Sexual selection and alternative reproductive tactics in the nursery-web mirabilis

Paolo Giovanni Ghislandi

PhD Thesis

Sexual selection and alternative reproductive tactics in the nursery-web spider

Paolo Giovanni Ghislandi

SpiderLab

Genetics, Ecology and Evolution, Department of Bioscience

Faculty of Science and Technology

Aarhus University

November 2017

SUPERVISORS:

Prof. Trine Bilde

Department of Bioscience, Aarhus University, Denmark

Dr. Cristina Tuni

Department Biologie II, Ludwig-Maximilians University München, Germany

CREDITS:

Photos: Paolo Ghislandi

SpiderLab Logo: Lena Grinsted

CITE AS:

Ghislandi P.G. (2017). Sexual selection and alternative reproductive tactics in the nursery- web spider Pisaura mirabilis. Ph.D. dissertation. Genetics, Ecology and Evolution, Department of Bioscience, Aarhus University, Denmark

NUMBER OF PAGES: 103

PRINTED BY:

Fællestrykkeriet – SUN, Aarhus University

Table of Contents

Acknowledgements……………………………………………………………………………...... 1

Summary…………………………………………………………………………………………….3

Resumé…………………………………………………………………………………………...... 5

State of the ART……………………………………………………………………………………7

Reproductive tactics in the nursery-web spider Pisaura mirabilis……………………...………11

Types of ARTs: fixed or plastic traits?...... 15

Which ecological drivers influence the expression and maintenance of ARTs?...... 20

References………………………………………………………………………………………….23

Chapter I - Behavioural plasticity and population variation in the expression of alternative reproductive tactics in the nuptial gift giving spider Pisaura mirabilis: are guys from the hood doing like we do? Manuscript...... 28

Chapter II - Resource availability, mating opportunity, and sexual selection intensity influence the expression of male alternative reproductive tactics. Re-submitted to The Journal of Evolutionary Biology...... 56

Appendix to Chapter II - The effect of environmental variation on the expression and maintenance of ARTs……………………………………………………………………………...89

Chapter III - Silk wrapping of nuptial gifts aids behaviour in male

Behavioral Ecology (2017), 00(00), 1–6. doi:10.1093/beheco/arx028……………………………93

Chapter IV - Extreme male mating behaviours: anecdotes in a nuptial gift-giving spider Arachnology (2015) 16 (8), 273–275……………………………………………………………..100 Acknowledgements

I‟ve always been the spider guy. I had my first pet spider when I was 13, as a present from my father, a young adult female Chilean rose ( rosea, Walckenaer, 1837). I had no clue it was actually a female, so I called it “Ranieri”, like the prince of Monaco, just because to me it looked so regal and majestic that it really deserved that name. I spent a massive amount of hours watching what Ranieri was doing (not much by the way, as most mygalomorphs..), as there was something special over her that I was not really able to decode or easily grasp. Something I never experienced before, in the face of the huge amount of whatever other I previously kept in a terrarium, cage or aquarium. To me, Ranieri seemed different. The growing interest and passion for that difference brought me to have raised and bred more than 60 spider from 20 different families and to be about to finish a PhD on spiders behaviour. Thanks, Ranieri.

Among all the people I want to thank for this great experience, the first one is my partner, Daniela

Boncinelli, who have always supported and encouraged me to pursue my PhD study far before I started it and continued doing that until today. And she does it with whatever project I decide to realize, from the smallest to the ones potentially changing our lives.

A huge, huge thanks goes to my parents and my entire family, for always supporting and feeding my interest and for having allowed me to keep whatever animal at home not to limit my passion, even when it came to black widows, brown recluse spiders or deadly poisonous centipedes. Grazie mamma, grazie papà.

Thanks to Myriam and Luciano Boncinelli, Daniela‟s parents, for all the support and lovely advice helping me figuring out which was the right direction to go.

Thanks to all the friends who were always with me, before, during and (hopefully) after this PhD.

Then thanks to myself, for having never given up on my passion and dedication, even when it seemed it wasn‟t really leading me anywhere, or when I had the feeling I could have had better things to do.

1 To my supervisor Trine Bilde, thanks for all the things I‟ve learnt, for helpful discussions, for

“cooling down” my love for spiders when it had to be more scientific and less emotional, and for all the kindness and human support which had nothing to do with being one‟s supervisor.

Thanks to my co-supervisor, Cristina Tuni, for being a friend more than a co-worker, and for involving me in a lot of supercool side projects taking time away from my real, massive Phd project.

Thanks to Virginia “Vivi” Settepani, Magnus Jacobsen and André Walter for being great friends and helpful colleagues, for all the fun in South Africa doing field work, for Pevestorf and for all the rather unfruitful fishing trips.

Thanks to Jesper Bechsgaard, Bram Vanthournout, Tommaso Manenti, Marie Rosenstand and “the

LMU crew”, Philipp Sprau, Petri Niemelä, Marco La Fortezza and Alex Hutfluss for all the beers, fun and also, sometimes, advice. Thanks to all the students who helped me collecting data.

I would have never forgotten to thank Fabio De Vita, the biggest spider lover and experienced breeder I have ever met, and Carlo Paoletti, the biggest amphibian lover and curiosity-driven- zoology fan I have ever met, for being two of my best friends and for walking me in the world of terraristics and when I was just a bit more than a kid. Grazie fratellone, grazie Carletto.

Last but not least, thanks to Emanuele Biggi and Francesco Tomasinelli, for all the hours at the phone talking about spiders when I was a kid, for being great friends and for the always too-few collecting trips together, grazie Ema and grazie Francio!

And obviously, thanks to something like 7.360 Pisaura mirabilis which “spontaneously” collaborated to my PhD project.

2 Summary

The evolution and maintenance of alternative reproductive tactics ARTs represent a topic of broad interest in evolutionary and behavioural biology. In this thesis, I studied the characteristics and factors that influence the expression of ARTs in the nuptial gift-giving spider Pisaura mirabilis, where males offer either genuine or worthless nuptial gifts but can also mate without a gift, and where males also facultatively perform thanatosis, a death feigning behaviour associated with courtship. I addressed the following questions: 1. Are ARTs fixed or plastic? I conducted a study in five different populations to investigate whether nuptial gift traits and thanatosis represent fixed or plastic tactics, and show that ARTs are plastic labile traits. 2. Is there variation in the expression of

ARTs display among populations? Populations may diverge in the expression of ARTs in response to variation in ecological factors. As a first step to gain insight into population differences, I determined variation in the expression of ARTs among populations, and found evidence for population variation in the frequencies of trait expression. 3. Which ecological factors influence the expression and maintenance of ARTs? Ecological factors may influence the expression of ARTs, and variation in ecological factors over the reproductive season may influence the maintenance of

ARTs. I conducted a field study to explore whether the nuptial gift tactics vary with ecological factors. I found patterns of correlation between sexual selection, male body condition and prey availability and the expression of tactic, suggesting that the interplay and variation in ecological factors over time influence the maintenance of ARTs in wild populations. 4. How do seasonal changes in ecological factors shape ARTs? Based on the empirical data collected in my thesis, I aimed to develop and parameterize a model to investigate how variation in ecological factors influence the maintenance of ARTs, this work is recently initiated and ongoing, and represented as a short sketch of the model. 5. What is the role of silk wrapping in alternative nuptial gift tactics?

Silk is used to wrap the nuptial gift, and can potentially be used to enlarge the gift or disguise gift content. I performed a study to test whether males make strategic use of silk when offering nuptial gifts, and found that males wrap worthless gifts with more silk than genuine gifts, possibly to

3 compensate for their lower nutritive value and to increase copulation time. Finally, while collecting data on male reproductive behaviour, I observed that males perform rather extreme behaviours in additions to those reported above, this is reported in a paper presenting extreme ARTs.

4 Resumé

Evolutionen og opretholdelse af alternative reproduktionstaktikker (ARTs) repræsenterer et emne af bred interesse indenfor evolutionær- og adfærdsbiologi. I denne afhandling, har jeg studeret karakteristika og faktorer der influerer udtrykket af ARTs hos edderkoppen, Pisaura mirabilis, hvor hanner oftest tilbyder enten spiselige eller værdiløse bryllupsgaver. Hanner kan dog også opnå parring uden en bryllupsgave. Hanner kan yderligere fakultativt udføre såkaldt thanatosis (spille død), en adfærd associeret med parring. Jeg adresserede følgende spørgsmål: 1) er ARTs fikserede eller plastiske? Jeg udførte et studie i 5 populationer for at undersøge om karakterer omkring bryllupsgaver og thanatosis repræsenterer fikserede eller plastiske taktikker, og finder at ARTs er plastiske labile træk. 2) er der variation i udtrykket af ARTs mellem populationer? Variation i

økologiske faktorer kan forårsage divergens i udtrykket af ARTs mellem populationer. Som et første skridt for at opnå indsigt i populationsforskelle, har jeg bestemt variationen i udtrykket af

ARTs mellem populationer, og fundet at variationen i frekvensen af de enkelte ARTs mellem populationer. 3) hvilke økologiske faktorer influerer udtrykket og opretholdelse af ARTs?

Økologiske faktorer kan influere udtrykket af ARTs, og variationen i økologiske faktorer henover den reproduktive sæson kan influere opretholdelsen af ARTs. Jeg udførte et feltstudie for at udforske om bryllupsgavetaktikker varierer sammen med økologiske faktorer. Jeg fandt korrelationer mellem seksuel selektion, hannernes kropskonditioner og bytte tilgængelighed, og udtrykket af taktik, hvilket tyder på at samspillet og variationen mellem økologiske faktorer over tid influerer opretholdelsen af ARTs i naturlige populationer. 4) hvorledes former ændringer i

økologiske faktorer henover sæsonen ARTs? Baseret på empiriske data indsamlet i min afhandling, har jeg startet udviklingen samt parametriseringen af en model til at undersøge hvorledes variationen i økologiske faktorer influerer opretholdelse af ARTs. Dette arbejde er lige indledende og præsenteret som en simpel skitse. 5) hvilken rolle spiller silkeindpakning for alternative bryllupsgavetaktikker? Hannerne burger silke til at pakke bryllupsgaverne ind, og kan potentielt udnyttes til at forstørre bryllupsgaven eller forklæde indholdet af gaven. Jeg udførte et studie for at

5 teste om hannerne udfører strategisk brug af silke i forbindelse med, at de tilbyder bryllupsgaver, og fandt, at hanner pakker værdiløse bryllupsgaver ind i mere silke end ægte bryllupsgaver, sandsynligvis for at kompensere for deres lavere ernæringsværdi og for at øge kopulationstid.

Endeligt har jeg under indsamlingen af data for hanners reproduktive adfærd observeret, at hanner hellere udfører ekstrem adfærd i tillæg til disse rapporteret herover, hvilket er præsenteret i en artikel om ekstreme ARTs.

6 State of the ART

Competition over limited resources is ubiquitous in nature, promoting selection for traits that facilitate access to and exploitation of the resource itself (Schluter, 1994; Rainey and Travisano,

1998; Dieckmann and Doebeli, 1999). If a single optimal response to such competition is present, then these traits are under directional selection, with the most advantageous trait outcompeting the others and becoming fixed (Darwin, 1859). When more than a single optimal response is present, selection can promote set of traits to evolve, with each trait providing individuals with equal fitness payoffs (Oliveira, Taborsky and Brockmann, 2008). Conspecific individuals might indeed adopt different traits to reach the same end, thus showing phenotypic plasticity (Oliveira, Taborsky and

Brockmann, 2008; Taborsky and Brockmann, 2010). When phenotypic plasticity happens in a sexual context, it may result in alternative reproductive tactics (ARTs) (Gross, 1996; Shuster and

Wade, 2003; Oliveira, Taborsky and Brockmann, 2008). ARTs consist of morphological or behavioural traits which both males and females can adopt, that are different from the prevailing trait that intra-sexual conspecific competitors display to obtain inseminations (Tuttle, 2003;

Oliveira, Taborsky and Brockmann, 2008). ARTs are highly variable from fixed genetically determined phenotypes to plastic and labile behavioural types (Brockmann and Penn, 1992; Gross,

1996; Engqvist, 2007; Locatello et al., 2007; Simmons and Kotiaho, 2007). In the case of fixed

ARTs, traits differ between individuals and are linked to early developmental stages of the individual, which determine into which alternate and exclusive phenotype to develop (Moczek and

Emlen, 2000; Simmons and Kotiaho, 2007; Cardoso, Teles and Oliveira, 2015). Plastic ARTs are expressed within individuals, and might be sequential or labile. Sequential traits are linked to a particular life stage of the individual, while labile plastic traits do not depend on any given life stage of the individual and can be multiply and reversibly displayed (Brockmann and Penn, 1992;

Locatello et al., 2007; Cardoso, Teles and Oliveira, 2015). Plastic traits are driven by the interaction between the internal state of the individual and environmental factors (Schradin, 2013; Cardoso,

Teles and Oliveira, 2015).

7 The study of ARTs represents a hot topic in behavioural and evolutionary biology.

Investigating ARTs provides a particularly interesting opportunity to study how phenotypic variation can evolve, which is a central question in evolutionary biology (Oliveira, Taborsky and

Brockmann, 2008). The great variety of organisms adopting ARTs also allows us to approach the topic from different perspectives, increasing even more the interest related to it and the possibilities to explore it.

Several questions regarding ARTs are yet to be investigated. First, in many , it is poorly understood whether and when ARTs are fixed or plastic. This aspect is crucial as it a better understanding on the flexibility of ARTs may provide insights into how behavioural phenotypic plasticity evolves, and on investigating the evolutionary processes maintaining ARTs (Oliveira,

Taborsky and Brockmann, 2008; Hendrickx, Vanthournout and Taborsky, 2015). Secondly, identifying the evolutionary forces and drivers that are involved in maintaining ARTs in wild populations remains a key unsolved issue, as this is only understood in a few species (Oliveira,

Taborsky and Brockmann, 2008). Sexual selection represents that major force driving the evolution of ARTs (Oliveira, Taborsky and Brockmann, 2008; Taborsky and Brockmann, 2010). Phenotypic plasticity may evolve in response to environmental heterogeneity, and variation in ecological factors is likely to shape behavioural flexibility (Oliveira, Taborsky and Brockmann, 2008; Monroe et al., 2016). Studies investigating population differences in the frequencies of ARTs are needed to understand how variation in environmental conditions shape the expression of ARTs.

During my PhD studies, I used the nuptial gift-giving spider Pisaura mirabilis, where males are known to perform alternative tactics to fertilize females (Bilde et al., 2006; Albo et al., 2011;

Ghislandi et al., 2014), to investigate what shapes the expression of ARTs and how they may be maintained. Different tactics in this species include the use of a nuptial gift that is offered to the female during courtship and mating, and which might contain genuine nutritious food or nutritionally worthless items such as food leftovers (Albo et al., 2011; Ghislandi et al., 2014).

Males also display another curios trait, they can perform thanatosis, a “death feigning” behaviour

8 prior to or during mating, which has been proposed to function as male mating effort (Bilde et al.,

2006; Hansen et al., 2008). In a previous study, we developed a theoretical model that propose that in Pisaura mirabilis the genuine and worthless gift traits might be maintained by the interplay of several ecological factors such as prey availability, female , sperm competition, and male body condition (Ghislandi et al., 2014), with these hypotheses still remaining unexplored.

Also, P. mirabilis males have been proposed to plastically make use of silk-wrapping, with this feature promoting males‟ deception by masking the worthless gift content to females (Albo et al.,

2011). If this hypothesis holds, it might further indicate the importance of behavioural plasticity in the expression of ARTs.

In order to examine the characteristics of ARTs and factors that influence the expression of ARTs in

P. mirabilis, I aimed to answer the following research questions in my PhD thesis:

1. Are ARTs fixed or plastic?

I conducted a study in five different populations to investigate whether nuptial gift traits and

thanatosis represent fixed or plastic tactics, which is presented in Chapter I. I show that

ARTs are plastic labile traits

2. Is there variation in the expression of ARTs among populations?

Populations may diverge in the expression of ARTs in response to variation in ecological

factors. As a first step to gain insight into population differences, I determined variation in

the expression of ARTs among populations, reported in Chapter I. I found evidence for

population variation in the frequencies of trait expression.

3. Which ecological drivers influence the expression and maintenance of ARTs?

Ecological factors may influence the expression of ARTs, and variation in ecological factors

over the reproductive season may influence the maintenance of ARTs. I conducted a field

study to explore whether nuptial gift tactics vary according to the interplay of ecological

factors ,which is presented in Chapter II. We found patterns of correlation between sexual

selection, male body condition and prey availability and the expression of gift (or no gift)

9 tactics, suggesting that the interplay and variation in ecological factors over time influence

the maintenance of ARTs in wild populations (Chapter II).

4. How do seasonal changes in ecological factors shape ARTs?

Based on the empirical data collected in my thesis, we aimed to develop and parameterize a

model to investigate how variation in ecological factors influence the maintenance of ARTs.

This work is recently initiated and ongoing, and the current stage of the study is developing

realistic parameter sets to test. Therefore, at present this model represents an appendix of the

study reported in Chapter II.

5. What is the role of silk wrapping in alternative nuptial gift tactics?

Silk is used to wrap the nuptial gift, and can potentially be used to enlarge the gift or

disguise gift content. I performed a study to test whether males make strategic use of silk

when offering nuptial gifts, which is reported in Chapter III. We found that males wrap

worthless gifts with more silk than genuine gifts, possibly to compensate for their lower

nutritive value and to increase copulation time.

6. Extreme ARTs

While collecting data on male reproductive behaviour, I observed that males perform rather

extreme behaviours in additions to those reported above. This resulted in a paper presenting

extreme ARTs in Chapter IV.

10 Reproductive tactics in the Pisaura mirabilis

During my PhD project, I used the araneomorph spider Pisaura mirabilis (Clerck, 1757) as study organism. Pisaura mirabilis is a paleartic ambush-hunting species belonging to the family

Pisauridae (The World Spider Catalog), which is commonly referred to as nursery-web spider due to maternal care by guarding broods in a dedicated tent-shaped nursery web. Pisaura mirabilis males are known to employ behavioural ARTs to acquire mates (Bilde et al., 2006; Albo et al.,

2011; Ghislandi et al., 2014). One of these tactics include the offer of a nutritious donation to females during courtship (Picture 1,A), a so called nuptial gift, which might occur in the form of a genuine or a worthless gift (Bristowe, 1958; Albo et al., 2011; Ghislandi et al., 2014). Genuine gifts consist of prey, caught and wrapped in silk by males, while worthless gift are often digested prey leftovers, plant buds, pollen or other inedible token items which are silk-wrapped in a similar way as a genuine gift (Picture 1, C-D) and then offered to the female (Albo et al., 2011;

Ghislandi et al., 2014). Silk wrapping has been proposed to have multiple functions. Males might benefit from enveloping gifts in silk as this could increase their control over the gift by allowing a better grip with their chelicera (Andersen et al., 2008), or make strategic use of silk by wrapping worthless gifts with more silk to better hide the content to females (Albo et al., 2011; Ghislandi et al., 2017). Males are also likely to face metabolic costs associated to gift wrapping, as it was shown that males in poor body condition are impaired in their wrapping abilities (Albo, Toft and Bilde,

2011). On the other hand, females might benefit from silk wrapping, because if it is a condition- dependent trait, it might be an honest signal of male quality, and also because silk represents a nutritious source, which females consume during gift consumption (Lang, 1996; Albo, Toft and

Bilde, 2011; Ghislandi et al., 2017). However, if males use silk to mask gift content, this could be detrimental to females, as they are not able to disclose the content prior to accepting the male and initiate feeding on the gift (Albo et al., 2011; Ghislandi et al., 2014). Females may therefore mate with males they would otherwise have rejected.

11 As a last tactic, males can also mate without offering any gift at all (Stålhandske, 2001; Albo et al.,

2011). Nuptial gifts are under sexual selection imposed by strong female preference for gift-giving males (Stålhandske, 2001; Bilde et al., 2007). Copulation duration is dictated by female gift consumption time, with males transferring sperm as long as the female is occupied with feeding on the gift (Drengsgaard and Toft, 1999) (Picture 1, B). Males offering worthless gifts are equally likely to be accepted as males that offer genuine gifts, i.e. around 90% of mating attempts, while males that offer no gifts have a much lower success rate at 40%, at least with virgin females (Albo et al., 2011). Worthless gifts have been suggested to be less costly to produce than genuine gifts, as males would likely save some of the costs associated to genuine gift production, as for instance missed foraging opportunities (Albo et al., 2011; Ghislandi et al., 2014). However, worthless gift- giving males experience at least one major cost, as the female interrupts copulation sooner, which results in reduced sperm transfer and consequently reduced fertilization opportunity (Albo et al.,

2011). As females are polyandrous (Tuni, Albo and Bilde, 2013), costs associated with offering worthless gifts are even amplified, as males are also likely to loose out in sperm competition by transferring a relatively smaller amount of sperm (Parker, 1970).

When courting, males vibrate their abdomen and repeatedly lift their first pair of legs, twitching around in the female‟s proximity (Bristowe, 1958). Courtship culminates with males touching the ground with the tip of their abdomen, keeping the first two pairs of legs open wide and often lifted, and maintaining this position in front of the female, offering the nuptial gift (Picture 1A) (Bristowe,

1958). Once the female accepts the gift by grabbing it in her chelicera (mouthparts) and starts feeding on it, mating usually begins.

12

Picture 1- A: A male P. mirabilis offering a nuptial gift to a female, adopting the typical courtship position, during an experimental trial at Ludwig-Maximilians University, München, Germany; B: A couple mating in the field during the mating season of 2015 in Trnava, Slovakia. The female consumes the gift while the male transfers sperm; C: The content of a genuine field-collected nuptial gift opened under a stereomicroscope in

München, Germany. Note the intact of Coleoptera, Family Coccinellidae; D: The content of a worthless field-collected nuptial gift opened under a stereomicroscope in München, Germany. The donation consists almost entirely of silk, with some arthropod exoskeleton parts.

Sometimes both sexes are holding the gift with their chelicera in a front-to-front position, either before copulation, or when males switch (sperm transfer organ), and it might then happen

13 that the female attempts to run away with the gift and avoid copulation. In this situation, Pisaura mirabilis males can also perform another ART called thanatosis, which is a “feigning death” tactic

(Bilde et al., 2006; Hansen et al., 2008). Thanatosis occurs in the form of a very stereotyped posture in which males stretch their legs out, hanging motionless onto the gift when the female attempts to flee (Bilde et al., 2006; Hansen et al., 2008) (Picture 2). This particular position allows males to be dragged around on the substrate, and in this way hanging on to the female via the gift. Once the female stops and starts feeding on the gift, males “revive” and resume copulation (Bilde et al.,

2006). Death feigning has been shown to function as a male mating effort, leading to longer copulations and increasing the probability of a male being accepted by the female (Hansen et al.,

2008). Thanatosis can also occur between mating insertions, i.e. when males return to the front-to- front position prior to switch pedipalp to continue copulation (Hansen et al., 2008; Ghislandi et al.,

2014). Furthermore, thanatosis has also been proposed as a male strategy to keep control of the gift not losing it to females: female spiders are often larger than males, with a struggle over the gift most likely resulting in the male losing it. Indeed, it is extremely unlikely that the female re-accepts the male they were mating with for further copulation if they manage to gain full control over the gift (Andersen et al., 2008; Ghislandi et al., 2014). Thus, males might perform thanatosis in order to increase their probability of mating. Hansen et al. (2008) showed that males are able to facultatively perform thanatosis during copulation, suggesting this trait to represent a plastic labile trait.

Considering the benefits provided by performing thanatosis, males are likely to face costs associated with displaying the trait, as it is not systematically performed (Hansen et al., 2008).

Males might for instance make strategic use of thanatosis when females are resistant to mating and/or to increase their chance when courting a high quality female.

14

Picture 2- A P. mirabilis male performing thanatosis during a laboratory trial at Aarhus University, Denmark.

The female drags the male over the substrate while he remains attached to the gift with his chelicera, and maintaining the typical motionless position.

Types of ARTs: fixed or plastic traits?

ARTs are highly variable, and can be genetically or morphologically fixed, or developmental or behaviourally plastic (Oliveira, Taborsky and Brockmann, 2008; Taborsky and Brockmann, 2010).

