THE EFFECTS OF SOCIAL EXPERIENCE ON SEXUAL SELECTION IN THREESPINE STICKLEBACKS ( G ASTEROSTEUS ACULEATUS ) By Emily G race Weigel A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Zoology - Doctor of Philosophy Ecology, Evolutionary Biology and Behavior - Dual Major 2015 ABSTRACT THE EFFECTS OF SOCIAL EXPERIENCE ON SEXUAL SELECTION IN THREESPINE STICKLEBACKS By Emily G race Weigel We investigate how social experience can alter traits, and thereby influence sexual selection, using threespine sticklebacks ( Gasterosteus aculeatus spp.). Specifically, we examine how female and male responses to social conditions affect their reproductive investments and mating success. We first focus on how mate availability impacts female reproductive investment as measured by the timing and amoun t of clutches and eggs. Using treatments that mimic high and low availability of reproductive males, we found that Paxton Lake females, although previously shown to adjust their mate choices, did not alter their reproductive investment strategies. Because females must make relatively fixed reproductive investments prior to courtship, but have plastic mate choice, the strength of sexual selection acting on males during mating is subject to change. This work therefore highlights the potential consequences whe n reproductive investment and female choice operate out of sync. We focus second on how males allocate resources into several traits, and when and how those traits are assessed in courtship. With Cranby Lake sticklebacks, we show that the sexual signal pr eferred by females in many populations (red throats) is neither condition - dependent, nor preferred in mating. Males indeed allocate resources to their morphological displays, courtship behavior, and nests, but only a subset of traits matter to females at e ach stage of the courtship progression. Specifically, females only prefer high male condition early in courtship, and they increasingly prefer more vigorous, display - oriented courtship behavior as courtship progresses. This work suggests that the relative importance of traits upon which females choose mates is not only a function of whether those traits indicate male quality, but also depends on what stage in courtship traits are assessed. Third, because population demography is likely to vary spatially a nd temporally for many organisms, we address how male traits might change in response to the intensity of ing traits, and mating success. We found males to display different coloration and build nests differently in response to a two - week exposure to either low or high competition. Given that these changes resulted from such a short - term exposure, this work cl arifies how rapid changes in demography affect the courtship and nesting strategies adopted by males. If females are assessing male traits throughout courtship, understanding whether and how male traits change with competitive conditions suggests how mate competition and female choice may interact to shape traits, and how acclimation to social conditions may influence adaptation. Finally, we inquire how experience and success in mating affect male traits. Because previous experience with other males alte red male traits, we were inspired to discover how male traits and mating success respond to prior experience and success in courtship with females. This is particularly important as changes to population demography often alter mating opportunities, competi tion, and the likelihood of success. Here we show that males alter their body coloration and court more quickly with experience, but males do not alter their nests when they experience mating success. The trait responses to experience suggest a cycle: the male traits used in mating are responsive to prior experiences and success, and these modified traits are used in future mating attempts. Thus, trait variation can be maintained in part by the responses to experiences individuals have across their lifetim e. iv ACKNOWLEDGEMENTS First, I thank Jenny Boughman, my research advisor, who has guided me throughout graduate school. Jenny always challenged me to focus on testing the important, interesting questions (to others, and not just me), and to prioritize my various interests. I am also appreciative of her advice for navigating academia beyond simply the science. I feel I now much better understand how to position myself for career success thanks to her. Along with Jenny, I also thank my commi ttee members Tom Getty, Kay Holekamp, Fred Dyer, and Kim Scribner. Like Jenny, they molded both my scientific questions and me as a scientist. Tom has always kept a much - appreciated open door to port has likewise been strong, particularly by imparting the importance of clear communication and work - life invaluable. Finally, thanks to Kim for extremely thoughtful and thorou gh reviews of my work, and for always reminding me to consider in what ways my experiments do (or do not) capture nature. In addition to being people I admire, I cannot be thankful enough to have had these wonderful mentors. n lab, but none of that matters. My lab mates are awesome. Thanks to Audra Chaput, Jason Keagy, Alycia Lackey, Liliana Lettieri, Robert Mobley, and Robin Tinghitella. They all have contributed significantly to my development and well - being through continuo us support and constructive feedback. They also are genuinely fun people to hang around and observe excelling in research v our collaborations and friendship for many y ears to come. who helped me complete my projects. Thanks to Ellyse Cipolla, Savannah Foster, Felicia Harmon, Angela Marchand, Anna Reh - Gingerich, Gavin Rienne, Marquita Til lotson, and Benjamin (Ben) Wurst. I appreciate your feedback and positivity, as well ch for the opportunity to be your mentor and to be mentioned in at least one of your Nobel Prize acceptance speeches. My work would not have been possible without assistance in fish collection and maintenance. Thanks to our lab technicians at MSU, Liz Racey, Jared Thompson, and Sonya Williams, and the team of undergraduates who kept our fish healthy and happy. these many years. For the difficult times, thank you to my friends for helping me stick through it. Amy Lark, Lisa Stelzner, Zachary Blount, Mike Wiser, Christina Ragan, Katie Feirer, Erin Garcia, and my cohort, thank you so much for being sources of advice, inspiration, and encouragement. To the Buckyballz (in p articular Pamela Roy, Sarah Luderer, and Nicole Hewlett), thank you for knocking around ideas and soccer balls with me for so many years. And to Louise Mead and Terri McElhinny, thank you for acting as my honorary committee members. You all have always bee n confident in my abilities; thank you for helping me to believe in myself. vi You might not be able to choose your family, but you can certainly be happy when life gives you a family like mine. Thanks to my parents, Cindi and Roger Weigel, for supporting th eir little scientist; to my Aunt Carla DeBord for broadening my horizons; this far without you. To the one family member I did get to choose, my husband D . Philip Reesma n, thank you for committing to excellence. vii TABLE OF CONTENTS . ..... . x ... .. x i .. . 1 LITERATURE . .. .... . 7 . . .. 1 1 No evidence for adjustment of maternal investment under alternative mate availability regimes . 1 1 .. 11 . . ... 11 . ... 18 18 1 9 Tracking reproductive investment 20 Statistical analys is. . . .. 21 . . ... 22 . ... .. 24 . .. 30 LITERATURE . . .. 34 .. .. 43 Doing the right thing, in the right amount, at the right time: Traits that predict courtship 43 . .. ... 43 . . ... 43 . ... 48 48 .. .. 48 Mea 49 . 5 1 .. 53 Is red throat coloration related to male condition? 53 How do other coloration?.............................................................................................. 53 Courtship behavior 53 54 . 54 DISCUSSION . .. . 54 .. 54 .... 57 viii AP ........ ... 60 LITERATURE CIT ...... .. .. . 66 CHAPTER THREE ... .. .. . ..75 .. . 75 ... . .. .. . 75 INTRODUCTIO ... ... .. 75 MATERIALS AND M . . . .. . .79 Collection of wild fish and laboratory conditions . ..7 9 .. . ... ...80 Establishing density treatments ..80 Enticing males to build nests ...80 Gathering male phenotypic data 3 3 Statistical analy . . . ...84 . ... . .. .. . 86 Which males nest?.................................. . .. ...86 Do male trait values differ between nesting males who experienced high versus low density?.............................................................................. .. . 87 ..... ..87 ..... ..88 . ..... ..89 Does density affect mating success? .. . .. .... ... . ..89 How does the variation in male traits compare between treatments? .....90 .. . . .... ...90 APPE .. . .... ..99 LITERATURE CITED ..... 105 CHAPTER FOUR ..... .113 Courtship Experience and Success Affect Male Traits .113 ... ... . . 113 ..... 113 MATERIALS AND METH ..... 117 Study population................................................................................... 117 Encouraging males to create and maintain nests .. ..... 118 Courtship trials .. . ... ................. .............. .. ...118 Male phenotypic data .. ... 12 1 12 1 Statistical anal . ...... 122 ...... 12 5 How does courtship experience affect male traits? ..... . . 12 5 How does courtship success affect male traits? .. 1 26 Does past courtship success predict future courtship success? .... . .. 12 7 ... ..... 127 ix Trait changes in response to experience ..... . 128 130 Does past success predict future success?.......................................... 132 AP ...... ... 1 36 LITERATURE CIT ..... 1 40 CONCLUSION . ...149 x LIST OF TABLES Table 1.1. Significant predictors of two metrics of courtship behavior from multiple regression models ( df =6,77). . . ....... ........ 61 Table 1.2. Male traits that predict female responsiveness and inspection rate. Multiple regression models with estimates and SE scaled by 100 ( df =68). Significant . ....... . ... 62 Table 2.1. Significant morphological predictors, all col or - related, from a linear model of male nest building ( df = 1, 233) ... .. 10 0 Table 2.2. Predictors of male mating success as measured by composite MANOVA df = 3, 94); and a linear model of preference score ( df = 15, 93). Negative coefficients for density and interactions indicate that the relationship is stronger for males from low than high density. Main effects, when not significant, are included to aid the i . 101 Table 3.1. Male traits that varied significantly with male age (trial date).. ... 137 Table 3.2. Male trait changes in response to courtship success, as represented by measures mixed models are displayed, and non - significant results are denoted - df =232). P - values are noted before and after FDR correction using asterisks, where p <0.05*, p <0.01**, and p (0.05< p .. 138 xi LIST OF FIGURES Figure 1.1. Comparison of the number of eggs per initial clutch between female - and male - biased mate availability treatments. The central boxes represent values from the lower to upper quartile (25 to 75 percentile; first and third quantiles), and are intended to give an c. 95% confide nce interval for differences in the two datasets. Bold lines represent the median, and extreme values are represented by open circles. ........ 31 Figure 1.2. Comparison of the number of days spent in a nongravid state between female - and male - biased mate availability treatments. The central boxes represent values from the lower to upper quartile (25 to 75 percentile; first and third quantiles), and are intended to give an c. 95% confidence interval for differences in the two da tasets. Bold lines represent the median, and extreme .. 32 Figure 1.3 . Comparison of the interclutch interval length (the number of consecutive days a female is completely nongravid, i.e neither carrying nor developing a clutch) between female - and male - biased mate availability treatments. The central boxes represent values from the lower to upper quartile (25 to 75 percentile; first and third quantiles), and are intended to give an c. 95% confidence interval for differences in the two datasets. Bold lines represent the median, and extreme values are represented by open circ les. ... .............................................................................. ............. ...... 33 Figure 2.1. More colorful males are faster to court when larger, but dull males are faster to court when smaller. Regression lines for male co lor represent categories of high and low coloration at 1.5 standard deviations above and below the mean (respectively). .. ... 63 Figure 2.2. As condition improves, low color males have heavier nests, whereas high color male s have lighter nests. Regression lines for male color represent categories of high and low coloration at 1.5 standard deviations above and below 64 Figure 2.3. Females respond differently to males based on interactions between male condition and vigor: Increased vigor leads to more responsiveness (follows/lead) for low - condition males, but less responsiveness for high - condition males. Regression lines for male condition represent categories of high and low c ondition at 1.5 standard deviations above and below the mean (respectively).. 65 xii Figure 3.1. After experiencing low density, males have more colorful bodies (t 1,233 = 5.28, p < 0.0001), whereas after experiencing high density, males are darker along their dorsal surface. Points indicate mean ± SE. Asterisks denote significance at * p <0.05 and *** p 1 02 Figure 3.2. Males build nests faster and larger after experiencing low than high density. Males who experienced low density took fewer days to build nests (t 6,114 = - 2.06, p= 0.042). Their nests were greater in both perimeter and area (t 6,114 = 3.21, p= 0.002 and t 6,114 = 2.67, p= 0.009, respectively), although the exposed nest area and perimeter (portion of the nest uncovered by sand ) was unaffected (all t 117 < 0.38, p >0.705). Asterisks denote significance at p <0.05 (*) and p <0.01 (**) 1 03 Figure 3.3. At early and late season, males who experienced high density took longer to build nests than males from low density. This effect was primarily driven by the lag to begin nest construction, rather than the days spent in construction. Lags were longer in May (t 6,114 = - 2.30, p = 0.023) and July (t 6,114 = - 2.60, p = 0.011), but males from both high and low density build nests equally quickly a t midseason. Points indicate mean ± SE. Asterisks denote significance at p 104 Figure 4.1 . Standardized parameter estimates and confidence intervals for the effect of courtship experience on male traits. Male traits that responded significantl y to courtship experience are denoted by asterisks (***, where after FDR adjusted p 1 39 1 INTRODUCTION Ever since Darwin (1871) proposed that the elaborate traits males possess function in competition with other males to gain the attention of females, we have been striving to understand how male traits are shaped, as well as the female preferences for these traits. In most populati generally implies that females make a larger investment than males, producing fewer 1948; Trivers, 1972). Therefore, females m ust make decisions about with whom they will mate, and as a consequence, males must compete with other males for mating opportunities. Because some males will enjoy greater reproductive success than others, the competition for mates can drive the evolution of traits that are useful to males both in the context of male - (Andersson, 1994). In the years since Darwin (1871), Bateman (1948), and Trivers (1972) published their seminal works on sexual sele ction, behavioral ecologists have recognized that sexual selection may be much more complex than imagined by these earlier investment made by males (Wedell et al., 2002) an d largely ignore the importance of Endler and Basolo, 1998; reviewed in Andersson, 1994). Evidence is mounting that the quality, quantity, and distribution of potential ma tes and competitors factor into mate choice decisions (Jennions and Petrie, 1997; Bailey and Zuk, 2009; Hebets and Vink, 2007; Tinghitella et al., 2015). In particular, we are beginning to ask how variables that 2 change within the lifespans of individual or ganisms might affect the course of evolution across multiple generations. There is currently strong interest in increasing our understanding of how population demography affects sexual selection. This is particularly important as population demography is u nlikely to remain constant across space or across time temporally that the factors that control mate availability, mate competition, and mating success are likely to vary as well. The importance of social information in informing then, do demography and social information affect the expression of traits and mating success? Organisms that are s ensitive to and capable of altering their trait expression in response to social conditions are likely to enjoy fitness advantages. These within - lifetime changes may alter rates of selection on expressed traits, facilitating their rapid spread or damping t hem out altogether. To understand the evolution of traits, we therefore need to understand to which factors organisms may respond within their lifetimes, in what ways, and to what degree. This is a timely and important area of inquiry, as the rapid environ mental changes we are experiencing now are increasing the frequency with which demography and associated social conditi ons are disturbed (reviewed in Candolin and Wong, 2012). To answer questions about the roles of various kinds of social experience in se xual selection, I use the stickleback fish ( Gasterosteus aculeatus ), which represent a well - known system for the study of sexual selection (Rowland, 1988; Bakker and 3 Milinski, 1993; Candolin, 1999a; Kraak and Bakker, 1998; Hendry et al., 2013; Tinghitella et al., 2013; Bolnick et al., 2015). Although my work primarily concerns the sticklebacks of Cranby Lake, I repeatedly draw connections to other populations that live in other freshwater lakes in British Columbia (McPhail et al., 1994; Bell and Foster, 199 4; Boughman, 2001). I examine the details of how sexual selection functions in response to specific demographic or social manipulations in this population and then discuss whether my results can be used to compare and contrast them with responses in other populations. Although Cranby sticklebacks are understudied relative to other populations, much is known about the biology of sticklebacks in general, and this knowledge can aid in our investigations. First, most stickleback fish in the wild live only one y ear, although some populations may live up to two years, but rarely longer (Baker, 1994). Within this time, and particularly during the breeding season, population density and adult sex ratios vary temporally and spatially (Tinghitella et al., 2013; Tinghi tella, Head and Boughman, unpublished data). Males are known to compete strongly for a territory within which to build nests to court females (Lackey and Boughman, 2013). Although populations vary in how they court (Wootton, 1976 and references therein; Ri dgway and McPhail, 1987; Foster, 1995; Scotti and Foster, 2007), sticklebacks do follow a characteristic pattern marked by an elaborate courtship dance (Tinbergen 1951), which, Males are then solely responsible for raising young, which includes oxygenating (fanning) the eggs and guarding fry (Wootton and Wootton, 1984). Males also produce a characteristic red throat coloration, which is relatively muted in Cranby compared to othe r populations 4 (Albert et al., 2007). Although exactly which traits are preferred by Cranby females (and to what relative extents) remains untested, females from many populations generally prefer larger, redder males (but see Boughman, 2001), and male color is often, but not always, positively correlated with physical condition (Wootton, 1976; Milinski and Bakker, 1990; Bakker and Milinski, 1993; Bakker and Mundwiler, 1994; Candolin, 1999b; Boughman, 2007). In general, males express a variety of traits to co mpete for mates, and choosy females must then evaluate male traits to select suitable mates. Here, I investigate how demography and social experience shape sexual selection. I first address how mate availability impacts female reproductive investment. In C hapter 1, I ask whether the availability of many or few potential mates impacts the reproductive state across the season and extract her clutches to determine whether investment varies with mate availability. This work sheds light on whether female reproductive investment is fixed or plastic in the face of changing social conditions. It also draws attention to the potential consequences when reproductive investment and female cho ice operate out of sync (e.g., when many eggs are invested in less - preferred males). Next, I turn the focus to males to investigate how and when male traits might be used in reproduction and whether these traits reliably indicate male condition. In Chapte r 2, I use no - choice mating trials (i.e., a single female with a single male) to investigate the male traits most favored by females as courtship progresses and determine which traits are linked to male condition. This work determines which male traits are preferred by females in this population, and at which stage in courtship that 5 preference is expressed. This work also explores the potential function of certain male traits in reproduction, and why they may differ between closely - related species. I then expression and female choice. In Chapter 3, I build on my findings in Chapter 2 to examine how male density (perceived level of mate competition) can affect male traits and mating success. Here I test whether morphological, behavioral, and nesting traits are responsive to competition, and if so, in what ways. This work relies on a temporary exposure to one of two density treatments, followed by individual tank placement, allowing us to teas e out what the lingering effects of density may be on male traits and mating success. This work suggests that short - term experiences can affect the variation seen in male traits upon which females will later select. Finally, I consider that males and fema les, as a consequence of different levels of mate availability and competition, will likely also vary in their amounts of experience important for males, as they are gene rally less assured to mate than females (Trivers 1972), and trait changes should have larger effects on their mating success. Thus, in Chapter 4, I ask how mating experience influences male traits and mating success. I build on this to also ask whether pas t mating success influences the traits expressed and mating success of males in subsequent mating attempts. This work therefore clarifies how mating opportunities and success may differentially impact male traits, and how experience - mediated trait changes may impact the strength of sexual selection on male traits. 