechner, A. Orth, and R. Fundele. 1999. Lack of correlation between placenta and offspring size in mouse interspecific crosses. Anatomy and embryology 200:335-343. Kwiatkowski, S., B. Dołegowska, E. Kwiatkowska, R. Rzepka, N. Marczuk, B. Loj, and A. Torbè. 2017. Maternal endothelial damage as a disorder shared by early preeclampsia, late preeclampsia and intrauterine growth restriction. Journal of perinatal medicine 45:793-802. Labov, J. B. 1981. Pregnancy blocking in rodents- adaptive advantages for the females. Am. Nat. 118:361-371. Labov, J. B., U. W. Huck, R. W. Elwood, and R. J. Brooks. 1985. Current problems in the study of infanticidal behaviour in rodents. Quart Rev Biol 60. Lack, D. 1947. The significance of clutch‐size. Ibis 89:302-352. Lack, D. 1954. The natural regulation of animal numbers. The Clarendon Press; Oxford. Lai, Z., S. Kalkunte, and S. Sharma. 2011. A Critical Role of Interleukin-10 in Modulating Hypoxia-Induced Preeclampsia-Like Disease in Mice. Hypertension 57:505-514. Larralde, C., J. Morales, I. Terrazas, T. Govezensky, and M. Romano. 1995. Sex hormone changes induced by the parasite lead to feminization of the male host in murine Taenia crassiceps cysticercosis. The Journal of steroid biochemistry and molecular biology 52:575-580.

157

Larsson, K. 1979. Features of the neuroendocrine regulation of masculine sexual behavior Raven Press New York, C Beyer (Ed.) Endocrine control of sexual behavior. Le Boeuf, B. J. and S. Mesnick. 1991. Sexual Behavior of Male Northern Elephant Seals: I. Lethal Injuries To Adult Females. Behaviour 116:143-162. Leighton, P. A., R. S. Ingram, J. Eggenschwiler, A. Efstratiadis, and S. M. Tilghman. 1995. Disruption of imprinting caused by deletion of the H19 gene region in mice. Nature 375:34. Leiser, R. and P. Kaufmann. 1994. Placental structure: In a comparative aspect. Experimental and Clinical Endocrinology and Diabetes 102:122-134. Lemaire, V., M. Koehl, M. Le Moal, and D. N. Abrous. 2000. Prenatal stress produces learning deficits associated with an inhibition of neurogenesis in the hippocampus. Proceedings of the National Academy of Sciences 97:11032- 11037. Lemaître, J.-F., S. A. Ramm, J. L. Hurst, and P. Stockley. 2012. Inbreeding avoidance behaviour of male bank voles in relation to social status. Animal behaviour 83:453-457. Leppänen, P. K., N. Ravaja, and S. Ewalds-Kvist. 2006. Twenty-three generations of mice bidirectionally selected for open-field thigmotaxis: selection response and repeated exposure to the open field. Behavioural processes 72:23-31. Lerch, S., C. Dormann, C. Brandwein, P. Gass, and S. Chourbaji. 2016. The scent of stress: environmental challenge in the peripartum environment of mice affects emotional behaviours of the adult offspring in a sex-specific manner. Laboratory Animals 50:167-178. Lessells, C. 2006. The evolutionary outcome of sexual conflict. Philosophical Transactions: Biological Sciences:301-317. Lessells, C. and G. Parker. 1999. Parent—offspring conflict: the full–sib—half–sib fallacy. Proceedings of the Royal Society of London B: Biological Sciences 266:1637-1643. Lewis, S. M. and S. N. Austad. 1994. Sexual selection in flour beetles: the relationship between sperm precedence and male olfactory attractiveness. Behavioral Ecology 5:223-224.

158

Lin, Y. C., Y. Rikihisa, H. Kono, and Y. Gu. 1990. Effects of larval tapeworm (Taenia taeniaeformis) infection on reproductive functions in male and female host rats. Experimental Parasitology 70:344-352. Linklater, W. L., E. Z. Cameron, E. O. Minot, and K. J. Stafford. 1999. Stallion harassment and the mating system of horses. Animal Behaviour 58:295-306. Liu, H. and E. Kubli. 2003. Sex-peptide is the molecular basis of the sperm effect in Drosophila melanogaster. Proceedings of the National Academy of Sciences 100:9929-9933. Lof, M., H. Olausson, K. Bostrom, B. Janerot-Sjöberg, A. Sohlstrom, and E. Forsum. 2005. Changes in basal metabolic rate during pregnancy in relation to changes in body weight and composition, cardiac output, insulin-like growth factor I, and thyroid hormones and in relation to fetal growth–. The American journal of clinical nutrition 81:678-685. Lofgren, J. L., S. E. Erdman, C. Hewes, C. Wong, R. King, T. E. Chavarria, A. R. Discua, J. G. Fox, and K. J. Maurer. 2012. Castration eliminates conspecific aggression in group-housed CD1 male surveillance mice (Mus musculus). Journal of the American Association for Laboratory Animal Science 51:594-599. Luckett, W. P. 1970. The fine structure of the placental villi of the rhesus monkey(Macaca mulatta). The Anatomical Record 167:141-163. Lukas, D. and E. Huchard. 2014. The evolution of infanticide by males in mammalian societies. Science 346:841-844. Luo, M., M. S. Fee, and L. C. Katz. 2003. Encoding pheromonal signals in the accessory olfactory bulb of behaving mice. Science 299:1196-1201. Lymbery, S. J. and L. W. Simmons. 2017. Males harm females less when competing with familiar relatives. Proceedings of the Royal Society B: Biological Sciences 284. Maccari, S., M. Darnaudery, S. Morley-Fletcher, A. Zuena, C. Cinque, and O. Van Reeth. 2003. Prenatal stress and long-term consequences: implications of glucocorticoid hormones. Neuroscience & Biobehavioral Reviews 27:119- 127. Macklon, N., J. Geraedts, and B. Fauser. 2002. Conception to ongoing pregnancy: the ‘black box’ of early pregnancy loss. Human Reproduction Update 8:333-343.