In the case of fixed traits, multiple intra-sexual phenotypes are simultaneously present, with each phenotype exclusively associated to a tactic (Oliveira, Taborsky and Brockmann, 2008; Taborsky and Brockmann, 2010). For instance, in the marine isopod Paracerceis sculpta (Holmes, 1904) 3 fixed genetically-determined male phenotypes are present : α males are large and defend territories adopting a fighting tactic, β males mimic females behaviour and morphology, and γ males adopt a sneaking tactic, favoured by their reduced body size (Shuster and Wade, 1991, 2003; Oliveira,

15 Taborsky and Brockmann, 2008). Each of these male phenotypes maintains the same tactic throughout its entire adult life, thus adopting a fixed tactic (Shuster and Wade, 1991; Taborsky and

Brockmann, 2010). When ARTs are performed by a single genotype able to adopt either tactic, then tactics rely on plastic traits. For example males of the scorpion fly Panorpa cognata (Rambur,

1842) might offer females a nuptial gift consisting either of salivary nutritious secretions, or a prey collected in the environment depending on male body condition (Vahed, 1998; Engqvist, 2007).

The same male phenotype can thus plastically perform either tactic, which may for example be condition dependent.

In the case of the nursery-web spider Pisaura mirabilis, previous studies have shown thanatosis to represent a plastic labile trait, that males can reversibly adopt (Hansen et al., 2008).

However, regarding the genuine, worthless and no-gift tactics, the question whether these males‟

ARTs depend on fixed or plastic traits is not well understood (Albo et al., 2011). Part of my PhD study therefore focused on understanding whether, in Pisaura mirabilis, males develop into multiple fixed phenotypes each of them performing a single gift tactic, or into a single phenotype able to plastically display either tactic. To address this question, I tested males from 5 different populations offering them the possibility to wrap a worthless and/or a genuine gift (Chapter I).

Individual males from all populations showed the capability of producing either gift type, confirming these traits to be plastic (Chapter I). Also, males were able to reversibly perform the investigated traits, meaning that they are not linked to any particular life stage of individuals, thus representing plastic labile traits (Chapter I).

Plasticity in male gift traits may not only be revealed by gift contents, but also by other aspects related to gift production. For example, previous studies outlined the possibility that males could plastically adjust silk investment depending on the type of donation they are offering (Albo et al., 2011; Ghislandi et al., 2014). This is due to the fact that males might wrap worthless gifts with additional silk to better mask the non-nutritious content and successfully deceive females into mating (Albo, Toft and Bilde, 2011; Ghislandi et al., 2014). I tested this assumption by combining a

16 field and a laboratory study (Ghislandi et al., 2017) (Chapter III), which indeed demonstrated that males plastically invest more in silk wrapping when offering a worthless donation (Ghislandi et al.,

2017) (Chapter III). However, it is possible that males wrap worthless gifts with more silk layers not necessarily to mask the gift content as previously proposed (Albo et al., 2011; Ghislandi et al.,

2014), but to compensate for the reduced quality of their donation. Since females also feed on the silk around gifts (Lang, 1996), males might deposit more silk in order to prolong copulation by improving the nutritional quality of their donation, keeping the female occupied feeding on the gift for longer to enable longer sperm transfer (Ghislandi et al., 2017) (Chapter III).

Independently from the reasons behind increased silk wrapping of worthless gifts, this finding further underlines male behavioural plasticity in performing gift tactics, with males being able to adjust their behaviour according to conditional contexts in which is expressed (Charmantier et al., 2008; Oliveira, Taborsky and Brockmann, 2008; Dingemanse et al., 2010). Furthermore, there may be no limit to the extreme variation in reproductive traits that behavioural plasticity enables. In fact, we also experienced unexpected male mating behaviours associated to gift wrapping during laboratory experimental trials (Ghislandi, Bilde and Tuni, 2015), where males deviated from the “standard” mode of courtship (Chapter IV). Males detached their own legs to wrap into a nuptial gift (Ghislandi, Bilde and Tuni, 2015) (Chapter IV). Interestingly, we also noticed this behaviour in the field, males carrying nuptial gifts containing their own legs (Ghislandi and Hutfluss, pers. obs.) (Pictures 3A-B), and males also allowed females to feed on their limbs while mating when offering no gifts (Ghislandi, Bilde and Tuni, 2015) (Chapter IV).

This might be a case referred in behavioural biology as “doing the best of a bad job”, meaning that individuals might invest in fitness enhancing traits even when the conditions to adopt this behaviour are not optimal (Dawkins, 1980). The ability of rapidly adopting a tactic and to be capable of suddenly modifying it, further underlines how crucial the role of behavioural plasticity is in performing the gift traits.

17

Picture 3- A: An adult male Pisaura mirabilis collected during the mating season of 2015 in Trnava, Slovakia.

Circled in red is the missing limb which has been self-autotomized and wrapped into a nuptial gift that he was carrying; B: nuptial gift of the male from panel A. The third left leg is visible through the silk.

ARTs have been shown to occur at different frequencies across populations (Emlen, 2008; Oliveira,

Taborsky and Brockmann, 2008; Monroe et al., 2016). This fact has been proposed to be related to differential genetic mechanisms and/or local environmental factors (Emlen, 2008; Monroe et al.,

2016). Tactics also return different fitness pay-offs depending on the environment in which they are displayed (Schluter, 2001), leading to variation in the frequencies of ARTs among populations in response to environmental heterogeneity (Charmantier et al., 2008; Oliveira, Taborsky and

Brockmann, 2008; Chevin, Lande and Mace, 2010).

In Pisaura mirabilis, empirical data suggest possible population differences in the expression of

ARTs. For instance, a study on a Slovakian population showed that all field-collected nuptial gifts consisted of intact fresh prey (Prokop and Maxwell, 2012), with males carrying exclusively genuine donations. By contrast, 41.6% of field-collected nuptial gifts from a German population consisted of worthless donations (Ghislandi et al., 2017). Similarly, studies of a Danish population found

18 thanatosis to be performed in 30-40% of mating encounters (Bilde et al., 2006; Hansen et al., 2008) while empirical data collected from a Slovakian population suggested thanatosis to occur at a lower frequency, approximately 13% (Pavol Prokop, pers. comm.).

In order to evaluate whether and to what extent population differences in the frequencies of performing a particular gift trait and/or thanatosis are present, I tested 5 different populations for the occurrence of these traits in standardized laboratory tests. We chose a British and an Italian population because of their likely isolation from the European mainland due to ecological barriers, consisting respectively of the British Channel and the Alps. The other 3 populations were chosen because of previous knowledge, consisting of a Danish, a German and a Slovakian population. We specifically included the Slovakian population studied in Prokop et al. (2012) in which males were found to only offer genuine gifts. Our results showed that populations differ in the frequency of thanatosis performance and in the frequency of producing a genuine nuptial gift (Chapter I).

However, no differences in the frequency of worthless gifts were found among populations

(Chapter I). Also, behavioural plasticity measured as the proportion of males performing all tactics did not differ among populations. Our study showed that variation in the frequency of P. mirabilis male ARTs is present among populations (Chapter I). Furthermore, we found that worthless gifts are produced also by Slovakian males, confirming this finding in both the mentioned laboratory study of Chapter I, and in a field study I performed in Slovakia, where they occurred at a natural frequency of 26%.

Our findings are consistent with the hypothesis that if individual implementation of a given tactic varies according to environmental heterogeneity, this may result in population differences in the expression of tactics (Emlen, 2008; Oliveira, Taborsky and Brockmann, 2008; Monroe et al., 2016).

19 Which ecological drivers influence the expression and maintenance of ARTs?

Different models have been proposed for the maintenance of ARTs in populations (Oliveira,

Taborsky and Brockmann, 2008; Taborsky and Brockmann, 2010). The mechanisms which is likely the most accredited for ARTs maintenance is frequency-dependent selection (Plaistow et al., 2004;

Oliveira, Taborsky and Brockmann, 2008; Taborsky and Brockmann, 2010). Theory predicts that

ARTs can be maintained when frequency-dependent selection results in each tactic in the population to be equally successful in term of fitness pay-off (Maynard Smith, 1982; Oliveira,

Taborsky and Brockmann, 2008). This results in an Evolutionarily Stable Strategy (ESS) wherein alternative tactics are maintained (Maynard Smith and Price, 1973; Gross, 1996). For instance, in the dwarf spider Oedothorax gibbosus (Blackwall 1841), two males morphs are present, a high- and a low-investing phenotype. The high investing phenotype experiences delayed sexual maturation due to the development of costly structures producing a secretion functioning as a nuptial gift for females (Hendrickx, Vanthournout and Taborsky, 2015). This maturation delay results in a vacant mating niche because females remain unmated in the beginning of the mating season. This vacant niche is occupied by a low-investing male phenotype that matures earlier without the secretion producing structure (Hendrickx, Vanthournout and Taborsky, 2015). The maintenance of these male

ARTs relies on negative frequency-dependent selection, with each morph favoured when rare

(Hendrickx, Vanthournout and Taborsky, 2015).

In some cases, ARTs maintenance have been shown to depend on demographic parameters such as density (Radwan, 2001). Social environment might indeed play an important role on the decision of which alternative tactic to adopt (Radwan, 2001; Oliveira, Taborsky and Brockmann,

2008; Taborsky and Brockmann, 2010). Males of the dragonfly Nannophya pygmaea

(Rambour,1842) for instance, adopt a territorial-defending tactic at lower densities, while they are more likely to switch to a satellite tactic when the number of individuals increases (Tsubaki and

Ono, 1986; Taborsky and Brockmann, 2010).

20 Variation in other ecological factors have also been identified as crucial for maintaining

ARTs in several study organisms (Oliveira, Taborsky and Brockmann, 2008; Taborsky and

Brockmann, 2010), as costs and benefits of tactics might vary in response to environmental changes

(Emlen, 2008; Oliveira, Taborsky and Brockmann, 2008). For instance, if environmental conditions are unpredictable or change to a negligible extent throughout individuals life, fixed ARTs may evolve (Shuster and Wade, 2003; Oliveira, Taborsky and Brockmann, 2008). If environmental changes happen in a more predictable way during an individual development, with the individual exposed to changing environments consecutively, sequential plastic ARTs might be selected

(Alonzo, Taborsky and Wirtz, 2000; Oliveira, Taborsky and Brockmann, 2008). If then unpredictability of changes in environmental factors occurs at a short-lived temporary scale (i.e. for instance number of potential mates or seasonal predator/prey abundance), selection should favour labile plastic ARTs, allowing individuals to respond more rapidly, and reversibly, to changing conditions (Engqvist, 2007; Oliveira, Taborsky and Brockmann, 2008).

During my PhD project, I performed field studies to investigate patterns of correlation between ecological factors and the occurrence of gift and worthless gift traits in P. mirabilis (Chapter II). We based our observations on a model we developed in 2014, where we hypothesized P. mirabilis males to adopt a gift-giving tactic and to offer worthless gifts depending on the interplay of sexual selection, body condition and prey availability (Ghislandi et al., 2014). We found some evidence for female availability influencing the expression of overall gift and worthless gift production, with the number of available females negatively correlating with worthless gift production (Chapter II). This result is best understood in the context of post-copulatory sexual selection, because sperm competition is likely to favour males that produce genuine gifts. OSR, measures as the relative proportion of females in the field positively correlated with males‟ likelihood to produce a nuptial gift and also with males‟ likelihood to switch to genuine gifts production (Chapter II). Possibly, also prey availability and male body condition might play a role. Males overall gift production (genuine and worthless) and worthless gift production positively correlated with prey availability (in one

21 year), and overall gift production also negatively correlated with male body condition (Chapter II).

Resource availability may therefore also influence trait expression.

Our findings underline the importance and also the complexity of ecological conditions in the shaping of Pisaura mirabilis ARTs, confirming what we previously hypothesized in our model, namely that gift and worthless gift production might possibly be effected by the interplay of sexual selection, male body condition and prey availability (Ghislandi et al., 2014). As is the case for several other systems showing plastic labile traits (Oliveira, Taborsky and Brockmann, 2008), nursery-web spider males appear to adjust their tactics depending on environmental conditions

(Chapter II).

22 References

Albo, M. J. et al. (2011) „Worthless donations: male deception and female counter play in a nuptial gift-giving spider‟, BMC Evolutionary Biology. BioMed Central Ltd, 11(1), p. 329. doi:

10.1186/1471-2148-11-329.

Albo, M. J., Toft, S. and Bilde, T. (2011) „Condition dependence of male nuptial gift construction in the spider Pisaura mirabilis (Pisauridae)‟, Journal of , 29(3), pp. 473–479. doi:

10.1007/s10164-011-0281-1.

Alonzo, S. H., Taborsky, M. and Wirtz, P. (2000) „Male alternative reproductive behaviours in a mediterranean wrasse, Symphodus ocellatus: Evidence from otoliths for multiple life-history pathways‟, Evolutionary Ecology Research, 2(8), pp. 997–1007.

Andersen, T. et al. (2008) „Why do males of the spider Pisaura mirabilis wrap their nuptial gifts in silk: Female preference or male control?‟, Ethology, 114(8), pp. 775–781. doi: 10.1111/j.1439-

0310.2008.01529.x.

Bilde, T. et al. (2006) „Death feigning in the face of .‟, Biology letters, 2(1), pp.

23–25. doi: 10.1098/rsbl.2005.0392.

Bilde, T. et al. (2007) „Nuptial gifts of male spiders: sensory exploitation of the female‟s maternal care or foraging motivation?‟, Animal Behaviour, 73(2), pp. 267–273.

Bristowe, W. S. (1958) „Bristowe, W. S. (1958). world of spiders..pdf‟.

Brockmann, H. J. and Penn, D. (1992) „Male mating tactics in the horseshoe crab, Limulus polyphemus‟, Animal Behaviour, 44(4), pp. 653–665. doi: 10.1016/S0003-3472(05)80293-3.

Cardoso, S. D., Teles, M. C. and Oliveira, R. F. (2015) „Neurogenomic mechanisms of social plasticity‟, The Journal of Experimental Biology, 218(Pt 1), pp. 140–149. doi: 10.1242/jeb.106997.

Charmantier, A. et al. (2008) „Adaptive Phenotypic Plasticity in Response to Climate Change in a

Wild Population‟, Science, 320(5877), pp. 800–803. doi: 10.1126/science.1157174.

Chevin, L. M., Lande, R. and Mace, G. M. (2010) „Adaptation, plasticity, and extinction in a changing environment: Towards a predictive theory‟, PLoS Biology, 8(4). doi:

23 10.1371/journal.pbio.1000357.

Darwin, C. (1859) The Origin of Species by means of Natural Selection; or the Preservation of

Favoured Races in the Struggle for Life. Murray,John, London. doi: 10.1038/005318a0.

Dawkins, R. (1980) „Good strategy or evolutionarily stable strategy?‟, : beyond nature/nurture , 331-367., p. 1980.

Dieckmann, U. and Doebeli, M. O. (1999) „On the origin of species by sympatric speciation‟,

Nature, 400(22), pp. 354–357. doi: 10.1038/22521.

Dingemanse, N. J. et al. (2010) „Behavioural reaction norms: animal personality meets individual plasticity‟, Trends in Ecology and Evolution, 25(2), pp. 81–89. doi: 10.1016/j.tree.2009.07.013.

Drengsgaard, I. and Toft, S. (1999) „Sperm Competition in a Nuptial Feeding Spider, Pisaura

Mirabilis‟, Behaviour, 136(7), pp. 877–897. doi: 10.1163/156853999501621.

Emlen, D. J. (2008) „The roles of genes and the environment in the expression and evolution of alternative tactics‟, Alternative reproductive tactics: An integrative approach. Cambridge

University Press Cambridge, pp. 85–108.

Engqvist, L. (2007) „Nuptial food gifts influence female egg production in the scorpionfly Panorpa cognata‟, Ecological Entomology, 32(3), pp. 327–332. doi: 10.1111/j.1365-2311.2006.00835.x.

Ghislandi, P., Bilde, T. and Tuni, C. (2015) „Extreme male mating behaviours : anecdotes in a nuptial gift-giving spider‟, 16, pp. 273–275.

Ghislandi, P. G. et al. (2014) „Evolution of deceit by worthless donations in a nuptial gift-giving spider‟, Current Zoology, 60(1), pp. 43–51.

Ghislandi, P. G. et al. (2017) „Silk wrapping of nuptial gifts aids cheating behaviour in male spiders‟, , 28(3), pp. 744–749. doi: 10.1093/beheco/arx028.

Gross, M. R. (1996) „Alternative reproductive strategies and tactics: diversity within sexes.‟, Trends in ecology & evolution (Personal edition), pp. 92–98. doi: 10.1016/0169-5347(96)81050-0.

Hansen, L. S. et al. (2008) „Thanatosis as an adaptive male mating strategy in the nuptial gift-giving spider Pisaura mirabilis‟, Behavioral Ecology, 19(3), pp. 546–551. doi: 10.1093/beheco/arm165.

24 Hendrickx, F., Vanthournout, B. and Taborsky, M. (2015) „Selection for costly sexual traits results in a vacant mating niche and male dimorphism‟, Evolution. Wiley Online Library, 69(8), pp. 2105–

2117.

Lang, A. (1996) „Silk investment in gifts by males of the nuptial feeding spider Pisaura mirabilis

(Araneae: Pisauridae)‟, Behaviour. Brill, 133(9), pp. 697–716.

Locatello, L. et al. (2007) „Variation pattern of sperm quality traits in two gobies with alternative mating tactics‟, Functional Ecology, 21(5), pp. 975–981. doi: 10.1111/j.1365-2435.2007.01314.x.

Maynard Smith, J. (1982) Evolution and the Theory of Games. Cambridge university press.

Maynard Smith, J. ohn and Price (1973) „Logic of animal conflict‟, Nature, 254(5499), pp. 463–

463. doi: 10.1038/254463b0.

Moczek, A. and Emlen, D. (2000) „Male horn dimorphism in the scarab beetle, Onthophagus taurus: do alternative reproductive tactics favour alternative phenotypes?‟, Animal behaviour, 59(2), pp. 459–466. doi: 10.1006/anbe.1999.1342.

Monroe, M. J. et al. (2016) „Seasonal variation in male alternative reproductive tactics‟, Journal of

Evolutionary Biology, 29(12), pp. 2362–2372. doi: 10.1111/jeb.12981.

Oliveira, R. F., Taborsky, M. and Brockmann, H. J. (2008) Alternative reproductive tactics: an integrative approach. Cambridge University Press.

Parker, G. A. (1970) „Sperm Competition and Its Evolutionary Consequences in the ‟,

Biological Reviews, 45(4), pp. 525–567. doi: 10.1111/j.1469-185X.1970.tb01176.x.

Plaistow, S. J. et al. (2004) „Evolution of alternative mating tactics: Conditional versus mixed strategies‟, Behavioral Ecology, 15(4), pp. 534–542. doi: 10.1093/beheco/arh029.

Prokop, P. and Maxwell, M. R. (2012) „Gift carrying in the spider Pisaura mirabilis: Nuptial gift contents in nature and effects on male running speed and fighting success‟, Animal Behaviour.

Elsevier Ltd, 83(6), pp. 1395–1399. doi: 10.1016/j.anbehav.2012.03.007.

Radwan, J. (2001) „Male Morph Determination in Rhizoglyphus Echinopus (Acaridae)‟,

Experimental and Applied Acarology · February 2001, 1621(June), pp. 36–43. doi: 10.1023/A.

25 Rainey, P. B. and Travisano, M. (1998) „Adaptive radiation in a heterogeneous environment‟,

Nature, 394(6688), pp. 69–72. doi: 10.1038/27900.

Schluter (1994) „Schluter_1994_science experimental evidence for character displacement.pdf‟.

Schluter, D. (2001) „Ecology and the origin of species‟, Trends in Ecology and Evolution, 16(7), pp.

372–380. doi: 10.1016/S0169-5347(01)02198-X.

Schradin, C. (2013) „Intraspecific variation in social organization by genetic variation, developmental plasticity, social flexibility or entirely extrinsic factors‟, Philosophical Transactions of the Royal Society B: Biological Sciences, 368(1618), p. 20120346. doi: 10.1098/rstb.2012.0346.

Shuster, S M, Wade, M. J. (2003) „Sexual Selection and the Sex Difference in Variance of

Reproductive Success‟, Book by Shuster and Wade, (Howard), pp. 1–35. Available at: http://press.princeton.edu/chapters/s7571.pdf.

Shuster, m and Wade (2003) „Sexual Selection and the Sex Difference in Variance of Reproductive

Success‟, in Book by Shuster and Wade.

Shuster, S. M. and Wade, M. J. (1991) „Equal Mating Success among Male Reproductive Strategies in a Marine Isopod‟, Nature, 350(6319), pp. 608–610. doi: 10.1038/350608a0.

Simmons, L. W. and Kotiaho, J. S. (2007) „The effects of reproduction on courtship, fertility and longevity within and between alternative male mating tactics of the horned beetle, Onthophagus binodis‟, Journal of Evolutionary Biology, 20(2), pp. 488–495. doi: 10.1111/j.1420-

9101.2006.01274.x.

Stålhandske, P. (2001) „Nuptial gift in the spider Pisaura mirabilis maintained by sexual selection‟,

Behavioral Ecology, 12(6), pp. 691–697. doi: 10.1093/beheco/12.6.691.

Taborsky, M. and Brockmann, H. J. (2010) Animal behaviour: Evolution and mechanisms, Animal

Behaviour: Evolution and Mechanisms. doi: 10.1007/978-3-642-02624-9.

The World Spider Catalog pp. http://research.amnh.org/iz/spiders/catalog/INTRO1.html

Tsubaki, Y. and Ono, T. (1986) „Competition for Territorial Sites and Alternative Mating Tactics in the Dragonfly , Nannophya pygmaea Rambur ( Odonata : Libellulidae ) Author ( s ): Yoshitaka

26 Tsubaki and Tomohiro Ono Reviewed work ( s ): Published by : BRILL Stable URL : http://www.jsto‟, 97(3), pp. 234–252.

Tuni, C., Albo, M. J. and Bilde, T. (2013) „Polyandrous females acquire indirect benefits in a nuptial feeding species‟, Journal of Evolutionary Biology, 26(6). doi: 10.1111/jeb.12137.

Tuttle, E. M. (2003) „Alternative reproductive strategies in the white-throated sparrow : behavioral and genetic evidence‟, 14(3), pp. 425–432.

Vahed, K. (1998) „The function of nuptial feeding in insects: a review of empirical studies‟,

Biological Reviews. Wiley Online Library, 73(1), pp. 43–78.

27

Chapter I

Behavioural plasticity and population variation in the expression of alternative reproductive tactics in the nuptial gift giving spider

Pisaura mirabilis: are guys from the hood doing like we do?

Ghislandi P.G., Lauridsen A., Hutfluss A., Tuni C. & T. Bilde

Manuscript

28 1 Behavioural plasticity and population variation in the expression of alternative

2 reproductive tactics in the nuptial gift giving spider Pisaura mirabilis: are guys

3 from the hood doing like we do?

4 Paolo Giovanni Ghislandi* a, Astrid Lauridsen a, Alexander Hutfluss b, Cristina Tuni b✚ & Trine

5 Bilde a✚

6

7 a Department of Bioscience, Aarhus University, Aarhus, Denmark

8 b Department of Biology, Ludwig Maximilian University, Munich, Germany

9 * Corresponding author: [email protected]; Ny Munkegade 116, 8000 Aarhus C, Denmark

10 ✚ These authors contributed equally.

11

12

13

29 14 Abstract

15 Behavioural plasticity allows animals to respond rapidly to variation in environmental

16 conditions. When behavioural plasticity occurs while competing for mates, it may result in

17 alternative reproductive tactics (ARTs). Males of the nursery-web spider Pisaura mirabilis

18 express alternative tactics to acquire mates, these involves the use of genuine or worthless

19 nuptial gifts offered to females during courtship, and thanatosis – a death feigning behaviour

20 that is displayed in a mating context. In this study, we examined i) variation in the display of

21 gift tactics and thanatosis among populations, ii) whether males can plastically adopt either of

22 the gift traits or thanatosis within and between populations. We collected data on the

23 frequency in tactic use and behavioural plasticity in tactic performance in laboratory trials of

24 P. mirabilis males originating from five different populations. We found variation in the

25 percentage of males that adopt the genuine gift tactic, but not the worthless tactic, among

26 populations. Males were much less inclined to construct a worthless than a genuine gift. We

27 also found variation in the percentage of males that perform thanatosis among populations.

28 Behavioural plasticity was pronounced, with some degree of variation among populations. If

29 populations vary in ecological and environmental factors that influence tactic expression, this

30 could explain the population differences observed in this study.