6 guidance and input of my research advisor, Janette Boughman. 7 LITERATURE CITED 8 LITERATURE CITED Albert AYK, Millar NP, Schluter D, 2007. Character displacement of male nuptial colour in threespine sticklebacks ( Gasterosteus aculeatus ). Biological Journal of the Linnean Society 91:37 - 48. doi: 10.1111/j.1095 - 8312.2007.00777.x. Ander sson M, 1994. Sexual Selection. Princeton, NJ: Princeton Univ. Press. pp. i xx., 1 - 599. Bailey NW, Zuk M, 2009. Field crickets change mating preferences using remember ed social information. Biology L etters 5:449 - 451. doi: 10.1098/rsbl.2009.0112. Baker JA, 1994. Life history variation in female threespine stickleback. In: Bell MA, Foster SA, editors. The Evolutionary Biology of the Threespine Stickleback Oxford , England. Oxford University Press. p. 144 - 187. Bakker TCM, Milinski M, 1993. The advantages of being red: Sexual selection in the stickleback. Marine and Freshwater Behaviour and Physiology 23:287 - 300. Bakker TCM, Mundwiler B, 1994. Female mate choice a nd male red coloration in a natural three - spined stickleback ( Gasterosteus aculeatus ) population. Behavioral Ecology 5:74 - 80. Bateman AJ, 1948. Intra - sexual selection in Drosophila . Heredity 2:349 - 368. Bell MA, Foster SA, 1994. The Evolutionary Biology o f the Threespine Stickleback: Oxford University Press. Bolnick DI, Shim KC, Brock CD, 2015. Female stickleback prefer shallow males: Sexual selection on nest microhabitat. Evolution 69:1643 - 1653. doi: 10.1111/evo.12682. Boughman JW, 2001. Divergent sexua l selection enhances reproductive isolation in sticklebacks. Nature 411:944 - 948. doi: 10.1038/35082064. Boughman JW, 2007. Condition - dependent expression of red colour differs between s tickleback species. Journal of Evolutionary B iology 20:1577 - 1590. doi: 10.1111/j.1420 - 9101.2007.01324.x. Candolin U, 1999a. Male - male competition facilitates female choice in sticklebacks. Proceedings of the Royal Society B - Biological Sciences 266:785 - 789. doi: 10.1098/rspb.1999.0706. Candolin U, 1999b. The relationship be tween signal quality and physical condition: Is sexual signaling honest in the three - spined stickleback? Animal Behaviour 58:1261 - 1267. 9 Candolin U, Wong BB, 2012. Behavioural responses to a changing wor ld: mechanisms and consequences. Oxford University Pr ess. Darwin C, 1871. The descent of man, and selection in relation to sex. United Kingdom , John Murray. Emlen ST, Oring LW, 1977. Ecology, Sexual Selection, and Evolution of Mating Systems. Science 197:215 - 223. doi: 10.1126/science.327542. Endler JA, Basolo AL, 1998. Sensory ecology, receiver biases and sexual selection. Trends in Ecology & Evolution 13:415 - 420 . Foster SA, 1995. Understanding the evolution of behavior in threespine stickleback: The value of geographic variation. Behaviour 132:1107 - 112 9. doi: 10.1163/156853995x00487. Hebets EA, Vink CJ, 2007. Experience leads to preference: experienced females prefer brush - legged males in a population of syntopic wolf spiders. Behavioral Ecology 18:1010 - 1020. doi: 10.1093/beheco/arm070. Hendry AP, Pei chel CL, Matthews B, Boughman JW, Nosil P, 2013. Stickleback research: The now and the next. Evol utionary Ecol ogy Res earch 15:111 - 141. Jennions MD, Petrie M, 1997. Variation in mate choice and mating preferences: A review of causes and consequences. Biolo gi cal R eviews of the Cambridge P hilosophical Society 72:283 - 327. Kraak SBM, Bakker TCM, 1998. Mutual mate choice in sticklebacks: attractive males choose big females, which lay big eggs. Animal Behaviour 56:859 - 866 . Lackey ACR, Boughman JW, 2013. Diverg ent sexual selection via male competition: ecology is key. Journal of Evolutionary B iology 26:1611 - 1624. doi: 10.1111/jeb.12173. McPhail JD, Bell MA, Foster SA, 1994. Speciation and the evolution of reproductive isolation in the sticklebacks ( Gasterosteus ) of south - western British Columbia. In: Bell MA, Foster SA, editors. The Evolutionary Biology of the Threespine Stickleback Oxford, England: Oxford University Press. p. 400 - 437 . Milinski M, Bakker TCM, 1990. Female sticklebacks use male coloration in mat e choice and hence avoid parasitized males . Nature 344:330 - 333. Ridgway M, McPhail J, 1987. Rival male effects on courtship behavior in the Enos Lake species pair of sticklebacks ( Gasterosteus ). Canadian Journal of Zoology 65:1951 - 1955. 10 Rowland WJ, 1988. Aggression versus courtship in threespine sticklebacks and the role of habituation to neighbours. Animal Behaviour 36:348 - 357. Scotti MAL, Foster SA, 2007. Phenotypic plasticity and the ecotypic differentiation of aggressive behavior in threespine stickl eback. Ethology 113:190 - 198. doi: 10.1111/j.1439 - 0310.2006.01311.x. Tinbergen N, 1951. The Study of Instinct. Oxford University Press. Tinghitella RM, Stehle C, Boughman JW, 2015. Females sample more males at high nesting densities, but ultimately obtain less attractive mates. BMC Evolutionary Biology 15:200. Tinghitella RM, Weigel EG, Head M, Boughman JW, 2013. Flexible mate choice when mates are rare and time is short. Ecology and Evolution 3:2820 - 2831. doi:10.1002/Ece3.666. Trivers RL, 1972. Parental investment and sexual selection. In: Campbell B, editor. Sexual selection and the descent of man, 1871 1971: Aldine, Chicago. p. 136 179. Wedell N, Gage MJG, Parker GA, 2002. Sperm competition, male prudence and sperm - limited females. Trends in Ecology & Evolution 17:313 - 320. Wootton RJ, 1976. The biology of the sticklebacks. Academic Press, London. Wootton RJ, Wootton RJ, 1984. A functional biology of sticklebacks. University of California Press. 11 C HAPTER ONE No evidence for adjustment of maternal investment under alternative mate availability regimes Published as: Weigel, EG, Tinghitella, RM, and Boughman, JW. ( 2015 ). No evidence for adjustment of maternal investment under alternative mate availability regimes. Journal of Fish Biology. DOI: 10.1111/jfb.12793 ABSTRACT Using treatments that mimic high and low availability of reproductive males, it was found that female threespine sticklebacks Gasterosteus aculeatus (L.1758), previously shown to adjust their mate choices when male mates we re rare, did not alter their reproductive investment strategies. These results suggest that plasticity in investment is perhaps limited by physiological requirements or dependent on relatively extreme mate availability regimes. The probability of becoming reproductive, number of clutches per season (per female), initial clutch size and mass, and the timing of reproduction were all independent of the experience a female had with mate availability. This suggests that pre - copulatory plasticity in reproductive strategies may contribute more to variation in the strength and direction of sexual selection than reproductive investment in offspring. INTRODUCTION Life histories of organisms are frequently shaped by trade - offs, whereby the fitness benefits of one trai t (or suite of traits) are linked to deficits in another trait (Stearns, 1992). Classic examples of life history trade - offs include whether to invest in growth or reproduction (Gadgil & Bossert, 1970), to care for offspring or court new mates (Lindstr ö m, 1998; Bjelvenmark & Forsgren, 2003; among others) and when to reproduce relative to mortality ( Zwaan et al ., 1995; Flatt, 2011 ). Because trade - offs are 12 intrinsically linked to the determinants of fitness, optimizing the costs and benefits of traits in a pa rticular environment is critical. Given that evolutionary trade - offs are fundamentally derived from responses to limited resources, trade - offs also shape the fitness of alternative strategies under fluctuating environmental conditions. This study considers trade - offs that may occur when mates are a limiting resource. Both male and female life histories are shaped by trade - offs between the quantity and quality of current and future reproductive opportunities. Here it is asked whether and how females alter th is investment in offspring when mates are rare vs. common. Population density and mate availability fluctuate dramatically, even within variation in the strength and direction of sexual selection via changes in pre - and post - copulatory mechanisms (Forsgren et al., 1996; Hebets, 2003; Cotton et al., 2006; Mück et al., 2013). For example, populations shrinking under high mortality are predicted to increase investment in current reproduction, because survival and therefore future opportunities to mate are uncertain (Gadgil & Bossert,1970; Stearns, 1992; Stearns, evidence supports the theory of early investment in reproduction under high mortality across taxa (Reznick, 1983; Tatar & Carey, 1995; Stearns, 2000) although counter - examples exist (Clutton - Brock, 1984; reviewed in Reznick, 1985). Populations of the two - spotted goby Gobiusculus flav escens (Fabricius 1779) shift from early - season male male competition to late - season female female competition as males become more scarce, reversing sex roles dynamically (Forsgren et al., 2004; Wacker et al., 2013; Wacker et al., 2014). Similarly, the female preference for large size in this species 13 disappears late in the season when males are rare (Borg et al., 2006) . Hence, male and female life histories are shaped by tradeoffs that depend on both quality and quantity of current and future reproductive opportunities. Therefore, existing evidence demonstrates that fluctuation in within - season mate availability leads to changes in the strength of sexual selection, and can select for phenotypically plastic responses in mating that m aximize reproductive success (Stearns, 1992; Roff, 1997; Pigliucci, 2001; DeWitt and Scheiner 2004). Pre - mating reproductive behaviours like mating competition and mate choice clearly respond to mate availability (e.g. Milinski & Bakker, 1992; Jirotkul, 1 999; Kokko & Mappes 2005; Shine et al., 2006; reviewed in Kokko & Rankin, 2006). Whether and how animals vary reproductive investment in response to mate availability, however, has received much less attention. In contrast to female choice, current reprodu ctive investment is predicted to increase when future mating opportunities are uncertain et al., 2009). Because females typically invest a great deal in offspring, selection for plastic reprodu ctive investment in response to mate availability should be strong, reflecting the potential loss of fitness if non - adaptive investments are made. Similarly, the greater the reproductive investment by males, the more exaggerated the effects should be on fe male reproductive success when mates are limited (Trivers, 1972; Emlen & Oring, 1977; Smith,1977; Halliday, 1978; Patterson et al., 1980; Westneat, 1988; Royer & McNeil, 1993). Indeed, there is growing evidence that mate availability influences female repr oductive success (Wedell et al., 2002; Smith & Reichard, 2005; Heubel et al., 2008; Carrillo et al., 2012), particularly in species with significant male investment in 14 parental care (Borg et al., 2002; Forsgren et al., 2004, Hopwood et al. 2015 ), or nutritional investment (Simmons & Kvarnemo, 2006; Simmons & Kotiaho, 2007; Scharf et al., 2013). For example, previous work in common gobies Pomatoschistus microps (Krøyer 1838), has shown that female - biased sex ratios lead to larger first clutches, signifying a trade - off of increased investment early in the season at a cost of producing fewer eggs across the season (Heubel et al., 2008). Although plasticity may come in more forms than previously studied, plastic reproductive investment could facilita te increased reproductive success in rapidly - changing environments, for instance when populations are small, highly fragmented, or experience high mortality. Threespine sticklebacks Gasterosteus aculeatus (L.1758) are an established model system in sexual selection research (Rundle et al., 2000; Boughman et al., 2005; Andersson & Simmons, 2006; Hendry et al., 2013). They offer an opportunity to test the effects of demography on female reproductive investme nt in a species in which pre - mating behaviour is known to respond to altered demography (Tinghitella et al., 2013). During the summer breeding season, males establish nesting territories on which to build nests, and females search amongst them for mates. Under these circumstances, females, both those not yet gravid and those ready to deposit eggs, are regularly exposed to many males and nesting sites before making mating decisions (Boughman, 2006). Mating occurs throughout the breeding season in this speci es and occurs only primarily of oxygenating eggs and defending fry. Each time a G. aculeatus female develops a clutch, she has the opportunity to make a mating decision . Time is, however, limited; females ovulate all of their mature 15 oocytes concurrently in any single clutch, thus females must invest an entire clutch with a single male and have limited time to do so (reviewed by McLennan, 2006). Therefore, the rate of ac cepting a mate in any one encounter is not only dependent on the long - term, perceived availability of mates throughout the breeding season, but also the short - term developmental time of the current clutch. Because body shape and size change dramatically wh ile developing a clutch of eggs, G.aculeatus offer an easily observable model in which to manipulate mate availability and quantitatively measure reproductive investment. In several post - glacial lakes in British Columbia, G. aculeatus demography varies wit hin the breeding season. Adult population density decreases through the breeding season, particularly for males (Boughman, Tinghitella, and Head, unpublished data), and there is spatial variation in the operational sex ratio across breeding sites within a given lake (Tinghitella et al., 2013). As the availability of males (proportion of males) can fluctuate by over 30% from location to location in a given lake, females do indeed experience drastic differences in mate availability as they search for mates. F urthermore, female mating decisions become relaxed at the end of the season (Tinghitella et al., 2013) as residual reproductive value declines (Kokko & Jennions, 2008; Lahti et al., 2009), but only for females who have experienced a female - biased adult sex ratio (ASR). These females spawn more quickly and are more accepting of males, regardless of their advertised quality. Because males also must care for young, father s are temporarily removed from the pool of potential mates, reducing availability in some areas, adding to the variation in mate availability a female experiences. 16 potential mates are rare, but this theory also ignores the possibility that reproductive investment and mating decisions may not be modified in parallel across the breeding season. If female mating decisions are relaxed when mates are rare, but females do not alter their investment in eggs or they increase current investment when the risk of not mating late r in the season is high (as in Forsgren et al. 2004), then sexual selection will be relatively weak. Lower quality males might sire a sizeable number of offspring if they wait to mate when competition is reduced, avoiding costly male competition and eventually acquiring a mating opportunity with a less choosy female. Alternatively, if investment decreases as mating decisions are relaxed, sexual selection is stronger than close. Theory also provides specific predictions whe n females must allocate whole clutches to an individual potential mate. Following the theoretical model in Heubel et al. (2008), it is first assumed that females can have many clutches across the season, but have a fixed total budget of eggs. Second, it is assumed that when mates are rare, females developing clutches face uncertain future mating opportunities. Because G. aculeatus limnetic females typically have a single season in which to spawn, they can neither delay reproduction to the next season for b etter mating options nor obtain enhanced fecundity by growing to larger size (unlike other fishes; reviewed in Koslow 1996; Koons et al. 2008; Secor 2008). Thus, consistent with life history theory, when mates are rare, reproductive effort should shift to favour investment into the first clutch(es) or to reduce the time between clutches (Gadgil & Bossert, 1970; Stearns, 17 2009). This would allow females to capitalise on currently - available mates and avoid the potential costs of waiting too long to spawn. Conversely, when mates are plentiful and females are more assured of future mating opportunities, investment may not shift to favour early clutches. Instead, females may produce m any clutches thereby dividing out their resources among many males they will encounter across the breeding season. The partitioning of reproductive effort across clutches could increase not only the genetic diversity of offspring, as different males may be nesting at different points in the season, but also limit the amount of resource competition between siblings by temporally separating clutches. To test whether females respond to mate availability by altering their reproductive investment, the mate av ailability (ASR) experienced by female G. aculeatus was experimentally manipulated throughout their reproductive lifespans. The following was then measured: (1) the number of clutches produced over the course of the breeding season, (2) the initial clutch size (number of eggs and clutch mass) and (3) the length of time females remained gravid per clutch, across mate availability treatments. Nearly all research to date has coupled the effects of limited mate availability with seasonal declines in mate avail ability. Because females encounter variation in mate availability across the breeding season, this study design teased apart different ways by which female reproductive investment may respond. The aim is to provide comprehensive knowledge of whether and ho w reproductive investment is plastic in response to mate availability, and how investment may modulate the strength of sexual selection in a species that responds behaviourally to mate availability. 18 MATERIALS AND METHODS Study p opulation and t reatments At the beginning of the 2011 breeding season, wild limnetic sticklebacks ( Gasterosteus aculeatus, NCBI Taxonomy ID: 481459) from Paxton Lake, Texada Island in British Columbia were collected using minnow traps. G. aculeatus were sexed using well - established differences in body shape (for males and females) and nuptial colour (for males) (McPhail, 1984; McPhail, 1992; Hatfield, 1997) and then transferred to single sex plastic bags at equal densities, which were loaded into coolers for transport by air to Mich igan State University. The social conditions for both sexes were, thus, the same prior to placement in treatment tanks. Immediately upon arrival in the laboratory, G. aculeatus were assigned to 283.9 L (75 - gallon) replicate tanks in one of two treatments: 16 G. aculeatus in a 3:1 (male - biased, early - season) or 1:3 (female - biased, late - season) ratio of males to females in 12 replicate tanks (6 of each treatment). These treatments represent extreme values of sex ratio variation found in the wild, particularly for the female - biased condition (Tinghitella et al. 2013). Although male G. aculeatus are capable of raising clutches from multiple females across their lifetime, they rarely do. Within each tank, males and females were size - matched by visual examination within sex and uniquely marked with elastomer (Northwest Marine Technology, Shaw Island WA; Jones et al., 2006) along the dorsal side to facilitate individual identification. After 1 day of acclimation, it was confirmed that there were no size differences between treatments or tanks in G. aculeatus length (treatment: t - test: t = 0.8084, P>0 05 ; tank: ANOVA: F 11,191 =0.892, P>0 05 ) and mass (treatment: t - test: t = 0.9998, P>0 05 ; tank: ANOVA: F 11,191 =0.103, P>0 05 ). 