159

Mahajan, S. and A. Menge. 1967. Influence of reproductive phase on the inflammatory response and rate of sperm removal in the uterus and oviduct of the cow. American journal of veterinary research 28:1037. Mainguy, J., S. D. Côté, M. Festa-Bianchet, and D. W. Coltman. 2009. Father–offspring phenotypic correlations suggest intralocus sexual conflict for a fitness-linked trait in a wild sexually dimorphic mammal. Proceedings of the Royal Society B: Biological Sciences 276:4067-4075. Maklakov, A. A., T. Bilde, and Y. Lubin. 2006. Inter‐sexual combat and resource allocation into body parts in the spider, Stegodyphus lineatus. Ecological entomology 31:564-567. Maklakov, A. A. and Y. Lubin. 2004. Sexual conflict over mating in a spider: increased fecundity does not compensate for the costs of polyandry. Evolution 58:1135-1140. Maklakov, A. A. and Y. Lubin. 2006. Indirect genetic benefits of polyandry in a spider with direct costs of mating. Behavioral Ecology and Sociobiology 61:31-38. Mallory, F. and F. Clulow. 1977. Evidence of pregnancy failure in the wild meadow vole, Microtus pennsylvanicus. Canadian journal of zoology 55:1-17. Mallory, F. F. and R. J. Brooks. 1980. Infanticide and Pregnancy Failure: Reproductive Strategies in the Female Collared Lemming (Dicrostonyx groenlandicus). Biol. Reprod. 22:192-196. Mangel, M. and S. B. Munch. 2005. A life‐history perspective on short‐and long‐term consequences of compensatory growth. The American Naturalist 166:E155- E176. Manning, C. J., D. A. Dewsbury, E. K. Wakeland, and W. K. Potts. 1995. Communal nesting and communal nursing in house mice, Mus musculus domesticus. Animal Behaviour 50:741-751. Manno, T. G., A. P. Nesterova, L. M. DeBarbieri, S. E. Kennedy, K. S. Wright, and F. S. Dobson. 2007. Why do male Columbian ground squirrels give a mating call? Animal Behaviour 74:1319-1327. Marashi, V. and T. Rülicke. 2012. The Bruce Effect in Norway Rats1. Biol. Reprod. 86:17, 11-15-17, 11-15. Marshall, D. and T. Uller. 2007. When is a maternal effect adaptive? Oikos 116:1957- 1963. 160

Martin-Coello, J., H. Dopazo, L. Arbiza, J. Ausió, E. R. Roldan, and M. Gomendio. 2009. Sexual selection drives weak positive selection in protamine genes and high promoter divergence, enhancing sperm competitiveness. Proceedings of the Royal Society of London B: Biological Sciences:rspb. 2009.0257. Matochik, J. A., M. L. Sipos, J. G. Nyby, and R. J. Barfield. 1994. Intracranial androgenic activation of male-typical behaviors in house mice: motivation versus performance. Behavioural brain research 60:141-149. Mattner, P. E. 1968. The distribution of spermatozoa and leucocytes in the female genital tract in goats and cattle. Journal of Reproduction and Fertility 17:253- 261. McCarthy, F. P., J. C. Kingdom, L. C. Kenny, and S. K. Walsh. 2011. Animal models of preeclampsia; uses and limitations. Placenta 32:413-419. McCarthy, M. M. and F. S. vom Saal. 1985. The influence of reproductive state on infanticide by wild female house mice (Mus Musculus). Physiol Behav 35. McDonald, L. E., W. Black, S. McNutt, and L. Casida. 1952. The response of the rabbit uterus to instillation of semen at different phases of the estrous cycle. American journal of veterinary research 13:419. McEachern, M., R. L. McElreath, D. H. Van Vuren, and J. M. Eadie. 2009. Another genetically promiscuous ‘polygynous’ mammal: mating system variation in Neotoma fuscipes. Animal Behaviour 77:449-455. McGrath, J. and D. Solter. 1986. Nucleocytoplasmic interactions in the mouse embryo. Journal of Embryology and Experimental Morphology 97:277-289. McGraw, L. A., S. S. Suarez, and M. F. Wolfner. 2015. On a matter of seminal importance. BioEssays 37:142-147. McGregor, W. G., W. J. Raymoure, R. W. Kuhn, and R. B. Jaffe. 1981. Fetal tissue can synthesize a placental hormone. Evidence for chorionic gonadotropin beta- subunit synthesis by human fetal kidney. The Journal of Clinical Investigation 68:306-309. McGuire, B., E. Parker, and W. E. Bemis. 2007. Sex differences, effects of male presence and coordination of nest visits in prairie voles (Microtus ochrogaster) during the immediate postnatal period. The American midland naturalist 157:187-201.

161

McKinney, T. D. and C. Desjardins. 1973. Postnatal Development of the Testis, Fighting Behavior, and Fertility in House Mice1. Biol. Reprod. 9:279-294. Mead, R. A. 1993. Embryonic diapause in vertebrates. Journal of Experimental Zoology Part A: Ecological Genetics and Physiology 266:629-641. Meek, L. R., K. M. Burda, and E. Paster. 2000. Effects of prenatal stress on development in mice: maturation and learning. Physiology & behavior 71:543-549. Mendl, M. and E. S. Paul. 1989. Observation of nursing and sucking behaviour as an indicator of milk transfer and parental investment. Animal Behaviour. Metcalfe, N. B. and C. Alonso-Alvarez. 2010. Oxidative stress as a life‐history constraint: the role of reactive oxygen species in shaping phenotypes from conception to death. Functional Ecology 24:984-996. Metcalfe, N. B. and P. Monaghan. 2001. Compensation for a bad start: grow now, pay later? Trends in Ecology & Evolution 16:254-260. Mochizuki, A., Y. Takeda, and Y. Iwasa. 1996. The evolution of genomic imprinting. Genetics 144:1283-1295. Monaghan, P., N. B. Metcalfe, and R. Torres. 2009. Oxidative stress as a mediator of life history trade‐offs: mechanisms, measurements and interpretation. Ecology letters 12:75-92. Moore, C. L. and K. L. Power. 1986. Prenatal stress affects mother–infant interaction in Norway rats. Developmental psychobiology 19:235-245. Moore, T. and D. Haig. 1991. Genomic imprinting in mammalian development: a parental tug-of-war. Trends in Genetics 7:45-49. Morales, J., C. Larralde, M. Arteaga, T. Govezensky, M. Romano, and G. Morali. 1996. Inhibition of sexual behavior in male mice infected with Taenia crassiceps cysticerci. The Journal of parasitology:689-693. Morgan, I., I. McCarthy, and N. Metcalfe. 2000. Life‐history strategies and protein metabolism in overwintering juvenile Atlantic salmon: growth is enhanced in early migrants through lower protein turnover. Journal of Fish Biology 56:637-647. Morison, I. M., J. P. Ramsay, and H. G. Spencer. 2005. A census of mammalian imprinting. Trends in Genetics 21:457-465.