31

30 32 Introduction

33 Genetic isolation caused by ecological barriers is considered a major driver in population

34 divergence, allowing beneficial mutations to become fixed via genetic drift and preventing

35 diluting effects by invading of exogenous individuals (Maynard Smith, 1966; Dillon, 1984;

36 Palumbi, 1994; Nosil, Vines and Funk, 2005; Hankison and Ptacek, 2008). However,

37 population divergence can also happen in the face of gene flow (Maynard Smith, 1966;

38 Maynard Smith and Price, 1973; Jiggins et al., 2001; Miyagi and Terai, 2013). Divergence of

39 traits display among populations that are not genetically segregated is often ascribed to

40 phenotypic plasticity, where a genotype give rise to multiple phenotypes under different

41 environmental conditions (Oliveira, Taborsky and Brockmann, 2008; Taborsky and

42 Brockmann, 2010). If environmental changes occur rapidly and unpredictably through an

43 individual‟s life-time, behavioural phenotypic plasticity should be beneficial, as it allows

44 individuals to rapidly adjust their responses (Oliveira, Taborsky and Brockmann, 2008;

45 Dingemanse et al., 2010; Wright et al., 2010). Moreover, costs and benefits associated to a

46 behavioural response have been shown to vary according to environmental conditions

47 (Oliveira, Taborsky and Brockmann, 2008). Individuals might thus develop a set of possible

48 responses they can adopt when exposed to rapid changes in the environment, and the relative

49 frequency of responses will be reflected in their behavioural reaction norm (Charmantier et

50 al., 2008; Oliveira, Taborsky and Brockmann, 2008; Chevin, Lande and Mace, 2010;

51 Dingemanse et al., 2010; Wright et al., 2010). For instance, male field crickets Teleogryllus

52 oceanicus have been shown to either sing from their burrows to attract females, or to behave

53 as satellites intercepting females approaching singing males (Rowell and Cade, 1993). Males

54 rapidly switch between behaviours depending on changes in environmental factors, such as

55 variation in female density (Rowell and Cade, 1993).

31 56 When behavioural plasticity occurs while competing for mates, it may result in alternative

57 reproductive tactics (ARTs) (Oliveira et al., 2008). Behavioural ARTs consist of traits that

58 differ from the prevailing trait adopted by same-sex competitors to obtain inseminations

59 (Tuttle, 2003; Oliveira, Taborsky and Brockmann, 2008). Such traits can be adopted

60 consistently throughout an individual reproductive life, or represent plastic traits,

61 conditionally expressed depending on internal and external environments (Brockmann and

62 Penn, 1992; Gross, 1996; Engqvist, 2007; Locatello et al., 2007; Simmons and Kotiaho,

63 2007). Plastic behavioural traits that are expressed within individuals might be sequential, i.e.

64 linked to a particular life stage of individuals (Brockmann and Penn, 1992; Locatello et al.,

65 2007; Cardoso, Teles and Oliveira, 2015), or labile, so they can be expressed facultatively and

66 reversibly (Engqvist, 2007; Oliveira, Taborsky and Brockmann, 2008). Since fitness pay-offs

67 associated to a particular tactic might vary depending on the environment in which this tactic

68 is adopted (Schluter, 2001; Schradin, 2013), ARTs are likely to occur at different frequencies

69 among populations (Oliveira, Taborsky and Brockmann, 2008; Monroe et al., 2016).

70 Differential selective pressures might therefore promote intra and inter-population phenotypic

71 divergence in tactic display, also under the presence of gene flow (Leinonen et al., 2006;

72 Hankison and Ptacek, 2008).

73 A study organism that shows behavioural ARTs is represented by the nursery-web spider

74 Pisaura mirabilis, where males express alternative tactics to acquire mates (Bilde et al., 2006;

75 Albo et al., 2011; Ghislandi et al., 2014). These tactics involve the use of a nuptial gift, which

76 might occur in the form of a genuine or a worthless gift offered to females during courtship

77 (Bristowe, 1958; Albo et al., 2011; Ghislandi et al., 2014). Genuine gifts consist of arthropod

78 prey wrapped in silk, whereas worthless gift are often prey leftovers, partially or entirely

79 consumed, plant buds, pollen or small pieces of wood which are offered to females after

80 wrapping in white silk (Albo et al., 2011; Ghislandi et al., 2014). Also, males can mate

32 81 without offering any gift at all (Stålhandske, 2001; Albo et al., 2011). Even if existing data

82 suggest that individual males may be able to perform either tactic (Albo et al., 2011), there is

83 limited quantitative evidence for the occurrence of behavioural plasticity in the use of

84 different tactics. Empirical data suggest that the frequency in male use of different gift types

85 may vary among populations (Prokop and Maxwell, 2012; Ghislandi et al., 2017). One

86 possibility is that trait expression varies in response to prey availability, sexual selection and

87 male body condition (Ghislandi et al., 2014; Chapter II). If different populations show

88 variation in the expression of male tactics, this could be in response to variation in such

89 ecological factors among populations, and their influence on pay-offs of male tactics.

90 Pisaura mirabilis males can also perform another interesting plastic trait, the so-called

91 thanatosis behaviour or “death feigning” (Bilde et al., 2006; Hansen et al., 2008). Thanatosis

92 consists of a very stereotyped posture in which males stretch their legs out hanging motionless

93 onto the nuptial gift when the female attempts to run away with the gift in her mouthparts

94 (Bilde et al., 2006; Hansen et al., 2008). This position allows the male to be dragged around

95 on the substrate, without impeding the female‟s movements, and once the female stops and

96 (re)start feeding on the gift, the male “reanimate” and initiate copulation (Bilde et al., 2006;

97 Ghislandi et al., 2014). Thanatosis behaviour has been shown to function as a male mating

98 effort, producing longer copulations and increasing male mating acceptance and re-acceptance

99 chances, with males also able to feign death between mating insertions of their copulatory

100 organs (Hansen et al., 2008; Ghislandi et al., 2014). Furthermore, thanatosis has been

101 proposed as a male strategy not to lose the gift to females: female spiders are often larger than

102 males and a struggle over the gift would most likely result in the male losing the donation. If

103 the female has full control over the gift, the male‟s likelihood of subsequently being accepted

104 or re-accepted is extremely low (Andersen et al., 2008; Ghislandi et al., 2014). Therefore,

105 thanatosis might play an important role for male mating success. Previous studies showed that

33 106 males in a Danish population facultatively express thanatosis during mating encounters,

107 suggesting that this is a plastic labile trait (Bilde et al., 2006; Hansen et al., 2008). Given the

108 benefits to males of enacting thanatosis and the fact that it is not systematically performed, it

109 is possible that there are costs associated with the performance of thanatosis (Bilde et al.,

110 2006; Hansen et al., 2008). If there is variation in potential costs in response to environmental

111 conditions, thanatosis is may be employed at different frequencies in different populations.

112 Also, the degree of plasticity in thanatosis display might vary among populations, with

113 individual males their plastic responses to external factors.

114 The aims of the present study are 1) to examine variation in the display of gift tactic and

115 thanatosis among populations, as this might be a consequence of variation in genetic, i.e.

116 internal, or ecological, i.e. external, factors, and 2) to understand whether males can

117 plastically adopt either of the gift traits or thanatosis within and between populations. We

118 collected data on the frequency in tactic use and behavioural plasticity in tactic performance

119 in laboratory trials of P. mirabilis males from five different populations.

120

121 Methods

122 Collecting and rearing

123 Immature Pisaura mirabilis specimens were collected from 5 European populations over a

124 period from April 2013 to April 2016. Spiders were collected from populations in 1) Italy

125 (Merlino, 45°25'18.0"N 9°27'34.8"E) 2) Germany (Grosshadern, 48°06'34.0"N 11°28'16.9"E)

126 3) Slovakia (Novaky, 48°43'25.9"N 18°32'18.0"E), 4) United Kingdom (Nottingham,

127 52°56'47.9"N 1°13'05.3"W) and 5) Denmark (Mols, 56°13'36.8"N 10°34'45.5"E). The

128 German and the Danish populations were chosen because of previous knowledge of these

129 spiders (i.e. populations‟ location, preliminary data on tactics display and average individuals

34 130 abundance in the field) (Albo et al., 2011; Ghislandi et al., 2017), the Italian and the British

131 because of their geographical isolation the Alps and the English Channel, and the Slovakian

132 population because of the reported difference from the Danish population in the frequency of

133 display of gift tactics and thanatosis (Prokop and Maxwell, 2012; Prokop, pers.comm.).

134 Spiders were raised in the laboratories at Ludwig Maximilian University, Munich, Germany

135 (German, Italian and Slovakian population), and at Aarhus University, Aarhus, Denmark

136 (Danish and British population) under standardised conditions.

137 Specimens were individually housed in plastic vials labelled with individual ID (5 cm

138 diameter X 10 cm height), with moss on the bottom to maintain humidity and topped with a

139 foam stopper in order to assure oxygen supply. Moss was sprayed with water twice a week

140 and spiders were located at room temperature of approximately 23C°, under natural

141 photoperiod. Animals were fed once every two days with prey consisting of Drosophila

142 spp. fruitflies, Musca spp. and Lucilia spp. flies and Gryllus spp. nymphs. Spider

143 moulting events were noted to keep track of individual age, and 12-14 days after the last

144 moult to adulthood, males were tested for gift their wrapping and thanatosis propensity in

145 standardised trials. Tactic display could be influenced by body condition. In Pisaura

146 mirabilis, body weight correlates positively with size (male cephalothorax width), used to

147 calculate residual body indexes in spiders (Jakob, Marshall and Uetz, 1996), and to residual

148 body index itself (Ghislandi, unpublished). We therefore used body mass as a proxy for male

149 body condition. Before starting each trial, we measured male body mass as a to the nearest

150 0.01 g using a digital scale (KERN PKT, KERN & SOHN GmbH, Balingen, Germany, and

151 SARTORIUS, Sartorius AG, Göttingen, Germany).

152

153

35 154 Gift tactics

155 We investigated the propensity of males for producing a genuine and a worthless gift. In total,

156 204 males were assayed, 26 from the German population, 67 from the Danish, 42 from the

157 Italian, 28 from the Slovakian and 41 from the British. Experiments were conducted in

158 transparent plastic terraria (7 cm height × 10 cm width × 10 cm depth) with a bottom layer of

159 absorbent paper. To elicit male gift-wrapping behaviour, we allowed a sexually mature female

160 belonging to the same population, to move freely in the experimental terrarium for 30

161 minutes, prior to male introduction. When walking around, females deposit silk draglines, to

162 which males respond by showing sexual excitement, often leading to gift construction

163 (Ghislandi et al., 2017). As soon as a male started vibrating their abdomen, indicating

164 initiation of courtship, we randomly offered the male either a genuine gift item consisting of a

165 live prey of known weight and size (a cricket nymph or a fly), or a worthless item consisting

166 of a dried prey (a cricket nymph exuvia or a dehydrated fly carcass). Fly carcasses were dried

167 out in an oven at 60 °C for 3 hours (UF30, MEMMERT GmbH, Schwabach, Germany and

168 THERMO SCIENTIFIC, Thermo Electron LED GmbH, Langenselbold, Germany) prior to

169 the experiment. This procedure mimicks what happens in nature, where worthless gifts often

170 consist of empty arthropod exoskeletons (Bristowe, 1958; Nitzsche, 1988; Ghislandi et al.,

171 2014). Body length of live prey and dried prey was measured to the nearest 0.01 mm using

172 digital callipers (AEROSPACE, China, and TOOLMATE, Denmark).

173 Live prey was simply released in the terrarium by placing it with forceps in the vicinity of the

174 male. If a male did not catch the prey within 20 minutes, the trial was interrupted and the male

175 returned to the vial. Also, trials were interrupted when males caught the prey but did not

176 construct a nuptial gift within 20 minutes. Dried prey was offered with forceps by gently

177 moving it in front of the male. This was repeated every 5 minutes. If a male did not accept the

178 worthless item within 20 minutes, thus after 4 offers, they were returned to their vial.

36 179 Gift wrapping occurred when a male accepted the offered item by grabbing it in their

180 mouthparts () and started depositing silk on it to produce a nuptial gift. Each

181 wrapping event was considered finished when males stopped circling around the gift to

182 envelop it and collected it from the substrate with their chelicera. After 20 minutes from the

183 last wrapping event, the experiment was terminated, gifts were gently removed with forceps

184 from the male‟s mouthparts, and males were returned to their vials.

185 Each male was tested twice per gift type (i.e. genuine and worthless) for the Italian, Danish

186 and Slovakian populations, up to 3 times per gift type for the German population and once per

187 gift type for the British population. Males were randomly assigned to produce a genuine or a

188 worthless gift in the first trial. Males tested for genuine gift wrapping were subsequently

189 tested for worthless gift wrapping and vice versa, systematically reversing trials order for all

190 populations. To estimate the percentage of males producing a gift for each population, and the

191 propensity of males to plastically employ both tactics, we used only the first trial for each

192 male for each gift type.

193

194 Thanatosis

195 In total, 128 males were assayed, 27 from the German population, 27 from the Danish, 21

196 from the Italian, 17 from the Slovakian and 36 from the British. Thanatosis behaviour was

197 examined in two successive mating trials, each conducted by pairing males with a random

198 novel female. Experiments were performed in transparent plastic terraria (19 × 13 × 14 cm)

199 provided with absorbent paper on the bottom. At first, a female belonging to the same

200 population as the male was placed in the terrarium allowing her to deposit draglines to elicit

201 male sexual excitement, as described above. After 30 minutes, the female was removed and

202 the male introduced in the mating arena. When the male started vibrating its abdomen, a live

37 203 cricket nymph or a fly of known weight and size was placed with forceps in the male‟s

204 proximity. Crickets body length was measured as described above. If the male did not catch

205 the prey or did not wrap it within 20 minutes the trial was terminated, and the male was

206 returned to its vial. When a male accepted the gift item, we allowed them to produce a nuptial

207 gift through one or multiple wrapping events. Wrapping events were considered concluded

208 when males stopped enveloping the gift and collected it from the bottom with their chelicera.

209

210 After 10 consecutive minutes from last gift wrapping event, the female was returned to the

211 mating arena. If a male initiated courtship and was not accepted within 1 hour, the trial was

212 interrupted, and each individual was returned to its vial to be tested again on the following

213 day. Once the female accepted the gift by fixing it in her mouthparts, we recorded thanatosis

214 occurrence (yes/no). Thanatosis was also recorded when occurring between mating

215 interruptions. When the mating ceased, i.e. when the couple spontaneously divided, males and

216 females were returned to their own vials. The trial was repeated with the same male and a

217 different female the following day, to get an estimation of individual plasticity. Each male

218 was tested up to 5 times for the Italian, German and Danish populations, 4 for the Slovakian

219 and 2 for the British.

220

221 Data analysis

222 The frequencies of males producing genuine and worthless gifts among populations was

223 compared with Chi-square models with gift production (Y/N) as response variable and

224 population as predictor, and male body mass and the interaction between male body mass and

225 population as co-variates. We calculated individual male behavioural plasticity as the

226 proportion of males facultatively performing thanatosis and compared male plasticity among

227 populations using a Chi-square test setting plasticity (Y/N) as response variable and

38 228 population, male body weight and the interaction between population and body weight as

229 explanatory variables. We used the 2 trials for each males in this analysis.

230 To study whether variation male behavioural plasticity varies with the number of trials

231 performed, we used a Chi-square model setting plasticity for thanatosis (Y/N) as response

232 variable and population, n° of trial, body weight and the interaction between population and

233 n° of trial as explanatory variables. For thanatosis, we included only the German, Danish and

234 British populations, due to low sample size for the Italian and Slovakian populations.

235 All statistical analysis were performed using the software JMP 13.0 from SAS Institute Inc.,

236 Cary, NC.

237

238 Results

239 Males assessed for gift production differed in their body mass across populations, with

240 individuals from the Slovakian population being the heaviest (one-way ANOVA, F=10.59,

241 p<0.0001, mean and SD “German” 96.08±2.99, “Danish” 82.43±1.82, “Italian” 86.70±2.35,

242 “Slovakian” 102.96±2.93, “British” 89.79±2.38).

243 Gift tactics

244 The percentage of males producing a genuine gift varied among populations (overall model:

245 χ2 test, χ2=22.19, p=0.0083). Genuine gift production did not vary significantly with body

246 mass (χ2 test, χ2=0.05, p=0.82), while high variation among populations was present (χ2 test,

247 χ2=19.09, p=0.0008, (Fig. 1). The lowest proportion of males producing a genuine gift was

248 seen in the German population, (60%), while the highest proportion was seen in the British

249 population (97,6%) (Fig.1). We found no significant body mass and population interaction (χ2

250 test, χ2=3.08, p=0.54).

39 251 No significant population differences in the percentage of males producing a worthless nuptial

252 gift was detected (χ2 test, χ2=11.16, p=0.26, Fig. 2). No significant effect of body mass (χ2

253 test, χ2=0.1909, p=0.66) and/or population (χ2 test, χ2=6.00, p=0.20, or their interaction χ2 test,

254 χ2=4.54, p=0.33) was detected.

255 In all populations, males showed behavioural plasticity in gift production (i.e. individual

256 males capable of producing genuine and worthless gifts), with marginally non-significant

257 variation in male plasticity among populations (Fig. 3, overall model, χ2 test, χ2=16.57,

258 p=0.055), effect of population (χ2 test, χ2=7.51, p=0.11), and body weight (χ2 test, χ2=0.2,

259 p=0.88), with a significant interaction between population and body weight (χ2 test, χ2=10.60,

260 p=0.0031).

261 Some males were tested more than once per gift type (German, Italian, Danish and Slovakian

262 populations). When we analysed the relationship between plasticity and trial number, we

263 found that males were less likely to show plasticity with increasing trial number (χ2 test,

264 χ2=8.32, p=0.0039, Fig. 4). Interestingly, the more trials we conducted, the less likely we were

265 to find that males are plastic for gift type production. Since this might be due to male

266 preference for producing genuine gifts, we tested this hypothesis using a χ2 model with male

267 acceptance Y/N as response variable and gift type Genuine/Worthless as explanatory variable.

268 The overall model was significant (χ2 test, χ2=153.746, p<0.0001), with males being much

269 more likely to produce a genuine (80%) rather than a worthless gift (20%).

270 Thanatosis

271 The overall model to analyse variation in the percentage of males performing thanatosis was

272 statistically significant (χ2 test, χ2=19.51, p=0.0073). The percentage of males performing

273 thanatosis varied significantly among populations (Fig. 5, χ2 test, χ2=12.27, p=0.0081), and

274 according to male body mass, with smaller males having a higher propensity of performing

40 275 thanatosis (χ2 test, χ2=4.14, p=0.038), but not with the interaction between population and

276 body mass (χ2 test, χ2=2.46, p=0.46).

277 In each of the investigated populations, males showed behavioural plasticity in performing

278 thanatosis (Fig. 6, overall model, χ2=9.3, p=0.097). Plasticity in performing thanatosis varied

279 among populations (χ2=7.84, p=0.019), but not with body weight (χ2 test, χ2=7.84, p=0.60)

280 and the interaction between population and body weight (χ2 test, χ2=0.78, p=0.67).

281 When we analysed the relationship between plasticity and trial number, we found that males

282 were more likely to show plasticity with increasing trial number (χ2 test, χ2=19.72, p=0.0006,

283 Fig. 7). The more trials we conducted, the more likely we were to find behavioural plasticity.

284 However, we did not detect significant variation among populations (χ2=4.05, p=0.13), or

285 with body weight (χ2=2.22, p=0.12), or their interaction (χ2=1.37, p=0.24).

286

287 Discussion

288 In our study, we investigated whether variation in the frequency of behavioral reproductive

289 tactics occurs among populations of the gift-giving spider Pisaura mirabilis. Our results show

290 that males expressed both genuine and worthless gift traits are in all the studied populations,

291 and that there is variation in tactics frequency occurs among populations. We found variation

292 in the percentage of males producing a genuine nuptial gift among populations. Our study

293 however, did not detect significant variation in the worthless gift tactic among populations.

294 This could perhaps be explained by a clear difference in male propensity to produce the two

295 different gift types. We found pronounced preference for genuine gifts, with 79.89% of all

296 males producing a genuine gifts, whereas worthless items were accepted for wrapping in

297 20.11% of cases among all males (all populations). Remarkably, the worthless gift trait also

298 occurred in the Slovakian population, where these were not previously reported (Prokop and

41 299 Maxwell, 2012). Frequencies of a tactic might vary to the extent of this tactic being displayed

300 by the largest majority of males in a population, with extremely few individuals performing

301 alternative tactics (Plaistow et al., 2004; Hendrickx, Vanthournout and Taborsky, 2015). The

302 relatively low incidence of worthless gift production may explain why previous studies on the

303 Slovakian population detected males exclusively carrying genuine fresh gifts (Prokop and

304 Maxwell, 2012). We also investigated whether male weight, as a proxy for male body

305 condition, affects male tactic, as males in good condition could be expected to be more likely

306 to produce a genuine gift. However, we did not find variation in the frequency of genuine and

307 worthless gifts to be explained by variation in male body weight. The choice of nuptial gift

308 tactic may therefore rely on other factors, possibly extrinsic to individuals, rather than

309 representing a condition-dependent tactic.

310 Genuine and worthless gift tactics were found to be plastic labile traits, which males can

311 alternatively and reversibly adopt (Oliveira, Taborsky and Brockmann, 2008; Taborsky and

312 Brockmann, 2010). Behavioural plasticity for nuptial gift production was indeed present in all

313 the populations included in our study. However, when investigating whether behavioural

314 plasticity varied among populations, we did not detect any significant relationship. However,

315 this could be because of lack of power, as population varied marginally non-significantly

316 between 10 and 30%. The stronger preference in males producing a genuine gift, is likely to

317 affect the expression of plasticity. Further support to this hypothesis is provided by the fact

318 that males did not show increasing plasticity in producing both gift types with additional

319 trials. In our study indeed, 77.5% of males tested never produced a worthless gift, while only

320 19.05% never produced a genuine gift. Among males that only produced a single gift type, the

321 percentage of males exclusively producing a worthless gift was 1.05%, while males chose to

322 solely produce a genuine gift in 64.01% of cases (Table 1). These numbers suggest that males

42 323 are much more reluctant to produce worthless than genuine gifts, perhaps suggesting that

324 worthless gift production is an opportunistic behaviour.

325 Our results also showed that males perform thanatosis in all studied population, and the

326 percentage males performing thanatosis varies among populations. Differently from the

327 nuptial gift tactics, we found some evidence for male body mass to affect thanatosis

328 performance, with lighter males being more likely to perform this behavioural tactic.

329 Although our results are not very conclusive on this point, it is possible that smaller or lighter

330 males use this tactic towards resistant females (Hansen et al 2008), where larger and heavier

331 males are able to directly overcome female resistance. Individual body condition has been

332 identified as an important factor in performing fitness enhancing behaviours (Lorch et al.,

333 2003; Webb, Shine and Christian, 2006), particularly when competing for mates (Burkhardt,

334 de la Motte and Lunau, 1994; Wiklund and Kaitala, 1995). Here we seem to have an opposite

335 effect, supporting the idea that thanatosis may indeed be an alternative reproductive tactic.

336 We detected variation in plasticity of performing thanatosis among populations, but we found

337 no effect of body mass on male plasticity for thanatosis. In contrast to plasticity in gift tactic,

338 thanatosis plasticity increased with number of trials. This shows that all males are capable of

339 performing thanatosis. The fact that not all males perform thanatosis, despite the benefits of

340 this behaviour (Bilde et al 2006), may nevertheless suggests costs associated to the trait‟s

341 expression. The expression of nuptial gift tactics in the field correlates with environmental

342 conditions that include demographic densities, with higher nuptial gift occurrence positively

343 correlating with increasing number of available females and increasing prey availability

344 (Chapter II). Since thanatosis has been proposed as a male mating effort (Bilde et al 2006),

345 and costs are likely associated to the trait, such tactic might be strategically displayed

346 depending on female densities. Under high competition for mates, males may be more likely

347 display a trait that increases mating success (Yoccoz et al., 2002; Oliveira, Taborsky and

43 348 Brockmann, 2008). Note, however, that thanatosis is performed in direct response to

349 encounters with females, in contrast to gift production, where gifts are produced prior to

350 encounters with females. While gift tactic may be expressed in response to extrinsic factors

351 (Chapter II), it is perhaps less likely that this is the case for thanatosis.

352 The fact that there is variation in tactic expression and also in plasticity of trait expression

353 among populations, suggests that populations vary in ecological or environmental factors that

354 may influence trait expression. (Oliveira, Taborsky and Brockmann, 2008; Taborsky and

355 Brockmann, 2010; Monroe et al., 2016). While we have some evidence that intrinsic factors

356 (male body weight) might influence the expression of tactic, other studies have recovered

357 little evidence for condition-dependent ARTs in P. mirabilis (Chapter II and III). A

358 population genetic study of P. mirabilis populations showed very little genetic variation

359 among populations (Lauridsen 2015). It is therefore unlikely that population differences are

360 genetically based. We showed in Chapter II that tactic expression within a population varies

361 with environmental factors along the reproductive season. If populations vary in ecological

362 and environmental factors that influence tactic expression, this could explain the population

363 differences observed in this study.