19 Treatment tanks were visually isolated from one another using opaque, white covers applied to the outside surfaces of the tank. The artificial tank habit at included ceramic caves made of halved flower pots and plastic plants for cover. No materials were added to aid males in building nests; however, males and females could and did otherwise freely court. Makeshift nests were common, but removed to prevent females from spawning. High rates of courtship and male competition still occurred even in the absence of nests (Tinghitella et al 2013). Tanks were maintained with 14 - hour day lengths at approximately 18 o C, mimicking the natural conditions of their native habitat. All G. aculeatus were fed defrosted brine shrimp ( Artemia sp.) and bloodworms ( Chironomus sp.) once per day ad libitum . Tracking daily female reproductive i nvestment To assess daily changes in female reproductive status, each female was visually assessed on a 0 - 5 gravidity scale, where 0 indicated a non - gravid female and 5 indicates a female ready to release her clutch (as evidenced by an open genital pore and a plump, swollen abdomen; as in Fro mmen et al. , 2012). A gravidity score for each calculate the number of days a given female was gravid, as well as the total number of gravid females per tank each day. Visu al assessment of females on the 0 to 5 gravidity scale were made by a single researcher (EGW). Additionally, changes from a gravidity score of 5 to 0 the following day indicated release of a clutch (either because the female dropped the clutch naturally or because the clutch was extracted as described below), and allowed calculation of the total number of clutches each female developed across 20 the season. These in - tank observations minimized the amount of handling of the G. aculeatus. Tracking reproductive i nvestment in initial clutches All four females in male - biased tanks and a randomly - determined subset of four females from female - biased tanks were identified as focal individuals (N=48) and used to obtain additional body size and clutch measurements. When these females attained a score of 5 on the gravidity scale, they were weighed to the nearest hundredth of a gram on an OHAUS Scout PRO SPE 202 balance and their body length was recorded to the nearest hundredth of a mm. To obtain standard length, females were photographed using a Canon G - 15 digital camera and the distance between digital landmarks at two extremes of the body (the anterior tip of upper lip and caudal border of the hypural plate at the lateral midline) was calcul ated using the program PAST ( http://folk.uio.no/ohammer/past ) according to established methods (Taylor et al. , 2006; Cooper et al., 2011). The female was then gently squeezed to extract her clutch. The clutch was weighed, the number of e ggs counted, and the female reweighed post - extraction. Because previous work (Heubel et al., 2008; Carrillo et al., 2012) suggests first clutches are most altered by sex ratio differences, and animal handling and egg extraction could decrease survivorship and reclutching rate (one of the key variables minimize adverse effects on G. aculeatus health and sample size. Note that t - tests conducted following the experiment showe d no effect of handling on reproduction (when comparing the 4 focal females to the other females within each female - biased tank; all p>0.05). 21 Statistical a nalysis Because the reproductive investment of multiple females from each replicate tank was assessed, mixed effect model analysis was appropriate to evaluate how perceived availability of mates impacts female reproductive investment. Measures of reproductive investment across mate availability treatments included the number of clutches a female h ad throughout the breeding season, the number of eggs per initial clutch, initial clutch mass, the number of days spent either nongravid (stage 0), developing eggs (stages 1 - 4), and fully gravid (stage 5) during the season, and finally the start and stop o f seasonal reproduction. Mixed models were conducted within R (R et al., 2013). Treatment (male - or female - bias) was entered as a fixed effect into all models. Intercepts for females (nested w ithin replicate treatment tanks) and number of clutches (for models concerning days gravid) were included as random effects. To account for covariances in daily measures of gravidity, repeated measures were incorporated into models measuring days gravid. P lots of residuals for each response variable were visually inspected to detect deviations from homoscedascity or normality, and generalized linear mixed models were used with a Poisson distribution when appropriate. The fit of the full model (with the fixe d treatment effect) was compared against a reduced null model without the fixed effect (only random effects) to test whether the fit of the model decreased significantly (p<0.05) using chi - squared tests (Winter, 2013). Finally, st for significant difference in variances between treatments. Data are presented as means ± SD. 22 RESULTS During the summer breeding season of this study, females under both male - and female - biased conditions had an equal likelihood of being reproductive (producing at least one clutch) ( X 2 = 0.112 , df =2, p >0.05 ). Females in both mate availability treatments had 1.188 +/ - 1.149 SD clutches on average (range = 0 - 4). As the focus was on whether and how reproductive investment changed in response to mate avai lability, for the following analyses, females who were never reproductive (N=8) were removed. Females in male - biased tanks produced 1.47 times the number of clutches produced in female - biased tanks. Despite differences in the number of clutches produced, due to greater variance within treatments, females from both treatments generated an equal number of clutches across the season and the variance did not differ between treatments ( X 2 = 2.148 , df=2, p>0.05 >0.05 ). Reproductive females (those which had at least one clutch) averaged 1.970 +/ - 0.825 clutches across the season, regardless of treatment, and further examinations of first clutch mass and first clutch egg number also revealed no difference in investment across treatments (firs t clutch mass: X 2 = 0.033 , df=1, p =>0.05 p= >0.05 and egg number: X 2 = 0.033 , df=1, p>0.05 >0.05 ; see Fig. 1 .1 ). Next, the timing of reproductive investment was assessed across mate availability treatments. Variables measured included the number of days females spent in a completely nongravid state (0 on the gravidity scale) both for consecutive days (spans) of nongr days spent in a nongravid state across the season. This method revealed how clutches 23 clutches which may not have developed fully. Interestingly, most days of the reproductive season were spent in a nongravid (0) state (avg 101.880 +/ - 12.248 days). Females spent 12.755 +/ - 9.352 days between 0 and other states of gravidity (the interclutch interval ). When considering differences between mate availability treatments, there was neither a difference in the total number of days spent in a nongravid state (Figure 1. 2; X 2 = 0.967 , df=1, p >0.05 >0.05 ), nor in interclutch interval leng th (Figure 1. 3; X 2 = 1.288 , df=1, p >0.05 >0.05 ). This indicates that females spent little time developing clutches across the season, with the majority of the days spent in a nongravid state. Next, the number of days a female spent developing a given clutch of eggs, before she was ready to release the clutch (stages 1 - 4) was examined. The consecutive number of days on which the gravidity score ranged from 1 - 4 was the measure of time spent devel oping a clutch. A drop to 0 gravidity before reaching a fully gravid state (5) The treatments did not differ significantly (X 2 = 0.076, df=1, p >0.05 >0.05 ); females spent an average of 9.690 +/ - 1.859 days of the season in a state of clutch development, and each span of development was relatively quick, lasting only 2.490 +/ - 0.212 days on average. Then, the number of days spent in a completely gravid state (5 on the gravidity scale) during the 116 day experiment was assessed across sex ratio treatments. Females under both male - and female - biased conditions spent 3.55 +/ - 0.64 days across the season in a completely gravid (5) state (X 2 = 1.955, df=1, p>0.05, 24 F=0.499 p >0.05), and each span of gravidity lasted < 2 days (1.670 +/ - 0.181 days; X 2 = 0.331, df=1, p>0.05 ). Finally, the time to first reproductive day (calculated from the day treatments were established) and time to last reproductive day (the end of reproduction for each female) were assessed. Given the importance of diet and time necessary to develop a clutch , females may be physiologically - constrained from modulating clutch numbers, size, or time to development; however, the time of season at which females begin reproducing may not be constrained in the same way. Thus, females may adjust when in the season th ey become reproductive in response to short term variation in available mates. Therefore, Julian dates were calculated to determine the first day on which each female began developing a clutch (score >0 on gravidity scale) as a measure of the beginning of reproduction for each female. Females, regardless of mate availability, started reproduction in early May and stop reproduction at the end of July with c. 1 week worth of variance during this breeding season experiment ( X 2 = 0.736 , df= 1, p >0.05 , Test F=0.209 p >0.05 ; and, X 2 = 0.722, df= 1, p >0.05, p >0.05 , respectively); that is, mate availability in female G. aculeatus does not appear to determine how early or late in the season a female is likely to become reproductive. DISCUSSION Both male and female life histories are shaped by trade - offs between the quality and quantity of current and prospective mates. This experiment investigated whether and how female reproductive investment responds to mate availability, and whether those res ponses are modified in parallel with known patterns of mate choice. The results show that limited plasticity in reproductive investment may keep females from 25 mitigating the effects of relaxed choosiness when mates are rare. Indeed, there was no evidence th at the number of reproductive females, clutches per season (per female), number or mass of eggs in initial clutches, or the timing of reproduction differed between mate availability treatments. When males are rare and time is short (females are approaching the end of their reproductive lifespans), female G. aculeatus relax their mating decisions (Tinghitella et al., 2013). As female mating decisions are relaxed under low mate availability, but reproductive investment appears fairly canalized, sexual selecti on should be relatively weakened under these conditions. Because females deposit an entire clutch in the nest of a single male when spawning, all chosen males, regardless of quality, can potentially receive the same reproductive investment because there is limited plasticity in how females allocate reproductive resources. In particular, if female choice is relaxed, allowing lower quality males to spawn, males of lower quality could enjoy the most reproductive success by avoiding costly male competition (par ticularly at higher - densities) through either 1. nesting in less dense areas, or 2. waiting to nest until other males are providing parental care to eventually acquire a mating opportunity once males become rare. Thus, less - preferred males could reap more benefits simply by being available in the right place (where there are fewer males) and at the right time (when females are ready to spawn). Plasticity in reproductive investment may also be limited by costs like predation risk and incomplete information about search costs. Specifically, spending more days predation (reviewed in Magnhagen, 1991). Because females have incomplete 26 information about the probability of success in continued mate search (Stephens & Krebs, 1986; Real, 1990; Wiegmann et al., 2010), they experience costs associated with risk (Raiffa 1970; Trimmer et al. 2011). Risks should increase with time unmated (as eggs may become unviable), rarity of mates, and pr evalence of predators. Although having developed eggs prepares a female should she encounter a rare mate, the risks of unviable eggs and expansive searches could counter selection for plasticity in increasing the expected duration of gravidity per clutch. Based on observations of short, repeated visits by females to male territories, Dale and Slagsvold (1996) suggested that the number of times a male is encountered on a territory might be an informative signal for females about future mating opportunities, and therefore the risk of continued search (Raiffa, 1970; Trimmer et al., 2011). However, males within this study were not allowed to maintain complete nests in their treatment tanks to prevent females from depositing eggs in them (because decisions were simultaneously being assessed in a parallel experiment using no - choice mating trials with nesting males; Tinghitella et al. 2013). Experience with both males and their nests may be necessary for females to assess and respond to differences in mate availability by adjusting reproductive investments; thus, from the perspective of searching females, although there was a difference in the adult sex ratio between treatments, the operational sex ratio could have been strongly female - biased in both treatments. However, given that the experimental set - up for the study used the same individuals and actually did find an effect of sex ratio treatment on female mating decisions (Tinghitella et. al 2013), it is likely that the experimental design was suff icient to produce an effect on female reproductive investment, if one exists. Similarly, had 27 females within a treatment tank been simultaneously gravid often, the OSRs could have been more variable than the ASRs would suggest. Multiple fully gravid females , however, scarcely occurred within a tank at the same time, thus our sex ratio remained largely consistent for the duration of the experiment. Although numerous examples exist demonstrating female plasticity in investment in response to her environment, including in response to predation (Giesing et al., 2011), conspecific brood parasitism (Lyon, 1998), and attractiveness or parenting ability of males (Sheldon, 2000; Kolm 2001; Stiver & Alonzo, 2009 (review); Kindsvater et al., 2013; Poisbleau et al., 2013, Soma & Okanoya, 2013), there is limited evidence to suggest plasticity in female reproductive investment due to demography. Perhaps the notabl e exception of plasticity in reproductive investment shown by Heubel et al. (2008) is facilitated by relatively extreme within - season variation in sex ratios in common gobies ( P. microps). This variation is substantially more dramatic than that observed in G. aculeatus , and thus the selection pressure favouring the evolution of plastic reproductive investment may simply be weak. The difference in effects between these species may be exaggerated by the manner in which clutches are partitioned between males. Female G. aculeatus must deposit all of their clutch in the nest of a single male, thus they must make investment decisions earlier with uncertainty as to whether any mating will be secured for the clutch under development. P. microps , in contrast, could s ecure a mating with a portion of a clutch, then delay to deposit the remaining eggs with another male. Thus, selection pressures on reproductive investment may differ in these species depending on clutch size, clutch partitioning, and the timing of mate en counters. 28 Likewise, limitations in observed plasticity may also be due to energetic constraints. Despite ample food, females in this study generally had few clutches, although clutch totals were within range of the expected reproductive output across popu lations of G. aculeatus (Wootton 1976). Therefore differences between clutches, particularly in timing, may be difficult to examine. Adding to this the fact that females are only 30% physiologically - efficient in producing eggs (i.e, resources are not very efficiently converted t o eggs; Wootton & Evans, 1976), specific reproductive investment trade - offs and reallocations of energy to other functions would need to be quite large to detectably alter days gravid, clutch sizes, clutch number, or egg number. Female G. aculeatus could also plastically adjust reproductive investment in response to mate availability in other ways than measured in this experiment. For instance, females could alter investment in offspring by varying their quality, not quantity, through increasing cortisol l evels and egg size to influence offspring survival (e.g. Giesing et al., 2011). Egg size was not measured here, but increasing egg size is known to positively affect feeding (Knutsen & Tilseth, 1985), swimming abilities (Ojanguren et al., 1996), and surviv al at both egg and larval stages (Lillelund & Lasker, 1971; Henrich, 1988; Leggett & Deblois, 1994), all of which may translate into fitness advantages in their current environment. Given that G. aculeatus are iteroperous but unlikely to live to a second breeding particularly for females experiencing an excess of mates (male - biased conditions). A more probable strategy might be to modify the final clutch ( e.g. in mass, egg nu mber, egg size, etc.), maximizing the terminal investment (Clutton - Brock, 1984). This 29 experiment did not capture variation in terminal investment, although characteristics of final clutches were not investigated because of the risk involved in extracting c lutches. This study highlights the importance of the interplay between mate availability, female mate choice, and female reproductive investment shaping variation in the strength and direction of sexual selection. Laboratory studies aimed toward estimating investment and behavioural interactions will likely benefit from considering that the demographic effects of mate choice (intersexual selection) and mate competition (intrasexual selection) for both males and females may be tempered by the degree of plast icity in female investment. As Moura & Peixoto (2013) point out in their meta - analysis, responses of females and males may differ from species to species as sex ratios become more male - biased; these responses may be exaggerated in species where males provi de care. Recent evidence in burying beetles shows increased care under increased reproductive competition (Hopwood et al. 2015), but more evidence is necessary to determine the extent to which this pattern exists across species, and whether it is exaggerat ed or ameliorated by other alterations in investment . Thus, future work should expand our understanding of how demography influences reproductive strategies of males, in response to mate competition and female choice, particularly in species where males provide care. 30 APPENDIX 31 AP PENDIX Figure 1.1. Comparison of the number of eggs per initial clutch between female - and male - biased mate availability treatments. The central boxes represent values from the lower to upper quartile (25 to 75 percentile; first and third quantiles), and are intended to give an c. 95% confidence interval for differences in the two datasets. Bold lines represent the median, and extreme values are represented by open circles. 32 Figure 1. 2. Comparison of the number of days spent in a nongravid state between female - and male - biased mate availability treatments. The central boxes represent values from the lower to upper quartile (25 to 75 percentile; first and third quantiles), and are intended to give an c. 95% confidence interval for d ifferences in the two datasets. Bold lines represent the median, and extreme values are represented by open circles. 33 Figure 1.3. Comparison of the interclutch interval length (the number of consecutive days a female is completely no ngravid, i.e neither carrying nor developing a clutch) between female - and male - biased mate availability treatments. The central boxes represent values from the lower to upper quartile (25 to 75 percentile; first and third quantiles), and are intended to give an c. 95% confidence interval for differences in the two datasets. Bold lines represent the median, and extreme values are represented by open circles. 