162

Morrow, E. H., G. Arnqvist, and S. Pitnick. 2003. Adaptation versus pleiotropy: why do males harm their mates? Behavioral Ecology 14:802-806. Mossman, C. A. and L. C. Drickamer. 1996. Odor preferences of female house mice (Mus domesticus) in seminatural enclosures. J. Comp. Psychol. 110:131. Mossman, H. W. 1987. Vertebrate fetal membranes. Rutgers University Press. Mühlhäuser, C. and W. U. Blanckenhorn. 2002. The costs of avoiding matings in the dung fly Sepsis cynipsea. Behavioral Ecology 13:359-365. Muir, J. L., H. P. Pfister, and A. Ivinskis. 1985. Effects of prepartum stress and postpartum enrichment on mother–infant interaction and offspring problem- solving ability in Rattus norvegicus. J. Comp. Psychol. 99:468. Muller, M. N., S. M. Kahlenberg, M. Emery Thompson, and R. W. Wrangham. 2007. Male coercion and the costs of promiscuous mating for female chimpanzees. Proceedings of the Royal Society B: Biological Sciences 274:1009-1014. Murrell, A., S. Heeson, L. Bowden, M. Constância, W. Dean, G. Kelsey, and W. Reik. 2001. An intragenic methylated region in the imprinted Igf2 gene augments transcription. EMBO reports 2:1101-1106. Naeye, R. L. 1981. Maternal blood pressure and fetal growth. American journal of obstetrics and gynecology 141:780-787. Nepomnaschy, P. A., K. B. Welch, D. S. McConnell, B. S. Low, B. I. Strassmann, and B. G. England. 2006. Cortisol levels and very early pregnancy loss in humans. Proc. Natl. Acad. Sci. U. S. A. 103:3938-3942. Nesse, R. and G. Williams. 1994. The New Science of Darwinian Medicine: Why We Get Sick. New York: Random House. Neubaum, D. M. and M. F. Wolfner. 1999. Mated Drosophila melanogaster females require a seminal fluid protein, Acp36DE, to store sperm efficiently. Genetics 153:845-857. Nicholls, R. D. and J. L. Knepper. 2001. Genome Organization, Function, and Imprinting in Prader-Willi and Angelman Syndromes. Annual Review of Genomics and Human Genetics 2:153-175. Niebyl , J. R. 2010. Nausea and Vomiting in Pregnancy. New England Journal of Medicine 363:1544-1550.

163

Nilsson, J.-A. and E. Svensson. 1996. The cost of reproduction: a new link between current reproductive effort and future reproductive success. Proceedings of the Royal Society of London B: Biological Sciences 263:711-714. Nobuyuki Yamaguchi, Hannah L Dugdale, and David W Macdonald. 2006. Female Receptivity, Embryonic Diapause, and Superfetation in the European Badger (Meles meles): Implications for the Reproductive Tactics of Males and Females. The Quarterly Review of Biology 81:33-48. Nussey, D. H., J. M. Pemberton, J. G. Pilkington, and J. D. Blount. 2009. Life history correlates of oxidative damage in a free‐living mammal population. Functional Ecology 23:809-817. O'Connell, S. M. and G. Cowlishaw. 1994. Infanticide avoidance, sperm competition and mate choice: the function of copulation calls in female baboons. Animal behaviour 48:687-694. Oakeshott, J. G. 1974. Social dominance, aggressiveness and mating success among male house mice (Mus musculus). Oecologia 15:143-158. Oftedal, O. T. 2000. Use of maternal reserves as a lactation strategy in large mammals. Proceedings of the Nutrition Society 59:99-106. Orr, A. and R. L. Rutowski. 1991. The function of the sphragis in Cressida cressida (Fab.)(Lepidoptera, Papilionidae): a visual deterrent to copulation attempts. Journal of Natural History 25:703-710. Ortiz, C. L., B. J. L. Boeuf, and D. P. Costa. 1984. Milk Intake of Elephant Seal Pups: An Index of Parental Investment. The American Naturalist 124:416-422. Ozanne, S. E. and C. N. Hales. 1999. The long-term consequences of intra-uterine protein malnutrition for glucose metabolism. Proceedings of the Nutrition Society 58:615-620. Palanza, P. L., K. L. Howdeshell, S. Parmigiani, and F. S. vom Saal. 2002. Exposure to a low dose of bisphenol A during fetal life or in adulthood alters maternal behavior in mice. Environmental health perspectives 110:415. Palme, R. and E. Möstl. 1997a. Measurement of cortisol metabolites in faeces of sheep as a parameter of cortisol concentration in blood. Zeitschrift fuer Saeugetierkunde (Germany).

164

Palme, R. and E. Möstl. 1997b. Measurement of cortisol metabolites in faeces of sheep as a parameter of cortisol concentration in blood. Int. J. Mammal. Biol. 62(Suppl. II): 19–197. . Palme, R., C. Touma, N. Arias, M. Dominchin, and M. Lepschy. 2013. Steroid extraction: get the best out of faecal samples. Wien Tierarztl Monatsschr 100:238-246. Pandya, I. J. and J. Cohen. 1985. The leukocytic reaction of the human uterine cervix to spermatozoa. Fertility and sterility 43:417-421. Parker, G. 1979. Sexual selection and sexual conflict. Sexual selection and reproductive competition in insects 123:166. Parker, G. A. 1970. Sperm competition and its evolutionary consequences in the insects. Biol. Rev. 45:525-567. Parker, G. A. 2006. Sexual conflict over mating and fertilization: an overview. Philosophical Transactions of the Royal Society of London B: Biological Sciences 361:235-259. Parkes, A. and H. M. Bruce. 1961. Olfactory stimuli in mammalian reproduction. Science 134:1049-1054. Parmigiani, S., P. Palanza, and P. Brain. 1989. Intraspecific maternal aggression in the house mouse (Mus domesticus): a counterstrategy to infanticide by male? Ethology Ecology & Evolution 1:341-352. Parmigiani, S., P. Palanza, J. Rodgers, and P. F. Ferrari. 1999. Selection, evolution of behavior and animal models in behavioral neuroscience. Neuroscience & Biobehavioral Reviews 23:957-970. Parmigiani, S. and F. S. Vom Saal. 1994. Infanticide and parental care. Taylor & Francis. Patten, M. M., L. Ross, J. P. Curley, D. C. Queller, R. Bonduriansky, and J. B. Wolf. 2014. The evolution of genomic imprinting: theories, predictions and empirical tests. Heredity 113:119-128. Peitz, B. and P. Olds-Clarke. 1986. Effects of seminal vesicle removal on fertility and uterine sperm motility in the house mouse. Biol. Reprod. 35:608-617. Pereira, M. E. 1983. Abortion following the immigration of an adult male baboon (Papio cynocephalus). American Journal of Primatology 4:93-98.