364

365

44 366 Figures and tables

367 Table 1- Percentages of males wrapping a genuine or a worthless gift. In line 1 is reported the percentage of males

368 that produced one or the other gift when provided with the possibility to produce a nuptial gift. Line 2 reports the

369 percentage of males never producing one or the other gift among all trials. Line 3 reports the percentage of males that

370 produced exclusively one gift type.

371

Genuine Worthless

Percentage of males producing a gift 79.89 20.11

Percentage of males never producing a gift type 19.05 77.5

Percentage of males exclusively producing a 64.01 1.05

single gift type

372

373

45 374 Figures

375

376 Fig.1-Percentage of males wrapping a genuine gift in different populations. Males differ significantly among

377 populations in producing a genuine nuptial gift

378

379

380 Fig.2-Percentage of males wrapping a worthless gift in different populations. Males did not differ significantly among

381 populations in producing a worthless nuptial gift

382

46 383

384 Fig.3-Percentage of males producing both nuptial gift among populations. Males did not differ significantly in

385 showing plasticity of producing both nuptial gifts among populations

386

387

388

389 Fig. 4-Percentage of males producing both nuptial gift in relation to trial number.

390

47 391

392 Fig. 5-Percentage of males performing thanatosis in different populations. Males differ significantly in adopting a

393 “death feigning” behaviour among populations.

394

395

396 Fig. 6-Percentage of males that were plastic for thanatosis among populations.

397

48 398

399 Fig. 7-Percentage of plastic males performing thanatosis in relation to trial number.

49 400 References

401 Albo, M. J. et al. (2011) „Worthless donations: male deception and female counter play in a

402 nuptial gift-giving spider‟, BMC Evolutionary Biology. BioMed Central Ltd, 11(1), p. 329.

403 doi: 10.1186/1471-2148-11-329.

404 Andersen, T. et al. (2008) „Why do males of the spider Pisaura mirabilis wrap their nuptial

405 gifts in silk: Female preference or male control?‟, Ethology, 114(8), pp. 775–781. doi:

406 10.1111/j.1439-0310.2008.01529.x.

407 Bilde, T. et al. (2006) „Death feigning in the face of sexual cannibalism.‟, Biology letters,

408 2(1), pp. 23–25. doi: 10.1098/rsbl.2005.0392.

409 Bristowe, W. S. (1958) „Bristowe, W. S. (1958). world of spiders..pdf‟.

410 Brockmann, H. J. and Penn, D. (1992) „Male mating tactics in the horseshoe crab, Limulus

411 polyphemus‟, Animal Behaviour, 44(4), pp. 653–665. doi: 10.1016/S0003-3472(05)80293-3.

412 Burkhardt, D., de la Motte, I. and Lunau, K. (1994) „Signalling fitness: larger males sire more

413 offspring. Studies of the stalk-eyed fly Cyrtodiopsis whitei (Diopsidae, Diptera)‟, Journal of

414 Comparative Physiology A, 174(1), pp. 61–64. doi: 10.1007/BF00192006.

415 Cardoso, S. D., Teles, M. C. and Oliveira, R. F. (2015) „Neurogenomic mechanisms of social

416 plasticity‟, The Journal of Experimental Biology, 218(Pt 1), pp. 140–149. doi:

417 10.1242/jeb.106997.

418 Charmantier, A. et al. (2008) „Adaptive Phenotypic Plasticity in Response to Climate Change

419 in a Wild Bird Population‟, Science, 320(5877), pp. 800–803. doi: 10.1126/science.1157174.

420 Chevin, L. M., Lande, R. and Mace, G. M. (2010) „Adaptation, plasticity, and extinction in a

421 changing environment: Towards a predictive theory‟, PLoS Biology, 8(4). doi:

50 422 10.1371/journal.pbio.1000357.

423 Dillon, R. T. (1984) „Geographic distance, environmental difference, and divergence between

424 isolated populations‟, Systematic Biology, 33(1), pp. 69–82. doi: 10.1093/sysbio/33.1.69.

425 Dingemanse, N. J. et al. (2010) „Behavioural reaction norms: animal personality meets

426 individual plasticity‟, Trends in Ecology and Evolution, 25(2), pp. 81–89. doi:

427 10.1016/j.tree.2009.07.013.

428 Engqvist, L. (2007) „Nuptial food gifts influence female egg production in the scorpionfly

429 Panorpa cognata‟, Ecological Entomology, 32(3), pp. 327–332. doi: 10.1111/j.1365-

430 2311.2006.00835.x.

431 Ghislandi, P. G. et al. (2014) „Evolution of deceit by worthless donations in a nuptial gift-

432 giving spider‟, Current Zoology, 60(1), pp. 43–51.

433 Ghislandi, P. G. et al. (2017) „Silk wrapping of nuptial gifts aids cheating behaviour in male

434 spiders‟, Behavioral Ecology, 28(3), pp. 744–749. doi: 10.1093/beheco/arx028.

435 Gross, M. R. (1996) „Alternative reproductive strategies and tactics: diversity within sexes.‟,

436 Trends in ecology & evolution (Personal edition), pp. 92–98. doi: 10.1016/0169-

437 5347(96)81050-0.

438 Hankison, S. J. and Ptacek, M. B. (2008) „Geographical variation of genetic and phenotypic

439 traits in the Mexican sailfin mollies, Poecilia velifera and P. petenensis‟, Molecular Ecology,

440 17(9), pp. 2219–2233. doi: 10.1111/j.1365-294X.2008.03736.x.

441 Hansen, L. S. et al. (2008) „Thanatosis as an adaptive male mating strategy in the nuptial gift-

442 giving spider Pisaura mirabilis‟, Behavioral Ecology, 19(3), pp. 546–551. doi:

443 10.1093/beheco/arm165.

51 444 Hendrickx, F., Vanthournout, B. and Taborsky, M. (2015) „Selection for costly sexual traits

445 results in a vacant mating niche and male dimorphism‟, Evolution. Wiley Online Library,

446 69(8), pp. 2105–2117.

447 Jakob, E. M., Marshall, S. D. and Uetz, G. W. (1996) „Estimating fitness: a comparison of

448 body condition indices‟, Oikos. JSTOR, pp. 61–67.

449 Jiggins, C. D. et al. (2001) „Reproductive isolation caused by colour pattern mimicry‟,

450 Nature, 411(6835), pp. 302–305. doi: 10.1038/35077075.

451 Leinonen, T. et al. (2006) „Contrasting patterns of body shape and neutral genetic divergence

452 in marine and lake populations of threespine sticklebacks‟, Journal of Evolutionary Biology,

453 19(6), pp. 1803–1812. doi: 10.1111/j.1420-9101.2006.01182.x.

454 Locatello, L. et al. (2007) „Variation pattern of sperm quality traits in two gobies with

455 alternative mating tactics‟, Functional Ecology, 21(5), pp. 975–981. doi: 10.1111/j.1365-

456 2435.2007.01314.x.

457 Lorch, P. D. et al. (2003) „Condition dependent sexual selection can accelerate adaptation.‟,

458 Evolution Ecology Research, 5(6), pp. 867–881.

459 Maynard Smith, J. (1966) „Sympatric speciation‟, American Naturalist, 100(916), pp. 637–

460 650. doi: 10.1086/282457.

461 Maynard Smith, J. ohn and Price (1973) „Logic of animal conflict‟, Nature, 254(5499), pp.

462 463–463. doi: 10.1038/254463b0.

463 Miyagi, R. and Terai, Y. (2013) „The diversity of male nuptial coloration leads to species

464 diversity in Lake Victoria cichlids.‟, Genes & genetic systems, 88(3), pp. 145–53. doi:

465 10.1266/ggs.88.145.

52 466 Monroe, M. J. et al. (2016) „Seasonal variation in male alternative reproductive tactics‟,

467 Journal of Evolutionary Biology, 29(12), pp. 2362–2372. doi: 10.1111/jeb.12981.

468 Nitzsche, R. O. M. (1988) „Brautgeschenk‟ und Umspinnen der Beute be Pisaura mirabilis,

469 fimbriatus und Thaumasia uncata (Arachnida, Araneida, Pisauridae)‟, Verh

470 Naturwiss Ver Hamburg, 30.

471 Nosil, P., Vines, T. H. and Funk, D. J. (2005) „Perspective: Reproductive isolation caused by

472 natural selection against immigrants from divergent habitats.‟, Evolution; international

473 journal of organic evolution, 59(4), pp. 705–719. doi: 10.1554/04-428.

474 Oliveira, R. F., Taborsky, M. and Brockmann, H. J. (2008) Alternative reproductive tactics:

475 an integrative approach. Cambridge University Press.

476 Palumbi, S. R. (1994) „Genetic divergence, reproductive isolation, and marine speciation‟,

477 Annual Review of Ecology and Systematics, 25, pp. 547–572. doi:

478 10.1146/annurev.es.25.110194.002555.

479 Plaistow, S. J. et al. (2004) „Evolution of alternative mating tactics: Conditional versus mixed

480 strategies‟, Behavioral Ecology, 15(4), pp. 534–542. doi: 10.1093/beheco/arh029.

481 Prokop, P. and Maxwell, M. R. (2012) „Gift carrying in the spider Pisaura mirabilis: Nuptial

482 gift contents in nature and effects on male running speed and fighting success‟, Animal

483 Behaviour. Elsevier Ltd, 83(6), pp. 1395–1399. doi: 10.1016/j.anbehav.2012.03.007.

484 Rowell, G. A. and Cade, W. H. (1993) „Simulation of alternative male reproductive behavior:

485 calling and satellite behavior in field crickets‟, Ecological Modelling, 65(3–4), pp. 265–280.

486 doi: 10.1016/0304-3800(93)90083-5.

487 Schluter, D. (2001) „Ecology and the origin of species‟, Trends in Ecology and Evolution,

53 488 16(7), pp. 372–380. doi: 10.1016/S0169-5347(01)02198-X.

489 Schradin, C. (2013) „Intraspecific variation in social organization by genetic variation,

490 developmental plasticity, social flexibility or entirely extrinsic factors‟, Philosophical

491 Transactions of the Royal Society B: Biological Sciences, 368(1618), p. 20120346. doi:

492 10.1098/rstb.2012.0346.

493 Simmons, L. W. and Kotiaho, J. S. (2007) „The effects of reproduction on courtship, fertility

494 and longevity within and between alternative male mating tactics of the horned beetle,

495 Onthophagus binodis‟, Journal of Evolutionary Biology, 20(2), pp. 488–495. doi:

496 10.1111/j.1420-9101.2006.01274.x.

497 Stålhandske, P. (2001) „Nuptial gift in the spider Pisaura mirabilis maintained by sexual

498 selection‟, Behavioral Ecology, 12(6), pp. 691–697. doi: 10.1093/beheco/12.6.691.

499 Taborsky, M. and Brockmann, H. J. (2010) Animal behaviour: Evolution and mechanisms,

500 Animal Behaviour: Evolution and Mechanisms. doi: 10.1007/978-3-642-02624-9.

501 Tuttle, E. M. (2003) „Alternative reproductive strategies in the white-throated sparrow :

502 behavioral and genetic evidence‟, 14(3), pp. 425–432.

503 Webb, J. K., Shine, R. and Christian, K. (2006) „The Adaptive Significance of Reptilian

504 Viviparity in the Tropics : Testing the Maternal Manipulation Hypothesis Author‟, 60(1), pp.

505 115–122.

506 Wiklund, C. and Kaitala, A. (1995) „Sexual selection for large male size in a polyandrous

507 : the effect of body size on male versus female reproductive success in Pieris

508 napi‟, Behavioral Ecology, 6(1), pp. 6–13. doi: 10.1093/beheco/6.1.6.

509 Wright, T. F. et al. (2010) „Behavioral flexibility and species invasions: The adaptive

54 510 flexibility hypothesis‟, Ethology Ecology and Evolution, 22(4), pp. 393–404. doi:

511 10.1080/03949370.2010.505580.

512 Yoccoz, N. G. et al. (2002) „Age- and density-dependent reproductive effort in male red

513 deer‟, Proceedings of the Royal Society B: Biological Sciences, 269(1500), pp. 1523–1528.

514 doi: 10.1098/rspb.2002.2047.

515

55

Chapter II

Resource availability, mating opportunity, and sexual selection intensity influence the expression of male alternative reproductive tactics

Ghislandi P.G., Pekár S., Matzke M., Schulte-Doinghaus S., Bilde T. & C. Tuni

Re-submitted to The Journal of Evolutionary Biology

56 1 Resource availability, mating opportunity, and sexual selection intensity influence the

2 expression of male alternative reproductive tactics

3

4 Paolo Giovanni Ghislandi* a, Stano Pekár b, Magdalena Matzke c, Sarah Schulte-

5 Doinghaus c, Trine Bilde a✚, Cristina Tuni c✚

6

7 a Department of Bioscience, Aarhus University, Aarhus, Denmark

8 b Department of Botany and Zoology, Masaryk University, Brno, Czech Republic

9 c Department of Biology, Ludwig Maximilian University, Munich, Germany

10 * Corresponding author: [email protected]; Ny Munkegade 116, 8000 Aarhus C,

11 Denmark

12 ✚ These authors contributed equally.

13

14 Running title: Alternative mating traits in gift-giving spiders

15

16

17 Acknowledgements

18 We thank Alexander Hutfluss for valuable help in collecting spiders, André Walter and 19 Frederik Hendrickx for thoughtful discussion, and Niels Dingemanse for logistic support. We 20 are grateful for helpful comments on a previous draft of the manuscript from Caitlin Stern, 21 Jessie Barker, André Walter and Thomas Rejsenhus. We also thank Rainar Nitzsche for 22 providing useful information. The study was supported by The Danish Council for 23 Independent Research grant number 4002-00328B to TB.

24

25

26 Conflict of interests

27 There are no conflict of interests for the present manuscript

57 28 Abstract

29 Variation in the expression of plastic alternative reproductive tactics may depend on

30 environmental cues that vary spatially or temporally. For example, variation in resources such

31 as food or competition for mates is expected to influence the pay-off of each tactic and shape

32 the decision of which tactic to employ. Males of the nuptial gift-giving spider Pisaura

33 mirabilis can adopt three tactics: offering a genuine prey gift, a ‘worthless’ non-nutritious

34 gift, or no gift. We propose that food resources and/or male body condition, mating

35 opportunities, and sexual selection intensity, vary over the course of the mating season to

36 shape the expression of alternative traits. We quantified these variables in the field over two

37 seasons, to investigate the correlation between environmental variables and tactic as the

38 mating season progresses. We hypothesize that 1) males become more likely to employ a gift-

39 giving tactic, and 2) the likelihood of switching from worthless to genuine gifts increases with

40 the availability of prey and /or mates. Prey availability increased over the season and co-

41 varied with the propensity of males to employ the gift-giving tactic, but we found no support

42 that males in good condition are more likely to produce genuine gifts. Males responded to an

43 increase in female availability by increasing their mating effort (gift production). The

44 frequency of genuine gift use increased with the progress of season, which is consistent with

45 increasing sperm competition intensity with time. Our results suggest that the frequency of

46 alternative tactics is shaped by seasonal changes in ecological factors and possibly post-

47 copulatory sexual selection. This leads to relaxed selection for the gift-giving tactic early in

48 the season when females are less choosy and resources more scarce, and increased selection

49 for genuine gifts later in the season driven by mating opportunity and risk of sperm

50 competition.

51

52 Key words: alternative reproductive tactics, nuptial gifts, opportunistic mating, sexual

53 selection, prey availability, female availability

58 54 Introduction

55

56 Intra-sexual competition over mate acquisition generates strong selection on morphological

57 and/or behavioural traits that increase male reproductive success, and is an important factor in

58 promoting phenotypic variation in sexual traits (Cogliati et al., 2014; Conroy & Gray, 2014;

59 Fraser et al., 2014; Outomuro et al., 2014). Phenotypic variation in a sexual context may

60 result in alternative reproductive tactics (ARTs), where several phenotypes are expressed

61 simultaneously in a population (Emlen, 2008; Oliveira et al., 2008; Outomuro et al., 2014).

62 ARTs can be genetically determined (Moczek & Emlen, 2000; Hunt & Simmons, 2001), or

63 context-dependent (plastic) and expressed in response to external or internal conditions

64 (Brockmann 2001; Smith et al. 2013). The maintenance of ARTs within populations is often

65 ascribed to frequency dependent selection or density dependence (Oliveira et al., 2008). For

66 example, a model developed by Hendrickx et al. (2015) showed that a high-investment male

67 phenotype is constrained in mating opportunities, because the costs of developing this

68 phenotype leads to extended development time and maturation later in the season. This

69 maturation delay leads to a vacant mating niche in the beginning of the mating season, which

70 favours a low-investment male phenotype that matures early to exploit the availability of

71 unmated females. The high-investment and low-investment strategies are maintained by

72 negative frequency-dependent selection (Hendrickx et al., 2015). Variation in ecological

73 factors is expected to play a key role in shaping the evolution and maintenance of alternative

74 traits, because changes in the environment alter the costs and benefits associated with trait

75 expression (Emlen, 2008; Oliveira et al., 2008). This may be particularly relevant for plastic

76 and simultaneous tactics, where fluctuating resource availability influences trait expression

77 (Oliveira et al., 2008; Bell, 2010).

78 The use of nuptial gifts in the form of nutritious donations offered to females to

79 facilitate male reproductive success is a common strategy, particularly in insects and spiders

59 80 (Boggs, 1995; Vahed, 1998; Lewis & South, 2012). Males of the nuptial gift-giving spider

81 Pisaura mirabilis (Clerck, 1757) adopt three different strategies: they can offer a genuine

82 nuptial gift consisting of an insect prey wrapped in silk (Bristowe & Locket, 1926; Nitzsche,

83 1988; Bilde et al., 2007), they can offer a ‘worthless’ gift consisting of prey remains, empty

84 exoskeletons, or even plant parts wrapped in silk (Nitzsche, 1988; Albo et al., 2011b;

85 Ghislandi et al., 2017), and they can mate without offering any gift at all (Stålhandske, 2001;

86 Albo et al., 2011b). Females exert strong preference for males that offer a nuptial gift

87 (Stålhandske, 2001; Albo et al., 2011b). While the donation of a nuptial gift dramatically

88 increases male mating success, gift production is likely to be costly (Simmons, 1990). This

89 could be, for instance, through increased exposure to predators while hunting for a prey, use

90 of venom to kill the prey, metabolic effort involved in prey capture and gift construction, lost

91 foraging opportunities, or energetic costs of carrying the gift while searching for a mate.

92 These potential costs might explain the coexistence of gift and no-gift tactics in the

93 population.

94 Interestingly, males that offer worthless gifts are equally likely to be accepted by

95 females as males offering a genuine gift, at least when females are unmated (Albo et al.,

96 2011b). If worthless gifts are less costly to produce, this should favour the worthless gift-

97 giving tactic in the population, as seen in dance flies (Empididae) (Preston-Mafham, 1999;

98 LeBas & Hockham, 2005). However, both gift-giving strategies are observed, and this is

99 likely because males with worthless gifts experience shorter copulations, as the female

100 interrupts the copulation sooner, which results in reduced sperm transfer and fertilization

101 success (Albo et al., 2011b). Since females are polyandrous (Tuni et al., 2013) the cost of

102 offering worthless gifts to mated females is amplified, as these males lose out in sperm

103 competition to males that offer genuine gifts and thereby acquire longer copulations to

104 transfer more sperm. This effect exerts selection for the genuine gift-giving tactic.

105 Interestingly, we observe all three mating tactics simultaneously in any given population

60 106 (Albo et al 2011; XY, unpublished data), prompting the question of what maintains the co-

107 existence of alternative strategies in this system (Ghislandi et al., 2014). Preliminary data

108 suggests that ARTs of P. mirabilis are plastic, as individual males are capable of producing

109 both genuine and worthless gifts (XY, unpublished data). It is therefore possible that males

110 switch among tactics in response to environmental or intrinsic factors, in order to adopt the

111 most advantageous tactic at any given time (Maynard Smith, 1982; Gross, 1996). The

112 interplay between environmental and/or intrinsic factors may result in differential pay-offs for

113 each of the male mating tactics at different times of the season thereby promoting the

114 simultaneous expression of tactics (Ghislandi et al., 2014; Hendrickx et al., 2015; Taborsky

115 and Brockmann 2010).

116 A number of external factors can influence gift tactic: Gift occurrence (gift vs. no gift) -

117 When considering the occurrence of a gift tactic (any gift, genuine or worthless) compared to

118 a no gift tactic, one possibility is that construction of a nuptial gift may depend on prey

119 availability, which provides males with the opportunity for gift construction. This generates

120 the expectation that the proportion of males producing any gift (genuine or worthless)

121 increases over the course of the mating season because prey becomes more abundant (Hatley

122 & Macmahon, 1980; Clarke & Gaston, 2006), and prey acquisition would require lower

123 effort. If gift construction is condition-dependent (Albo et al., 2011a), and male condition

124 improves with increasing prey availability over the season, we would expect a higher

125 frequency of the no-gift strategy early in the season, as only males in good condition can bear

126 the cost of producing a gift. If gift construction depends on sexual selection intensity, we

127 expect that the no-gift tactic is predominantly used early in the mating season. Firstly, because

128 sperm competition is lower at the beginning of the season, and should favour low sperm

129 investment tactics (Parker & Pizzari, 2010). Females are known to mate with multiple males

130 in the wild and broods often represent at least 4 sires (Lauridsen et al., 2016), justifying the

131 assumption that sperm competition increases as the season progresses. Second, preliminary

61 132 data show that only virgin females accept males courting without a gift (XY, unpublished

133 data), suggesting a possible role for variation in female choice over the season in influencing

134 tactic use.

135 Gift type (genuine vs. worthless gifts) - Once males construct a nuptial gift, they can produce

136 either a genuine or worthless gift. We expect the proportion of males producing worthless

137 gifts to be higher earlier in the mating season. When prey availability is low, males may face

138 greater effort in prey capture and rely on prey for their own feeding prior to gift construction.

139 Moreover, males may be in lower body condition and unable to bear the costs of genuine gift

140 construction (Albo et al., 2011a). Finally, post-copulatory sexual selection is expected to

141 increase as the mating season progresses, creating selection on males to offer genuine gifts

142 that facilitate longer sperm transfer and improved performance in sperm competition

143 (Drengsgaard & Toft, 1999).

144 To examine how male reproductive tactics vary in relation to several ecological factors

145 we performed a 2 year field study where we i) determined the relative use of three male

146 tactics (no gift, genuine gift, worthless gift) over the mating season, and ii) investigated how

147 male tactic expression co-varies with prey availability, male body condition, and intensity of

148 sexual selection (measured as operational sex ratio (OSR), and changes in the availability of

149 adult females) over the season (Ghislandi et al., 2014). Overall these data allow us to

150 investigate 1) whether males become more likely to employ a gift-giving tactic compared to

151 no-gift as the season progresses, and 2) whether the likelihood of switching from worthless to

152 genuine gifts increases as the mating season progresses.

153

154 Materials and Methods

155 Pisaura mirabilis

156 The nursery-web spider Pisaura mirabilis belongs to the Family Pisauridae (World Spider

157 Catalog) and has a Paleartic distribution. The species typically occurs in grassy meadows and

62 158 forest areas and is a generalist predator, feeding on arthropod prey (Nentwig, 1986). The life

159 cycle is annual (Bristowe & Locket, 1926), with spiderlings hatching from late spring to early

160 summer depending on the latitude (XY, pers. obs.), and reaching adulthood the following

161 spring. The mating season occurs in May-June and females produce eggsacs in June-July,

162 which they carry in their mouthparts. Prior to hatching, females produce a tent shaped

163 nursery-web, where they suspend the eggsac and guard their brood until dispersal.

164

165 Field-collected spiders and nuptial gifts

166 Pisaura mirabilis spiders were collected in XY, during spring 2015 and 2016 over the

167 duration of the mating season, from the 11th of May to the 12th of June in 2015 and from the

168 21st of May to the 22nd of June in 2016. Spiders were primarily collected from grassy

169 meadows and forest edge, consisting mainly of Urtica spp. bushes. On each day of collection,

170 all specimens found were caught and placed in transparent plastic vials (10 x 4 cm or 10 x 2

171 cm) closed with foam lids. Once collecting ended, we scored the developmental stage and sex

172 of all individuals (juveniles, subadult males and females, and adult males and females) and

173 counted the total number of individuals collected per stage. If adult males carried a nuptial

174 gift, the gift was immediately removed using forceps, placed in an Eppendorf tube, and

175 tagged with the male ID. Males and gifts were brought to the laboratory at XY University,

176 XY. Juveniles, subadult and adult females were released after identification. A few

177 individuals that escaped during the survey were counted if their sex and life history stage

178 could be correctly identified. We avoided the risk of multiple counting by sampling in one

179 direction only.