34 LITERATURE CITED 35 LITERATURE CITED Andersson, M., & Simmons, L. W. (2006). Sexual selection and mate choice. Trends in Ecology & Evolution 21(6), 296 - 302. doi:10.1016/j.tree.2006.03.015 Bjelvenmark, J., & Forsgren E. (2003). Effects of mate attraction and male - male competition on paternal care in a goby. Behaviour 140 , 55 - 69. doi:10.1163/ 156853903763999890 Borg, Å. A., Forsgren, E., & Magnhagen, C. (2002). Plastic sex roles in the common goby the effect of nest availability. Oikos 98(1), 105 - 115. doi:10.1034/j.1600 - 0706.2002.980111.x Borg, Å.A., Forsgren, E., & Amundsen, T. (2006). Seasonal change in female choice for male size in the two - spotted goby. Animal Behaviour 72, 763 771.doi:10.1016 /j.anbehav.2005.11.025 Boughman, J. W. (2006) Speciation in sticklebacks. In: The Biology of the Threespine Stickleback, (Ostlund - Nilsson, S., Mayer, I. & Huntingford, F. A., eds.), pp. 83 - 126. London: Taylor and Francis Group. doi: 10.1201/9781420004830.ch3 Boughman, J. W., Rundle, H.D., & Schluter, D. (2005). Parallel evolution of sexual isolation in sticklebacks. Evoluti on 59(2), 361 - 373. doi:10.1554/04 - 153 Carrillo, J., Danielson - Francois, A., Siemann, E., & Meffert, L. (2012). Male - biased sex ratio increases female egg laying and fitness in the housefly, Musca domestica . Journal of Ethology 30(2), 247 - 254. doi:10.1007/s10164 - 011 - 0317 - 6 specific reproductive rates may result from ageing and optimal resource allocation. Journal of Evolutionary Biology 14(1), 180 - 185. doi:10.1046/j.1420 - 9101.2001.00243.x Clutton - Brock, T. H. (1984). Reproductive effort and terminal investment in iteroparous animals. The American Naturalist 123(2), 212 - 229. doi:10.1086/284198 Cooper, I. A., R. T. Gilman & Boughman, J. W. (2011). Sexual dimorphism and s peciation on two ecological coins: patterns from nature and theoretical predictions. Evolution 65(9), 2553 - 2571. doi:10.1111/j.1558 - 5646.2011.01332.x Cotton, S., Small, J., & Pomiankowski, A. (2006). Sexual selection and condition - dependent mate preferenc es. Current Biology 16(17), R755 - R765. doi:10.1016/j.cub.2006.08.022 36 Dale, S., & Slagsvold, T. (1996). Mate choice on multiple cues, decision rules and sampling strategies in female pied flycatchers. Behaviour 133(11 - 12), 903 - 944. doi:10.1163/156853996X00 305 DeWitt, T. J., & Scheiner, S. M. (2004). Phenotypic Plasticity: Functional and Conceptual Approaches. New York; Oxford: Oxford University Press. Emlen, S. T., & Oring, L. W. (1977). Ecology, sexual selection, and the evolution of mating systems. Science 197(4300), 215 - 223. doi:10.1126/science.327542 Flatt, T. (2011). Survival costs of reproduction in Drosophila. Experimental Gerontology 46(5), 369 - 375. doi: 10.1016/j.exger.2010.10.008 Forsgren, E., Kvarnemo, C., & Lindström, K. (1996). Mode of s exual selection determined by resource abundance in two sand goby populations. Evolution 50(2), 646 - 654. doi:10.2307/2410838 Forsgren, E., Amundsen, T., Borg, A. A., & Bjelvenmark, J. (2004). Unusually dynamic sex roles in a fish. Nature 429(6991), 551 - 55 4. doi:10.1038/nature02562 Gadgil, M., & Bossert, W. H. (1970). Life historical consequences of natural selection. The American Naturalist 104(935), 1 - 24. doi:10.1086/282637 Giesing, E.R., Suski, C.D., Warner, R.E., & Bell, A.M. (2011). Female sticklebacks transfer information via eggs: effects of maternal experience with predators on offspring. Proceedings of the Royal Society of London B: Biological Sciences 278, 1753 - 1759. doi:10.1098/rspb.2010.1819 Halliday, T.R. (1978). Sexual selection an d mate choice. In: Behavioural ecology. An evolutionary approach (Krebs, J.R., Davies, N.B., eds.). pp 180 - 213. Oxford: Blackwell Scientific Publications. Hatfield, T. (1997). Genetic divergence in adaptive characters between sympatric species of stickl eback. The American Naturalist 149,1009 - 1029. doi:10.1086 /286036 Hebets, E. A. (2003). Subadult experience influences adult mate choice in an arthropod: Exposed female wolf spiders prefer males of a familiar phenotype. Proceedings of the National Academy of Sciences of the United States of America 100(23), 13390 - 13395. doi:10.1073/pnas.2333262100 Hendry, A.P., Peichel, C.L., Matthews, B., Boughman, J.W. & Nosil, P. (2013). Stickleback research: the now and the next. Evolutionary Ecology Research 15, 111 141. 37 Henrich, S. (1988). Variation in offspring sizes of the poeciliid fish Heterandria formosa in relation to fitness. Oikos 51(1), 13 - 18. doi:10.2307/3565801 Heubel, K. U., Lindström, K., & Kokko, H. (2008). Females increase current reproductive effort when future access to males is uncertain. Biology Letters 4(2), 224 - 227. doi:10.1098/rsbl.2007.0630 Hopwood, P. E., Moore, A. J., Tregenza, T., & Royle, N. J. (2015). Male burying beetles extend, not reduce, parental care duration when reproductive compe tition is high. Journal of Evolutionary Biology. 58, 1169 1175, doi: 10.1111/jeb.12664 Jirotkul, M. (1999). Population density influences male - male competition in guppies. Animal Behaviour 58, 1169 - 1175. doi:10.1006/anbe.1999.1248 Jones, F. C., Brown, C ., Pemberton, J. M., & Braithwaite, V. A. (2006). Reproductive isolation in a threespine stickleback hybrid zone. Journal of Evolutionary Biology 19(5), 1531 - 1544. doi:10.1111/j.1420 - 9101.2006.01122.x Kindsvater, H. K., Simpson, S. E., Rosenthal, G. G., & Alonzo, S. H. (2013). Male diet, female experience, and female size influence maternal investment in swordtails. Behavioral Ecology 24(3), 691 - 697. doi: 10.1093/beheco/ars213 Knutsen, G. M., & Tilseth, S. (1985). Growth, development, and feeding success of atlantic cod larvae Gadus morhua related to egg size. Transactions of the American Fisheries Society 114(4), 507 - 511. doi:10.1577/1548 - 8659(1985)114 <507:GDAFSO>2.0.CO;2 Kokko, H. & Jennions, M.D. (2008). Parental investment, sexual selection and sex ratios. Journal of Evolutionary Biology 21, 919 - 948. doi:10.1111/j.1420 - 9101.2008.01540.x Kokko, H., & Mappes, J. (2005). Sexual selection when fertilization is not guaranteed. Evoluti on 59(9), 1876 - 1885. doi:10.1111/j.0014 - 3820.2005.tb01058.x Kokko, H., & Rankin, D. J. (2006). Lonely hearts or sex in the city? Density - dependent effects in mating systems. Philosophical Transactions of the Royal Society B: Biological Sciences 361(1466), 319 - 334. doi:10.1098/rstb.2005.1784 Kolm, N. (2001). Females produce larger eggs for large males in a paternal mouthbrooding fish. Proceedings of the Royal Society of London B: Biological Sciences 268, 2229 2234. doi:10.1098/rspb.2001.1792 Koons, D. N., Metcalf, C. J. E., & Tuljapurkar, S. (2008). Evolution of delayed reproduction in uncertain environments: A life history perspective. The American Naturalist, 172(6), 797 - 805. 38 Koslow, J. A. (1996). Energetic and life history patterns of deep sea benthic, benthopelagic and seamount associated fish. Journal of Fish Biology, 49(sA), 54 - 74. Lahti, D. C., Johnson, N. A., Ajie, B. C., Otto, S. P., Hendry, A. P., Blumstein, D. T., Coss, R.G., Donohue, K., & Foster, S. A. (2009). Relaxed selection in the wild. T rends in Ecology & Evolution 24(9), 487 - 496. doi:10.1016/j.tree.2009.03.010 Leggett, W. C., & Deblois, E. (1994). Recruitment in marine fishes: Is it regulated by starvation and predation in the egg and larval stages? Netherlands Journal of Sea Research 3 2(2), 119 - 134. doi:10.1016/0077 - 7579(94)90036 - 1 Lillelund K., & Lasker R. (1971). Laboratory studies of predation by marine copepods on fish larvae. Fishery Bulletin 69, 655 667. Lindström, K. (1998). Energetic constraints on mating performance in the sa nd goby. Behavioral Ecology 9(3), 297 - 300. doi:10.1093/beheco/9.3.297 Lyon, B. E. (1998). Optimal clutch size and conspecific brood parasitism. Nature 392, 380 383. doi:10.1038/32878 Magnhagen, C. (1991). Predation risk as a cost of reproduction. Trends in Ecology and Evolution 6, 183 185. doi:10.1016/ 0169 - 5347(91)90210 - O McLennan, D. A. (2006) The umwelt of the threespine stickleback. In: The Biology of the Threespine Stickleback, (Ostlund - Nilsson, S., Mayer, I. & Huntingford, F. A., eds.), pp. 180 - 21 1. London: Taylor and Francis Group. doi: 10.1201/ 9781420004830.ch6 McNamara, J. M., Houston, A. I., Barta, Z., Scheuerlein, A., & Fromhage, L. (2009). Deterioration, death and the evolution of reproductive restraint in late life. Proceedings of the Royal Society B: Biological Sciences 276(1675), 4061 - 4066. doi:10.1098/rspb.2009.0959 McPhail, J. D. (1984). Ecology and evolution of sympatric sticklebacks ( Gasterosteus ): morphological and genetic evidence for a species pair in Enos Lake, British Columb ia. Canadian Journal of Zoology 62,1402 - 1408. doi:10.1139/z84 - 201 McPhail, J. D. (1992). Ecology and evolution of sympatric sticklebacks ( Gasterosteus ): evidence for a species - pair in Paxton Lake, Texada Island, British Columbia. Canadian Journal of Zoolo gy 70, 361 - 369. doi:10.1139/z92 - 054 Milinski, M. & T. C. M. Bakker (1992). Costs influence sequential mate choice in sticklebacks, Gasterosteus aculeatus . Proceedings of the Royal Society of London B: Biological Sciences 250, 229 - 233. doi:10.1098/rspb.199 2.0153 39 Moura, R. R., & Peixoto, P. E. C. (2013). The effect of operational sex ratio on the opportunity for sexual selection: A meta - analysis. Animal Behaviour 86(4), 675 683. doi: 10.1016/j.anbehav.2013.07.002 Mück, I., Wacker, S., Myhre, L. C., & Amund sen, T. (2013). Nest distribution affects behaviour and mating success in a marine fish. Behavioral Ecology and Sociobiology 67(4), 609 - 619. doi:10.1007/s00265 - 013 - 1480 - 2 Ojanguren, A. F., Reyes - Gavilán, F. G., & Braña, F. (1996). Effects of egg size on o ffspring development and fitness in brown trout, Salmo trutta . Aquaculture 147(1), 9 - 20. doi:10.1016/S0044 - 8486(96)01398 - 1 Patterson, C. B., Erckmann, W. J., & Orians, G. H. (1980). An experimental study of parental investment and polygyny in male blackbi rds. The American Naturalist 116(6), 757 - 769. doi:10.1086/283668 Pigliucci, M. (2001). Phenotypic Plasticity: Beyond Nature and Nurture. Baltimore: Johns Hopkins University Press. Poisbleau, M., Dehnhard, N., Demongin, L., Parenteau, C., Quillfeldt, P., & Eens, M. (2013). Females paired with new and heavy mates reduce intra - clutch differences in resource allocation. PloS One 8(8), e72136. doi: 10.1371/ journal.pone.0072136 Raiffa, H. (1970). Decision Analysis. Reading, MS: Addison - Wesley. Real, L. A. ( 1990). Search theory and mate choice. I. Models for single - sex discrimination. American Naturalist 136(3), 376 - 405. doi:10.1086/285103 Reznick, D. N. (1983). The structure of guppy life histories: The tradeoff between growth and reproduction. Ecology 64, 862 873.doi: 10.2307/1937209 Reznick, D. (1985). Costs of reproduction: An evaluation of the empirical evidence. Oikos 44(2), 257 - 267. doi: 10.2307/3544698 Roff, D. A. (1997). Evolutionary Quantitative Genetics. New York: Chapman & Hall. doi: 10.1007/978 - 1 - 4615 - 4080 - 9 Royer, L., & McNeil, J. N. (1993). Male investment in the European corn borer, Ostrinia nubilalis (Lepidoptera: Pyralidae): Impact on female longevity and reproductive performance. Functional Ecology 7(2), 209 - 215. doi:10.2307/23 89889 Rundle, H. D., L. Nagel, & Boughman, J. W. (2000). Natural selection and parallel speciation in sympatric sticklebacks. Science 287, 306 - 307. doi:10.1126/ science.287.5451.306 40 Scharf, I., Peter, F., & Martin, O. Y. (2013). Reproductive trade - offs a nd direct costs for males in arthropods. Evolutionary Biology 40(2), 169 - 184. doi:10.1007/s11692 - 012 - 9213 - 4 Secor, D. H. (2008). Influence of Skipped Spawning and Misspecified Reproductive Schedules on Biological Reference Points in Sustainable Fisheries. Transactions of the American Fisheries Society 137(3): 782 - 789. doi: 10.1577/T07 - 105.1 Sheldon, B. C. (2000). Differential allocation: tests, mechanisms and implications. Trends in Ecology & Evolution 15, 397 - 402. doi:10.1016/S0169 - 5347(00)01953 - 4 Shine , R., Webb, J. K., Webb, A., & Mason, R.T. (2006). Flexible mate choice: a male snake's preference for females is modified by the sizes of females encountered. Animal Behaviour 71(1), 203 - 209. doi:10.1016/j.anbehav.2005.04.005 Simmons, L. W., & 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), 488 - 495. doi:10.1111/j.1420 - 9101.2006.01274. x Simmons, L. W., & Kvarnemo, C. (2006). Costs of breeding and their effects on the direction of sexual selection. Proceedings of the Royal Society B: Biological Sciences 273, 465 470. doi:10.1098/rspb.2005.3309 Smith, J. M. (1977). Parental investment: A prospect ive analysis. Animal Behaviour 25, 1 - 9. doi:10.1016/0003 - 3472(77)90062 - 8 Smith, C. & Reichard, M. (2005). Females solicit sneakers to improve fertilization success in the bitterling fish ( Rhodeus sericeus ). Proceedings of the Royal Society B: B iological Sciences 272, 1683 1688. doi:10.1098/rspb.2005.3140 Soma, M., & Okanoya, K. (2013). Differential allocation in relation to mate song quality in the Bengalese finch. Behaviour 150(13), 1491 - 1508. doi: 10.1163/1568539X - 00003106 Stearns, S. C. (19 92). The Evolution of Life Histories. Oxford; New York: Oxford University Press. Stearns, S. C. (2000). Life history evolution: Successes, limitations, and prospects. Die Naturwissenschaften 87(11), 476 - 486. doi:10.1007/s001140050763 Stephens, D. W. & Krebs, J. R. (1987). Foraging Theory. Princeton, NJ: Princeton University Press. 41 Stiver, K. A., & Alonzo, S. H. (2009). Parental and mating effort: Is there necessarily a trade - off?: Invited review. Ethology 115(12), 1491 - 1508. doi:10.1111/j.1439 - 0310.200 9.01707.x Tatar, M., & Carey, J. R. (1995). Nutrition mediates reproductive trade - offs with age - specific mortality in the beetle Callosobruchus maculatus . Ecology 76(7), 2066 - 2073. doi: 10.2307/1941681 Taylor, E. B., Boughman, J.W., Groenenboom, M. Sniat ynksi, M. & Schluter, D. (2006). Speciation in reverse: Morphological and genetic evidence of the collapse of a threespine stickleback ( Gasterosteus aculeatus ) species pair. Molecular Ecology 15, 343 - 355. doi:10.1111/j.1365 - 294X.2005.02794.x Tinghitella, R.M., Weigel, E., Head, M., & Boughman, J.W. 2013. Flexible mate choice when mates are rare and time is short. Ecology and Evolution 3(9), 2820 - 2831. doi:10.1002/ece3.666 Trimmer, P.C., Houston, A.I., Marshall, J.A.R., Mendl, M.T., Paul, E.S. & McNamara, J.M. (2011). Decision - making under uncertainty: biases and Bayesians. Animal Cognition 14, 465 - - 76. doi:10.1007/s10071 - 011 - 0387 - 4 Trivers, R.L. (1972). Parental investment and sexual selection. In: Sexual Selection and the Descent of Man, 1871 - 19 71, (Campbell B., ed.), pp. 136 - 179. Chicago: Aldine Publishing Company. Wacker, S., Mobley, K., Forsgren, E., Myhre, L. C., Jong, K., & Amundsen, T. (2013). Operational sex ratio but not density affects sexual selection in a fish. Evolution, 67(7), 1937 - 1949. doi:10.1111/evo.12092 Wacker, S., Amundsen, T., Forsgren, E., & Mobley, K. B. (2014). Within season variation in sexual selection in a fish with dynamic sex roles. Molecular Ecology, 23(14), 3587 - 3599. doi:10.1111/mec.12826 Wedell, N., Gage, M. J. G. & Parker, G. A. (2002). Sperm competition, male prudence and sperm - limited females. Trends in Ecology & Evolution 17(7), 313 - 320. doi:10.1016/S0169 - 5347(02)02533 - 8 Westneat, D. F. (1988). Male parental care and extrapair copulations in the indigo bun ting. The Auk 105(1), 149 - 160. Wiegmann, D. D., Weinersmith, K.L., & Seubert, S.M. (2010). Multi - attribute mate choice decisions and uncertainty in the decision process: a generalized sequential search strategy. Journal of Mathematical Biology 60(4), 543 - 572. doi:10.1007/s00285 - 009 - 0274 - 7 42 Wootton, R. J. (1976). The Biology of the Sticklebacks. New York; London: Academic Press Inc. Wootton, R. J., & Evans, G. W. (1976). Cost of egg production in the three spined stickleback ( Gasterosteus aculeatus L.). Journal of Fish Biology, 8(5), 385 - 395. doi:10.1111/j.1095 - 8649.1976.tb03967.x Zwaan, B., Bijlsma, R., & Hoekstra, R. F. (1995). Direct selection on life span in Drosophila melanogaster . Evolution 49(4), 649 - 659. doi: 10.2307/2410820 Electronic Res ources Bates, D., Maechler, M., Bolker, B. & Walker, S. (2013). lme4: Linear mixed - effects models using Eigen and S4. R package version 1.0 - 5. Available at http://CRAN.R - project.org/package=lme4 (last accessed 6 September 2013). Winter, B. (2013). Linear models and linear mixed effects models in R with linguistic applications. arXiv:1308.5499. Available at http://arxiv.org/pdf/1308.5499.pdf (last accessed 6 September 2013). 43 CHAPTER TWO Doing the right thing, in the right amount, at the right time: Traits that predict courtship success vary among courtship stages ABSTRACT To prepare for reproduction, males can allocate resources into several traits including morphological display, courtship behavior, and external structures built or acquired f or mating. However, the value and variation in male traits can depend heavily on how these traits affect mating success and when during courtship traits are assessed by females. Here we show that, although male threespine stickleback ( Gasterosteus aculeatu s ) may display a variety of traits, only some of these are directly associated with mating success. Females from this population appear to assess and prefer high male condition only early in courtship, whereas late in courtship, females strongly prefer mor e vigorous, display - oriented courtship behavior, even more than they prefer the characteristic red nuptial coloration (found to be unrelated to male condition in this population). Our study suggests that the relative importance of several male traits upon which females choose is related not only to whether those traits indicate male quality, but also when during the courtship process traits are assessed. INTRODUCTION When females assess potential mates, they often have a variety of traits to consider (Can dolin 2003; Hebets and Papaj 2005). When and how to evaluate these traits in courtship is then a challenge, as traits may or may not convey useful information about a mate and may not be accessible at all stages of courtship (Uy and 44 Safran 2013). Thus, to better understand how male traits evolve, we need to consider their information - signaling utility, as well as when in courtship females may select them. Exaggerated male traits preferred by females can often result from competition for mates (Kirkpatrick and Ryan 1991; Andersson 1994 ). These male traits can convey Rowe and Houle 1996; Cotton et al. 2004; Tomkins et al. 2004 ), to which a female attends in order to increase her fitness and that of her offspring ( Iwasa et al. 1991 ; Iwasa and Pomiankowski 1994 ; Iwasa and Pomiankowski 1999 ; Houle and Kondrashov 2002 ). Males of good condition can signal their quality through enlarged, elaborated, and/or elongated versions of traits which females prefer (rev iewed in Andersson 1994), whereas males of poor condition may be Pomiankowski 1987; Iwasa et al. 1991; Iwasa and Pomiankowski 1994; Iwasa and Pomiankowski 1999; Cotton et al. 2004) . The relative differences in tra it expression between high and low - condition males may thus provide females with a direct and reliable indicator of male quality (McLean et al. 2012 and references therein). Although many species display mate preferences for such condition - dependent traits (reviewed in Møller and Alatalo 1999 ), not all mate choice involves condition - dependent traits, and may instead have mate choice dependent on traits which indirectly signal male quality (reviewed in Andersson 1994). To produce offspring, males may alloca te resources to the expression of several different traits, including morphological display traits to attract females, courtship behaviors to entice them to mate, and structures for mating or raising young. For example, sexually - selected morphological dis play traits, such as tail length in barn swallows ( Hirundo rustica) ( Møller and Nielsen 1997) or coloration in guppies ( Endler 45 1980) , are conspicuous, costly traits thought to signal quality to females (Andersson 1994 ; Grether 1997; Zuk and Kolluru 1998 ; a s reviewed in Kotiaho 2001) . Likewise, courtship behaviors, such as the auditory calls of frogs, lizards, and insects, can function in the location and evaluation of mates (Ryan et al. 1981 ; Ryan et al. 1982; Kotiaho et al. 1998 ; Mappes et al. 1996; Hoback and Wagner 1997; Saino et al. 1997 ; Bailey and Haythornthwaite 1998; Reinhold et al. 1998 ; Kotiaho 2000; Ophir et al. 2010 ; Voituron et al. 2012) . Finally, males may choose to allocate resources to the acquisition, 2012), such as burrows and bowers (Borgia 1985; Kim et al. 2007 ), or in structures used to care for y oung, such as nests ( Hill et al. 2006 ; Schaedelin and Taborsky 2006) , should a successful mating occur. Moreover, females may not assess traits simultaneously, but sequentially as courtship proceeds (reviewed in Uy and Safran 2013). Males may need to init ially attract a female, and the traits used may differ by the distance at which signals are being assessed (Borgia 1995; Candolin 2003; Hebets and Papaj 2005). For example, nd 2005). However, the traits that function in initial advertisement may differ from the courtship and display traits that convince a female to mate or traits that may s ignal a - courtship, such as nest structure (Endler 1992; Bradbury and Vehrencamp 1998). Because females often make mating decisions using male traits, it is useful to assess whether the traits assessed 1) convey different or