165

Perrigo, G., L. Belvin, and F. S. Vom Saal. 1991. Individual variation in the neural timing of infanticide and parental behavior in male house mice. Physiology & Behavior 50:287-296. Perrigo, G., W. C. Bryant, and F. S. vom Saal. 1990. A unique neural timing system prevents male mice from harming their own offspring. Animal Behaviour 39:535-539. Petry, C. J., M. L. Evans, D. L. Wingate, K. K. Ong, W. Reik, M. Constância, and D. B. Dunger. 2010. Raised Late Pregnancy Glucose Concentrations in Mice Carrying Pups With Targeted Disruption of H19. Diabetes 59:282-286. Petry, C. J., K. K. Ong, B. J. Barratt, D. Wingate, H. J. Cordell, S. M. Ring, M. E. Pembrey, W. Reik, J. A. Todd, and D. B. Dunger. 2005. Common polymorphism in H19 associated with birthweight and cord blood IGF-II levels in humans. BMC genetics 6:22. Pichaud, N., E. H. Chatelain, J. W. O. Ballard, R. Tanguay, G. Morrow, and P. U. Blier. 2010. Thermal sensitivity of mitochondrial metabolism in two distinct mitotypes of Drosophila simulans: evaluation of mitochondrial plasticity. Journal of Experimental Biology 213:1665-1675. Pichaud, N., J. Pellerin, M. Fournier, S. Gauthier-Clerc, P. Rioux, and É. Pelletier. 2008. Oxidative stress and immunologic responses following a dietary exposure to PAHs in Mya arenaria. Chemistry Central Journal 2:1. Pizzari, T., H. Lø, and C. K. Cornwallis. 2004. Sex–specific, counteracting responses to inbreeding in a bird. Proceedings of the Royal Society of London B: Biological Sciences 271:2115-2121. Pizzari, T. and R. R. Snook. 2003. Perspective: sexual conflict and sexual selection: chasing away paradigm shifts. Evolution 57:1223-1236. Platek, S. M. and T. K. Shackelford. 2006. Female infidelity and paternal uncertainty: Evolutionary perspectives on male anti-cuckoldry tactics. Cambridge University Press. Polak, M., L. L. Wolf, W. T. Starmer, and J. Barker. 2001. Function of the mating plug in Drosophila hibisci Bock. Behavioral Ecology and Sociobiology 49:196-205. Pollux, B., M. Pires, A. Banet, and D. Reznick. 2009. Evolution of placentas in the fish family Poeciliidae: an empirical study of macroevolution. Annu. Rev. Ecol. Evol. Syst. 40:271-289. 166

Pontier, D., J. M. Gaillard, D. Allaine, J. Trouvilliez, I. Gordon, and P. Duncan. 1989. Postnatal-growth rate and adult body-weight in mammals - a new approach. Oecologia 80:390-394. Pradhan, G. and C. van Schaik. 2008. Infanticide-driven intersexual conflict over matings in primates and its effects on social organization. Behaviour 145:251-275. Price, T. and D. Hosken. 2007. Evolution: good males are bad females. Current biology 17:R168-R170. Profet, M. 1988. The evolution of pregnancy sickness as protection to the embryo against Pleistocene teratogens. Evolutionary Theory 8:177-190. Prowse, N. and L. Partridge. 1997. The effects of reproduction on longevity and fertility in male Drosophila melanogaster. Journal of Insect Physiology 43:501-512. Pusey, A. E. and C. Packer. 1994. Infanticide in lions: consequences and counterstrategies. Infanticide and parental care:277-299. Qazi, B. 2003. A potential mechanism for cryptic female choice in a flour beetle. Journal of evolutionary biology 16:170-176. Queen, K., C. Dhabuwala, and C. Pierrepoint. 1981. The effect of the removal of the various accessory sex glands on the fertility of male rats. Journal of reproduction and fertility 62:423-426. Rabinovitz, S., Y. Kaufman, G. Ludwig, A. Razin, and R. Shemer. 2012. Mechanisms of activation of the paternally expressed genes by the Prader-Willi imprinting center in the Prader-Willi/Angelman syndromes domains. Proceedings of the National Academy of Sciences 109:7403-7408. Radespiel, U., V. Dal Secco, C. Drögemüller, P. Braune, E. Labes, and E. Zimmermann. 2002. Sexual selection, multiple mating and paternity in grey mouse lemurs, Microcebus murinus. Animal Behaviour 63:259-268. Rajendren, G. and C. Dominic. 1993. The male-induced implantation failure (the Bruce effect) in mice: effect of exogenous progesterone on maintenance of pregnancy in male-exposed females. Experimental and Clinical Endocrinology & Diabetes 101:356-359.

167

Ramsey, E. M., M. L. Houston, and J. W. Harris. 1976. Interactions of the trophoblast and maternal tissues in three closely related primate species. American Journal of Obstetrics & Gynecology 124:647-652. Rankin, D. J., U. Dieckmann, and H. Kokko. 2011. Sexual conflict and the tragedy of the commons. The American Naturalist 177:780-791. Réale, D., P. Boussès, and J.-L. Chapuis. 1996. Female-biased mortality induced by male sexual harassment in a feral sheep population. Canadian Journal of Zoology 74:1812-1818. Reik, W., M. Constancia, W. Dean, K. Davies, L. Bowden, A. Murrell, R. Feil, J. Walter, and G. Kelsey. 2000. Igf2 imprinting in development and disease. Pp. 93-104. Chromosomes today. Springer. Reik, W., M. Constância, A. Fowden, N. Anderson, W. Dean, A. Ferguson‐Smith, B. Tycko, and C. Sibley. 2003. Regulation of supply and demand for maternal nutrients in mammals by imprinted genes. The Journal of physiology 547:35- 44. Reik, W. and J. Walter. 2001. Genomic imprinting: parental influence on the genome. Nature Reviews Genetics 2:21. Renfree, M. B., S. Suzuki, and T. Kaneko-Ishino. 2013. The origin and evolution of genomic imprinting and viviparity in mammals. Philosophical Transactions of the Royal Society B: Biological Sciences 368:20120151. Reznick, D. 1985. Costs of reproduction: an evaluation of the empirical evidence. Oikos:257-267. Rice, W. and A. Chippindale. 2001. Intersexual ontogenetic conflict. Journal of Evolutionary Biology 14:685-693. Rice, W. R. 1992. Sexually antagonistic genes: experimental evidence. Science 256:1436-1439. Rice, W. R. 1998. Intergenomic conflict, interlocus antagonistic coevolution, and the evolution of reproductive isolation. Endless forms: species and speciation:261-270. Rich, T. J. and J. L. Hurst. 1998. Scent marks as reliable signals of the competitive ability of mates. Animal Behaviour 56:727-735.

168

Rioux-Paquette, E., M. Festa-Bianchet, and D. W. Coltman. 2010. No inbreeding avoidance in an isolated population of bighorn sheep. Animal Behaviour 80:865-871. Risco, C., G. Donovan, and J. Hernandez. 1999. Clinical mastitis associated with abortion in dairy cows. Journal of dairy science 82:1684-1689. Robbins, L. W., G. D. Hartman, and M. H. Smith. 1987. Dispersal, reproductive strategies, and the maintenance of genetic variability in mosquitofish (Gambusia affinis). Copeia:156-164. Roberts, E. K., A. Lu, T. J. Bergman, and J. C. Beehner. 2012. A Bruce effect in wild geladas. Science 335:1222-1225. Roberts, M. L., K. L. Buchanan, and M. Evans. 2004. Testing the immunocompetence handicap hypothesis: a review of the evidence. Animal Behaviour 68:227- 239. Robertson, S. A. 2005. Seminal plasma and male factor signalling in the female reproductive tract. Cell and tissue research 322:43-52. Robertson, S. A. 2007. Seminal fluid signaling in the female reproductive tract: Lessons from rodents and pigs1. Journal of Animal Science 85:E36-E44. Rodriguez-Sosa, J. R., H. Dobson, and A. Hahnel. 2006. Isolation and transplantation of spermatogonia in sheep. Theriogenology 66:2091-2103. Rodríguez‐Martínez, H., U. Kvist, J. Ernerudh, L. Sanz, and J. J. Calvete. 2011. Seminal plasma proteins: what role do they play? American Journal of Reproductive Immunology 66:11-22. Rohrbach, C. 1982. Investigation of the Bruce effect in the Mongolian gerbil (Meriones unguiculatus). Journal of reproduction and fertility 65:411-417. Rollo, C. D. 2002. Growth negatively impacts the life span of mammals. Evolution & development 4:55-61. Ronkainen, K., A. Kaitala, and R. Huttunen. 2005. The effect of abdominal spines on female mating frequency and fecundity in a water strider. Journal of insect behavior 18:619-631. Rönn, J., M. Katvala, and G. Arnqvist. 2007. Coevolution between harmful male genitalia and female resistance in seed beetles. Proceedings of the National Academy of Sciences 104:10921-10925.