180

181 Determination of gift type

182 In the laboratory, nuptial gifts from 2015 were opened under a Zeiss stereomicroscope W-PI

183 10 x 23 binocular using a Zeiss KL 1500 LCD light by gently removing the silk cover with

63 184 micro-dissecting forceps. Gift content was inspected and the items found inside counted. The

185 gift was defined as genuine if it contained fresh and intact prey, and worthless if it contained

186 dried prey or prey leftovers (i.e. pieces of body parts or exoskeletons), seeds, pollens or plant

187 parts. Gifts that contained more than one prey were classified as genuine when at least one of

188 the items consisted of fresh prey.

189 Nuptial gift type in 2016 was determined based on their weight, as we used data from 2015 to

190 estimate the relationship between gift type (genuine or worthless) and gift weight using a logit

191 linear calibration model fitted by GLM with binomial error structure. We determined the

192 threshold between genuine and worthless gifts based on 155 field-collected gifts (93 from

193 2015 and 62 collected from other experiments) that were weighed and opened. The obtained

194 model of the form ln(p/1-p)=1.983+-0.287weight, where p is the probability of genuine type,

195 estimated a 50% probability of the genuine gift type to be at 6.9095 mg. Thus, gifts heavier

196 than this value were considered genuine, and gifts lighter than this value were considered

197 worthless.

198

199 Male condition

200 In the laboratory, male body mass was measured using a digital scale KERN PKT (KERN-

201 SOHN GmbH, Germany) accurate to 0.001 g. Cephalothorax width was measured using a

202 digital calliper (AEROSPACE, China) accurate to 0.01 mm by gently covering males with a

203 transparent plastic sheet or a mosquito net to immobilize them. The measure was taken three

204 times per male. To evaluate male body condition we calculated Residual Body Index (RBI),

205 by saving residuals from a simple linear regression of body weight on mean cephalothorax

206 width, calculated on the three repeated measures (Jakob et al., 1996). All males were released

207 either immediately or, on a few occasions on the following day, in the same location as they

208 were collected.

209

64 210 Prey availability

211 To measure prey availability in the mating season of 2015, we used three different capture

212 methods: pitfall traps, sticky sheets, and sweeping net. In 2016, we did not use sweep nets as

213 sweeping yielded the lowest prey capture.

214 Pitfall traps were used to catch prey moving on the ground. They consisted of plastic cups

215 (0.2 l, 10x7 cm) placed approximately 10 m from each other, covered by a plastic mesh (grid

216 6.44 mm x 12.81mm). Each cup was placed in a hole dug in the ground and filled with a

217 solution of water and liquid soap that reduces the water surface tension, preventing

218 from walking on the fluid and escaping the trap. Due to its size, the plastic mesh prevented

219 larger organisms from falling in while permitting relevant-sized prey to slide down the waxed

220 knitting. We placed a total of 72 pitfall traps in 2015 and 40 in 2016. This was because in

221 2015 we sampled both a forest and a meadow area, but found that spiders were mainly present

222 in the meadow, therefore the forest area was not sampled in 2016. Traps were replaced every

223 2-3 days in 2015, and every 4 days in 2016. Sticky sheets were used to catch flying prey.

224 They consisted of transparent A4 format sheets placed approximately 5 m from one another.

225 Each sheet was covered with a thin layer of odourless and pheromone free, transparent,

226 weather-resistant insect-glue (Temmen Insekten-Leim, Temmen GmbH, Germany). In the

227 field, they were hung on branches of bushes and trees (1.5 m above the ground on average)

228 with sisal ropes (3 mm diameter) fixed on two pierced holes in the corner of the sheet.

229 Removed sticky traps were covered with clean wrapping plastic film, brought to the

230 laboratory and frozen until the content was scored. We placed a total of 74 sticky sheets in

231 2015 and 32 in 2016 (see explanation above).

232 Sweeping was used to catch prey sitting on grass blades or flying immediately over the

233 vegetation level and was performed with a sweeping net (32 cm diameter, 72.5 cm length,

234 grid 0.72 mm x 0.72 mm). We used a standardized procedure consisting of sweeping once

235 every step moved, for a total of five sequential steps walking in one direction. This was

65 236 repeated three times. The net content was emptied in plastic bags and prey samples frozen

237 until identification was performed. We swept a total of 70 times.

238 All collected arthropods were sorted and identified to order in the laboratory using a

239 Zeiss stereomicroscope W-PI 10 x 23 binocular and a Zeiss KL 1500 LCD light. Arthropods

240 groups scored as prey were Diptera, Coleoptera, Hymenoptera, Hemiptera, Arachnida,

241 Lepidoptera and minor Pterygota groups. We excluded from our analysis the groups that are

242 known not to represent a trophic resource for Pisaura mirabilis, such as Mollusca, Chilopoda,

243 Myriapoda, Odonata, Hymenoptera (Formicidae), and Hemiptera (Aphididae) (Toft, 1997;

244 Bilde & Toft, 2001). We discarded Isopoda in the 2015 analysis as we did not find them

245 wrapped in nuptial gifts, but we considered them in 2016 as they occurred in a few cases as

246 wrapped prey.

247

248 Measures of sexual selection intensity

249 To evaluate the relationship between operational sex ratio and gift-trait, we measured the

250 Operational Sex Ratio (OSR) on each day of sampling, calculated by dividing the total

251 number of adult males by the sum of adult males and females, only including individuals with

252 reproductive capability. Therefore, adult females with eggsacs or with a nursery-web were

253 excluded from the analysis, since these females are not sexually receptive while taking care of

254 their eggs/broods (Nitzsche, pers. comm.). We also calculated how the change in availability

255 of mature females over the season co-varied with male reproductive tactic, by estimating the

256 proportion of adult females relative to the sum of adult females plus sub-adult females during

257 the mating season.

258

259 Statistical Analyses

260 Prey availability, male condition, sexual selection intensity over the mating season

66 261 We first explored how prey availability, OSR, proportion of maturing females, proportion of

262 males with a gift (N males with a gift/total number of males), and proportion of worthless

263 gifts (N males with a worthless gift/total number of males with a gift) varied over the mating

264 season using Generalized Additive Models (GAM), that allow fitting both linear and non-

265 linear relationships to temporal trends (Wood, 2006). When investigating seasonal trends,

266 because the effects of many (not measured) environmental variables, such as temperature,

267 precipitation, sun isolation, etc., fluctuate and have strong effect on the study response

268 variables, smoothed curves to temporal trends (non-linear relationships) provide a better fit.

269 Thin-plate splines with automatic selection of their dimension were used with non-linear

270 terms. Depending on the distribution and variance-to-mean ratio of the response variable, we

271 used Gamma error structure (GAM-g) to analyse effects of season on amount of prey, and

272 binomial errors (GAM-b) for OSR, proportion of maturing females, proportion of males with

273 a gift, and proportion of worthless gifts.

274 When investigating variation of Residual Body Index (RBI) over the season we conducted a

275 GAM model using Gaussian errors (GAM), with both season and prey availability as

276 explanatory variables. We also used standard linear model (LM) to explore the effect of prey

277 availability on RBI.

278

279 Proportion of males with a gift (genuine or worthless), and with a worthless gift

280 To explore the relationships between prey availability, OSR and proportion of mature females

281 respectively with the proportion of males with a gift and with the proportion of males with a

282 worthless gift we used Generalized Linear Models (GLMs) with binomial errors structure.

283

284 Analyses were performed separately for each year because there were major differences

285 in prey number and other factors among years, but the effect of year per se was not an

286 objective of the study. Adequacy of models was evaluated by assessing the distribution of

67 287 residuals and homogeneity of variance using standard diagnostic tools. All analyses were

288 performed in R 3.0.0 (R Core Team 2015) using functions from the mgcv package (Wood,

289 2006).

290

291 Results

292 In total we collected 2511 spiders, 1327 in 2015 and 1184 in 2016, of which 387 were adult

293 males in 2015 and 241 in 2016 (Figure S1). In 2015, 93 carried a gift (24%), of which 42

294 were worthless (45.1%), while in 2016, we collected 56 males carrying a gift (23.2%) of

295 which 25 were classified as worthless (44.6%).

296

297 Prey availability, male condition, sexual selection intensity over the mating season

298 We caught, counted and identified 14729 arthropods, 12851 collected in 2015 and

299 1878 collected in 2016. In 2015, there was on average 5.4 times more prey per male available

300 than in 2016 excluding data from sweep nets in 2015. In 2015, the amount of prey (suitable

301 prey/male/day) increased significantly during the mating season (GAM-g, F1,19 = 4.6, P =

302 0.045, Figure S2a). In 2016, the effect of season on prey availability was also significant, but

303 showing a different pattern, prey availability fluctuated and increased only slightly over

304 season (GAM-g, F5.8,6.0 = 3.7, P = 0.02, Figure S2b).

305 ‘Season’ (time) had significant effect on Male Residual Body Index (RBI) in both

306 years. In 2015 RBI fluctuated with no clear trend with the progress of the mating season

307 (GAM, F7.2,8.2 = 6.2, P < 0.0001, Figure S3a). In 2016 a slightly negative overall relationship

308 was present (GAM, F6, 7.2 = 6.9, P < 0.0001, Figure S3b), with male body condition

309 decreasing as the season progressed. Male RBI increased significantly with prey availability

310 in 2015 (LM, F1,378 = 3.8, P = 0.02, Figure S4a), and increased significantly initially with

311 prey availability and then decreased in 2016 (LM, F1,173 = 16.5, P < 0.0001, Figure S4b).

68 312 In 2015, OSR (the number of adult males divided by the sum of adult males and

313 females) became more female-biased over the mating season but showing a fluctuating

2 314 pattern (GAM-b, X 4.75 = 19.6, P = 0.003, Figure S5a). In 2016, no significant relationship

2 315 between season and OSR was found (GAM-b, X 5.6 = 11.1, P = 0.12, Figure S5b). The

316 proportion of maturing females showed a significant increase over the season in both years

2 2 317 (2015: GAM-b, X 3.8= 5.8, P = 0.0035, Figure S6a; 2016: GAM-b, X 2.2 = 18.7, P < 0.0001,

318 Figure S6b).

319

320 Proportion of males with a gift (genuine or worthless)

321 We found a significant increase over time in the proportion of males carrying a gift in 2015

2 322 (GAM-b, X 1 = 6.6, P = 0.018, Figure 1a) and in 2016 (GAM-b, F1,16 = 5.2, P = 0.04, Figure

323 1b). The proportion of males carrying a gift and prey availability were significantly positively

2 324 related in 2015 (GLM-b, X 1 = 7.7, P = 0.005, Figure 2a), whereas we did not observe the

2 325 same relationship for 2016 (GLM-b, X 1 < 0.1, P = 0.95, Figure 2b). The proportion of males

326 carrying a gift was not significantly affected by male body condition in 2015 (GLM-b, F1,86 =

327 0.4, P = 0.52) but it was negatively affected in 2016 (GLM-b, F1,159 = 5.4, P = 0.021; Figure

328 S7). We found no effect of OSR on the proportion of males carrying a gift for both 2015

2 2 329 (GLM-b, X 1 = 1.5, P = 0.21) and 2016 (GLM-b, X 1 < 0.5, P = 0.50). We found a positive

330 significant relationship between the proportion of males carrying a gift and the proportion of

2 331 maturing females, for both 2015 (GLM-b, X 1 = 4.3, P = 0.03, Figure 3a) and 2016 (GLM-b,

2 332 X 1 = 4.4, P = 0.036, Figure 3b).

333

334 Proportion of males with a worthless gift

335 In 2015, the proportion of males carrying a worthless gift varied over time (GAM-b, F1.9,2.0 =

336 7.0, P = 0.043, Figure 1c): early in the season the proportion was relatively high, then it

337 declined with males offering proportionally more genuine gifts, and at the end of season the

69 338 proportion of worthless gifts again increased. In 2016, the proportion of worthless gifts had a

339 rather similar U-shaped pattern (GAM-b, F1.9,2.0 = 7.5, P = 0.041, Figure 1d).

340 For the mating season of 2015, we found a significant positive relationship between

2 341 the proportion of males carrying a worthless gift and prey availability (GLM-b, X 1 = 4.7, P =

342 0.03, Figure 2c), while this relationship was not statistically significant for 2016 (GLM-b,

343 F1,12 = 0.2, P = 0.67, Figure 2d). RBI did not predict the type of gift used by males either in

344 2015 (GAM-b, F1,86 = 0.4, P = 0.52) or in 2016 (GAM-b, F1,41 = 3.7, P = 0.063). We also

345 found a significant positive effect of OSR on the proportion of males carrying a worthless gift

346 for 2015, with a higher proportion of males carrying these donations as the relative number of

2 347 males increased (GLM-b, X 1 = 5.5, P = 0.019, Figure 4a). However, we did not detect the

348 same relationship for 2016, where the proportion of males carrying a worthless gift was not

349 significantly affected by OSR (GLM-b, F1,15 = 0.2, P = 0.63, Figure 4b). A significant

350 negative relationship between the proportion of males carrying a worthless gift and the

2 351 proportion of mature females was present in 2015 (GLM-b, X 1 = 4.8, P = 0.029, Figure 3c),

352 but this relationship was not significant for 2016, although a similar trend existed (GLM-b,

353 F1,12 = 1.7, P = 0.22, Figure 3d).

354

355 Discussion

356

357 Our result reveal that variation in prey availability, female availability, OSR and male body

358 condition show a degree of correlation with the proportion of males with a gift or a worthless

359 gift, suggesting that variation in external factors may influence the expression of male tactics.

360 This can happen if variation in resource availability, mating opportunity, and sexual selection

361 intensity over the mating season generates variation in the pay-off of different male

362 reproductive tactics, which leads to shifts in the relative expression of different tactics over

363 the season (Ghislandi et al., 2014).

70 364 We found a tendency for prey availability to increase over the mating season, in

365 accordance with the expectation of seasonal variation in temperate environments (Hatley &

366 Macmahon, 1980; Clarke & Gaston, 2006). Prey availability could influence male gift-use in

367 two ways. First, if male body condition depends on prey availability, and gift production is a

368 condition dependent trait, we expect body condition (Residual Body Index, RBI) to improve

369 with increasing prey availability, and the likelihood of males producing a nuptial gift to

370 increase with increasing RBI. We found a tendency for RBI to increase with prey availability

371 in both years (only initially for 2016, Figure S4), however, this did not translate into a general

372 improvement of male body condition over the mating season (Figure S3). In contrast to the

373 condition dependent model, we found no relationship (2015), and a negative relationship

374 82016) between the expression of nuptial gifts in the field and male RBI (Figure S7). So

375 while there is some evidence for prey availability to influence RBI positively, we have no

376 evidence that gift production is related to good body condition. If anything, we find the

377 opposite effect that males in lower body condition are more likely to produce a nuptial gift

378 (2016). Secondly, prey availability could directly influence the likelihood of obtaining a

379 nuptial gift, as we expect the effort of gift acquisition to decrease when prey becomes more

380 abundant. The data from 2015 showed that the likelihood of finding a male with a nuptial gift

381 was positively correlated with prey availability (Figure 2a). This result supports the prediction

382 that opportunity, in terms of prey availability, influences male tactic at least in some years.

383 The likelihood of prey capture and gift production may therefore co-vary, so increasing prey

384 availability over the mating season reduces the effort of acquiring a gift, thereby favouring the

385 gift-giving tactic.

386 We also expected that variation in prey availability influences which gift type (genuine

387 or worthless) males produce. We hypothesized that under low prey availability, there is higher

388 incentive to partially consume the prey and use the remains as a worthless gift, whereas males

389 would shift to producing genuine gifts when prey availability increases (Ghislandi et al.,

71 390 2014). However, the proportion of genuine gifts did not increase with prey availability, in

391 contrast the likelihood of producing a worthless gift increased with prey availability in 2015

392 but not in 2016 (Figure 2c,d). Overall, it seems that while opportunity in terms of prey

393 availability influences the likelihood of producing a gift (any gift), apparently it does not

394 predict the increased production of genuine nuptial gifts. This leads us to suggest that prey

395 availability does not have major influence on use of nuptial gift tactic.

396

397 In addition to food resources, we proposed that sexual selection plays an important role for

398 which tactic a male employs (Ghislandi et al., 2014). We assume that post-copulatory sexual

399 selection intensifies as the mating season progresses, because mated females become more

400 resistant to additional matings (Halliday, 1983; Gabor & Halliday, 1997; Bateman et al.,

401 2001; Tuni & Bilde, 2010), and this resistance may be overcome by males offering nuptial

402 gifts (Albo et al., 2011b; Hendrickx et al., 2015; Toft & Albo, 2016). Indeed, males were

403 increasingly more likely to carry a nuptial gift as the season progressed. Males without gifts

404 occurred at higher frequency early in the season, where it is possible to mate with a virgin

405 female without a gift (Stålhandske, 2001). We note that these males cannot necessarily all be

406 assumed to have chosen a no-gift tactic, as it is possible that some males are not yet ready to

407 mate, or that a male that recently mated lost his gift to the female. Nevertheless, this pattern is

408 consistent with the prediction that the use of the no-gift tactic is only successful early in the

409 season. Later in the season, the risk of sperm competition imposes strong selection on the gift-

410 giving trait (Parker & Pizzari, 2010), as males with a nuptial gift experience substantially

411 longer copulations than males with no gift (Stålhandske, 2001; Albo et al., 2011b). This is

412 because copulation duration is determined by the time it takes females to consume the gift

413 (Lang, 1996; Stålhandske, 2001), and it takes longer time to consume a genuine gift than a

414 worthless gift (Albo et al., 2011b).

72 415 We also found that males were more likely to carry a nuptial gift when the likelihood of

416 encountering a female increased. This indicates that males respond to female availability by

417 increasing their effort on gift construction. To adjust tactic according to female density in the

418 environment, males must be capable of assessing fluctuations in the number of potential

419 mates. Males spiders often use chemical cues produced by females and deposited on silk to

420 find their mates (Schulz & Toft, 1993; Roberts & Uetz, 2005; Wyatt, 2014). We know from

421 laboratory trials that male P. mirabilis respond to cues in female silk by displaying

422 behavioural sexual excitement and initiate the construction of nuptial gifts (Ghislandi et al.,

423 2017). Higher female densities in the field are likely to increase the intensity of sexual

424 chemical cues, providing a cue for males to assess female availability and/or female mating

425 status (virgin or mated) (Maklakov et al., 2003; Gaskett, 2007; Tuni & Berger-Tal, 2012).

426 Finally, we predicted the use of genuine over worthless gifts to increase as the season

427 progresses, in response to increased risk of sperm competition. Pisaura mirabilis females are

428 polyandrous (Tuni & Bilde, 2010; Tuni et al., 2013); microsatellite genotyping of offspring

429 showed that wild caught females could mate with at least four males (Lauridsen, 2016), and in

430 the laboratory females accepted up to 16 males depending on their nutritional status (Toft &

431 Albo, 2015). This degree of polyandry imposes huge sperm competition risk on males,

432 favouring strategies that prolong copulation duration and maximize sperm transfer (Parker,

433 1970; Simmons, 2001; Parker & Pizzari, 2010). In P. mirabilis, copulation duration correlates

434 positively with the amount of sperm transferred (Drengsgaard & Toft, 1999; Albo et al.,

435 2013). Female nuptial gift consumption was shown to determine copulation duration (Lang,

436 1996; Stålhandske, 2001), and copulation duration was significantly longer when males

437 offered genuine gifts compared to worthless gifts (Albo et al., 2011b). In accordance with a

438 response to increased sperm competition, we found an increase in the frequency of genuine

439 gifts over the season. Worthless gifts were relatively more frequent early in the season when

73 440 virgin females are available, while the frequency of genuine gifts increased as the mating

441 season progressed.

442

443 The co-existence of ARTs in P. mirabilis

444 We found that variation in resources affects the propensity of males to employ the gift-giving

445 strategy, as increasing prey availability over the season in one year co-varied with the

446 frequency of males producing a gift. However, we found no support for gift production to be a

447 condition-dependent trait in the sense that males in good condition are more likely to produce

448 a gift, in contrast, the result for 2016 suggest an opposite effect. Since we did not detect a

449 correlation between RBI and the probability to offer a worthless gift, we suggest that male

450 decisions underlying tactic use is more likely to be opportunistic rather than driven by

451 nutritional needs that favour prey consumption over gift production. Variation in female

452 encounter rate also influences male tactic, as males responded to an increase in female

453 availability by increasing their mating effort (producing gifts). Finally, our study supports the

454 prediction that the frequency of the genuine gift tactic increases with sexual selection

455 intensity, based on the assumption that sperm competition increases with time and favours

456 males with genuine gifts, which enhances their chances in sperm competition. Our results

457 thereby provide some support for a previously proposed model that the simultaneous

458 expression of alternative reproductive tactics is maintained by seasonal changes in ecological

459 factors and variation in sexual selection (Ghislandi et al., 2014). This leads to relaxed

460 selection for the gift-giving tactic early in the season when females are less choosy and

461 resources more scarce. Worthless gifts are also more common early in the season, where any

462 gift type facilitates male mating success. Finally, our study suggests that elevated mating

463 opportunities and sperm competition favours the genuine gift-giving tactic late in the season.

464

465 References

74 466

467 Albo, M.J., Bilde, T. & Uhl, G. 2013. Sperm storage mediated by cryptic female choice for 468 nuptial gifts. Proc. R. Soc. London B Biol. Sci. 280: 20131735. 469 Albo, M.J., Toft, S. & Bilde, T. 2011a. Condition dependence of male nuptial gift 470 construction in the spider Pisaura mirabilis (Pisauridae). J. Ethol. 29: 473–479. 471 Albo, M.J., Winther, G., Tuni, C., Toft, S. & Bilde, T. 2011b. Worthless donations: male 472 deception and female counter play in a nuptial gift-giving spider. BMC Evol. Biol. 11: 473 329. BioMed Central Ltd. 474 Bateman, P.W., Gilson, L.N. & Ferguson, J.W.H. 2001. Male size and sequential mate 475 preference in the cricket Gryllus bimaculatus. Anim. Behav. 61: 631–637. Elsevier. 476 Bell, G. 2010. Fluctuating selection: the perpetual renewal of adaptation in variable 477 environments. Philos. Trans. R. Soc. London B Biol. Sci. 365: 87–97. The Royal Society. 478 Bilde, T. & Toft, S. 2001. The value of three cereal aphid species as food for a generalist 479 predator. Physiol. Entomol. 26: 58–68. Wiley Online Library. 480 Bilde, T., Tuni, C., Elsayed, R., Pekar, S. & Toft, S. 2007. Nuptial gifts of male spiders: 481 sensory exploitation of the female’s maternal care instinct or foraging motivation? Anim. 482 Behav. 73: 267–273. 483 Boggs, C.L. 1995. Male nuptial gifts: phenotypic consequences and evolutionary 484 implications. 485 Bristowe, W.S. & Locket, G.H. 1926. The courtship of British Lycosid spiders, and its 486 probable significance. Proc Zool Soc 22. 487 Brockmann, H.J. 2001. The evolution of alternative strategies and tactics. Adv. Study Behav. 488 30: 1–51. 489 Clarke, A. & Gaston, K.J. 2006. Climate, energy and diversity. Proc. Biol. Sci. 273: 2257– 490 2266. 491 Cogliati, K.M., Balshine, S. & Neff, B.D. 2014. Competition and cuckoldry: estimating 492 fitness of alternative reproductive tactics in plainfin midshipman. Behaviour 151: 1209– 493 1227. Brill. 494 Conroy, L.P. & Gray, D. a. 2014. Forced copulation as a conditional alternative strategy in 495 camel crickets. Behav. Ecol. Sociobiol. 68: 1431–1439. 496 Drengsgaard, I. & Toft, S. 1999. Sperm Competition in a Nuptial Feeding Spider, Pisaura 497 Mirabilis. Behaviour 136: 877–897. 498 Emlen, D.J. 2008. The roles of genes and the environment in the expression and evolution of 499 alternative tactics. Altern. Reprod. tactics An Integr. approach 85–108. Cambridge 500 University Press Cambridge. 501 Fraser, B. a, Janowitz, I., Thairu, M., Travis, J. & Hughes, K. a. 2014. Phenotypic and 502 genomic plasticity of alternative male reproductive tactics in sailfin mollies. Proc. R. 503 Soc. B Biol. Sci. 281: 20132310. 504 Gabor, C.R. & Halliday, T.R. 1997. Sequential mate choice by multiply mating smooth 505 newts: females become more choosy. Behav. Ecol. 8: 162–166. ISBE. 506 Gaskett, A.C. 2007. Spider sex pheromones : emission , reception , structures , and functions. 507 82: 27–48. 508 Ghislandi, P.G., Albo, M.J., Tuni, C. & Bilde, T. 2014. Evolution of deceit by worthless 509 donations in a nuptial gift-giving spider. Curr. Zool. 60: 43–51. 510 Ghislandi, P.G., Beyer, M., Velado, P. & Tuni, C. 2017. Silk wrapping of nuptial gifts aids 511 cheating behaviour in male spiders. Behav. Ecol. 28: 744–749. 512 Gross, M.R. 1996. Alternative reproductive strategies and tactics: diversity within sexes. 513 Halliday, T.R. 1983. The study of mate choice. Mate choice 1: 462. 514 Hatley, C.L. & Macmahon, J.A. 1980. Spider community organization: seasonal variation and 515 the role of vegetation architecture. Environ. Entomol. 9: 632–639. The Oxford