red undant information about the male (Partan et al. 2005; Uetz et al. 2009; Leonard and Hedrick 46 2010), 2) are restricted to a particular proximity or stage in courtship at which to convey information for evaluation (Hebets and Papaj 2005), or 3) whether the t raits function to communicate quality during courtship. Whether traits are condition - dependent or not, understanding how and at what stage in courtship traits are ass essed can provide insight into potential function of male traits in securing matings. We examined different male traits and their potential utility during courtship in Cranby Lake threespine sticklebacks ( Gasterosteus aculeatus ). Cranby sticklebacks, in c ontrast to other stickleback populations, are relatively muted in the characteristic red throat coloration ( Albert et al. 2007 ) and are phenotypically highly variable ( Robinson 2000 ). The high amount of variation in male traits and decreased redness in thi s population is useful in determining the relative importance of various male traits in mating which have been largely ignored compared to red throat coloration. Although Cranby sticklebacks are relatively understudied with respect to female preferences co mpared to other populations, sticklebacks generally represent a well - known system for the study of sexual selection (Rowland 1988; Bakker and Milinski 1993; Kraak et al. 1999; Candolin 1999; Tinghitella et al. 2013) . Because of their well - documented morpho logy, courtship, and nesting behavior, sticklebacks offer the opportunity to study how females evaluate many different male traits throughout courtship. Male sticklebacks establish territories on which to build nests, court females, and solely raise young (Wootton 1976) . Females from many populations generally prefer larger and redder males, and male color is often, but not always, positively correlated with physical condition ( Wootton 1976 ; Milinski and Bakker 1990 ; Bakker and Milinski 47 1993 , Bakker and Mun dwiler, 1994; Candolin, 1999; Boughman 2007) . Although populations vary in how they court (Wootton 1976 and references therein; Ridgway and McPhail, 1987, Foster 1995; Scott and Foster 2007), less is known about courtship in Cranby sticklebacks specifically, and more broadly, about how and when traits in behavior, morphology, and nests are used in stickleback courtship. Evidence suggests that males across populations experience strong sexual selection through female choice, which we will address here, although we acknowledge that other sources of selection may also act on male reproductive traits, including competition with other males or demands of sole parental care ( Wootton and Wootton 1984) . To understand which male traits females prefer, and how those preferences may change as courtship progresses, we first confirm the relationship between red throat coloration (the characteristic sexual signal) and male condition in this population. We then ask how other morphological, behavioral, and nestin g traits depend on condition and color. Finally, we address which traits are important both in initiating female responsiveness (early courtship) and nest inspections (late courtship). These two stages of courtship serve as early - and late - must pass on the way to a successful mating. We predict that conspicuous traits (e.g., color, size, and vigorous courtship) should be involved early in courtship to attract females; however, nest characteristics should be important only in late courtship, as nests can be more closely inspected at the stage. Alternatively, traits important early in courtship need not wane in importance as courtship proceeds. Because males typically possess multiple traits, identifying which traits c ommunicate information about the male, and when during courtship they are 48 assessed, can provide clues as to their function and relative importance in securing matings. MATERIALS AND METHODS Collection of wild fish and laboratory conditions We collected reproductive sticklebacks ( Gasterosteus aculeatus ) from Cranby Lake, British Columbia at the beginning of the 2012 summer breeding season and transported them to Michigan State University in East Lansing, MI. Once in the laboratory, males and females were placed into separate, visually isolated, 75 - gallon (~284 L) holding tanks. Tanks were maintained under summer conditions of 14 - hour day lengths at 18°C, and fish were fed defrosted brine shrimp ( Artemia sp.) and bloodworms ( Chironomus spp.) once daily thr oughout the breeding season. Study design At two - week intervals throughout the summer 2012 breeding season, 20 males were removed from the male holding tank and placed into individual, 29 - gallon tanks and allowed to nest. Each individual tank included a 9 00 - mL tray of sand and 12 g wet - weight Chara spp. (nest - building materials), a ceramic pot and plastic plants (for cover), and external, opaque covers (to visually isolate the tanks). Each male was exposed daily for 10 minutes to a free - swimming, gravid f emale to encourage nest building. Males were allotted a maximum of 14 days to complete a nest, and males who completed nests (N=85) then underwent courtship trials. No - choice mating trials are standard for measuring mating preferences in sticklebacks ( Nagel and Schluter 1998 ) and allow males and females to interact all the way to spawning. With this method, we measured female preference at various stages in 49 courtship, as well as the strength of those preferences ( Wagner 1998 ), while excluding male - male competition. following a 5 - minute acclimation period within an opaque holding container. Trials began when the male and female first saw each other and lasted 20 minutes, or u ntil each male and was removed from a tank at the completion of the trial. All courtship behaviors during the trials were documented using an event recorder (Observer: N oldus Technologies, Wageningen, The Netherlands). Measuring male traits and their impact on mating success Morphological traits, such as body size and male nuptial color, are two traits that may be important in mate preferences, as seen in some sticklebac k species ( Head et al. 2013 immediately following courtship trials using a standardized color index scale developed by our laboratory group ( Boughman 2001; Boughman 2007; Lewand owski and Boughman 2008) . We then obtained an average color index for the amount of area and intensity of redness for each male. Post - trial, males were immediately weighed (to the nearest 0.01 gram) and photographed in side view using a Canon G - 15 digital camera. A second observer placed digital landmarks on the photographs of the males and calculated male centroid size (size) according to established methods (Taylor et al. - t rial as his residual from the regression of male weight on male size. 50 Male courtship behavior is also comprised of many traits expressed in a complex, conspicuous courtship sequence. Initially, males attract a female via either display (zig - - pricking, and biting) courtship behaviors. Then males lead females back to the nest (lead), whereupon the male shows the female the nest opening (show). Females may then examine, enter, and deposit eggs within the nest of a preferred male; however, females may terminate courtship with non - preferred males at various points during the courtship sequence (for further descriptions of these behaviors, see Rowland 1994 ). To quantify male behavior, we measured the presence an d duration of the above courtship behaviors. To correct for trial duration, p er - minute rates for all measured behaviors were calculated by dividing by the duration of the trial. We then calculated rt, nature of courtship, and courtship vigor. We calculated latency to court as the difference between the start of the trial and the first behavior conducted. We summarized the nature of courtship by first categorizing courtship behaviors as either displa y - oriented (zig - zags, leads) or aggressive (bites, chases, and dorsal - pricks; as in Kozak et al. 2009) . The nature of courtship, an inverse relationship between aggressive and display courtship, was then calculated as proportion of courtship which was disp lay - oriented in nature, where 0 would indicate all courtship was aggressive and 1 all display. To estimate courtship vigor, we summed all courtship behaviors conducted by a male, following Kozak et al. ( 2009 ). Male sticklebacks build and maintain nests in which young are raised; the nest not only serves as a structure conducive to oxygenating embryos, but potentially also an 51 - veg etation and sand bound together by a glue (spiggin) secreted by the male, are the structures through which females pass to deposit eggs to be fertilized and later solely reared by the male (for more detail, see Rowland 1994; Wootton 1976 ). Because of the c lear importance of nests in successful fertilization and raising of young, we recorded several nest characteristics that indicate the time and resources invested in constructing and maintaining nests. Nest measurements included the number of days from beg as in Barber et al. 2001 ) and nesting behaviors (fanning, creeping through, boring) conducted while within one body length of the nest during the courtship trial (as in McKinnon et al. 2012) . W e recorded ne sting behavior within courtship trials as the number of visits a male made to his nest (nesting bouts). Because nests are variable, require energy (both in producing spiggin and gathering materials to construct nests), and have been suggested as indicators of male quality (Rushbrook et al. 2008), we also measured nest weights. To obtain weights, nesting trays were removed and dried, whereupon dry nests were separated from nesting trays and weighed to the nearest 0.01 gram. A second observer who was blind to male identity took nest weight measurements. S tatistical a nalysis We considered how male redness and condition are related both to one another and to three categories of potential resource allocation (morphological display traits, courtship behavior, nests and nesting behavior). Significant factors from these models were then combined with evidence from the literature to develop an overall model of 52 male traits that predict success at different stages of courtship with a female throughout the breeding season. s willingness to mate and on ultimately carrying out mating behaviors, we used rates of responsiveness (female receptiveness to mating; the number of times a female followed a male when he ference for this male; the number of times a male showed a female the nest that resulted in a female ( 2009) . Because females are known under certain conditions to relax choosiness late in the breeding season (Tinghitella et al. 2013), all models also included a time variable (biweekly period in which the courtship trial occurred); we refer to this variable as Analyses of linear models were conducted in R (R Core Team, 2013, version 3.0.2) using the packages car ( Fox and Weisberg 2010) , AER ( Kleiber and Zeileis 2008) and rsm (Lenth 2009 ). We visually inspected plots to detect deviations from homoscedascity and normality, and multicolinearity of variables was investigated through Variance Inflation Factor (VIF) of each model. A VIF above 4 was used as the threshold reference value; no tests indicated a VIF above these levels, therefore the redundancies between predictor variables were judged as acceptably low enough to justify their inclusion as separate predictor variables. Models included both predictor variables and their interactions, and models were reduced stepwise by AIC to provide the best model. Note that the c first analyzed using a generalized linear model with a Poisson link; however, similar results were achieved 53 with a better model fit (as determined by AIC) using a linear model with Gaussian distribution. To avoid over - fitting the model, select two - way int eraction terms were added to the models based on hypothesized relationships between main effects. disproportionately affecting estimates. No more than two outliers were ever fou nd in any model, and running the models with and without the outliers did not change the significance of effects, so all data points were included in our analyses. Finally, to visualize i nteractions, graphs were generated following the methods described in Aiken and West (1991, p. 12) . In short, to illustrate significant interaction effects, separate regression lines are computed and plotted on an X - Y graph by dividing individuals into two groups; one 1.5 SD above and one 1.5 SD below the mean for a given p redictor variable Z. RESULTS Is red throat coloration related to male condition? Stickleback males of many populations use condition - dependent red nuptial coloration to attract mates. However, redness in this population from Cranby Lake was not predicted by male condition nor size, but only by seasonal time (slope ± SE = 0.46 ± 0.136, t 3,80 = - 3.37, p= 0.0012, R 2 = 0.12). Specifically , males were redder earlier in the season . Cour tship b ehavior Males varying in color and condition do not differ in courtship vigor, but do court females differently (nature of courtship and latency to court). Both color and condition 54 affect the latency to court (Table 1 .1 ), and these variables interact in their effects (Figure 2. 1). More colorful males are faster to court when larger, but dull males are faster to court when smaller. Males in better condition perform more display - oriented courtship (Table 1. 1). Neither seas onal time, male size, male condition, male color, nor interactions between these variables were found to influence how vigorously males court females (all t 6,77 < - 1.17, p>0.24, N=85). Nests and nesting b ehavior High - condition males have heavier nests (17.4 ± 6.57, t 6,74 = 2.65, p= 0.0097, R 2 = 0.077, N=85), and the interaction between condition and color also significantly predicted nest weight (6.7 ± 2.33, t 6,74 = 2.88, p= 0.0052, R 2 = 0.090, N=85 ; Figure 2. 2). Neither seasonal time, male size, male condition, male color, nor their interactions predicted days until nest (All t 6,73 < - 1.37, p>0.17, N=85) or number of nest bouts during courtship (All t 6,77 < - 1.66, p>0.10, N=85). Which traits predict courtship success? For models of female responsiveness (early courtship) and inspection rate (late courtship), the nature of courtship and courtship vigor appear to be the strongest factors predicting success with females, and females appear to be evaluating males on many ch aracteristics early in courtship, but only a few factors as courtship proceeds (vigor and the nature of courtship; see Table 1. 2 and Figure 2. 3). DISCUSSION Male mating traits Our results suggest that several traits are involved in mate selection, but not all preferred traits depend on condition. Both the nature of courtship and vigor were 55 not how vigorously he courted. Specifically, males in better condition were not more vigorous, but did perform the preferred display - oriented courtship behaviors more often. Display - oriented courtship may therefore provide females with some information on the rait throughout courtship, consistent with a direct benefits model of sexual selection. Contrary to many populations, male Cranby sticklebacks do not have condition - dependent red throats, nor do females have a preference for redness in courtship. In gener al, Cranby sticklebacks are not as red ( Albert et al. 2007; Boughman 2007 ), likely due to a combination of both diet (Schluter and McPhail 1992; Frischknecht 1993 ) and unfavorable light conditions for red expression in these lakes ( Albert et al. 2007 ). Bec ause the water in Cranby Lake is more red - shifted, red mating signals do not contrast well against the water and are less visible because females have low perceptual sensitivity to red ( Boughman 2001 ). Thus, the expression of red coloration may not be as effective in ensuring mating success. This aligns with predictions of both condition - dependent models, where an indicator trait should be both informative and preferred by females to evolve, and also sensory drive, which suggests biased sensory perception drives male trait evolution (reviewed in Andersson 1994). If the trait is not easily detected, both models are applicable and suggest the trait should not be used in mate selection. Furthermore, if pr imarily female choice acts to maintain the relationship between redness and condition, then the lack of relationship here may help to explain the variability of color preferences across stickleback populations (Boughman 2001) and their reliability as a con dition - dependent trait ( Wootton 1976 ; Milinski and Bakker 1990 ; 56 Bakker and Milinski 1993 , Bakker and Mundwiler 1994 ; Boughman 2007; Candolin 1999 ) . Other signals, such as courtship behaviors, may be more reliable indicators of quality. Females should use t hese quality - indicative traits to assess males for the fitness - increasing benefits (direct or indirect) they provide to their offspring. Here, females prefer vigor and display - oriented courtship. The reliance on behavior may also be due to the fact that be condition, rather than the past condition reflected in more fixed traits, such as size (Sullivan 1994). Even where behavior does not directly represent male quality, behaviors likely also encode some information about the male (such as age or courtship experience; Weigel, Chapter 4) or enhance the detection of his other signals ( Hailman 1977 ; Smith 1977 ; Endler 1992 ; Rowland et al. 1995) , which helps to explain why mating success based on behavior is so common across species ( Gibson and Bradbury 1985 ; Andersson 1991 ). More broadly, as suggested by Candolin (2003 ), and supported in a variety of species (Zuk et al., 1990 ; Houde 1997 ; Møller et al. 1998 ), this reliance on multiple traits could mean the se lection of higher - quality males in female mate choice ( Møller and Pomiankowski 1993 ). condition or offer a direct advantage in mating (see Table 2). Although some extended phenotypes, such as nests, have been suggested to signal information useful in mate assessment ( Barber et al. 2001 ), other selective pressures, such as those from parental care or survival, may be more responsible for the variation seen in these traits. F uture studies are needed to clearly establish whether extended phenotypes reflect condition 57 and whether they come at a cost to, or enhance other aspects of, courtship, survival, and parental care. Stages of c ourtship Our results suggest that some traits may function early in courtship to initially attract a female, but other traits are more critical at later stages of courtship and potentially in raising young. Females responded to only a few traits and grew more discr iminating as courtship proceeded. Early in courtship, females responded to condition, the nature of courtship, and vigor (plus interactions), but late in courtship they only responded strongly to the nature of courtship and vigor. Because these behaviors are likely assessed at the same distance from the male and for similar duration, assessment based on these different aspects of courtship behavior suggests redundancy in these signals (Partan et al. 2005; Uetz et al. 2009; Leonard and Hedrick 2010) rather than differing propagation properties of signals in our system (Hebets and Papaj 2005). However, neither explanation rules out the possibility that the information garnered from early signals could simply have been used and no longer aid in further narrowi ng the pool of potential mates, or that females may also need more time to evaluate behavior than is required for other traits, such as size (Sullivan 1994). Two consequences of sequentially evaluating traits could be either to switch traits as more useful information becomes available (perhaps at a closer range or because the male starts a new type of signal), or to appear more discriminating if signals lose their information - providing utility as courtship proceeds. Our work is consistent with findings in several species where early, typically long - range courtship is governed by different traits than is later, short - range courtship. These 58 studies include dark - eyed juncos and field crickets ( Gryllus lineaticeps ), where many male traits are assessed differently during long - and short - range courtship (reviewed in Reichard et al. 2011), and squirrel tree frogs and house crickets, whose courtship relies on acoustic signals early, but later switches to visual signals (Taylor et al. 2007; Stoffer and Walker 2012). As suggested by Uy and Safran (2013), the sensory environment is which it can be assessed; our results add that the duratio n of expression (particularly important for behavior) may play a role in signal use. Many male traits may not be preferred by females directly, but lead to more mating opportunities because of their function in male competition. For example, quickly beginn ing to court may help males gain access to females when in the presence of rivals (i.e. in the face of male competition), which is typical in the wild but not captured in the no - choice design of our study ( Rowland 1988 ). s attention, quickly initiating courtship to avoid interruption can be a mechanism by which courtship success is achieved. Because male competition can change how females select mates ( Trail, 1985 ; Zimmerer and Kallman 1988 ; Greenfield 1994 ; Morris et al. 1995 ; Galeotti et al. 1997; Petersson et al., 1999 ), future studies should consider the role of courtship latency in gaining courtship opportunities under male - male competition. Our study demonstrates that females assess many male traits to differing degr ees as courtship progresses. Behavior, not the red throat coloration characteristic of many stickleback populations, appears to be the strongest target of selection by females throughout the courtship process. By studying many male traits and the stages 59 in courtship at which they may be assessed, we can gain insight into the functional importance of traits. 