169

Ross, C. 1988. The intrinsic rate of natural increase and reproductive effort in primates. Journal of Zoology 214:199-219. Rossiter, M. 1996. Incidence and consequences of inherited environmental effects. Annual review of ecology and systematics 27:451-476. Rottenstein, F. 1989. Superfetation in a pig. Magyar Allatorvosok Lapja 44:42-42. Rowe, L. 1994. The costs of mating and mate choice in water striders. Animal Behaviour 48:1049-1056. Rubenstein, D. 1986. 13. Ecology and Sociality in Horses and Zebras. Rueda-Clausen, C., J. Stanley, D. Thambiraj, R. Poudel, S. Davidge, and P. Baker. 2014. Effect of prenatal hypoxia in transgenic mouse models of preeclampsia and fetal growth restriction. Reproductive Sciences 21:492-502. Saller, D. N. J. and J. A. Canick. 2008. Current Methods of Prenatal Screening for Down Syndrome and Other Fetal Abnormalities. Clinical Obstetrics and Gynecology 51:24-36. Samuels, S. E. and V. E. Baracos. 1995. Tissue protein turnover is altered during catch-up growth following Escherichia coli infection in weanling rats. The Journal of nutrition 125:520. Sapolsky, R. M., L. M. Romero, and A. U. Munck. 2000. How do glucocorticoids influence stress responses? Integrating permissive, suppressive, stimulatory, and preparative actions. Endocrine reviews 21:55-89. Scantlebury, M., A. Russell, G. McIlrath, J. Speakman, and T. Clutton-Brock. 2002. The energetics of lactation in cooperatively breeding meerkats Suricata suricatta. Proceedings of the Royal Society of London B: Biological Sciences 269:2147- 2153. Schellinck, H. M., E. Rooney, and R. E. Brown. 1995. Odors of individuality of germfree mice are not discriminated by rats in a habituation-dishabituation procedure. Physiology & behavior 57:1005-1008. Schjenken, J. and S. Robertson. 2014. Seminal fluid and immune adaptation for pregnancy–comparative biology in mammalian species. Reproduction in domestic animals 49:27-36. Schjenken, J. E. and S. A. Robertson. 2015. Seminal Fluid Signalling in the Female Reproductive Tract: Implications for Reproductive Success and Offspring

170

Health. Pp. 127-158 in R. Bronson, ed. The Male Role in Pregnancy Loss and Embryo Implantation Failure. Springer International Publishing, Cham. Schoröder, I. 1993. Concealed ovulation and clandestine copulation: a female contribution to human evolution. Ethology and Sociobiology 14:381-389. Schwagmeyer, P. 1979. The Bruce effect: an evaluation of male/female advantages. The American Naturalist 114:932-938. Scott, J. L., N. Ketheesan, and P. M. Summers. 2006. Leucocyte population changes in the reproductive tract of the ewe in response to insemination. Reproduction, Fertility and Development 18:627-634. Scott, R. J. and M. Spielman. 2006. Genomic imprinting in plants and mammals: how life history constrains convergence. Cytogenetic and genome research 113:53-67. Sferruzzi-Perri, A., O. Vaughan, P. Coan, M. Suciu, R. Darbyshire, M. Constancia, G. Burton, and A. Fowden. 2011. Placental-specific Igf2 deficiency alters developmental adaptations to undernutrition in mice. Endocrinology 152:3202-3212. Shackelford, R. M. 1952. Superfetation in the ranch mink. The American Naturalist 86:311-319. Sheldon, B. C. and S. A. West. 2004. Maternal dominance, maternal condition, and offspring sex ratio in ungulate mammals. The American Naturalist 163:40-54. Sibley, C. P., P. M. Coan, A. C. Ferguson-Smith, W. Dean, J. Hughes, P. Smith, W. Reik, G. J. Burton, A. L. Fowden, and M. Constância. 2004. Placental-specific insulin- like growth factor 2 (Igf2) regulates the diffusional exchange characteristics of the mouse placenta. Proc. Natl. Acad. Sci. U. S. A. 101:8204-8208. Silasi, M., B. Cohen, S. A. Karumanchi, and S. Rana. 2010. Abnormal placentation, angiogenic factors, and the pathogenesis of preeclampsia. Obstetrics and gynecology clinics of North America 37:239-253. Simmons, L. and M. Siva-Jothy. 1998. 10 Sperm Competition in Insects: Mechanisms and the Potential. Sperm competition and sexual selection:341. Simmons, L. W. 2001. Sperm competition and its evolutionary consequences in the insects. Princeton University Press.

171

Simmons, L. W. 2005. The evolution of polyandry: sperm competition, sperm selection, and offspring viability. Annual Review of Ecology, Evolution, and Systematics 36. Simon, N. G. and S. Lu. 2006. Androgens and aggression. Biology of aggression, ed. RJ Nelson:211-230. Sims, D. 2003. Tractable models for testing theories about natural strategies: foraging behaviour and habitat selection of free‐ranging sharks. Journal of Fish Biology 63:53-73. Sims, D., E. Southall, V. Wearmouth, N. Hutchinson, G. Budd, and D. Morritt. 2005. Refuging behaviour in the nursehound Scyliorhinus stellaris (Chondrichthyes: Elasmobranchii): preliminary evidence from acoustic telemetry. Marine Biological Association of the United Kingdom. Journal of the Marine Biological Association of the United Kingdom 85:1137. Smith, J. M. and D. Harper. 2003. Animal signals. Oxford University Press. Smith, K., A. Blunden, K. Whitwell, K. Dunn, and A. Wales. 2003. A survey of equine abortion, stillbirth and neonatal death in the UK from 1988 to 1997. Equine veterinary journal 35:496-501. Smits, G., A. J. Mungall, S. Griffiths-Jones, P. Smith, D. Beury, L. Matthews, J. Rogers, A. J. Pask, G. Shaw, and J. L. VandeBerg. 2008. Conservation of the H19 noncoding RNA and H19-IGF2 imprinting mechanism in therians. Nature genetics 40:971. Smultea, M. A. 1994. Segregation by humpback whale (Megaptera novaeangliae) cows with a calf in coastal habitat near the island of Hawaii. Canadian Journal of Zoology 72:805-811. Smuts, B. B. and R. W. Smuts. 1993. Male aggression and sexual coercion of females in nonhuman primates and other mammals: evidence and theoretical implications. Advances in the Study of Behavior 22:1-63. Soini, H. A., D. Wiesler, S. Koyama, C. Féron, C. Baudoin, and M. V. Novotny. 2009. Comparison of urinary scents of two related mouse species, Mus spicilegus and Mus domesticus. Journal of chemical ecology 35:580-589. Somers, M. J., O. A. E. Rasa, and B. L. Penzhorn. 1995. Group structure and social behaviour of warthogs Phacochoerus aethiopicus. Acta Theriologica 40:257- 281. 172