75 516 University Press. 517 Hendrickx, F., Vanthournout, B. & Taborsky, M. 2015. Selection for costly sexual traits 518 results in a vacant mating niche and male dimorphism. Evolution (N. Y). 69: 2105–2117. 519 Wiley Online Library. 520 Hunt, J. & Simmons, L.W. 2001. Status-dependent selection in the dimorphic beetle 521 Onthophagus taurus. Proc. R. Soc. London B Biol. Sci. 268: 2409–2414. The Royal 522 Society. 523 Jakob, E.M., Marshall, S.D. & Uetz, G.W. 1996. Estimating fitness: a comparison of body 524 condition indices. Oikos 61–67. JSTOR. 525 Lang, A. 1996. Silk investment in gifts by males of the nuptial feeding spider Pisaura 526 mirabilis (Araneae: Pisauridae). Behaviour 133: 697–716. Brill. 527 LeBas, N.R. & Hockham, L.R. 2005. An invasion of cheats; the evolution of worthless 528 nuptial gifts. Curr. Biol. 15: 64–7. 529 Lewis, S. & South, A. 2012. The evolution of animal nuptial gifts. Adv Study Behav 44: 53– 530 97. 531 Maklakov, A.A., Bilde, T. & Lubin, Y. 2003. Vibratory courtship in a web-building spider: 532 signalling quality or stimulating the female? Anim. Behav. 66: 623–630. Elsevier. 533 Maynard Smith, J. 1982. Evolution and the Theory of Games. Cambridge university press. 534 Moczek, A.P. & Emlen, D.J. 2000. Male horn dimorphism in the scarab beetle, Onthophagus 535 taurus: do alternative reproductive tactics favour alternative phenotypes? Anim. Behav. 536 59: 459–466. Elsevier. 537 Nentwig, W. 1986. Non-webbuilding spiders: prey specialists or generalists? Oecologia 69: 538 571–576. Springer. 539 Nitzsche, R.O.M. 1988. Brautgeschenk’ und Umspinnen der Beute be Pisaura mirabilis, 540 Dolomedes fimbriatus und Thaumasia uncata (Arachnida, Araneida, Pisauridae). Verh 541 Naturwiss Ver Hambg. 30. 542 Oliveira, R.F., Taborsky, M. & Brockmann, H.J. 2008. Alternative reproductive tactics: an 543 integrative approach. Cambridge University Press. 544 Outomuro, D., Rodríguez-Martínez, S., Karlsson, A. & Johansson, F. 2014. Male wing shape 545 differs between condition-dependent alternative reproductive tactics in territorial 546 . Anim. Behav. 91: 1–7. Elsevier Ltd. 547 Parker, G.A. 1970. Sperm Competition and Its Evolutionary Consequences in the Insects. 548 Biol. Rev. 45: 525–567. 549 Parker, G. a. & Pizzari, T. 2010. Sperm competition and ejaculate economics. Biol. Rev. 85: 550 897–934. 551 Preston-Mafham, K.G. 1999. Courtship and mating in Empis (Xanthempis) trigramma Meig., 552 E. tessellata F. and E. (Polyblepharis) opaca F. (Diptera: Empididae) and the possible 553 implications of “cheating” behaviour. J. Zool. 247: 239–246. Cambridge Univ Press. 554 Roberts, J.A. & Uetz, G.W. 2005. Information content of female chemical signals in the wolf 555 spider, Schizocosa ocreata: male discrimination of reproductive state and receptivity. 556 Anim. Behav. 70: 217–223. Elsevier. 557 Schulz, S. & Toft, S. 1993. Identification of a sex pheromone from a spider. Sci. 260: 1635– 558 1637. 559 Simmons, L.W. 1990. Nuptial feeding in tettigoniids male costs and the rates of fecundity 560 increase. Behav. Ecol. Sociobiol. 27: 43–47. Springer. 561 Simmons, L.W. 2001. Sperm competition and its evolutionary consequences in the insects. 562 Princeton University Press. 563 Smith, M.D., Schrank, H.E. & Brockmann, H.J. 2013. Measuring the costs of alternative 564 reproductive tactics in horseshoe crabs, Limulus polyphemus. Anim. Behav. 85: 165– 565 173. Elsevier. 566 Stålhandske, P. 2001. Nuptial gift in the spider Pisaura mirabilis maintained by sexual

76 567 selection. Behav. Ecol. 12: 691–697. 568 Toft, S. 1997. Acquired food aversion of a to three cereal aphids: intra-and 569 interspecific effects. Entomophaga 42: 63–69. 570 Toft, S. & Albo, M.J. 2015. Optimal numbers of matings: The conditional balance between 571 benefits and costs of mating for females of a nuptial gift-giving spider. J. Evol. Biol. 28. 572 Toft, S. & Albo, M.J. 2016. The shield effect: nuptial gifts protect males against pre- 573 copulatory sexual cannibalism. Biol. Lett. 12: 43–78. 574 Tuni, C., Albo, M.J. & Bilde, T. 2013. Polyandrous females acquire indirect benefits in a 575 nuptial feeding species. J. Evol. Biol. 26. 576 Tuni, C. & Berger-Tal, R. 2012. Male preference and female cues: Males assess female sexual 577 maturity and mating status in a web-building spider. Behav. Ecol. 23: 582–587. 578 Tuni, C. & Bilde, T. 2010. No preference for novel mating partners in the polyandrous 579 nuptial-feeding spider Pisaura mirabilis (Araneae: Pisauridae). Anim. Behav. 80: 435– 580 442. 581 Vahed, K. 1998. The function of nuptial feeding in insects: a review of empirical studies. 582 Biol. Rev. 73: 43–78. Wiley Online Library. 583 Wood, S. 2006. Generalized additive models: an introduction with R. CRC press. 584 Wyatt, T.D. 2014. Pheromones and animal behavior: chemical signals and signatures. 585 Cambridge University Press. 586 587

77 588 Figure Legends

589

590 Figure. 1. Proportion of P. mirabilis males with a nuptial gift (genuine or worthless) over the

2 591 mating season of in (a) 2015 (X 1 = 6.6, P = 0.018) and (b) 2016 (F1,16 = 5.2, P = 0.04), and

592 with a worthless gift in (c) 2015 (F1.9,2.0 = 7.0, P = 0.043) and (d) 2016 (F1.9,2.0 = 7.5, P =

593 0.041). Lines are the fitted GAM-b models.

594

595 Figure. 2. Relationship between prey availability and the proportion of P. mirabilis males

2 596 carrying a nuptial gift (genuine or worthless) in (a) 2015 (GLM-b, X 1 = 7.7, P = 0.005) and

2 597 (b) 2016 (GLM-b, X 1 < 0.1, P = 0.95), and the proportion of P. mirabilis males carrying a

2 598 worthless gift in (c) 2015 (GLM-b, X 1 = 4.7, P = 0.03) and (d) 2016 (GLM-b, F1,12 = 0.2, P =

599 0.67).

600

601 Figure 3. Relationship between the proportion of adult females and (a-b) the proportion of P.

2 602 mirabilis males with a nuptial gift (genuine or worthless) in (a) 2015 (GLM-b, X 1 = 4.3, P =

2 603 0.03) and (b) 2016 (GLM-b, X 1 = 4.4, P = 0.036), and (c-d) proportion of males with a

2 604 worthless gift in (c) 2015 (GLM-b, X 1 = 4.8, P = 0.029) and (d) 2016 (GLM-b, F1,12 = 1.7, P

605 = 0.22).

606

607 Figure. 4. The proportion of P. mirabilis males with a worthless gift in relation to OSR in

608 2015 (a) and 2016 (b). The line is the fitted GAM model. No significant relationship was

609 found in 2016.

610

78 Figures

()a) ()b) t if g a h t ales wi ales m f o n o i t r o p o r P 0.0 0.2 0.4 0.6 0.8 0.0 0.1 0.2 0.3 0.4 0.5 0.6

0 5 10 15 20 25 30 0 5 10 15 20 25 30 () c) () d) t if g less h t r o a w a h t ales wi m f o

n o i t r o p o r P 0.00.20.40.60.81.0 0.0 0.2 0.4 0.6 0.8 1.0

0 5 10 15 20 25 30 0 5 10 15 20 25 30 Season (day) Season (day)

Figure. 1. The proportion of P. mirabilis males over the mating season (a-b) with a nuptial gift

2 (genuine or worthless) in (a) 2015 (X 1 = 6.6, P = 0.018) and (b) 2016 (F1,16 = 5.2, P = 0.04), and

(c-d) with a worthless gift in (c) 2015 (F1.9,2.0 = 7.0, P = 0.043) and (d) 2016 (F1.9,2.0 = 7.5, P =

0.041). Lines are the fitted GAM-b models.

79

a) b)

c) d)

Figure. 2. The relationships between prey availability and (a-b) the proportion of P. mirabilis males

2 carrying a nuptial gift (genuine or worthless) in (a) 2015 (GLM-b, X 1 = 7.7, P = 0.005) and (b)

2 2016 (GLM-b, X 1 < 0.1, P = 0.95), and (c-d) the proportion of P. mirabilis males carrying a

2 worthless gift in (c) 2015 (GLM-b, X 1 = 4.7, P = 0.03) and (d) 2016 (GLM-b, F1,12 = 0.2, P = 0.67).

80 a) b) t f i g a h t i w

s male f o

n o i t r o p o r P 0.0 0.2 0.4 0.6 0.8 0.00.10.20.30.40.50.6 0.20.40.60.81.0 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 c) d)

Figure 3. The relationships between the proportion of mature females and (a-b) the proportion of P.

2 mirabilis males with a nuptial gift (genuine or worthless) in (a) 2015 (GLM-b, X 1 = 4.3, P = 0.03)

2 and (b) 2016 (GLM-b, X 1 = 4.4, P = 0.036), and (c-d) proportion of males with a worthless gift in

2 (c) 2015 (GLM-b, X 1 = 4.8, P = 0.029) and (d) 2016 (GLM-b, F1,12 = 1.7, P = 0.22).

81

Figure 4. The proportion of P. mirabilis males with a worthless gift in relation to OSR in (a) 2015

2 (GLM-b, X 1 = 5.5, P = 0.019) and (b) 2016 (GLM-b, F1,15 = 0.2, P = 0.63).

82 Supplementary Figures

(a) (b)

Figure S1. The proportion of P. mirabilis males, subadult females and adult females over the mating season of 2015 (a) and 2016 (b). The bars show the percentage of collected individuals per day of collection, the total number of individuals collected per day is given on the top of the bars. Spiders were collected on 21 days in the period of 11 May-12 June in 2015, and on 18 days from 22 May-

22 June in 2016.

83

(a) (b)

y a d / e al m / y e r p

e l b vaila A 0 5 10 15 020406080100

0 5 10 15 20 25 30 0 5 10 15 20 25 30 Season (days) Season (days)

Figure S2. Variation in prey availability over the mating season of P. mirabilis in (a) 2015 (F1,19 =

4.6, P = 0.045) and (b) 2016 (F5.8,6.0 = 3.7, P = 0.02). Lines are the fitted GAM-g models.

84 (a) (b) x e d n i y d o b al u d i s e R -0.2 -0.1 0.0 0.1 0.2 0.3 -0.4 -0.2 0.0 0.2 0.4

0 5 10 15 20 25 30 0 5 10 15 20 25 30 Season (days) Season (days)

Figure S3. Male Residual Body Index (RBI) over the mating season of (a) 2015 (F7.2,8.2 = 6.2, P <

0.0001) and (b) 2016 (F6, 7.2 = 6.9, P < 0.0001). Lines are the fitted GAM models.

85 (a) (b) x e d n i

y d o b

l ua d i es R -0.2 -0.1 0.0 0.1 0.2 0.3 -0.4 -0.2 0.0 0.2 0.4

0 20 40 60 80 100 2 4 6 8 10 12 14 Available prey/male/day Available prey/male/day

Figure S4. Relationship between P. mirabilis male Residual Body Index (RBI) and prey availability in (a) 2015 (F1,378 = 3.8, P = 0.02) and (b) 2016. (F1,173 = 16.5, P < 0.0001). Lines are the fitted LM models.

86 (a) (b) R S O 0.2 0.3 0.4 0.5 0.6 0.20 0.25 0.30 0.35 0.40 0.45 0.50

0 5 10 15 20 25 30 0 5 10 15 20 25 30 Season (day) Season (day)

2 Figure S5. Variation in OSR over the mating season of P. mirabilis in (a) 2015 (X 4.75 = 19.6, P =

2 0.003) and (b) 2016 (X 5.6 = 11.1, P = 0.12). Lines are the fitted GAM-b models.

87 (a) (b) ales m fe t f adul o

n o i t r o p o r P 0.4 0.6 0.8 1.0

0 5 10 15 20 25 30 Season (day) Season (day)

Figure S6. The proportion of adult females over the mating season of P. mirabilis in (a) 2015 (X2

2 3.8= 5.8, P = 0.0035) and (b) 2016 (X 2.2 = 18.7, P < 0.0001). Lines are the fitted GAM-b models.

Figure S7. Relationship between residual body index and the proportion of males with a nuptial gift

(genuine or worthless) in 2016. Logit GLM model is fitted.

88 Appendix

The effect of environmental variation on the expression and maintenance of ARTs

Paolo Ghislandi and Caitlin Stern, work in progress

Introduction

Theory predicts that stable ARTs can co-exist each tactic in the population is equally successful

(Maynard Smith, 1982; Gross, 1996). Evolutionarily Stable Strategy (ESS) models show, that if the average fitness return for each tactic is equal, alternative strategies cane maintained at different frequencies (Maynard Smith & Price, 1973, Gross, 1996). The empirical data collected in my thesis allows us to propose that the 3 gift tactics performed by Pisaura mirabilis are likely maintained by the interplay of intrinsic and extrinsic (environmental) factors, which we identified as male body condition, prey availability, number of receptive females and operational sex ratio. In collaboration with Dr. Caitlin Stern, I have initiated the development of a model to study under which set of conditions ARTs are expressed (individual tactic expression) and can be maintained (ESS situations for ARTs) in Pisaura mirabilis. The aim is to parameterize the model with empirical data collected in my PhD thesis, and to first model individual tactic expression based on the expected fitness outcome of each tactic depending on female encounter rates and females mating status. Female mating status has been found to play an important role for the no-gift strategy, as laboratory experiments showed that recently mated females never accept males without a nuptial gift

(Ghislandi, unpublished). Since we believe it is clear, that sperm competition plays a profound role for determining the fitness function of male tactic, this factor will also be incorporated in the model.

This work was recently initiated and is still and ongoing, therefore it is briefly presented with a short sketch of the model. (Caitlin gave pre-mature birth to twins, so the modelling work is currently on hold during maternity leave…).

89 Model

We modelled three alternative reproductive tactics that can be employed by male spiders: offering no gift, offering a worthless gift, and offering a genuine gift. We assume that any given male can express any of the three tactics (Chapter I), and can alter which tactic to use in response to environmental feedback (Chapter II).

Specifically, we consider the behaviour of males in response to different seasonal regimes, which are defined by two parameters: 1) the proportion of females encountered that are virgin rather than mated, and 2) the rate at which males encounter females. Within a seasonal period s, the proportion of females that are unmated (virgin) is given by (and thus the proportion of females that are mated is given by ). Note that this parameter is fixed for a given seasonal period. The number of females a male encounters on average within the seasonal period is given by . For example, if a seasonal period lasts for 10 days, and the encounter rate for that period is one female per day, .

Because males can change tactics in response to cues from the social environment (Chapter

II), the frequencies of each tactic can vary among seasonal periods. In seasonal period s, the probabilities that a male offers no gift, a worthless gift, or a genuine gift are , , and , respectively. Note that no other tactics are available to males, and thus .

The payoffs of each male tactic are determined by female mating status. A male that gives no gift, a worthless gift, or a genuine gift to a virgin female obtains the payoff , , or , respectively. A male that gives no gift, a worthless gift, or a genuine gift to a mated female obtains the payoff , , or , respectively.

Each male’s fitness is composed of the payoff from matings with virgin females,

, and the payoff from matings with mated females,

. Weighted by the frequencies of each female mating status type, the fitness contributions the male gains are and ( ) . The total fitness that a male gains from a given seasonal

90 period is these payoffs multiplied by the number of females encountered in that period:

( ) .

Model analysis

Using the fitness function , we can ask how the seasonal period parameters and influence the proportion of effort a male is selected to invest in each of the three alternative reproductive tactics. First, we insert biologically realistic values for the payoffs , , , , , and .

Next, we select the partial derivatives ⁄ , ⁄ , and ⁄ . By studying how and

affects these partial derivatives, we can determine how the social environment influences the optimal proportion of effort a male should invest in offering no gift, offering a worthless gift, and offering a genuine gift.

Results

We studied different sets of possible payoff values, which are summarized in Table 1 (to be extended and expanded). receives a fitness payoff of 10 as represents a male offering no gift to a virgin female, thus male is accepted in 40% of cases (Albo et al., 2011). receives a fitness pay-off of 15 as offering a worthless gift to a virgin female returns high acceptance rates, approximately of 90%. receives a pay-off of 15 as genuine gift-offering male experience high acceptance from virgin females, around 90% as well (Albo et al., 2011).

represents a male offering no gift to a mated female, which never accepts a male when recently mated. However, males might be likely to be subsequently accepted if female mating rates are low, therefore the fitness return is 5 (Ghislandi, unpublished). obtain a fitness pay-off of 10 as worthless gift-givers experience shorter copulations with already mated females as they interrupt copulation sooner when discovering gift content. They however experience the same acceptance as genuine gift-givers (Ghislandi, unpublished). receives a fitness pay-off of 15, as offering a genuine

91 gift provides prolonged copulations and high acceptance rates on mated females (Drengsgaard and Toft,

1999; Ghislandi, unpublished)

For set 1, ⁄ ( ), showing that the likelihood of offering no gift increases with both the proportion of virgin females and the number of females encountered. Similarly,

⁄ ( ), showing that the likelihood of offering a worthless gift also increases with both the proportion of virgin females and the number of females encountered. However,

⁄ , showing that the likelihood of offering a genuine gift increases with the number of females encountered but is not influenced by the proportion of virgin females.

Set

1 10 15 15 5 10 15

2

3

4

5

Table 1. Sets of mating payoff parameter values to be studied (to be expanded). , the payoff gained by a male that gives no gift to a virgin female; , the payoff gained by a male that gives a fake gift to a virgin female; , the payoff gained by a male that gives a genuine gift to a virgin female; , the payoff gained by a male that gives no gift to a mated female; , the payoff gained by a male that gives a fake gift to a mated female; , the payoff gained by a male that gives a genuine gift to a mated female.

92

Chapter III

Silk wrapping of nuptial gifts aids cheating behaviour in male spiders

Ghislandi P.G, Beyer M., Velado P. & C. Tuni

Behavioral Ecology (2017), 00(00), 1–6. doi:10.1093/beheco/arx028

93 Behavioral The official journal of the ISBE Ecology International Society for Behavioral Ecology

Behavioral Ecology (2017), 00(00), 1–6. doi:10.1093/beheco/arx028

Original Article Silk wrapping of nuptial gifts aids cheating behaviour in male spiders

Paolo Giovanni Ghislandi,a,* Michelle Beyer,b,* Patricia Velado,b and Cristina Tunib aDepartment of Bioscience, Aarhus University, Ny Munkegade 1540, DK-8000, Aarhus, Denmark, and bDepartment of Biology, Ludwig Maximilian University, Grosshaderner Strasse 2, 82152 Planegg- Martinsried, Germany *These authors share equal contribution to the work. Received 21 July 2016; revised 19 January 2017; editorial decision 20 January 2017; accepted 7 February 2017.

Sexual traits, such as nuptial gifts, are costly and often condition-dependent. Males should be under selection to reduce these costs without impairing their reproductive success. Spider gifts consist of silk-wrapped food, but may also consist of worthless (non-nutri- tive) donations that successfully lead to mating, despite yielding shorter copulations. Worthless gifts may either represent a cheaper cheating strategy or the inability to produce genuine gifts due to resource limitations (i.e. poor body condition). Unless energetic con- straints limit expenditure in silk, males should apply more silk to worthless gifts to compensate for their lower reproductive value. We ask whether in Pisaura mirabilis 1) worthless gifts are condition-dependent and 2) males strategically use silk based on gift type (genu- ine vs worthless). We tested whether male body condition explains the gift-giving strategy and compared silk amounts covering each gift type, in gifts collected from the field and produced in the laboratory by males given different feeding regimes. Our findings show that worthless gifts are not promoted by poor body condition or limited resources. They rather result from a cheating strategy evolved to opportunistically reduce the costs of genuine gifts while ensuring nutritional advantages, with cheaters gaining body mass. Males applied more silk to worthless gifts regardless of their body condition or feeding state, suggesting they can strategically adjust silk expenditure despite its costs. By masking gift contents and prolonging female feeding, silk is crucial for the maintenance of cheating, likely resulting from an evolutionary arms race between male deception and female assessment. Key words: deception, mating strategies, nuptial gifts, sperm competition.