60 APPENDIX 61 APPENDIX Table 1. 1. Significant predictors of two metrics of courtship behavior from multiple regression models ( df =6,77). Slope ± SE T p Partial R 2 Latency to Court Condition 5.18 ± 1.110 4.66 0.000013 0.19 Color 1.78 ± 0.880 2.04 0.045 0.036 Color: Size 0.028 ± 0.0128 - 2.19 0.031 0.042 Nature of Courtship Condition - 0.013 ± 0.00 570 - 2.22 0.030 0.059 62 Table 1. 2. Male traits that predict female responsiveness and inspection rate. Multiple regression models with estimates and SE scaled by 100 ( df =68). Significant findings are bolded. Responsiveness (Early in Courtship) Inspection Rate (Late in Courtship) Slope ± SE t p Partial R 2 Slope ± SE t p Partial R 2 Morphology Size 0.08 ± 0.043 1.8 0.080 0.013 0.041 ± 0.047 0.9 0.38 0.0029 Condition 7.2 ± 2.30 3.1 0.0028 0.039 4.80 ± 2.49 1.9 0.058 0.014 Color 2.4 ± 1.27 1.7 0.097 0.011 2.07 ± 1.37 1.5 0.14 0.0086 Courtship Behavior Vigor 37.4 ± 13.50 2.7 0.0075 0.031 33.68 ± 14.70 2.3 0.025 0.02 Latency to Court - 0.002 ± 0.0015 - 1.5 0.150 0.009 0.00034 ± 0.0016 0.2 0.83 0.00018 Nature of Courtship 8330 ± 1060 7.8 <0.001 0.250 7430 ± 1150 6.4 <0.001 0.16 Nest Characteristics Nest weight - 0.04 ± 0.026 - 1.4 0.170 0.008 - 0.016 ± 0.029 - 0.6 0.58 0.0012 # Nest Bouts - 0.55 ± 0.93 - 0.6 0.560 0.0014 - 0.270 ± 1.01 - 0.3 0.78 0.00027 Interactions Color:Size - 0.03 ± 0.018 - 1.7 0.089 0.012 - 0.03 ± 0.02 - 1.5 0.14 0.0087 Nature of Courtship:Vigor - 35820 ± 18300 - 2.0 0.054 0.016 - 2215 ± 19700 - 0.11 0.91 0.000048 Vigor:Condition - 87 ± 35.9 - 2.4 0.018 0.024 - 72.24 ± 38.80 - 1.9 0.067 0.013 Seasonal Time 0.15 ± 0.19 0.76 0.45 0.002 0.20 ± 0.21 0.95 0.35 0.00 63 Figure 2.1. More colorful males are faster to court when larger, but dull males are faster to court when smaller. Regression lines for male color represent categories of high and low coloration at 1.5 standard deviations above and below the mean (respectively). 64 Fig ure 2. 2. As condition improves, low color males have heavier nests, whereas high color males have lighter nests. Regression lines for male color represent categories of high and low coloration at 1.5 standard deviations above and below the mean (respectively). 65 Fig ure 2. 3. Females respond differently to males based on interactions between male condition and vigor: Increased vigor leads to more responsiveness (follows/lead) for low - condition males, but less responsiveness for high - condition males. Regression lines for male condition represent categories of high and low condition at 1.5 standard deviations above and below the mean (respectively). 66 LI TERATURE CITED 67 LITERATURE CITED Aiken LS, West SG (1991) Multiple regression: Testing and interpreting interactions. Sage, Newbury Park, CA. Albert AYK, Millar NP, Schluter D (2007). Character displacement of male nuptial colour in threespine sticklebacks ( Gasterosteus aculeatus ). Biol J Linn Soc 91:37 - 48. Andersson M B, 1994. Sexual selection. Princeton University Press. Andersson S (1991) Bowers on the savanna: Display courts and mate choice in a lekking widowbird. Behav Ecol 2:210 - 218. Bailey NW (2012) Evolutionary models of extended phenotypes. TREE 27:561 - 569. do i: 10.1016/j.tree.2012.05.011. Bailey WJ, Haythornthwaite S (1998) Risks of calling by the field cricket Teleogryllus oceanicus ; potential predation by Australian long - eared bats. J Zool 244:505 - 513. Bakker TCM, Milinski M (1993) The advantages of being red: Sexual selection in the stickleback. Mar Behav Physiol 23:287 - 300. Bakker TCM, Mundwiler B (1994) Female mate choice and male red coloration in a natural three - spined stickleback ( Gasterosteus aculeatus ) population. Behav Ecol 5:74 - 80. Barber I, Nai rn D, Huntingford FA (2001) Nests as ornaments: Revealing construction by male sticklebacks. Behav Ecol 12:390 - 396. doi: 10.1093/beheco/12.4.390. Borgia G (1985) Satin bowerbirds ( Ptilonorhynchus violaceus ): An experimental analysis. Anim Behav 33:266 - 271 . Borgia, G (1995) Complex male display and female choice in the spotted bowerbird: specialized functions for different bower decorations. Anim Behav 49(5): 1291 - 1301. Boughman JW (2007) Condition dependent expression of red color differs between stickleback species J Evolution Biol 20:1577 - 1590. Bradbury JW, Vehrencamp SL (1998) Principles of animal communication. Sinauer, Sunderland 68 Candolin U (1999) The relationship between signal quality and phys ical condition: Is sexual signa ling honest in t he three - spined stickleback? Anim Behav 58:1261 - 1267. Candolin U (2003). The use of multiple cues in mate choice. Biol Rev 78(4): 575 - 595. Cooper IA, Gilman RT, Boughman JW (2011) Sexual dimorphism and speciation on two ecological coins: patterns from na ture and theoretical predictions. Evolution 65:2553 - 2571. doi: 10.1111/j.1558 - 5646.2011.01332.x. Cotton S, Fowler K, Pomiankowski A (2004) Do sexual ornaments demonstrate heightened condition - dependent expression as predicted by the handicap hypothesis? P Roy Soc B - Biol Sci 271:771 - 783. Demary K, Michaelidis CI, Lewis SM (2005) Firefly courtship: Behavioral and morphological predictors of male mating success in Photinus greeni . Ethology 485 - 492. doi: 10.1111/j.1439 - 0310.2005.01176.x. Doucet SM, Montgomer ie R (2003) Multiple sexual ornaments in satin bowerbirds: ultraviolet plumage and bowers signal different aspects of male quality. Behav Ecol 14: 503 - 509. Endler JA (1980) Natural selection on color patterns in Poecilia reticulata . Evolution 34:76 - 91. E ndler JA (1992) Signals, signal conditions, and the direction of evolution. Am Nat 139:S125 S153. Foster SA (1995). Understanding the evolution of behavior in threespine stickleback: the value of geographic variation. Behaviour 132(15): 1107 - 1129. Fox J, Weisberg HS (2010) An R companion to applied regression: Sage Publications. Frischknecht, M (1993) The breeding colouration of male three - spined sticklebacks ( Gasterosteus aculeatus ) as an indicator of energy investment in vigor. Evol Ecol 7(5):439 - 450. Galeotti P, Saino N, Sacchi R, Møller AP (1997) Song correlates with social context, testosterone and body condition in male barn swallows. Anim Behav 53:687 - 700. Gibson RM, Bradbury JW (1985) Sexual selection in lekking sage grouse: Phenotypic correlate s of male mating success. Behav Ecol and Sociobiol 18:117 - 123. Greenfield MD (1994) Cooperation and conflict in the evolution of signal interactions. Annu Rev Ecol Syst 97 - 126. 69 Grether GF (1997) Survival cost of an intrasexually selected ornament in a dam selfly. P Roy Soc B - Biol Sci 264:207 - 210. Hailman JP (1977) Optical signals: Animal communication and light. Indiana University Press, Bloomington, IN Head, ML, Kozak, GM, Boughman, JW (2013). Female mate preferences for male body size and shape promote sexual isolation in threespine sticklebacks. Ecol Evol 3(7): 2183 2196. Hebets EA, Papaj DR (2005) Complex signal function: developing a framework of testable hypotheses. Behav Ecol Sociobiol 57:197 214. Hill PSM, Wells H, Shadley JR (2006) Singing from a constructed burrow: Why vary the shape of the burrow mouth? JOR 15:23 - 29. Hoback WW, E Wagner W (1997) The energetic cost of calling in the variable field cricket, Gryllus lineaticeps . Physiol Entomol 22:286 - 290. Houde AE (1997) Sex, color, and mate c hoice in guppies. Princeton University Press, Princeton, NJ. Houle D, Kondrashov AS (2002) Coevolution of costly mate choice and condition - dependent display of good genes. P Roy Soc B - Biol Sci 269: 97 - 104. Iwasa Y, Pomiankowski A (1994) The evolution of mate preferences for multiple sexual ornaments. Evolution 48: 853 - 867. Iwasa Y, Pomiankowski A (1999) Good parent and good genes models of handicap evolution. J Theor Biol 200: 97 - 109. Iwasa Y, Pomiankowski A, Nee S (1991) The evolution of costly mate p references II. The 'handicap' principle. Evolution 45:1431 - 1442. Kennedy CEJ, Endler JA, Poynton SL, McMinn H (1987) Parasite load predicts mate choice in guppies. Behav Ecol Sociobiol 21:291 - 295. Kim TW, Christy JH, Choe JC (2007) A preference for a sexual signal keeps females safe. PLoS One 2:e422. Kirkpatrick M, Ryan MJ (1991) The evolution of mating preferences and the paradox of the lek. Nature 350:33 - 38. Kleiber C, Zeileis A (2008) Applied econometrics with R. Springer. 70 Korzan WJ, Øverli Ø, Su mmers CH (2006) Future social rank: Forecasting status in the green anole ( Anolis carolinensis ). Acta Ethol 9:48 - 57. Kotiaho J, Alatalo RV, Mappes J, Parri S, Rivero A (1998) Male mating success and risk of predation in a wolf spider: A balance between se xual and natural selection? J Anim Ecol 67:287 - 291. Kotiaho JS (2000). Testing the assumptions of conditional handicap theory: costs and condition dependence of a sexually selected trait. Behav Ecol Sociobiol 48:188 - 194. Kotiaho JS (2001) Costs of sexual traits: A mismatch between theoretical considerations and empirical evidence. Biol Rev 365 - 376. Kozak GM, Reisland M, Boughmann JW (2009) Sex differences in mate recognition and conspecific preference in species with mutual mate choice. Evolution 63:353 - 365. doi: 10.1111/j.1558 - 5646.2008.00564.x. Kraak SBM, Bakker TCM, Mundwiler B (1999) Sexual selection in sticklebacks in the field: correlates of reproductive, mating, and paternal success. Behav Ecol 10 (6): 696 - 706 doi:10.1093/beheco/10.6.696. Leonard AS, Hedrick AV (2010) Long - distance signals influence assessment of close range mating displays in the field cricket, Gryllus integer . Biol J Linn Soc 100:856 865. Lenth RV (2009) Response - Surface Methods in R, using rsm. J Stat SoftW 32:1 - 17. Lewandows ki E, Boughman JW (2008) Effects of genetics and light environment on colour expression in threespine sticklebacks. Biol J Linn Soc 94:663 - 673. Loyau A, Saint Jalme M, Cagniant C, Sorci G (2005) Multiple sexual advertisements honestly reflect health status in peacocks ( Pavo cristatus ). Behavioral Ecology and Sociobiology 58:552 - 557 Mappes J, Alatalo RV, Kotiaho J, Parri S (1996) Viability costs of condition - dependent sexual male display in a drumming wolf spider. P Roy Soc B - Biol Sci 263:785 - 789. Mc Lean MJ, Bishop PJ, Nakagawa S (2012). Male quality, signal reliability and female choice: assessing the expectations of inter - sexual selection. J Evol Biol 25(8):1513 - 1520. McKinnon JS, Hamele N, Frey N, Chou J, McAleavey L, Greene J, Paulson W (2012) Ma le choice in the stream - anadromous stickleback complex. PLoS One 7. 71 Milinski M, Bakker TCM (1990) Female sticklebacks use male coloration in mate choice and hence avoid parasitized males. Nature 344:330 - 333. Møller AP, Pomiankowski A (1993) Why have bir ds got multiple sexual ornaments? Behav Ecol Sociobiol 32: 167 - 176. Møller AP, Nielsen JT (1997) Differential predation cost of a secondary sexual character: Sparrowhawk predation on barn swallows. Anim Behav 54:1545 - 1551. Møller AP, Saino N, Taramino G, Ferrario S, Galeotti P (1998) Paternity and multiple signaling: effects of a secondary sexual character and song on paternity in the barn swallow. Am Nat 151:236 - 242. Møller AP, Alatalo RV (1999) Good - genes effects in sexual selection. P Roy Soc B - Biol S ci 266:85 - 91. Morrell LJ, Hentley WT, Wickens VJ, Wickens JB, Rodgers GM (2012) Artificial enhancement of an extended phenotype signal increases investment in courtship by three - spined sticklebacks. Anim Behav 84:93 - 101. doi: 10.1016/j.anbehav.2012.04.013 . Morris MR, Mussel M, Ryan MJ (1995) Vertical bars on male Xiphophorus multilineatus : a signal that deters rival males and attracts females. Behav Ecol 6:274 - 279. perfor mance indicates survival in guppies. Funct Ecol 24:1045 - 1053. Ophir AG, Schrader SB, Gillooly JF (2010) Energetic cost of calling: General constraints and species - specific differences. J Evolution Biol 23:1564 - 1569. doi: 10.1111/j.1420 - 9101.2010.02005.x. Ord TJ, Stamps JA. 2008. Alert signals enhance animal communication in 18835. Ormond CI, Rosenfeld JS, Taylor EB (2011) Environmental determinants of threespine stickleback species pair evolution and persistence. Canadian Journal of Fisheries and Aquatic Sciences 68:1983 - 1997. Partan S, Yelda S, Price V, Shimizu T (2005) Female pigeons, Columba livia, respond to multisensory audio/video playbacks of male courtship behaviour. Anim Behav 70: 957 966 Petersson E, Järvi T, Olsén H, Mayer I, Hedenskog M (1999) Male male competition and female choice in brown trout. Anim Behav 57:777 - 783. 72 Pomiankowski A (1987) Sexual selection: The handicap principle does work -- sometimes. P Roy Soc B - Biol Sci 231:123 - 1 45. Reichard, DG, Rice, RJ, Vanderbilt, CC, Ketterson, ED (2011). Deciphering information encoded in birdsong: male songbirds with fertile mates respond most strongly to complex, low - amplitude songs used in courtship. Am Nat 178(4): 478 - 48. Reinhold K, G reenfield MD, Jang Y, Broce A (1998). Energetic cost of sexual attractiveness: ultrasonic advertisement in wax moths. Anim Behav 55:905 - 913. Ridgway MS, McPhail JD (1987) Rival male effects on courtship behavior in the Enos Lake species pair of threespine sticklebacks (Gasterosteus). Can. J. Zool 65:1951 1955. Robinson BW (2000) Trade offs in habitat - specific foraging efficiency and the nascent adaptive divergence of sticklebacks in lakes. Behaviour 137:865 - 888. doi: 10.1163/156853900502501. Rowe L, Houl e D (1996) The lek paradox and the capture of genetic variance by condition dependent traits. P Roy Soc B - Biol Sci 263: 1415 - 1421. Rowland WJ (1988) Aggression versus courtship in threespine sticklebacks and the role of habituation to neighbours. Anim Beh av 36:348 - 357. Rowland WJ (1994) Proximate determinants of stickleback behaviour: An evolutionary perspective. In: Bell MA, Foster SA, editors. The Evolutionary Biology of the Threespine Stickleback. Oxford University Press pp 297 - 344. Rowland WJ, Bolyard KJ, Jenkins JJ, Fowler J (1995) Video playback experiments on stickleback mate choice: Female motivation and attentiveness to male colour cues. Anim Behav 49:1559 - 1567. Rushbrook B.J, Dingemanse NJ, Barber, I (2008) Repeatability in nest construct ion by male three - spined sticklebacks. Anim Behav 75(2): 547 - 553. Ryan MJ, Tuttle MD, Rand AS (1982) Bat predation and sexual advertisement in a neotropical anuran. Am Nat 119: 136 - 139. Ryan MJ, Tuttle MD, Taft LK (1981) The costs and benefits of frog ch orusing behavior. Behav Ecol Sociobiol 8:273 - 278. Saino N, Bolzern AM, Møller AP (1997) Immunocompetence, ornamentation, and viability of male barn swallows ( Hirundo rustica ). P Natl Acad Sci USA 94:549 - 552. 73 Schaedelin FC, Taborsky M (2006) Mating crater s of Cyathopharynx furcifer (Cichlidae) are individually specific, extended phenotypes. Anim Behav 72:753 - 761. Schluter D, McPhail JD (1992) Ecological character displacement and speciation in sticklebacks. Am Nat 140:85 - 108. Schneider JM, Lubin Y (1998) Intersexual Conflict in Spiders. Oikos 83:496 - 506. Scott ML, Foster SA (2007) Phenotypic plasticity and the ecotypic differentiation of aggressive behavior in threespine stickleback. Ethology 113:190 198. Smith WJ (1977) The behavior of communicating: a n ethological approach. Harvard University Press, Cambridge, MA. Stoffer B, Walker SE (2012) The use of multimodal communication in mate choice decisions by female house crickets, Acheta domesticus . Anim Behav 83:1131 1138. Sullivan MS (1994) Mate choice as an information gathering process under time constraint: implications for behaviour and signal design. Anim Behav, 47: 141 - 151. Tack EJ, Putland DA, Robson TE, Goldizen AW (2005) Geographic variation in vocalisations of Satin Bowerbirds, Ptilonorynchus violaceus , in south - eastern Queensland. Emu 105:27 31 Taylor EB, Boughman JW, Groenenboom M, Sniatynski M (2006) Speciation in reverse: Morphological and genetic evidence of the collapse of a three - spined stickleback ( Gasterosteus aculeatus ) species pair . Mol Ecol 15:343 - 355. Taylor RC, Buchanan BW, Doherty JL (2007) Sexual selection in the squirrel treefrog Hyla squirella : the role of multimodal cue assessment in female choice. Anim Behav 74:1753 1763. Tinghitella RM, Weigel EG, Head M, Boughman JW (2013) Flexible mate choice when mates are rare and time is short. Ecol Evol 3:2820 - 2831. doi: Doi 10.1002/Ece3.666. Tomkins JL, Radwan J, Kotiaho JS, Tregenza T (2004) Genic capture and resolving the lek paradox. TREE 19:323 - 328. Trail PW (1985). Courts hip disruption modifies mate choice in a lek - breeding bird. Science 227:778 - 780. 74 Uetz GW, Roberts JA, Taylor PW (2009) Multimodal communication and mate choice in wolf spiders: female response to multimodal versus unimodal signals. Anim Behav 78:299 305. Uy JAC, Safran RJ (2013) Variation in the temporal and spatial use of signals and its implications for multimodal communication. Behav Ecol Sociobiol 67(9): 1499 - 1511. Voituron Y, Brepson L, Richardson C, Joly P, Lengagne T (2012) Energetics of calling i n the male treefrog Hyla arborea : When being large means being sexy at low cost. Behaviour 149:775 - 793. doi: 10.1163/1568539X - 00003004. von Hippel FA (2000) Vigorously courting male sticklebacks are poor fathers. Acta Ethol 2:83 - 89. doi: 10.1007/s10211990 0010. Wagner WE (1998) Measuring female mating preferences. Anim Behav 55:1029 - 1042. Wootton RJ (1976) The biology of the sticklebacks. Academic Press, London. Wootton RJ (1984) A functional biology of sticklebacks. University of California Press, Berke ley, CA. Zimmerer EJ, Kallman KD (1988) The inheritance of vertical barring (aggression and appeasement signals) in the pygmy swordtail, Xiphophorus nigrensis (Poeciliidae, Teleostei). Copeia 1988 (2):299 - 307. Zuk M, Kolluru GR (1998) Exploitation of sex ual signals by predators and parasitoids. Q Rev Biol 73 (4):415 - 438. Zuk M, Thornhill R, Ligon JD, Johnson K, Austad S, Ligon SH, Thornhill NW, Costin C (1990) The Role of Male Ornaments and Courtship Behavior in Female Mate Choice of Red Jungle Fowl. Am Nat 136:459 - 473. doi: 10.1086/285107. 