Speakman, J. R. 2008. The physiological costs of reproduction in small mammals. Philosophical Transactions of the Royal Society B: Biological Sciences 363:375-398. Speakman, J. R. and M. Garratt. 2014. Oxidative stress as a cost of reproduction: Beyond the simplistic trade‐off model. Bioessays 36:93-106. Spindler, K. 1988. Parasites and hormones. Parasitology in focus. Berlin Heidelberg New York: Springer:465-473. Spruijt, B. M., J. A. v. Hooff, and W. H. Gispen. 1992. Ethology and neurobiology of grooming behavior. Physiological Reviews 72:825-852. Spruijt, B. M., P. Welbergen, J. A. N. Brakkee, and W. H. Gispen. 1988. An Ethological Analysis of Excessive Grooming in Young and Aged Rats. Annals of the New York Academy of Sciences 525:89-100. Staff, A. C., S. J. Benton, P. von Dadelszen, J. M. Roberts, R. N. Taylor, R. W. Powers, D. S. Charnock-Jones, and C. W. G. Redman. 2013. Redefining Preeclampsia Using Placenta-Derived Biomarkers. Hypertension 61:932-942. Stearns, S. C. 1992. The evolution of life histories. Oxford University Press Oxford. Steck, T. and S. Bussen. 1997. Conception during pregnancy (superfetation). Human reproduction (Oxford, England) 12:1835. Stehn, R. and F. Jannett. 1981. Male-induced abortion in various microtine rodents. Journal of Mammalogy 62:369-372. Stockley, P. 2003. Female multiple mating behaviour, early reproductive failure and litter size variation in mammals. Proceedings of the Royal Society of London B: Biological Sciences 270:271-278. Stockley, P., J. B. Searle, D. W. Macdonald, and C. S. Jones. 1993. Female Multiple Mating Behaviour in the Common Shrew as a Strategy to Reduce Inbreeding. Proceedings of the Royal Society of London. Series B: Biological Sciences 254:173-179. Stone, G. N., P. M. Loder, and T. M Blackburn. 1995. Foraging and courtship behaviour in males of the solitary bee Anthophora plumipes (Hymenoptera: Anthophoridae): thermal physiology and the roles of body size. Ecological Entomology 20:169-183.

173

Storey, A. E. 1986. Influence of sires on male-induced pregnancy disruptions in meadow voles (microtus-pennsylvanicus) differs with stage of pregnancy. J. Comp. Psychol. 100:15-20. Stormshak, F., M. B. Zelinski-Wooten, and S. E. Abdelgadir. 1987. Comparative aspects of the regulation of corpus luteum function in various species. Pp. 327-360. Regulation of Ovarian and Testicular Function. Springer. Sundaresan, S. R., I. R. Fischhoff, and D. Rubenstein. 2007. Male harassment influences female movements and associations in Grevy's zebra (Equus grevyi). Behavioral Ecology 18:860-865. Surani, M., S. C. Barton, and M. Norris. 1984. Development of reconstituted mouse eggs suggests imprinting of the genome during gametogenesis. Nature 308:548-550. Suzuki, S., R. Ono, T. Narita, A. J. Pask, G. Shaw, C. Wang, T. Kohda, A. E. Alsop, J. A. M. Graves, and Y. Kohara. 2007. Retrotransposon silencing by DNA methylation can drive mammalian genomic imprinting. PLoS Genetics 3:e55. Svare, B. B. 1981. Maternal aggression in mammals. Parental care in mammals:179- 210. Swedell, L. and J. Saunders. 2006. Infant mortality, paternity certainty, and female reproductive strategies in hamadryas baboons. Pp. 19-51. Reproduction and fitness in baboons: Behavioral, ecological, and life history perspectives. Springer. Szulkin, M., K. V. Stopher, J. M. Pemberton, and J. M. Reid. 2013. Inbreeding avoidance, tolerance, or preference in animals? Trends in Ecology & Evolution 28:205-211. Szulkin, M., P. Zelazowski, G. Nicholson, and B. C. Sheldon. 2009. Inbreeding avoidance under different null models of random mating in the great tit. Journal of Animal Ecology 78:778-788. Szuran, T., E. Zimmerman, V. Pliska, H. P. Pfister, and H. Welzl. 1991. Prenatal stress effects on exploratory activity and stress-induced analgesia in rats. Developmental Psychobiology 24:361-372. Takami, Y., M. Sasabe, N. Nagata, and T. Sota. 2008. Dual function of seminal substances for mate guarding in a ground beetle. Behavioral Ecology 19:1173-1178. 174

Tallarek, A. and H. Stepan. 2012. Preeclampsia and HELLP syndrome as an obstetric emergency. Medizinische Klinik, Intensivmedizin und Notfallmedizin 107:96- 100. Tallquist, M. D. and P. Soriano. 2000. Epiblast-restricted Cre expression in MORE mice: A tool to distinguish embryonic vs. extra-embryonic gene function. genesis 26:113-115. Tamashiro, K. L., C. E. Terrillion, J. Hyun, J. I. Koenig, and T. H. Moran. 2009. Prenatal stress or high fat diet increases susceptibility to diet-induced obesity in rat offspring. Diabetes. Tavares, M., A. Carraro, A. V. Favaretto, S. Petenusci, R. Lopes, R. Ribeiro, and T. L. Carvalho. 1994. The male reproductive organs in experimental Chagas' disease: III. Plasma testosterone and accessory sex glands in the acute phase of the disease. Experimental and Toxicologic Pathology 46:243-246. Teichroeb, J. A., E. C. Wikberg, I. Bădescu, L. J. Macdonald, and P. Sicotte. 2012. Infanticide risk and male quality influence optimal group composition for Colobus vellerosus. Behavioral Ecology:ars128. Terborgh, J. and A. W. Goldizen. 1985. On the mating system of the cooperatively breeding saddle-backed tamarin (Saguinus fuscicollis). Behavioral Ecology and Sociobiology 16:293-299. Thompson, M. E., M. N. Muller, S. M. Kahlenberg, and R. W. Wrangham. 2010. Dynamics of social and energetic stress in wild female chimpanzees. Hormones and Behavior 58:440-449. Thomson, R., M. Taha, A. Napier, and K. Welkesa. 2013. The induction of pregnancy block in mice by bodily fluids via the vomeronasal organ. Biochem Physiol 2:2. Thonhauser, K. E., S. Raveh, A. Hettyey, H. Beissmann, and D. J. Penn. 2013. Why do female mice mate with multiple males? Behavioral ecology and sociobiology 67:1961-1970. Thornhill, R. 1983. Cryptic female choice and its implications in the scorpionfly Harpobittacus nigriceps. The American Naturalist 122:765-788. Thornhill, R. and J. Alcock. 1983. The evolution of insect mating systems. Harvard University Press.