INTRODUCTION (direct or indirect) from their mate choice, with males saving energy Theory predicts that sexual traits (e.g. courtship displays) are honest at their own reproductive advantage (Johnstone and Grafen 1993; indicators of males’ underlying quality due to the costs associated LeBas and Hockham 2005). Alongside cost-reduction in males, to their production and maintenance (Zahavi 1975). The extent to selection should favour accurate female mate assessment (Kokko which males invest in reproduction is therefore often affected by et al. 2003), leading to a potential evolutionary arms race between energetic resources, with males of higher body condition being bet- selection for cheaper deceptive traits in males and resistance to ter at expressing sexual traits (van Noordwijk and de Jong 1986). such traits in females (Arnqvist and Rowe 2013). As it is in the male’s best interest to optimize investment in current Nuptial gifts, consisting of male donations of nutritious substances reproduction in order to allocate resources to future mating or other to females (Lewis and South 2012), are sexually selected traits partic- activities, males should be under selection to reduce some of these ularly prone to male cheating (LeBas and Hockham 2005; Ghislandi costs without impairing their overall reproductive success. Males, et al. 2014). The gift may have several, non-mutually exclusive, func- for example, can use cheaper mating strategies if there are limits to tions: it is used by males to attract females, it secures sperm trans- female accurate assessment of the quality of the information they fer, and may act as parental investment through nutrient-derived convey (Dawkins and Guilford 1991). These strategies lead to male resources (Thornhill 1976a; Boggs 1995; Vahed 1998; Gwynne cheating behaviour if females do not receive the expected benefits 2008). Invertebrates show a rich variation of gift types, ranging from oral or seminal male secretions (endogenous gifts) to food items cap- tured or collected from the environment (exogenous gifts) (Lewis Address correspondence to C. Tuni, Department of Biology, Ludwig et al. 2014). Exogenous gifts are often polymorphic and may also Maximilian University, Grosshaderner Strasse 2, 82152 Planegg- consist of items with little or no nutritional value to females (i.e. plant Martinsried, Germany. E-mail: [email protected] parts, prey remains, and empty silk balloons), which nevertheless

© The Author 2017. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: [email protected] 94 Page 2 of 6 Behavioral Ecology successfully elicit female mate acceptance (Preston-Mafham 1999; condition-dependent and 2) whether males modulate silk investment Albo et al. 2011b). These types of donations generally reflect a reduc- based on their gift giving strategy (genuine vs worthless gifts). In a tion of male investment into reproduction from more costly to less first study, we collected gifts from a natural population of spiders valuable donations (from nutritious prey, to prey remains and worth- and scored them as genuine or worthless based on their contents, less donations) (LeBas and Hockham 2005; Ghislandi et al. 2014). consisting respectively of intact prey (genuine) or prey leftovers and Hence, cheaper (worthless) gifts may either represent an opportu- plant parts (worthless). We tested whether the use of worthless gifts nistic cheating strategy used by males to reduce the costs of mating is explained by variation in male body condition and compared the or a consequence of facing unfavourable ecological conditions, for amounts of silk covering each gift type (genuine and worthless). In example when resources are lacking (Engqvist 2007; Albo et al. 2014; a second study we obtained gifts during laboratory trials by pro- Ghislandi et al. 2014). Through cheaper donations, males enhance viding males with live prey, and scored whether a gift is worthless their reproductive success while minimizing the costs associated with based on the extent males fed on prey prior to silk wrapping (i.e. producing a genuine gift, such as increased risk of predation while prey weight loss). We experimentally manipulated male body condi- hunting for a prey to offer, missed foraging opportunity, or transpor- tion by allocating males to high and low feeding regimes and tested tation costs (Boggs 1995). On the other hand, males attempting to whether food availability explained worthless gift production, and use worthless gifts may incur fitness costs, derived by female mate whether silk investment varied based on gifts’ worthlessness. Unless rejection or reduced sperm transfer due to shorter copulation time the use of worthless gifts itself is condition-dependent and limits (LeBas and Hockham 2005; Albo et al. 2011b). silk investment we expect males to cover worthless gifts in greater The evolution of worthless gifts in spiders remains particu- silk amounts compared to genuine gifts to compensate for their larly intriguing. Male spiders offer a prey gift to the female dur- lower reproductive value. ing courtship and mating. Gift-offering enhances mate acquisition (Stålhandske 2001; Albo and Costa 2010) by exploiting female for- aging motivation (Bilde et al. 2007). It also increases male fertiliza- METHODS tion success by prolonging female gift-consumption and concurrent Field collected nuptial gifts sperm transfer (Drengsgaard and Toft 1999; Albo and Costa 2010; Collecting Albo et al. 2011b; Albo et al. 2013). In certain ecological contexts Males carrying gifts were collected from a natural population of the gift may also function as paternal investment (Toft and Albo Pisaura mirabilis surrounding Ludwig Maximilian University of 2015). Males wrap their gifts in dense silk layers, an advantageous Munich (Germany) during May and June 2014 and 2015. On each trait for males because it prolongs mating by increasing female day of collection, which occurred over a total period of 4 weeks feeding duration (Lang 1996; Albo et al. 2011b) and allows the at variable intervals, males were captured, placed in a vial (5 cm male to strengthen hold of the gift, preventing mating interruptions diameter, 10 cm height) covered with a sponge lid and given an (Andersen et al. 2008). Gifts may however range from edible intact individual ID. In order to avoid males’ gift manipulation after cap- prey, to worthless prey remains and plant parts (Stålhandske 2001; ture, the gift was immediately removed from the males’ mouthparts Albo et al. 2011b; Albo et al. 2014). Limited food availability and using forceps and placed in an Eppendorf tube labelled with the poor male body condition have been proposed to be the possible male ID. Males were brought to the laboratory, their body mass ecological drivers promoting the use of these worthless gifts (Albo was weighted to the nearest 0.01 g using a digital scale (KERN et al. 2014; Ghislandi et al. 2014). PKT, KERN & SOHN GmbH, Balingen, Germany) and their Disguising gift content with silk to lure females into accepting a cephalothorax width (used as a proxy for body size) measured to gift that does not confer the expected food reward is suggested to the nearest 0.01 mm using digital callipers (AEROSPACE, China). play a key role in facilitating the use of worthless gifts (Albo et al. Animals were then released back to the study site. Gifts were frozen 2011b; Ghislandi et al. 2014). In the species Pisaura mirabilis, males at −4 °C. offering worthless gifts are known to achieve similar mating success as of those offering genuine prey gifts, indicating that during court- Scoring of gift content ship females are not able to assess the value of the gift content (Albo Gifts were weighed to the nearest 0.001 g using a microbalance et al. 2011b). These males however experience on average 20% (Sartorius Supermicro S4, Sartorius AG, Göttingen, Germany.) shorter copulations and reduced fertilization success, most likely and then placed under a stereomicroscope where the silk was gen- due to female interruption of copulation once the silk layers cover- tly opened using micro-dissecting forceps. The content of field- ing inedible gift contents are consumed and the content unmasked collected gifts was inspected and the gift was defined as genuine (Albo et al. 2011b). Due to the importance of silk in serving the if containing intact prey, which did not appear to be dried out or male’s reproductive interest (Lang 1996; Andersen et al. 2008), eaten, and worthless if containing dried prey and prey leftovers (i.e. males are expected to use silk strategically. Enhancing silk invest- pieces of body parts or exoskeletons) or items such as seeds or plant ment when offering worthless gifts would be particularly advanta- parts. Gifts that contained more than one prey were classified as geous for males, not only to effectively mask worthless gift contents, genuine where at least one of the items consisted of intact prey. If but also to prolong female feeding and compensate for the otherwise possible the Taxonomical group of the arthropod prey was defined. shorter copulation achieved (Lang 1996; Andersen et al. 2008; Albo Gift contents were scored by the same observer (P.G.G). et al. 2011b). However, if gift-wrapping entails substantial energetic requirements due to costly silk proteins (Craig et al. 1999; Nentwig Experimentally produced gifts and Kuhn-Nentwig 2013), we would expect higher silk investment only in the case worthless gifts occur as a male cheating strategy. Animal rearing and treatment groups In contrast, if worthless gifts are condition-dependent limiting ener- Sub-adult spiders were collected during April 2014 from grass getic resources may also constraint silk production. meadows described above, and were brought to the laboratory. By taking advantage of the silk-wrapped nuptial gifts of male They were reared at room temperature (approx. 25 °C) under nat- Pisaura mirabilis we investigated 1) whether worthless gifts are ural photoperiod and were kept individually in plastic vials (5 cm

95 Ghislandi et al. • Males cover worthless gifts in more silk Page 3 of 6 diameter, 10 cm height) covered with a sponge lid and with a bot- nearest 0.001 g using a Sartorius microbalance. All measurements tom layer of moss that was kept wet by spraying water every sec- were taken by the same observer (M.B.). ond day. Spiders were fed regularly twice a week with 5–10 fruit flies (Drosophila spp.) or one prey consisting of a field cricket nymph Statistical analyses (Gryllus spp.) or a housefly (Musca domestica). The occurrence of We estimated male body condition by calculating a body condition molted exoskeletons in the vials was noted every second day. Upon index measured as the residuals of a regression of body mass on emergence to adulthood males were randomly assigned to one of body size (cephalothorax width) (Jakob et al. 1996). the two feeding treatments, with males in the high food treatment (HF, n = 28) being fed an insect prey (fly or cricket) twice a week Field collected nuptial gifts and those in the low food treatment (LF, n =17) being fed once a We used Logistic regression to test whether male body condition week with the same prey type. has an effect on the gift-giving strategy used (genuine vs worthless gifts). To understand whether silk investment depends on the gift Silk-wrapping trials giving strategy used by males and is affected by body resources we Males were used in experimental trials 12–14 days after reaching used ANCOVA, testing for the effects of male body condition and adulthood in order to ensure sexual maturity. Prior to each trial, gift type (genuine or worthless) on silk amounts covering gifts. male body mass and size was measured (as above). In order to elicit gift-wrapping behaviour we exposed males to female silk-draglines: Experimentally produced gifts we placed males into housing vials of sexually receptive females We tested whether the applied food treatments (LF and HF) gen- that were temporarily removed, or into transparent plastic boxes erated differences in male body condition index by using a t-test. (7 cm height × 10 cm width × 10 cm depth) with a bottom layer To analyse whether feeding treatment (LF and HF) affected silk of paper in which a female was previously housed for a minimum wrapping latency we used ANCOVA, including prey body mass as of 30 min. As soon as males showed signs of sexual stimulation a covariate in the model to control for possible effects of prey size on (i.e. abdomen jerking and rubbing of the first pair of legs against male behaviour. We estimated weather a gift is worthless in terms of each other) we placed a live cricket nymph of known weight nearby the extent to which males consumed the prey prior to silk wrapping. the spider using forceps and waited for the male to catch the prey. We therefore calculated prey weight loss as the differences in weight If males did not catch the prey within 20 min the trial was inter- between prey body mass before the trial and the content of the nup- rupted, males were returned to their own vial and the trial was tial gift after the trial (calculated as gift weight—silk weight). We repeated on the following day, for a maximum of 3 times. Similarly, used ANCOVA to test whether prey weight loss is affected by the trials were interrupted if males caught a prey but did not wrap it food treatment (LF and HF), controlling for initial prey body mass. into a nuptial gift. Overall we obtained one gift from each male, 10 To ascertain whether regression-to-mean effects explain prey weight from the LF and 24 from the HF treatment. loss in respect to prior prey weight value we used Pitman’s test for For each male that produced a gift we measured latency to the equality of variances in paired samples (Kelly and Price 2005). gift-wrapping, as the time spent from prey capture to start of silk To understand whether silk investment depends on the gift-giving wrapping, to account for possible manipulations of the prey prior strategy (worthlessness of the gift) and is affected by body resources gift construction. During gift-wrapping males move their body we used ANCOVA, testing for the effect of feeding treatment (LF in rounds keeping their silk spinning organs (spinnerets) that are and HF) and prey weight loss on silk amounts covering gifts. Linear located at the underside posterior of their abdomen attached to regression was used to assess whether the total duration of silk wrap- the gift. Males may often interrupt this process, remaining motion- ping behaviour related to the amount of silk deposited by males. We less and/or picking up the gift it their chelicerae, producing distinct also estimated the costs of gift production for males by measuring wrapping episodes. We measured the total gift-wrapping duration male body mass loss and used Linear Regression to assess whether (seconds) as the sum of the duration of all wrapping episodes to such cost is related to cheating (prey weight loss). Finally, for gifts correlate behavioural observations of silk wrapping to quantified produced overnight without direct observations we tested differences silk measures. All time durations were measured using a CG-501 in mean silk amount between LF and HF males using t-test. stopwatch (Genutek Electronics Co., Ltd, China). The trial was Statistical tests were performed using the software JMP, Version ended when males did not wrap for 10 consecutive minutes fol- 11 (SAS Institute Inc., Cary, NC). Appropriate data transforma- lowing the end of the last wrapping episode. The gift was then tions were applied when residuals did not meet assumptions of nor- removed from the male using forceps and frozen at −4 °C. mal distribution. Means are reported ± SE. At the end of the experiments, in order to evaluate male silk pro- duction without time constraints imposed by our experimental tri- als, a cricket nymph was placed inside the male’s housing vial and RESULTS on the following day we collected silk-wrapped gifts constructed overnight and stored them at −4 °C. Field collected nuptial gifts We collected a total of 113 nuptial gifts from the field (21 in 2014 Silk measurements and 92 in 2015), of which 41.6% (47) contained intact prey and Silk of nuptial gifts was measured from field-collected and experi- were scored as genuine gifts, whereas 58.4% (66) contained dried mentally produced samples. Silk was gently separated from the gift and fragmented prey leftovers and plant parts and were scored as content using micro-dissecting forceps and placed in an Eppendorf worthless gifts. Male body condition, calculated in 110 males, did tube. Silk measurements were conducted by modifying an existing not affect the mating strategy adopted (genuine vs worthless gifts) protocol for this species (Lang 1996). The removed silk was heated (Logistic regression, χ2 = 0.12, n = 110, P = 0.72). for 1–2 h in a heating block at 100 °C and then stored in a desic- Gift contents ranged from 1 to 4 items, with most gifts contain- cator (in dry atmosphere) for 12–36 h until it was weighed to the ing 1 (84%) and 2 (11.5%) items. In 79.7% of the cases, we found

96 Page 4 of 6 Behavioral Ecology arthropods (26.3% Dipteran, 21.8% Spiders, 19.1% Coleopteran, with the total amount of silk measured (log-transformed, Linear

16.4% Lepidoptera and 16.4% others), in 4.4% of plant parts regression, R2 = 0.42, F1, 27 = 21.10, P = <0.0001). and 15.9% of not identifiable items, the latter being scored as one We estimated the costs of gift construction for males in terms of item. The number of items found in gifts did not differ significantly body mass loss (male weight prior—after gift construction). After between those scored as genuine and worthless (mean N items ± gift construction 53.33% (n = 16) of the males increased their SE, genuine 1.19 ± 0.07 and worthless 1.25 ± 0.09; Wilcoxon, weight and 46.66% (n = 14) showed a decrease in body mass. We Z = 0.74; n = 113, P = 0.45). Gift weight was measured in 111 sam- found that male body mass increase correlated positively with prey ples and we found that genuine gifts were significantly heavier than weight loss (Linear regression, R2 = 0.27, F1,21 = 9.11, P = 0.0065). worthless gifts (mean ± SE (mg), genuine 13.93 ± 1.59 (n = 64) and Of the total of 45 males, 40 constructed gifts in their housing worthless 5.76 ± 0.91 (n = 47); log-transformed, t-test, t = −5.28, vials overnight, without direct observations. Gifts of HF males were df = 109, P = <0.0001). covered in higher amounts of silk compared to LF males (mean ± Due to practical unfeasibility in isolating silk in 3 genuine gifts, silk SE (mg), HF 0.38 ± 0.05 (n = 23), LF 0.17 ± 0.03 (n = 17), t-test, weight was measured from 110 samples. We found that the amount DF = 38, t = 2.68, P = 0.016). We acknowledge that our experi- of silk covering gifts varied significantly according to gift type, with mental set up may explain the differences in average silk amounts higher amounts of silk being found on worthless compared to measured from gifts produced during experimental observations genuine gifts (log-transformed, ANCOVA, Gift type F1, 104 = 4.39, (0.10 mg) and those produced without direct observations, in the P = 0.038; Body condition F1, 104 = 1.68, P = 0.19; Figure 1). laboratory (0.29 mg) and in the field (0.39 mg). Experimentally produced nuptial gifts The food treatment applied had a significant effect on male body DISCUSSION condition (mean body condition index ± SE, LF −7.49 ± 3.60 Energetic limitations have been suggested to play a major role in (n = 10), HF 3.98 ± 2.66 (n = 24); t-test; t = 2.41, DF = 32, the production of nuptial gifts (Engels and Sauer 2006; Immonen P = 0.022). Neither feeding treatment nor prey body mass (mea- et al. 2009; Albo et al. 2011a). However, unlike suggested for male sured in 27 trials) affected male latency to silk wrapping (log trans- spiders (Albo et al. 2014; Ghislandi et al. 2014), our study shows that formed, ANCOVA, Food Treatment F1, 24 = 0.06, P = 0.79; Prey limitation in male body condition or food resources does not pro- weight F1, 24 = 0.01, P = 0.93). We measured prey weight loss in 23 mote the use of worthless gifts, consisting of less valuable food dona- gifts. Whereas prey generally lost weight, in 3 cases we measured tions to females. In the field, males varying in their body condition an increase of prey body mass. This is most likely due to spiders’ were equally likely to carry genuine and worthless gifts. Similarly, injection of digestive fluids that serve for liquefying the prey during males raised under high and low food regimes in the laboratory fed feeding. We found that the feeding treatment given to males did on prey prior to gift construction to an equal extent. We can there- not significantly affect prey weight loss (worthlessness of the gift) fore overall reject the hypothesis of worthless gifts deriving from the but heavier prey lost more weight (ANCOVA, log-transformed, inability of poor quality males to produce genuine gifts. Worthless

Food Treatment F1, 20 =1.25, P = 0.27, Prey weight F1, 20 = 35.23, food gifts rather appear to result from a cheating strategy, evolved P < 0.0001) and the correlation between change in prey mass and to opportunistically reduce the costs of mating while ensuring nutri- initial prey mass was not entirely driven by regression-to-mean tional advantages. We indeed show that males producing worthless effects (Pitman’s test, t = 18.98, DF = 21, P > 0.0001). gifts increase their body mass from partially feeding on prey prior to We were not able to isolate silk from prey in 4 gifts, therefore silk gift construction. Short-term access to nutrients may promote male weight was measured from 30 samples. We found higher amounts mating success through strengthened courtship effort and vigour, of silk on the gifts that were scored as worthless (higher prey weight providing males with mating advantages that go beyond gift quality loss), but not according to male feeding treatment (LF and HF) (Kotiaho 2002; Hunt et al. 2004; Shamble et al. 2009; Albo et al.

(square-root transformed, ANCOVA, Food Treatment F1, 20 = 2.54, 2012). Using a cheaper strategy would then be particularly valu- P = 0.12, Prey weight loss F1, 20 = 4.41, P = 0.048; Figure 2). The able with increasing energy requirements necessary to overcome total time males spent wrapping a gift in silk correlated significantly resistance of already mated females (Gabor and Halliday 1997;

0.25 0.6 0.5 0.2 0.4 0.15 0.3

Silk (mg) 0.1 Silk (mg) 0.2 0.05 0.1 0 0 Genuine Worthless -5 515253545 Gift Type Prey weight loss (mg) Figure 1 Figure 2 Amount of silk (mg) covering genuine (prey) and worthless (prey leftovers The amounts of silk (mg) covering experimentally produced nuptial gifts and plants) nuptial gifts collected from a natural population of the spider correlates linearly with the extent on which males fed on the prey prior to Pisaura mirabilis. *Denotes statistical significance. silk wrapping (prey weight loss).

97 Ghislandi et al. • Males cover worthless gifts in more silk Page 5 of 6

Maklakov et al. 2003; Aisenberg and Costa 2005; Tuni and Bilde Once the gift quality is reduced, females are likely to suffer fitness 2010), or to outcompete competitors (Hack 1997; Nitzsche 2011). costs from being deceived owing to reduced direct benefits (i.e. The finding that males producing worthless gifts fed to a greater missed foraging opportunity), and are expected to be under selec- extent on prey of higher body mass, further suggests that cheat- tion to distinguish genuine from worthless gifts. In return, males ing may function to efficiently feed on larger and/or higher quality might evolve more efficient traits, such as silk wrapping to prevent prey providing males of greater hunting abilities with the oppor- accurate female assessment. Females may instead become more tunity to gain nutritional benefits, while still being able to produce resistant to the information carried by silk (Albo et al. 2012), poten- a gift that allows successful reproduction. This may also represent tially basing their reproductive decisions on other traits. Overall, an effective male strategy to optimize the decreasing reproductive this would generate an antagonistic arms race between males under payoff of prolonged copulations in relation to the costs of not feed- selection to deceive and females under selection to evolve resistance ing on the prey. Males would indeed benefit from offering larger to deception (Arnqvist and Rowe 2013). and more nutritive prey to females only if the extra copulation time Our study shows that the use of worthless gifts is well estab- achieved provides substantial fertilization advantages. If paternity lished in the population, occurring in 58.4% of the males in gain is instead a decreasing payoff function of copulation duration our field study (and 85% in laboratory trials). Similar frequen- (Parker 1970; Thornhill 1976b) feeding on larger prey and using cies are also described in natural populations of the Neotropical prey of reduced size for mating, may allow males to outweigh the gift-giving spider ornata (Albo et al. 2014). These costs of a missed meal while ensuring sufficient sperm transfer for high frequencies of worthless gifts are not explained by theory, reproduction. which instead predicts cheating to occur at low incidence as the We also found that males producing worthless gifts apply more benefits of the strategy will decline with increasing frequency of silk regardless of their body condition or feeding state, suggest- cheats in the population (Dawkins and Guilford 1991; Johnstone ing that they are not energetically limited in their ability to invest and Grafen 1993). Our following reasoning may provide a pos- in the gift and can strategically adjust silk expenditure despite sible explanation for such high levels of worthless gifts. The its costs (Albo et al. 2011a). The silk cover is instrumental for non-trivial nutritional benefits gained by cheaters together with a successful cheating strategy as it prevents female pre-mating the effectiveness of the strategy derived by silk-wrapping (worth- evaluation of gift contents; greater silk amounts covering worth- less gifts are not rejected at higher rates), may allow males using less gifts will effectively disguise distasteful or inedible items to worthless gifts to maintain overall high mating rates (LeBas and females. Additional silk wrapping may also be a means of extend- Hockham 2005; Albo et al. 2011b) at the costs of reduced pater- ing inseminations in the face of the shorter matings yielded from nity gained per mating (Albo et al. 2011b). Despite the fact that worthless gifts, as extra silk layers will prolong female feeding larger inseminations promote male success in sperm competition duration and hence sperm transfer (Lang 1996). Silk itself is likely when females mate with multiple partners (Drengsgaard and Toft an important component of the nuptial gift as females ingest it, 1999; Simmons 2001), shorter copulations may not necessarily hence cheaters may provide extra silk to improve the nutritive affect male lifetime reproductive success negatively. Theoretical value of worthless items (Craig et al. 1999; Wilder 2011). Finally, models of sperm allocation predict that males should increase silk may also facilitate handling and control of smaller lower qual- sperm investment when facing sperm competition risk, but should ity gifts (Andersen et al. 2008). Because of the lack of consen- reduce sperm investment under high sperm competition intensity sus on silk acting as a visual signal enhancing mate acceptance (Parker 1970; Parker and Pizzari 2010). Therefore under environ- (Stålhandske 2002; Bilde et al. 2007; Albo et al. 2012), greater silk mental conditions in which sperm competition is most intense (i.e. layers are suggested not to promote male mating success. For all male biased sex ratios, high female re-mating rates) it may pay to the above-mentioned reasons, we suggest that males use higher reduce investment in gifts and hence sperm, as larger insemina- silk investment to compensate for the lower reproductive value tions achieved through high quality gifts may not confer the fertil- of worthless gifts. The finding that males that feed opportunisti- ization benefits necessary to counterbalance the costs of mating. cally on the prey to produce a worthless gift gain weight despite Under these circumstances males could enhance their reproduc- wrapping their gifts in more silk, further suggests that adding silk tive success by ensuring mate acquisition through cheaper mat- is not a costly compensation energy wise. We expect such com- ings (worthless gifts, shorter inseminations) rather than investing pensation in gift quality however not to be complete, as achieving largely in fewer costly encounters (i.e. genuine gifts, longer insemi- copulations equivalent in frequency and duration to those with nations), with selection favouring cheating. genuine gifts would indicate that the worthless gifts may in turn To conclude, genuine and worthless gifts may be maintained as have the potential to invade and replace genuine gifts (LeBas and alternative reproductive strategies, each of which maximizes male Hockham 2005). reproductive fitness under a specific set of circumstances (Oliveira Nuptial gifts most likely evolved in the context of female forag- et al. 2008). Whether these reflect changes in the competitive social ing, with males providing extra nutrients to their mating partners environment (e.g. sperm competition intensity) or the quality of while enhancing their reproductive success through improved fertil- available prey remain interesting avenues for research. Silk wrap- izations and increased female fecundity (Boggs 1995; Vahed 1998; ping, which increases assessment costs for females and compensates Bilde et al. 2007). Male spiders may have originally benefited from for the lower reproductive value of worthless gifts, facilitates the investing in costly nutritive gifts at their own feeding expenses (Albo maintenance of the cheating strategy. et al. 2009), with gift-giving becoming a trait strongly selected by females (Stålhandske 2001; Albo et al. 2013). Only subsequently, males may have evolved means to reduce the nutritional costs of ACKNOWLEDGEMENTS gift production by lowering content quality in order to facilitate We thank Jingzi Xu for assistance in maintaining spiders during the labora- their own reproductive interests, yet still fulfilling female mate tory trials and Sarah Schulte-Döinghaus for help in collecting spiders from choice expectations (Sakaluk 2000; LeBas and Hockham 2005). the field. Trine Bilde and anonymous reviewers gave valuable comments