75 CHAPTER THREE Short term shifts in population density influence male reproductive strategies ABSTRACT Population demography, although not a constant feature of populations, can change the traits favored by sexual selection. Using a captive population of threespine sticklebacks, we experimentally manipulated the male density (and thereby competition perceiv ed by males), and examined the effects on male nesting probability, mating traits, and mating success. Males from low density were more colorful and built larger nests, whereas males from high density were less conspicuously colored and built smaller nests . Although there was no main effect of density on mating success, density interacted with vigor to favor vigorous males from low density. Given that these changes resulted from short, two - week long density treatments, this work clarifies how rapid changes in demography can affect the courtship and nesting strategies adopted by males, and how those strategies are shaped by female mate choice. INTRODUCTION Population demography is important to how populations function and can alter both natural and sexual se lection, yet the demography of any given species can vary widely over time and space. For example, the number of available mates, competitors, impact fitness. Individuals who respond appropriately to the demographic conditions of their social environment may have a fitness advantage. In particular, the ability to perceive and respond to changes in density can aid individuals in competing for resources, including mates (Em len and Oring 1977; Kokko 76 and Rankin 2006; Kaiser et al. 2013; Clutton - Brock and Huchard 2013; Holveck et al. 2015). Density - induced competition can alter individual fitness (e.g., Grant 1997; Rodenhouse et al. 1997; Sutherland and Norris 2002; reviewed in Candolin and Wong 2012). However, responses can also manifest in the physical and behavioral traits individuals display (e.g. Møller et al. 2006; Candolin and Selin 2012; Guyer et al. 2012; Sprenger et al. 2011; Niemela et al. 2012) and are likely to be i mportant for survival and reproduction, particularly if density affects not just mate availability (i.e., sex ratio), but also other resources used in mating, such as territory. Because males and females experience differing evolutionary pressures, respon ses to altered densities may differ by sex. Females may, for example, relax their mating requirements (Tinghitella et al. 2013) or change their mate search strategies (Lehmann 2007) when opportunities to mate are rare. Males, in contrast, may change both h ow they compete with other males and how they attract females (reviewed in Moura and Peixoto 2013). Males may use a variety of traits to compete with rivals and attract females. Conspicuous and often costly morphological display traits are used to attract females (Møller and Nielsen 1997; Endler 1980). Courtship behaviors, such as mating calls, are also frequently used both to attract females and during courtship itself, as are locomotor displays (Bradbury and Vehrencamp 2011). In some taxa, males may also build structures as such as burrows and bowers for mating (Borgia 1985; Kim et al. 2007) or nests used to care for young (reviewed in Schaedelin and Taborsky 2009). These are considered extended phenotypes, which are external to the animal but can relay information about their builder that is useful during courtship (Dawkins 1999; Bailey 77 2012). As such, these extended phenotypes can impact courtship, parental care, or both. Males may alter reproductive traits in response to changes in male density, which can reflect both altered operational sex ratio and general population density. Density affects the level of competition, and thus, males may perceive more intense competition at high density than at low density. Because male density may reflect the percei ved intensity of competition for mates, if males respond to perceived competition, then we competitiveness. Morphologically and behaviorally, we might expect males to increase their display intensity to attract females or defeat rivals, or conversely, reduce their display to decrease negative male - male aggression (Bertin and Cezilly 2005). Similarly, the use and size of extended phenotypes (bowers or nests) could either be alte red to increase success with females or simply to maintain the nest or bower if males compete may be affected by density, examining how male traits may change together, rather than in isolation, elucidates how combinations of traits, trait means, and trait variation may respond to density at a population - level. Ultimately, selection on these male traits, both through mate selection and male competition, might be altered b y the density of competitors present (Kokko and Rankin 2006; Shuster and Wade 2003, p. 174). Thus far, studies that have addressed density - induced changes in traits have considered the impact of density directly. That is, the impact of density on traits is measured concurrent with density experience. However, a male may perceive density at one time - point and exhibit a delayed response at a later time - point. Past experiences 78 can influence later behavior, as in the case of pre - mating density experience and fu ture mate guarding behavior (Oku 2009), but it remains to be seen how changes in traits displayed before, during, and after mating are affected by perceived density. Investigating the perception of density and delayed or lingering responses, rather than pe rception and responses concurrently, can help to identify when mismatches occur response to prior experience. This is particularly important and timely, as rapid environmental changes are increasing the frequency with which demography is disturbed (reviewed in Candolin and Wong 2012), increasing the frequency at which phenotypic mismatches might occur. Here we examine how male traits in a population of threespine stickleback f ish (Gasterosteus aculeatus) change in response to male density; we experimentally modify demography to manipulate the perceived level of mate competition. Threespine stickleback males are under strong sexual selection and produce a distinctive nuptial col oration, engage in competitive and courting behaviors, and build nests for raising young (Wootton 1976). Although these traits are common among sticklebacks, variation exists both between and among populations in morphology, particularly redness (Boughman 2001), behavior (reviewed in Ward and McLennan 2008), and nests (Wootton 1976). Additionally, population density is known to vary both temporally and spatially (Tinghitella et al. 2013; Tinghitella, Head and Boughman, unpublished data), and males are known to compete strongly for territory and mating opportunities (Lackey and Boughman 2013a). The temporal and spatial variation in levels of competition suggest that selection might strongly favor responsiveness to changes in density. We 79 aits will be modified in response to density experienced with potential effects on male reproductive success in attracting females. To test our ideas, we investigate the potential induced differences in suites of morphological - , behavioral - , and extended - p henotypic traits in response to a temporary shift in male density, and whether those differences alter mating success. To set the stage, we first address the question of which morphological, behavioral, and extended - phenotypic traits predict which males bu ild nests, and whether this varies across the breeding season or with perceived mate competition. We then ask which traits predict mating success, and whether the density experienced alters the suite of traits that do so. MATERIALS AND METHODS Collection o f wild fish and laboratory conditions At the beginning of the 2013 breeding season, we collected reproductive sticklebacks ( Gasterosteus aculeatus ) from Cranby Lake, British Columbia (49.7 N, - 124.5 W). We then transported the collected fish to Michigan S tate University in East Lansing, MI, whereupon males and females were placed into separate, visually - isolated, 75 - gallon holding tanks. The fish acclimated to laboratory conditions for 48 hours before assignment to treatment tanks. We maintained all tanks under conditions of 14 hours of simulated daylight and 10 hours of darkness at 18°C and fed the fish defrosted brine shrimp ( Artemia sp.) and bloodworms ( Chironomus spp.) once daily. 80 Study design Establishing density treatments We removed individual males from the 75 - gallon holding tank and randomly 2 ) or 2 ). These treatments fall within the ranges of density found in wild stick leback populations (Gislason et al. 1998; Wootton and Smith 2000), although densities are known to fluctuate strongly within and between years (Wootton and Smith 2000) . Males in the two treatments did not initially differ in size, condition, or coloration (all t<0.024, all p>0.36). Males were not provided nest building materials while in treatment tanks, but they were free to interact, and they exhibited male - male competition behaviors, including territoriality, chasing, tussling, and biting. After a 2 - week exposure to their density treatment, we then placed individual males into 29 - gallon tanks in which to build nests for courtship trials (N=238; with N High =114, N Low =124). Enticing males to build nests - gallon tank contained nest building materials (a 900 - mL tray of sand and 12 g wet - weight of the aquatic plant Chara spp.) and a ceramic pot and plastic plants for cover. We used removable opaque white covers to visually isolate male tanks fr om one another, and we encouraged nest building by enticing males daily with a free - swimming, gravid female for 10 minutes. We recorded male courtship and nesting behavior following the introduction of the female and visually surveyed tanks for signs of ne st building (holes in the sand, use of plant material, signs of glue, and sand removed from nests and tray box). Once we identified a nest as complete, either because the 81 male swam through the nest or the nest had a visible opening, males underwent a court ship trial the following day. Conducting courtship trials Courtship in sticklebacks is complex and comprised of many steps. Males engage in display (zig - - pricking, and biting) to attract females, and then males lead females back to the nest. Once a male shows a female his nest, she may examine it, whereupon she may either enter the nest or abandon the male (for further descriptions of these behaviors, see Rowland 1994 ). Because females may choose to te rminate courtship with a male at any step during this process, males that are considered preferable by females should proceed further in courtship, and some males may receive no response from the female at all. In this study, males that had built nests (N =121 with N High =52, N Low =62) were used in no - - gallon tank. No - choice mating trials are standard for measuring mating preferences in sticklebacks (Nagel and Schluter 1998) and allow for the measurement of female preference, while excluding male - male competition. Each male was used only once, and females were always unfamiliar to males to ensure that familiarity did not alter female preferences. A single observer (EGW) scored all mating trials. In each cour tship trial, we allowed the gravid female to acclimate within an freely interact with the male. Trials began when the male and female first saw each other and lasted 20 minu 82 occurred first. We used an event recorder (Observer 2007: Noldus Technologies, Wageningen, The Netherlands) to document all courtship behaviors that occurred. We recorded the number, time of occu rrence, and duration of each courtship behavior, correcting for trial duration by using p er - minute rates, calculated by dividing counts by the duration of the trial. We then characterized male courtship style by calculating summary indices to describe a ma courtship, and courtship vigor. We calculated latency to court as the difference between the start of the trial and the first courtship behavior conducted. To calculate the nature of courtship, we first categorized behaviors as either display - oriented (zigzags, leads) or aggressive (bites, chases, and dorsal - pricks; as in Kozak et al. 2009 ). We then calculated the nature of courtship as the proportion of courtship that was display - oriented in nature, where 0 indicate d all courtship was aggressive and 1 was all display. Finally, we summed all courtship behaviors conducted by a male and again corrected for trial duration to estimate courtship vigor (Kozak et al. 2009) . Our measures of female choice included rates of female approach, response (a measure of motivation to mate; calculated as the number of times a female followed a male per times he led her to the nest), and nest inspection (the number of times a female insp score, which measures both how attractive a male is to a female and how far courtship progresses (on a scale from 0 - 4, from 0 = no response to 4 = attempted spawning; Kozak and Boughman 2009 ; Kozak et al. 2009 ). 83 Gathering male phenotypic data Morphological traits, such as body size and male nuptial color are important in mate preferences of some stickleback populations and species ( Nagel and Schluter 1998 ; Boughman 2001; Boughman et al. 2005) were examined at several time - points in the experiment: directly after the male was moved from the density treatment to the nest ing tank (after density exposure), once a color using a standardized color scoring me thod ( Boughman 2001; Boughman 2007; Lewandowski and Boughman 2008 ) that closely matches reflectance data (Albert et al. 2007; Boughman, 2007) . We also collected eye color, body color, and darkness following established methods (Lewandowski and Boughman 200 8; Lackey and Boughman 2013b) (to the nearest 0.01 mm). To determine male condition, we calculated the residuals of a regression of male weight on length. Generating nest measurements Mal e sticklebacks build and maintain tunnel - like nests in which females deposit eggs and young are raised. In addition to providing a structure conducive to oxygenating and protecting embryos, nests may also serve as an assessment criterion by females. Thus, in addition to recording the time in days that it took males to begin building and to complete nests (Barber et al. 2001) , we also recorded the frequency and duration of nesting behavior during courtship trials, including fanning, creeping through, 84 and bor ing, conducted while within one body length of the nest (McKinnon et al. 2012) . Each visit to a nest denoted a nesting bout. Following the completion of courtship trials, nests were carefully removed from tanks to dry, weighed to the nearest 0.01 g, and ph otographed from above at a standard height and with a ruler. Photos were later analyzed blindly with regard to treatment conditions using IMAGEJ (Abràmoff et al. 2004) to extract both the overall and exposed nest area and perimeter. Overall nest area and perimeter reflect the area and outline of the entire nest, whereas exposed area and perimeter reflect only the portion of the nest not covered by sand. Statistical anal ysis We considered four general categories in which males can experience changes in response density: morphological and color traits, nesting behavior, courtship behavior, and mating success. Within each category, we visually inspected each variable to det ect deviations from homoscedascity, normality, and homogeneity of covariance. Variables were standardized by calculating z - scores for use in statistical tests. We also included the predictor variable of seasonal time, divided into months during the breedin g season, which was incorporated in the models to account for known seasonal changes in both female choosiness (Tinghitella et al. 2013) and male coloration (Weigel, Chapter 2; Candolin 2000a) . All analyses were conducted in R (version 3.0.2) using the pac kages car (Fox and Weisberg 2010) , lmSupport (Curtin et al. 2014) and multcomp (Hothorn et al. 2008) . To investigate whether nesting rates are affected by density, we used linear models to determine the variables (and their interactions) that most influenc e which males build 85 nests. The model included density, time of season, male throat color, eye color intensity, body color intensity, body darkness, condition, and length as predictors. We next conducted multivariate analyses to investigate whether density affects the morphology, behavior, nests, or mating success of nesting males. Significant MANCOVA models were then followed with univariate models examining each predictor, with significance determined using Bonferroni corrections for the number of predicto rs tested in the MANCOVA (Wright 1992) . Significant results were examined individually post hoc using Tukey adjustments on all pairwise comparisons. In an effort to maximize power and find the best model, we reduced models stepwise using AIC to include onl y variables and interactions hypothesized to be of biological importance. Finally, two density treatments for any of these traits, and the significant results are reported belo w. When considering male morphological traits in aggregate, we chose two models: might have occurred directly after density exposure but before nesting. The Initial Respo nse model uses the measurements taken directly after the male was removed morphological factors might further change after males have nested and courted. The Changed Response m odel examines the change in each variable that occurs between measurements made directly after density exposure (but prior to nesting) and the final measurements taken after courtship trials. In both morphology models, our predictors were seasonal time and density, and response variables were male condition, length, 86 throat color index, eye intensity, body color, and body darkness. Models for male behavior and nesting also used the predictors of seasonal time and male density and added the morphological resp onse variables above as predictors. The response variables for the behavior model included the nature of courtship, latency to court, and courtship vigor. The response variables for the nest model of days until nest completion (pre - building lag (in days) + construction days), nest weight, nest perimeter, nest area, exposed nest area, and exposed nest perimeter. To determine whether the density males experienced affected their mating success, we ran two models: a MANCOVA model and a generalized linear model . In treatment, but also including male traits which differ between treatments and/or traits widely supported in the literature as factors in female choice. The MANCOVA model used response variables of rates of female approach, response, and nest inspection. The predictors were male throat color index, body color, body darkness, body length, condition, vigor, the nature of courtship, nest area, days to construct a nest, seasona l time, and density treatment, as well as 2 - way interactions with density (reduced stepwise by AIC to only the interactions of density with vigor, nature of courtship, and body length). The second linear model included only the preference score, a measure that includes eventual spawning, as a response to these predictors. RESULTS Which males nest? Approximately half of the males introduced to individual tanks completed nests, and males are most likely to build nests at midseason (t 2,235 = 4.56, p<0.001 ). Nesting is 87 best predicted by intense coloration (Table 2.1 ), not male condition, length, or density experienced (all t 1,233 < 1.38, p> 0.17). Note that males with redder throats were more likely to nest, whether the intensity, area, or brightness of red c o loration increased (see Table 2.1 ), thus throat color is hereafter collapsed to a single variable, color index (color area + intensity; as in Boughman, 2001; Boughman 2007; Lewandowski and Boughman 2008) , but models using all throat color variables individ ually and in combination yielded similar results. Do male trait values differ between nesting males who experienced high versus low density? We established which males nest, and then used MANCOVA models to ask whether nesting male traits differed by densi ty. As females could only select among those males who build nests, density - induced differences in trait values and variation should be important to the eventual outcomes of female choice. Here we examined how density affected three broad trait categories: male morphology, behavior, and nest construction. Morphology F 12,700 = 2.72, p=0.0006; Density Tre atment: F 8, 102 = 5.50, p < 0.0001). In contrast, our morphological changes from the time males exited the density treatment to the time l Time: F 16, 192 = 1.99, p=0.015). 88 With respect to specific morphological traits among nesters, males from low density treatments had more colorful bodies, whereas high density males were darker along their dorsal side (Figure 3. 1). However, density did no t affect male condition, length, throat color index, or eye color intensity (all F 1,233 <1.10, p>0.27). Thus, among nesting males, whether males experienced high or low density affected only the body coloration they displayed. In addition, although male condition was not affected, density did affect both the mean initial weight of males immediately after leaving the density treatment (t 112 = - 2.07, p = 0.041) and the change in weight observed between leaving the density treatment and completing courtship trials with females (t 102 = 2.09, p = 0.039). Males were on average heavier after leaving the low density treatment (Low density: 1.33 g; High Density: 1.22 g), but then lost weight during the nesting and courtship period (1.31 g), whereas high density mal es gained weight during nesting and courtship (1.33 g), ending up at weights equal to those of low density males. Behavior Behavior while constructing nests did not differ by any model predictor (all F 3,98 <1.30, p> 0.10). Considering behavior during cour tship trials, we found that males 4,101 =2.96 p=0.023). No one single courtship variable was responsible for the condition - dependent courtship change (all F 1,116 < 3.58, all p>0.06 1), although the strongest effects seemed to be driven by more display - oriented courtship as condition improves. 89 Nests F 7,94 , F 112,700 =1.69, p< 0.0001), and male 7,94 =2.79, p=0.011). Nests were built faster and with bigger area and perimeter by males from low density (Figure 3. 2). Total time to construct nests varied with both time of season and dens ity treatment (Figure 3. 