175

Thornton, J. G. and J. L. Onwude. 1992. Convulsions in pregnancy in related gorillas. American Journal of Obstetrics & Gynecology 167:240-241. Thorvaldsen, J. L., K. L. Duran, and M. S. Bartolomei. 1998. Deletion of the H19 differentially methylated domain results in loss of imprinted expression of H19 and Igf2. Genes & development 12:3693-3702. Thünken, T., T. C. Bakker, S. A. Baldauf, and H. Kullmann. 2007. Active inbreeding in a cichlid fish and its adaptive significance. Current Biology 17:225-229. Tilbrook, A., A. Turner, and I. Clarke. 2000. Effects of stress on reproduction in non- rodent mammals: the role of glucocorticoids and sex differences. Reviews of reproduction 5:105-113. Tong, S., E. M. Wallace, and L. Rombauts. 2006. Association between low day 16 hCG and miscarriage after proven cardiac activity. Obstetrics & Gynecology 107:300-304. Touma, C., R. Palme, and N. Sachser. 2004. Analyzing corticosterone metabolites in fecal samples of mice: a noninvasive technique to monitor stress hormones. Hormones and Behavior 45:10-22. Touma, C., N. Sachser, E. Möstl, and R. Palme. 2003. Effects of sex and time of day on metabolism and excretion of corticosterone in urine and feces of mice. General and comparative endocrinology 130:267-278. Tregenza, T. and N. Wedell. 2000. Genetic compatibility, mate choice and patterns of parentage: Invited Review. Molecular Ecology 9:1013-1027. Tregenza, T., N. Wedell, and T. Chapman. 2006. Introduction. Sexual conflict: a new paradigm? The Royal Society. Tremellen, K. P., D. Valbuena, J. Landeras, A. Ballesteros, J. Martinez, S. Mendoza, R. J. Norman, S. A. Robertson, and C. Simón. 2000. The effect of intercourse on pregnancy rates during assisted human reproduction. Human Reproduction 15:2653-2658. Trivers, R. 1972. Parental investment and sexual selection. Sexual Selection & the Descent of Man, Aldine de Gruyter, New York:136-179. Tycko, B. and I. M. Morison. 2002. Physiological functions of imprinted genes. Journal of cellular physiology 192:245-258. Úbeda, F. and A. Gardner. 2012. A model for genomic imprinting in the social brain: elders. Evolution 66:1567-1581. 176

Ubeda, F. and D. Haig. 2003. Dividing the child. Genetica 117:103-110. Van Cleve, J. and M. W. Feldman. 2007. Sex-specific viability, sex linkage and dominance in genomic imprinting. Genetics 176:1101-1118. Van Noordwijk, M. 2000. Reproductive patterns in eutherian mammals: Adaptations against infanticide? Infanticide by males and its implications:322-360. Van Schaik, C. P. and C. H. Janson. 2000a. Infanticide by males and its implications. Cambridge University Press. Van Schaik, C. P. and C. H. Janson. 2000b. Social counterstrategies against infanticide by males in primates and other mammals. In: Kappeler PM, editor. Primate males, Cambridge University Press 34-54. Van Schaik, C. P. and P. M. Kappeler. 1997. Infanticide risk and the evolution of male–female association in primates. Proceedings of the Royal Society of London B: Biological Sciences 264:1687-1694. Van Schaik, C. P., G. R. Pradhan, and M. A. van Noordwijk. 2004. Mating conflict in primates: infanticide, sexual harassment and female sexuality. Sexual selection in primates: new and comparative perspectives. Cambridge University Press, Cambridge:131-150. Vandenbergh, J. G. 1967. Effect of the presence of a male on the sexual maturation of female mice. Endocrinology 81:345-349. Vasilaki, A., A. Mansouri, H. Van Remmen, J. Van Der Meulen, L. Larkin, A. Richardson, A. McArdle, J. Faulkner, and M. Jackson. 2006a. Free radical generation by skeletal muscle of adult and old mice: effect of contractile activity. Aging cell 5:109-117. Vasilaki, A., A. Mansouri, H. Van Remmen, J. H. Van Der Meulen, L. Larkin, A. G. Richardson, A. McArdle, J. A. Faulkner, and M. J. Jackson. 2006b. Free radical generation by skeletal muscle of adult and old mice: effect of contractile activity. Aging Cell 5:109-117. Vogel, P. 2005. The current molecular phylogeny of Eutherian mammals challenges previous interpretations of placental evolution. Placenta 26:591-596. Vom Saal, F. S. 1985. Time-contingent change in infanticide and parental behavior induced by ejaculation in male mice. Physiology & behavior 34:7-15. Vrana, P. B., X.-J. Guan, R. S. Ingram, and S. M. Tilghman. 1998. Genomic imprinting is disrupted in interspecific Peromyscus hybrids. Nature Genetics 20:362. 177

Wake, N., Y. Shiina, and K. Ichinoe. 1978. A further cytogenetic study of hydatidiform mole, with reference to its androgenetic origin. Proceedings of the Japan Academy, Series B 54:533-537. Wang, X., D. C. Miller, R. Harman, D. F. Antczak, and A. G. Clark. 2013. Paternally expressed genes predominate in the placenta. Proc. Natl. Acad. Sci. U. S. A. 110:10705-10710. Waser, P. M., S. N. Austad, and B. Keane. 1986. When should animals tolerate inbreeding? The American Naturalist 128:529-537. Waterland, R. A. and C. Garza. 1999. Potential mechanisms of metabolic imprinting that lead to chronic disease. The American Journal of Clinical Nutrition 69:179-197. Watson, P., Robert R. Stallmann, and Göran Arnqvist. 1998a. Sexual Conflict and the Energetic Costs of Mating and Mate Choice in Water Striders. The American Naturalist 151:46-58. Watson, P. J., R. R. Stallmann, and G. Arnqvist. 1998b. Sexual conflict and the energetic costs of mating and mate choice in water striders. The American Naturalist 151:46-58. Weber, E. M., B. Algers, J. Hultgren, and I. A. S. Olsson. 2013. Pup mortality in laboratory mice – infanticide or not? Acta Veterinaria Scandinavica 55:83. Weber, E. M. and I. A. S. Olsson. 2008. Maternal behaviour in Mus musculus sp.: an ethological review. Applied Animal Behaviour Science 114:1-22. Wedell, N., C. Kvarnemo, and T. Tregenza. 2006. Sexual conflict and life histories. Animal Behaviour 71:999-1011. Wee, B. E., D. R. Weaver, and L. G. Clemens. 1988. Hormonal restoration of masculine sexual behavior in long-term castrated B6D2F1 mice. Physiology & behavior 42:77-82. Weir, B. J. 1974. Reproductive characteristics of hystricomorph rodents. Pp. 265- 301. Symposia of the Zoological Society of London. Wells, D. A., D. N. Jones, D. Bulger, and C. Brown. 2014. Male brush-turkeys attempt sexual coercion in unusual circumstances. Behavioural Processes 106:180- 186. West, S. A., I. Pen, and A. S. Griffin. 2002. Cooperation and competition between relatives. Science 296:72-75. 178