98 Page 6 of 6 Behavioral Ecology to the manuscript. C.T. was funded by the “LMUMentoring” program. Jakob EM, Marshall SD, Uetz GW. 1996. Estimating fitness: a comparison P.G.G. and P.V. conducted the field study, M.B. conducted the experimental of body condition indices. Oikos:61–67. trials, P.V. and C.T. performed statistical analysis; C.T. conceived the study Johnstone RA, Grafen A. 1993. Dishonesty and the handicap principle. and wrote the manuscript. Anim Behav. 46:759–764. Kelly C, Price TD. 2005. Correcting for regression to the mean in behavior Data accessibility: Analyses reported in this article can be reproduced using and ecology. Am Nat. 166:700–707. the data provided by Tuni (2017). Kokko H, Brooks R, Jennions MD, Morley J. 2003. The evolution of mate choice and mating biases. Proc R Soc London B Biol Sci. Handling editor: Francisco Garcia-Gonzalez 270:653–664. Kotiaho JS. 2002. Sexual selection and condition dependence of court- ship display in three species of horned dung beetles. Behav Ecol. REFERENCES 13:791–799. Lang A. 1996. Silk investment. Behaviour 133:697–716. Aisenberg A, Costa FG. 2005. Females mated without sperm transfer main- Lebas NR, Hockham LR. 2005. An invasion of cheats; the evolution of tain high sexual receptivity in the wolf spider Schizocosa malitiosa. Ethology. worthless nuptial gifts. Curr Biol. 15:64–67. 111:545–558. Lewis S, South A. 2012. The evolution of animal nuptial gifts. Adv Study Albo M, Toft S, Bilde T. 2011a. Condition dependence of male nup- Behav 44:53–97. tial gift construction in the spider Pisaura mirabilis (Pisauridae). J Ethol. Lewis SM, Vahed K, Koene JM, Engqvist L, Bussière LF, Perry JC, 29:473–479. Gwynne D, Lehmann GUC. 2014. Emerging issues in the evolution of Albo MJ, Bilde T, Uhl G. 2013. Sperm storage mediated by cryptic female animal nuptial gifts. Biol Lett. 10:20140336. choice for nuptial gifts. Proc R Soc London B Biol Sci. 280:20131735. Maklakov AA, Bilde T, Lubin Y. 2003. Vibratory courtship in a web-build- Albo MJ, Costa-Schmidt LE, Costa FG. 2009. To feed or to wrap? Female ing spider: signalling quality or stimulating the female? Anim Behav. silk cues elicit male nuptial gift construction in a semiaquatic trechaleid 66:623–630. spider. J Zool. 277:284–290. Nentwig W, Kuhn-Nentwig L. 2013. Spider Ecophysiology. Berlin/ Albo MJ, Costa FG. 2010. Nuptial gift giving behaviour and male mating Heidelberg (Germany): Springer Science & Business Media. effort in the Neotropical spiderParatrechalea ornata (). Anim Nitzsche ROM. 2011. Courtship, mating and agonistic behaviour in Pisaura Behav. 79:1031–1036. mirabilis (Clerck, 1757). Bull Br Arachnol Soc. 15:93–120. Albo MJ, Melo-González V, Carballo M, Baldenegro F, Trillo MC, Costa van Noordwijk AJ, de Jong G. 1986. Acquisition and allocation of resources: FG. 2014. Evolution of worthless gifts is favoured by male condition and their influence on variation in life history tactics. Am Nat. 128:137–142. prey access in spiders. Anim Behav. 92:25–31. Oliveira RF, Taborsky M, Brockmann HJ. 2008. Alternative reproductive Albo MJ, Toft S, Bilde T. 2012. Female spiders ignore condition-dependent tactics: an integrative approach. Cambridge (UK): Cambridge University information from nuptial gift wrapping when choosing mates. Anim Press. Behav. 84:907–912. Parker GA. 1970. Sperm competition and its evolutionary effect on copula Albo MJ, Winther G, Tuni C, Toft S, Bilde T. 2011b. Worthless donations: duration in the fly Scatophaga stercoraria. J Insect Physiol. 16:1301–1328. male deception and female counter play in a nuptial gift-giving spider. Parker GA. 1970. Sperm competition and its evolutionary consequences BMC Evol Biol. 11:329. in the insects. Biol Rev. 45:525–567. [cited 7 Jul 2015]. http://doi.wiley. Andersen T, Bollerup K, Toft S, Bilde T. 2008. Why do males of the spider com/10.1111/j.1469-185X.1970.tb01176.x. Pisaura mirabilis wrap their nuptial gifts in silk: Female preference or male Parker GA, Pizzari T. 2010. Sperm competition and ejaculate economics. control? Ethology. 114:775–781. Biol Rev. 85:897–934. Arnqvist G, Rowe L. 2013. . Princeton (NJ): Princeton Preston-Mafham KG. 1999. Courtship and mating in Empis (Xanthempis) University Press. trigramma Meig., E. tessellata F. and E. (Polyblepharis) opaca F. (Diptera: Bilde T, Tuni C, Elsayed R, Pekar S, Toft S. 2007. Nuptial gifts of male Empididae) and the possible implications of “cheating” behaviour. J spiders: sensory exploitation of the female’s maternal care instinct or for- Zool. 247:239–246. aging motivation? Anim Behav. 73:267–273. Sakaluk SK. 2000. Sensory exploitation as an evolutionary origin to nuptial Boggs CL. 1995. Male nuptial gifts: phenotypic consequences and evolu- food gifts in insects. Proc R Soc London B Biol Sci. 267:339–343. tionary implications. In: Insect Reproduction (Ed. by S. R. Leather & J. Shamble PS, Wilgers DJ, Swoboda KA, Hebets EA. 2009. Courtship effort Hardie), pp. 215–242. Boca Raton (FL): CRC Press. is a better predictor of mating success than ornamentation for male wolf Craig CL, Hsu M, Kaplan D, Pierce NE. 1999. A comparison of the spiders. Behav Ecol. 20:1242–1251. composition of silk proteins produced by spiders and insects. Int J Biol Simmons LW. 2001. Sperm competition and its evolutionary consequences Macromol. 24:109–118. in the insects. Princeton (NJ): Princeton University Press. Dawkins MS, Guilford T. 1991. The corruption of honest signalling. Anim Stålhandske P. 2001. Nuptial gift in the spider Pisaura mirabilis maintained by Behav. 41:865–873. sexual selection. Behav Ecol. 12:691–697. Drengsgaard I, Toft S. 1999. Sperm competition in a nuptial feeding spider, Stålhandske S. 2002. Nuptial gifts of male spiders function as sensory traps. Pisaura mirabilis. Behaviour. 136:877–897. Proc R Soc London B Biol Sci. 269:905–908. Engels S, Sauer KP. 2006. Resource-dependent nuptial feeding in Panorpa Thornhill R. 1976a. Sexual selection and parental investment in insects. vulgaris: an honest signal for male quality. Behav Ecol. 17:628–632. Am Nat. 110:153–163. Engqvist L. 2007. Sex, food and conflicts: nutrition dependent nuptial Thornhill R. 1976b. Sexual selection and nuptial feeding behaviour in feeding and pre-mating struggles in scorpionflies. Behav Ecol Sociobiol. Bittacus apicalis (Insecta:). Am Nat. 110:529–548. 61:703–710. Toft S, Albo MJ. 2015. Optimal numbers of matings: the conditional bal- Gabor CR, Halliday TR. 1997. Sequential mate choice by multiply mat- ance between benefits and costs of mating for females of a nuptial gift- ing smooth newts: females become more choosy. Behav Ecol. 8:162–166. giving spider. J Evol Biol. 28:457–467. Ghislandi PG, Albo MJ, Tuni C, Bilde T. 2014. Evolution of deceit by Tuni C, Bilde T. 2010. No preference for novel mating partners in the poly- worthless donations in a nuptial gift-giving spider. Curr Zool. 60:43–51. androus nuptial-feeding spider Pisaura mirabilis (Araneae: Pisauridae). Gwynne DT. 2008. Sexual conflict over nuptial gifts in insects. Annu Rev Anim Behav. 80:435–442. Entomol. 53:83–101. Vahed K. 1998. The function of nuptial feeding in insects: a review of Hack MA. 1997. The energetic costs of fighting in the house cricket, Acheta empirical studies. Biol Rev. 73:43–78. domesticus L. Behav Ecol. 8:28–36. Tuni C. 2017. Data from: Silk wrapping of nuptial gifts aids cheating Hunt J, Brooks R, Jennions MD, Smith MJ, Bentsen CL, Bussière LF. 2004. behaviour in male spiders. Dryad Digital Repository. http://dx.doi. High-quality male field crickets invest heavily in sexual display but die org/10.5061/dryad.kb1fh. young. Nature. 432:1024–1027. Wilder SM. 2011. Spider nutrition: an integrative perspective. Adv In Insect Immonen E, Hoikkala A, Kazem AJN, Ritchie MG. 2009. When are vom- Phys. 40:87. iting males attractive? Sexual selection on condition-dependent nuptial Zahavi A. 1975. Mate selection-a selection for a handicap. J Theor Biol. feeding in Drosophila subobscura. Behav Ecol. 20:289–295. 53:205–214.

99

Chapter IV

Extreme male mating behaviours: anecdotes in a nuptial gift-giving spider

Ghislandi P.G., Bilde T. & C. Tuni

Arachnology (2015) 16 (8), 273–275

100 Arachnology (2015) 16 (8), 273–275 273 Extreme male mating behaviours: anecdotes in a increase sperm transfer (Vahed 1998). Gifts may consist of nuptial gift-giving spider captured prey, seminal substances transferred during copu- lation, or even somatic gifts, consisting of specialized body structures on which females feed. Male ground crickets, for Paolo Ghislandi example, possess spurs on their hind legs that females chew Trine Bilde Department of Bioscience, on during mating (Mays 1971), whereas male dwarf spiders Aarhus University, possess a dorsal cephalic elevation from which secretory Ny Munkegade 1540, glands produce a substance taken up by females during DK-8000 Aarhus C, Denmark mating (Kuntz, Garbe & Uhl 2012). Although females in email: [email protected] some species may benefit from nuptial gifts, there is much scope for sexual conflict, as males may manipulate females Cristina Tuni Ludwig Maximilians University Munich, into re-mating rates below their optima (Arnqvist & Rowe Department Biology II, 2005). Großhaderner Str. 2, Male nuptial gifts in the form of captured prey are rela- 82152 Planegg-Martinsried, Germany tively rare in spiders (Albo, Toft & Bilde 2013), and occur in the Palaeartic Pisaura mirabilis Clerck, 1757 (Araneae: Pisauridae). Males of this species offer a silk-wrapped Summary insect prey during courtship. Upon female gift consump- Nuptial gifts may serve to increase male mating success, tion, the male enters the mating position and inserts his copulation duration, and fertilization success, as is known for pedipalp into the female’s . The gift, which exploits the nuptial feeding spider Pisaura mirabilis Clerck, 1757. In the female foraging motivation in a sexual context (Bilde et this species, strong sexual selection for the gift-giving trait may al. 2007) is under strong sexual selection, as gift-less males lead to male strategies, such as gift enlargement or thanatosis behaviour (death feigning), which ultimately maximize male suffer from decreased mating success (Stålhandske 2001; fitness. In laboratory trials, we observed male gift enlargement Albo et al. 2011). The gift functions as a male mating effort by inclusion of an autosomal (self-amputated) limb, female by increasing male mating success, assuring sperm transfer, consumption of male soma during copulation, and high-injury and prolonging copulation duration, which correlates posi- risk thanatosis in which a male feigned death while still in copula and only attached to the female with his pedipalp, instead tively with male fertilization success (Stålhandske 2001; of hanging onto the gift with the chelicerae as is performed in a Albo et al. 2011). Control over the gift during mating is of typical thanatosis. Although the observations are anecdotal, we crucial importance as females often attempt to steal the gift propose functional hypotheses for these traits in the context of and then avoid or interrupt copulation. Silk wrapping facili- extreme male mating effort and cannibalism avoidance, which are characteristics of the mating system of this species. Sexual tates male gift control during courtship and copulation, and selection and sexual conflict are major evolutionary forces that reduces the risk of losing the gift to the females (Andersen shape a variety of morphological, behavioural, and physiological et al. 2008). Males may also overcome female resistance by traits (Parker 1979). Males must compete intra-sexually over performing thanatosis (death feigning) (Bilde et al. 2006; access to females or inter-sexually to attract females, and have evolved traits that serve to increase their reproductive success Hansen et al. 2008), a stereotyped behaviour in response (Andersson 1994). to female attempts to escape with the gift. When females cease eating and try to take the prey, males stretch-out their legs and hang motionless onto the gift with their chelicerae Introduction (mouthparts), while being dragged by the female. Once females resume gift consumption, males insert their pedi- In mating systems in which access to females is limited palp and resume sperm transfer (Bilde et al. 2006; Hansen and competition among males intense, males may adopt et al. 2008). These behaviours strongly suggest conflicting extreme mating strategies, in which female monopolization interests over mating rate and mating duration between the and paternity assurance is often achieved at the expense of sexes (Arnqvist & Rowe 2005). This scenario is further the male’s own survival (Schneider & Fromhage 2010). supported by the male ability to deceive females into mating Self-sacrifice, in which males feed themselves to females using worthless or inedible silk-wrapped items as gifts is one of the most severe forms of extreme male mating (Albo et al. 2011). behaviour. Male Australian red-back spiders, for example, perform a 180° somersault into the female’s mouthparts during mating, a behaviour that facilitates cannibalism, and Material and methods is adaptive for males as it increases their paternity (Andrade 1996). Breakage of male copulatory organs during mating, Here, we describe the methods adopted to raise the spiders resulting in male death or functional sterility, is also common and the observations recorded during experimental trials. in various spider species (Schneider & Fromhage 2010). We collected Pisaura mirabilis spiders at juvenile stages in Similarly, in bees and certain ants, males die by breaking April 2013 from Mols, near Aarhus (Denmark), and from off their abdomen during mating (Boomsma, Baer & Heinze Merlino, near Lodi (Italy). Specimens were housed individ- 2005). ually in 30 ml vials containing moist Sphagnum moss in the To out-compete rivals, males may adopt other mating laboratory at Aarhus University, Denmark. Individuals were strategies, such as providing nuptial gifts, which are fed three times per week with green-bottle flies Lucilia ( nutritional donations offered to females during courtship ssp.), house flies Musca( domestica) and crickets (Gryllus or mating that attract partners, facilitate copulation, and ssp.). Water was provided daily. Spiders were raised at room

101 274 Extreme male mating behaviours temperature (on average 20°C) and natural photoperiod. through the experimental arena. Notably, these males did Once individuals reached sexual maturity (approximately not disengage their pedipalp from the female’s epigyne and, 14 days after their final moult) they were assigned to mating once the female started moving, they remained suspended experiments. Mating trials took place in transparent plastic from the female’s body with the pedipalp still inserted. One terraria (19 × 19 × 15 cm) in which females were allowed of the males did not fix his chelicerae to the gift, which is to walk and leave draglines for approximately 1 hour before how males usually remain attached during thanatosis (Bilde the start of the trial in order to stimulate a sexual response et al. 2006), and was dragged solely attached by his palp. In in males. Prior to the mating experiments, males were both cases, once females stopped moving and restarted gift given a fly (Lucilia sp.) to wrap with silk as a nuptial gift, consumption, copulation was resumed and sperm transfer and were subsequently introduced to the female. During a was clearly occurring, as the haematodocha of the inserted total number of 250 experimental trials we observed the palp was pulsating. following behaviours. However, it was not possible to determine whether during these unusual thanatosis males were transferring sperm while dragged in the mating arena or if they interrupted Results sperm transfer until the females stopped moving.

Gift enlargement via male limb autotomy Discussion An adult male was observed performing autotomy (self-amputation) of one of its legs (3rd pair) by literally Limb or caudal self-amputation (autotomy) is known pulling its leg off with its chelicerae, and fixing it with as a defensive strategy to counteract predation or nega- multiple silk threads to the nuptial gift it had previously tive consequences of appendage damage in several taxa, constructed by wrapping a fly in silk. This behaviour including arthropods (Maginnis 2006; Fleming, Muller & occurred when the male was alone in its vial prior to contact Bateman 2007). In spiders, autotomy is recorded across with a female. several families and is known to have very marginal costs, Since it was not possible to determine whether the leg had by not reducing locomotor or prey-capture ability (Brueseke previous damage or was completely functional, different et al. 2001), competitive fights over prey, or development costs could have been involved in the described autotomy. (Johnson & Jakob 1999), and male’s mating success (Brau- tigam & Persons 2003). These relatively negligible costs, suggesting that one or two limbs can be sacrificed with very Soma-consumption by females during copulation little negative effect, have led to the spare limb hypothesis for this taxonomic group (Fleming, Muller & Bateman During an experimental trial in which we paired a 2007). Although autotomy has been observed in spiders non-virgin female with a gift-less virgin male, we observed in a sexual scenario, as described for Tidarren cuneolatum the female feeding on the male’s leg during copulation. As males which amputate one of their palps a few hours after previously reported for gift-less males (Albo et al. 2011), the penultimate moult and insert a single palp during copu- the male courted the female by repeatedly vibrating his lation (Knoflach & Van Harten 2000), it has never been abdomen, bouncing towards and around the female, and documented in the context of nuptial gift construction. finally pushing up the female to reach her ventral epigyne. In addition to the insect prey, Pisaura mirabilis males After four minutes of courtship the male inserted his right are known to include several elements in the nuptial gift palp in the female’s epigyne. At this stage, the female bit via silk wrapping, e.g. nutritionally worthless items such the male’s leg, and started feeding on a droplet of emerging as sucked-out insect carcasses, plant parts or seeds, or a hemolymph that was clearly visible on the female’s cheli- combination of these (Albo et al. 2011). This creates the cerae. The male did not escape but remained motionless and potential to enlarge the gift prior to encounters with females. continued copulation for 10 minutes; during this time the Since gift consumption predicts copulation duration, larger female fed on the male’s limb. gifts lead to longer copulations and consequently larger transfer of ejaculate, representing a remarkable advantage in sperm competition (Lang 1996; Stålhandske 2001; Albo Pedipalp insertion during thanatosis et al. 2011). Males should, therefore, be under strong selec- tion pressures to donate large gifts. Due to the low costs of In two mating trials we observed gift-giving males in leg loss, inclusion of an autotomized leg to the nuptial gift copula with virgin females performing thanatosis without should increase gift size and, ultimately, lead to fertilization removing their pedipalp from the female’s epigyne. In the advantages at low expense for the male. This prediction, first trial, the female attempted to avoid copulation and however, relates to a hypothesis that remains to be tested. escape with the gift once the male assumed the mating posi- What triggered leg autotomy in our laboratory trial (e.g. tion and inserted his pedipalp. In the second trial, the female disease, chemical stimulation from cues of conspecific, attempted to take the gift and interrupt copulation after inadequate size perception of the nuptial gift) also remains 20 minutes of male pedipalp insertion. In both cases, the unknown and should be subject to further investigations. female’s movements triggered thanatosis: males stretched Low costs of appendages loss may also explain female out their legs and hung motionless, anchored on the gift limb consumption during copulation. During our observa- with their chelicerae, and were transported by the females tion, leg consumption seemed to replace the function of gift

102 P. Ghislandi, T. Bilde & C. Tuni 275 consumption by allowing the male to copulate while the References female was motionless and occupied by feeding. Gift-less males are known to suffer from reduced mating success ALBO, M. J., TOFT, S. & BILDE, T. 2013: Sexual selection, ecology and evolution of nuptial gifts in spiders. In R. H. Macedo & G. and shorter copulations, which translates into lower sperm Machado (eds.), Sexual selection: perspectives and models from the transfer and fertilization success (Stålhandske 2001; Albo Neotropics. London: Academic Press: 183–200. et al. 2011). Thus, allowing the female to feed on one leg ALBO, M. J., WINTHER, G., TUNI, C., TOFT, S. & BILDE, T. 2011: may serve to prolong copulation duration, and ultimately to Worthless donations: male deception and female counter play in a increase male reproductive success. nuptial gift-giving spider. BMC Evolutionary Biology 11: 329. ANDERSEN, T., BOLLERUP, K., TOFT, S. & BILDE, T. 2008: Why do Providing the female with a nuptial gift may also males of the spider Pisaura mirabilis wrap their nuptial gifts in silk: function to reduce the risk of cannibalism (Fromhage & female preference or male control? Ethology 114: 775–781. Schneider 2005). In Pisaura mirabilis, sexual cannibalism, ANDERSSON, M. B. 1994: Sexual selection. Princeton, New Jersey: in which a female kills and consumes the male, occurs in Princeton University Press. 1.2–4% of laboratory trials (Bilde et al. 2006; Hansen et ANDRADE, M. C. B. 1996: Sexual selection for male sacrifice in the Australian redback spider. Science 271: 70–72. al. 2008), suggesting that the trait may have evolved as a ARNQVIST, G. & ROWE, L. 2005: Sexual conflict. Princeton, New cannibalism avoidance strategy. Cannibalistic attacks may Jersey: Princeton University Press. result in the male surviving while the female consumes BILDE, T., TUNI, C., ELSAYED, R., PEKAR, S. & TOFT, S. 2006: Death part of the male, most commonly a leg (Bilde et al. 2006). feigning in the face of sexual cannibalism. Biology Letters 2: 23–25. BILDE, T., TUNI, C., ELSAYED, R., PEKAR, S. & TOFT, S. 2007: Nuptial However, cannibalistic attacks usually occur either pre- or gifts of male spiders: sensory exploitation of the female’s maternal post-mating and not concurrent with sperm transfer. To our care instinct or foraging motivation? Animal Behaviour 73: 267– knowledge, sperm transfer during soma consumption that 273. does not include killing and consuming the mate as observed BOOMSMA, J. J., BAER, B. & HEINZE, J. 2005: The evolution of male traits in social insects. Annual Review of Entomology 50: 395–420. here, has so far only been seen in species with specialized BRAUTIGAM, S. E. & PERSONS, M. H. 2003: The effect of limb loss body structures such as dorsal cephalic secretions (Mays on the courtship and mating behavior of the wolf spider Pardosa 1971; Kuntz, Garbe & Uhl 2012). Intriguingly, autotomy milvina (Araneae: Lycosidae). Journal of Insect Behavior 16: in Pisaura mirabilis may serve both as a substitute for a 571–587. nuptial gift to facilitate sperm transfer and as a cannibalism BRUESEKE, M. A., RYPSTRA, A. L., WALKER, S. E. & PERSONS, M. H. 2001: Leg autotomy in the wolf spider Pardosa milvina: a avoidance behaviour, in which the male’s leg has acquired common phenomenon with few apparent costs. American Midland alternative functions in a sexual context. Naturalist Journal 146: 153–160. Thanatosis is an adaptive male mating strategy that FLEMING, P. A., MULLER, D. & BATEMAN, P. W. 2007: Leave it allows males to increase their reproductive success by all behind: a taxonomic perspective of autotomy in invertebrates. Biological Reviews 82: 481–510. remaining in contact with resistant females (Hansen et al. FOELIX, R. 2011: Biology of spiders, third edition. Oxford: Oxford 2008). During thanatosis the male fixes his chelicerae to the University Press. gift and remains motionless, with his legs stretched out and FROMHAGE, L. & SCHNEIDER, J. M. 2005: Safer sex with feeding un-inserted raised. Not removing the pedipalp females: sexual conflict in a cannibalistic spider. Behavioral from the female epigyne may be advantageous for males Ecology 16: 377–382. HANSEN, L. S., GONZALEZ, S. F., TOFT, S. & BILDE, T. 2008: if it allows sperm transfer during female movements, or to Thanatosis as an adaptive male mating strategy in the nuptial gift- resume sperm transfer more promptly once the female stops giving spider Pisaura mirabilis. Behavioral Ecology 19: 546–551. moving. The suggested mechanism of sperm transfer relies JOHNSON, S. A. & JAKOB, E. M. 1999: Leg autotomy in a spider has on capillary forces (Foelix 2011), and whether transfer is minimal costs in competitive ability and development. Animal Behaviour 57: 957–965. possible during thanatosis remains unclear. Nevertheless, by KNOFLACH, B. & van HARTEN, A. 2000: Palpal loss, single palp engaging in such behaviour, males may face severe injury copulation and obligatory mate consumption in Tidarren cuneolatum risks, as pedipalp damage or breakage could limit their (Tullgren, 1910) (Aranea, ). Journal of Natural History current and future reproductive success. Whether death 34: 1639–1659. feigning while the male is in copula with the female is an KUNTZ, K., GARBE, S. & UHL, G. 2012: The function of the secretory cephalic hump in males of the dwarf spider Oedothorax retusus adaptive strategy that increases sperm transfer, similar to the (: Erigoninae). Animal Behaviour 83: 511–517. somersault behaviour of red-back spiders (Andrade 1996), LANG, A. 1996. Silk investment in gifts by males of the nuptial feeding remains to be explored. spider Pisaura mirabilis (Araneae: Pisauridae). Behaviour: 697- To conclude, whilst the functional significance of the 716. MAGINNIS, T. L. 2006: The costs of autotomy and regeneration in extreme male mating behaviours reported here awaits animals: a review and framework for future research. Behavioral further study, we believe our observations represent Ecology 17: 857–872. intriguing examples of male alternative mating strategies, MAYS, D. L. 1971: Mating behavior of nemobiine crickets Hygronemobius, providing a basis for further investigations. We hypothesize Nemobius and Pteronemobius (Orthoptera: Gryllidae). Florida that the reproductive advantages of the gift-giving trait and Entomologist 54: 113–126. PARKER, G. A. 1979: Sexual selection and sexual conflict.In M. S. Blum thanatosis favour the evolution of extreme male mating & N. A. Blum (eds.), Sexual selection and reproductive competition behaviours such as those reported here. in insects New York: Academic Press: 123–166. SCHNEIDER, J. & FROMHAGE, L. 2010: Monogynous mating strategies in spiders. In P. Kappeler (ed.), Animal behaviour: evolution and Acknowledgments mechanisms. Berlin & Heidelberg: Springer: 441–464. STÅLHANDSKE, P. 2001: Nuptial gift in the spider Pisaura mirabilis maintained by sexual selection. Behavioral Ecology 12: 691–697. We thank Anjia Junghaans for helping collecting spiders. VAHED, K. 1998: The function of nuptial feeding in insects: review of Paolo Giovanni Ghislandi was hosted by Trine Bilde’s lab. empirical studies. Biological Reviews 73: 43–78.

103