Whitten, W. 1966. Pheromones and mammalian reproduction. Advances in reproductive physiology 1:155-&. WHO and UNICEF. 2012. UNFPA, The World Bank. Trends in maternal mortality: 1990 to 2010. WHO, UNICEF. UNFPA, and The World Bank Estimates. Wiegand, G. and S. J. Remington. 1986a. Citrate synthase - structure, control, and mechanism. Annual Review of Biophysics and Biophysical Chemistry 15:97- 117. Wiegand, G. and S. J. Remington. 1986b. Citrate synthase: structure, control, and mechanism. Annual review of biophysics and biophysical chemistry 15:97- 117. Wielgus, R. B. and F. L. Bunnell. 2000. Possible negative effects of adult male mortality on female grizzly bear reproduction. Biological Conservation 93:145-154. Wigby, S. and T. Chapman. 2004. Female resistance to male harm evolves in response to manipulation of sexual conflict. Evolution 58:1028-1037. Wigby, S. and T. Chapman. 2005. Sex peptide causes mating costs in female Drosophila melanogaster. Current Biology 15:316-321. Wilkin, F., J. Paquette, E. Ledru, C. Mamelin, M. Pollak, and C. L. Deal. 2000. H19 sense and antisense transgenes modify insulin-like growth factor-II mRNA levels. European Journal of Biochemistry 267:4020-4027. Wilkins, J. F. and D. Haig. 2003. What good is genomic imprinting: the function of parent-specific gene expression. Nature Reviews Genetics 4:359-368. Williams, G. C. 1966. Adaptation and natrual selection: a critique of some current evolutionary thought. Willis, C. and R. Poulin. 2000. Preference of female rats for the odours of non- parasitised males: the smell of good genes? Folia parasitologica 47:6-10. Wingfield, J. and R. Sapolsky. 2003. Reproduction and resistance to stress: when and how. Journal of neuroendocrinology 15:711-724. Winn, V. D., M. Gormley, and S. J. Fisher. 2011. The impact of preeclampsia on gene expression at the maternal–fetal interface. Pregnancy Hypertension: An International Journal of Women's Cardiovascular Health 1:100-108. Wolf, J. and R. Hager. 2009. Selective abortion and the evolution of genomic imprinting. Journal of evolutionary biology 22:2519-2523. 179

Wolf, J. B. 2013. Evolution of genomic imprinting as a coordinator of coadapted gene expression. Proceedings of the National Academy of Sciences 110:5085-5090. Wolf, J. B. and R. Hager. 2006. A maternal–offspring coadaptation theory for the evolution of genomic imprinting. PLoS biology 4:e380. Wolf, J. B., G. Kauermann, and F. Trillmich. 2005. Males in the shade: habitat use and sexual segregation in the Galapagos sea lion (Zalophus californianus wollebaeki). Behavioral Ecology and Sociobiology 59:293-302. Wolff, J. O. and D. W. Macdonald. 2004. Promiscuous females protect their offspring. Trends in Ecology & Evolution 19:127-134. Wolff, J. O. and P. W. Sherman. 2008. Rodent societies: an ecological and evolutionary perspective. University of Chicago Press. Wolfner, M. 2002. The gifts that keep on giving: physiological functions and evolutionary dynamics of male seminal proteins in Drosophila. Heredity 88:85. Wooding, P. and G. Burton. 2008. Comparative placentation: structures, functions and evolution. Springer Science & Business Media. Woodside, B. 2007. Prolactin and the hyperphagia of lactation. Physiology & Behavior 91:375-382. Xiao, K., Y. Kondo, and Y. Sakuma. 2004. Sex-specific effects of gonadal steroids on conspecific odor preference in the rat. Hormones and behavior 46:356-361. Yamaguchi, N., H. L. Dugdale, and D. W. Macdonald. 2006. Female Receptiveity, Embryonic Diapause, and Superfetation in the European Badger (Meles Meles: Implications for the Reproductive Tactics of Males and Females. The Quarterly Review of Biology 81:33-48. Yang, J., H. Han, J. Cao, L. Li, and L. Xu. 2006. Prenatal stress modifies hippocampal synaptic plasticity and spatial learning in young rat offspring. Hippocampus 16:431-436. Yasui, Y. 1998. Thegenetic benefits' of female multiple mating reconsidered. Trends in Ecology & Evolution 13:246-250. Zahavi, A. 1975. Mate selection—a selection for a handicap. Journal of theoretical Biology 53:205-214. Zala, S. M., W. K. Potts, and D. J. Penn. 2004. Scent-marking displays provide honest signals of health and infection. Behavioral Ecology 15:338-344. 180

Zeh, D. W. and J. A. Zeh. 2000. Reproductive mode and speciation: the viviparity‐ driven conflict hypothesis. Bioessays 22:938-946. Zeh, J. A. and D. W. Zeh. 1996. The evolution of polyandry I: intragenomic conflict and genetic incompatibility. Pp. 1711-1717. Proc. R. Soc. Lond. B. The Royal Society. Zeh, J. A. and D. W. Zeh. 1997. The evolution of polyandry II: post–copulatory defenses against genetic incompatibility. Proceedings of the Royal Society of London B: Biological Sciences 264:69-75. Zeh, J. A. and D. W. Zeh. 2006. Outbred embryos rescue inbred half-siblings in mixed- paternity broods of live-bearing females. Nature 439:201. Zemel, S., M. S. Bartolomei, and S. M. Tilghman. 1992. Physical linkage of two mammalian imprinted genes, H19 and insulin–like growth factor 2. Nature Genetics 2:61. Zhou, Q., B. O'Brien, and J. Relyea. 1999. Severity of Nausea and Vomiting During Pregnancy: What Does It Predict? Birth 26:108-114. Zielinski, W. J. and J. G. Vandenbergh. 1993. Testosterone and competitive ability in male house mice, Mus musculus: laboratory and field studies. Animal Behaviour 45:873-891. Zimmer, C. 2006. Silent struggle: A new theory of pregnancy. The New York Times 14:1-4.

181

Appendices

D o m in a n t m a le

C a s tra te d m a le s c e n t

2 .0 N o O d o u r S u b o rd in a te m a le

1 .5

n 1 .0

f o 0 .5

0 .0 D a y 2 D a y 5 D a y 8 D a y 1 1 D a y 1 4 D a y 1 7 D a y 2 0

D a y

Figure 17: Mean number of total scans (± 1 s.e.m) of sniffing substrate behaviour recorded every three days for each odour treatment.

182