3). Nests were heavier later in the season (June: t 6,114 = 4.69, p <0.0001; and July, t 6,114 = 2.96, p= 0.004). The exposed area of the nest also increased across the season, marginally so by June (t 6,114 = - 1.94, p= 0.055), and significantly by July (t 6,114 = - 2.22, p= 0.028). Moreover, males of better condition had more exposed nests (t 6,114 = 2.78, p= 0.006) which were heavier (t 6,114 = 1.99, p= 0.049). Does density affect mating success? We measured trait impacts on mating succe ss in two ways: a composite MANOVA model based on rates of approach, response, and nest inspection by females; and a linear model predicting preference score (how far males progress in courtship, including spawning; Table 2.2 ). Density is coded such that n egative values refer to low density; for example, males from high density have a greater composite mating score in the MANOVA, whereas males from low density have greater preference scores in the linear model. Our models revealed that mating success is pre dicted by courtship behaviors, as well as the interactions of courtship behavior with density. These models also revealed that rates of female responses to males and whether a stage of courtship was reached were affected by different factors. In particular , vigorous males who experienced high density treatments had high rates of female responses 90 throughout courtship; however, vigorous males who experienced low density treatments had higher preference scores than less - vigorous males from the s ame treatment ( see Table 2.2 ). How does the variation in male trai ts compare between treatments? We examined the above traits with respect to differences in variance between morph ological traits, male growth, as measured by change in length, was more variable at low density (F 116 =6.60, p=0.012), and throat color index was more variable among males who experienced high density prior to construction of a nest (F 117 = 4.03, p= 0.047). The number of days required to construct a nest was significantly more variable among males from high density (F 119 = 6.01, p= 0.016). Finally, the latency of response of a female to a male was significantly more variable among males who experienced low den sity (F 119 =6.24, p=0.015). Variation in all other male traits and rates of success were not significantly different between treatments (all F 119 <1.99, p>0.160). DISCUSSION Density should affect sexual selection by increasing same - sex competitive interactio ns (Emlen and Oring 1977 ; Eshel 1979 ; Kokko and Rankin 2006) . Consistent with this expectation, we found that prior experience with either high or low density altered a variety of male traits known from other studies to be important to male competition and mate choice, including coloration (Baube et al. 1995) , behavior (reviewed in Bell and Foster 1994) , and nest construction (Barber et al. 2001; Ostlund - Nilsson and Holmlund 2003) . Moreover, while different traits were impacted to different 91 degrees, suites of traits often responded to the density treatments similarly and in complementary ways. Males experienced body coloration changes differently depending on density treatment. Males from high density treatments were both less colorful and darker on their do rsal side than those that experienced low density treatments, both before constructing nests and during courtship. As prior work has shown that experience with other males or females alters male throat color (Guderley 2009) , our study suggests male body co lor may also communicate information in this population, perhaps about the perceived level of competition. For example, when competition was perceived to be low, males displayed brighter coloration, which may be ideal for attracting mates at a distance (Mc Lennan, 2007). Alternatively, when the level of perceived competition was high, darker dorsal coloration may serve, as it does in many fish, to make males less conspicuous (Price et al. 2008 ; Johnsen 2002) , which could potentially reduce aggressive interac tions during competition for nesting territory, decrease egg - cannibalism while raising young (FitzGerald 1991) , or decrease predation (Stevens and Merilaita 2009) , all of which may be high in dense areas. Directional color plasticity for background - matchin g has been found in other populations of sticklebacks (Clarke and Schluter 2011) . However, it is also possible that the darker coloration that males display is not to avoid competition, but instead to enhance their competitive ability (Tinghitella et al. 2 015) and references therein). Darkened coloration in this case could signal increased aggressiveness, as it does in tortoises and many birds (Gonzalez et al. 2001; Mafli et al. 2011) . Responses may also vary among individuals, where quality males, who are better able to compete, may increase signaling, whereas lower quality may 92 decrease signal coloration (Candolin 2000b) . Future studies would be needed to determine whether body color is condition - dependent in this population, and how male - male competition, female choice, and predation integrate to influence rapid body color changes. Contrary to expectations, we did not observe significant differences in mean red throat coloration between males from high density versus ones from low density, although the var iation in red throat coloration was greater in males who experienced high density prior to nesting (but variation differences disappeared after exposure to females). The greater variation in red throat coloration among males who perceived more mate competi tion (i.e. who experienced high density) is perhaps because red serves as a threat signal among males (Baube 1997; Maynard - Smith and Harper 2003) . Red throat coloration is important in male competition for territories (Bakker and Sevenster 1983) and influe nces the frequency and intensity of antagonistic attacks between males (Pelkwijk and Tinbergen 1937 ; Tinbergen 1948; Rowland 1982; Rowland 1994; Rowland et al. 1995;Baube 1997) . Under competition, better - quality males tend to be brighter, and lower quality males may decrease color expression to reduce the risk of fights with superior males (Candolin 2000b) . Our results agree with prior findings that Cranby females have weak preference for red coloration (Boughman 2001) , however variation in red coloration appears to be strongly affected by male - male competition in this population. Although we expected perceived male competition to result in appreciable differences in behavior (Andersson 1994), it did not. The same behavi ors predict mating success, but the strength of that relationship varies with density. At both densities, 93 females prefer more vigorous males. Specifically, vigorous males from high density had higher response rates from females than did less vigorous males from high density. More vigorous males from low density treatments, in contrast, had higher preference scores than did less vigorous males from low density. Therefore, males from high density were able to elicit responses from females more often compared to males from low density, but males from low density were able to progress further with females in courtship than could males from high density. Courtship behavior between treatments might therefore have interacted with differences in other traits, such a s male color, to produce these different outcomes. For example, males who are vigorous and antagonistically - colored, versus colored for advertisement, may experience different responses from females. These responses could result in different levels of mati ng success, but this remains to be tested. Given that studies across species have established that male - male competition can change how females select mates (Trail 1985; Zimmerer and Kallman 1988; Greenfield 1994 ; Galeotti et al. 1997; Petersson et al. 199 9) , our work provides evidence that prior experience with male - male competition may impact female choice later on, even when females are not directly exposed to males competing. We found that aspects of nests and nest building behavior were strongly affec ted by differences in male density exposure. Nests of males who experienced high density were constructed more slowly and were of smaller size, which suggests risk mitigation: potential vigilance towards male competitors may slow nest building, and smaller nests are perhaps both less likely to be noticed by competitors and less costly if destroyed. It is important to note here, however, that these changes resulted from a 2 - week exposure 94 to a certain density. The two - week duration represents about only one - e ighth of the breeding season, but a male could build a nest, successfully court a female, and raise young within this time (Wootton 1976) . Furthermore, demography is expected to change substantially across the breeding season (Tinghitella et al. 2013) , whi ch may render a nest vulnerable if no longer ideally structured under new levels of competition. Because males are expected to mate across the breeding season, the mismatch of nests to the competitive environment could have consequences for the mating succ ess of males and survival of young. Previous studies have shown within - individual repeatability of nest construction under constant conditions, suggesting a strong genetic component (Rushbrook et al. 2008) , but sensitivity to environmental variation also s uggests potential rapid responses to altered environments (Raeymaekers et al. 2009; Tuomainen and Candolin 2011; Wong et al. 2012) . Future studies should examine within - individual consistency in nests when environments both social and physical -- change. R egardless of nest structure, males were just as likely to construct a nest following exposure to either male density regime. Once constructed, nests may provide nest weight and area are impacted by condition. This is supported by prior work which (Barber et al. 2001), and that nests may serve as honest signals of genetic and physical quality in an environment (Raeymaekers et al. 2009) . It is also likely that nest structure may provide important information for parental care, as in the related 15 - spine stickleback, Spinachia spinachia (Ostlund - Nilsson 2000; Ostlund - Nilsson 2001) . Females in our 95 st udy did generally prefer smaller nests, consistent with expectations from these other findings. Although recent evidence in burying beetles has shown increased parental care under increased reproductive competition (Hopwood et al. 2015) , more evidence is n ecessary to determine whether parental care responds to competition in sticklebacks and whether this occurs in parallel with variation in nest structure. These questions are intriguing, particularly given that males in this species are sole caregivers (Wootton 1976) . Given our experimental design, an interesting hypothesis follows from our observations regarding male competition in nesting. Acquiring a territory is essential before males can build a nest (Wootton 1976) , and construction also requires th at a minimum healthy body condition is achieved (Rushbrook et al. 2007) . Our males, by virtue of being in a laboratory setting, likely all developed enough body condition to attempt building nest. However, in the wild, male competition for territory and it s defense would mean only a subset of males (much smaller than the half of males who constructed nests in our experiment) would be able to initially construct a nest, and the nest structure would likely reflect the competitive conditions under which it was built. In response to initial housing with rivals and subsequent nesting in isolation, males built nests that reflected the perceived competition they experienced, even in the absence of current, direct competition. Given the importance of male competitio n in establishing nesting territories (Wootton 1976), we therefore hypothesize that direct competition from rivals may determine who is successful in initially building a nest, but the perceived level of competition experienced then dictates how quickly an d large build it. Direct competition by rival males may prevent other males from securing a territory or may 96 result in their nests being destroyed. However, nest size or the pace at which it is constructed may be controlled by the availability of supplies and the likelihood of a nest being destroyed once built, which may be signaled to males by the level of competition they perceive. Our work suggests that aspects of nest construction, such as pace and compactness, may be subject to some level of control by interactions with other males prior to nesting, even in the absence of direct competitive actions such as the destruction of nests or takeover of territory by rival males. We found that the seasonal timing of nest construction was also important in deter mining nest characteristics. Nests were most likely to be constructed at midseason, with increasing weight and area as the season progressed. Because male density decreases toward the end of the mating season in the wild (Tinghitella et al. 2013) , increase d area may be possible at late season, as fewer males remain on the nesting grounds defending territories. This same logic applies to our low density treatment: when fewer competitors exist, nests should be larger. Because low density is likely to exist la te in the season in the wild, this may also be when males are approaching the end of their lives. Because males build nests faster both at late season and at low density, there is some support for the idea that male competition or natural responses to ligh t cycle changes (Wootton and Wootton 1984) may control how the speed and size at which males build nests to avoid missed mating opportunities later in life (Cluttonbrock 1984) . Females may then benefit if they relax their mate choice on other traits, like redness (Tinghitella et al. 2013) , and instead use other traits, such as nests, to assess males later in the season. Because females prefer more compact nests, as previously suggested by Barber et al. (2001) , females may be able to assess 97 variation in nest size later in the season to detect which mates are of better condition and thus more likely to survive to raise their young. This relationship remains to be explicitly tested; however, our findings provide some empirical support that female mate choice de cisions take nest area into account. Changes in male traits in response to perceived mate competition have important implications for sexual selection. First, the delayed changes displayed in male traits and nests after density exposure can mean a mismatch between a male and his current social environment. The once - disfavored under new selection pressures when conditions, such as male competition or mate availability, change during his lifetime. For example, larger nests built under low competition may be more vulnerable if competition increases. Secondly, when an traits in that individual. For instance, if body color darken s, red coloration may be accentuated (Flamarique et al. 2013) , which may modify how males are perceived, in both male competition and female choice. Third and ultimately, when male traits respond intra - generationally and evolutionarily to selection, the va riation in traits reflects defined by degree of male competition). Because evolution acts on phenotypes, the degree to which a trait reflects underlying genes is reduced in t he face of trait plasticity. Therefore, responding to perceived changes in mate competition may allow for rapid acclimation but also ultimately retard trait adaptation. Population demography plays a major role in determining which traits are favored by sexual selection, but demography often varies. We have shown that the perceived 98 level of mate competition can alter the expression of multiple traits and alter female choic e, even when females are not directly exposed to such density differences themselves. Therefore, if females select on traits that change at different perceived male densities, sexual selection may favor males who not only have the preferred traits inherent ly, but also those that can best adjust traits to cope with competition. Of course, increasing density of rivals is only one of many ways in which population demography may change. Interactions and experiences with the opposite sex may also modulate respon ses to density changes, so more work is needed to determine whether mating experience and success can alter traits. Another area of interest is the question of the effects. U nderstanding the various ways organisms can respond to demographic changes is critical to elucidating whether and how organisms can cope with new social conditions consequent to environmental change. 99 APPENDIX 100 APPENDIX Table 2 .1. Significant morphological predictors, all color - related, from a linear model of male nest building ( df = 1, 233) Morphology of Nest - Building Males Slope ± SE t p Throat Color Throat Intensity 0.09 ± 0.030 3.17 0.002 Throat Area 0.11 ± 0.030 3.57 0.0004 Throat Brightness 0.80 ± 0.027 2.95 0.004 Eye Color Intensity 0.15 ± 0.036 4.30 < 0.0001 Body Color Intensity 0.13 ± 0.023 5.75 < 0.0001 Body Darkness 0.37 ± 0.064 5.74 < 0.0001 101 Table 2 .2. Predictors of male mating success as measured by composite MANOVA df = 3, 94); and a linear model of preference score ( df = 15, 93). Negative coefficients for density and interac tions indicate that the relationship is stronger for males from low than high density. Main effects, when not significant, are included to aid the interpretation of interactions. Significant results are bolded. MANOVA (Composite) Linear Model (Includes Spawning) F p Slope ± SE t p Density Treatment 0.30 0.825 - 4.27 ± 2.56 - 1.67 0.099 Nature of Courtship 0.43 < 0.0001 9.54 ± 2.07 4.58 < 0.0001 Vigor 11.47 < 0.0001 0.97 ± 0.59 1.65 0.102 Nest Area 0.82 0.483 - 0.007 ± 0.004 - 2.03 0.046 Condition 1.53 0.212 0.44 ± 0.57 0.77 0.443 Body Length 0.31 0.820 - 0.06 ± 0.04 - 1.34 0.184 Throat Color Index 0.16 0.923 - 0.05 ± 0.07 - 0.77 0.442 Body Color Intensity 0.91 0.441 0.005 ± 0.06 0.086 0.932 Body Darkness 1.21 0.311 - 0.11 ± 0.16 - 0.68 0.499 Seasonal Time 0.44 0.848 0.67 ± 0.59 1.14 0.259 Body Length * Density Treatment 0.35 0.787 0.09 ± 0.05 1.77 0.080 Vigor * Density Treatment 4.62 0.005 - 1.31 ± 0.60 - 2.19 0.031 Nature of Courtship * Density Treatment 8.99 < 0.0001 - 4.55 ± 2.30 - 1.99 0.051 102 Figure 3. 1. After experiencing low density, males have more colorful bodies (t 1,233 = 5.28, p < 0.0001), whereas after experiencing high density, males are darker along their dorsal surface. Points indicate mean ± SE. Asterisks denote significance at * p <0.05 and *** p <0.0001. 103 Males build nests faster and larger after experiencing low than high density. Males who experienced low density took fewer days to build nests (t 6,114 = - 2.06, p= 0.042). Their nests were greater in both perimeter and area (t 6,114 = 3.21, p= 0.002 and t 6,114 = 2.67 , p= 0.009, respectively), although the exposed nest area and perimeter (portion of the nest uncovered by sand) was unaffected (all t 117 < 0.38, p>0.705). Asterisks denote significance at p <0.05 (*) and p <0.01 (**). 104 At early and late season, males who experienced high density took longer to build nests than males from low density. This effect was primarily driven by the lag to begin nest construction, rather than the days spent in construction. Lags were longer in May (t 6,114 = - 2.30, p = 0.023) and July (t 6,114 = - 2.60, p = 0.011), but males from both high and low density build nests equally quickly at midseason. Points indicate mean ± SE. Asterisks denote significance at p <0.05 105 LITERATURE CITED 1 06 LITERATURE CITED Abràmoff MD, Magalhães PJ, Ram SJ (2004) Image processing with ImageJ. Biophotonics International 11:36 - 42. Albert AYK, Millar NP, Schluter D (2007) Character displacement of male nuptial colour in threespine sticklebacks ( Gasterosteus aculeatus ). Biologi cal Journal of the Linnean Society 91:37 - 48. Bailey NW (2012) Evolutionary models of extended phenotypes. Trends in Ecology & Evolution 27:561 - 9. Bakker TCM, Sevenster P (1983) Determinants of dominance in male sticklebacks ( Gasterosteus aculeatus L). Be haviour 86:55 - 71. Barber I, Nairn D, Huntingford FA (2001) Nests as ornaments: Revealing construction by male sticklebacks. Behavioral Ecology 12:390 - 396. Baube CL (1997) Manipulations of signalling environment affect male competitive success in three - sp ined sticklebacks. Animal Behaviour 53:819 - 833. Baube CL, Rowland WJ, Fowler JB (1995) The mechanisms of colour - based mate choice in female threespine sticklebacks: Hue, contrast and configurational, cues. Behaviour 132:979 - 996. Bell MA, Foster SA (1994) The Evolutionary Biology of the Threespine Stickleback. Oxford University Press. Bertin A, Cezilly F (2005) Density - dependent influence of male characters on mate - locating efficiency and pairing success in the waterlouse Asellus aquaticus : an experimenta l study. Jounal of Zoology 265:333 - 338. Bolton J, Backwell PRY, Jennions MD (2013) Density dependence and fighting in species with indeterminate growth: a test in a fiddler crab. Animal Behaviour 85:1367 - 1376. Borgia G (1985) Bower quality, number of dec orations and mating success of male satin bowerbirds ( Ptilonorhynchus violaceus ): an experimental analysis. Animal Behaviour 33:266 - 271. Boughman JW (2001) Divergent sexual selection enhances reproductive isolation in sticklebacks. Nature 411:944 - 948. Boughman JW (2007) Condition - dependent expression of red colour differs between stickleback species. Journal of Evolutionary Biology 20:1577 - 90. 107 Boughman JW, Rundle HD, Schluter D (2005) Parallel evolution of sexual isolation in sticklebacks. Evolution 59: 361 - 373. Bradbury JW, Vehrencamp SL (2011) Principles of Animal Communication, 2 edn. Sinauer Associates, Inc. Candolin U (2000a) Changes in expression and honesty of sexual signalling over the reproductive lifetime of sticklebacks. Proceedings of the Ro yal Society B - Biological Sciences 267:2425 - 2430. Candolin U (2000b) Male - male competition ensures honest signaling of male parental ability in the three - spined stickleback ( Gasterosteus aculeatus ). Behavioral Ecology and Sociobiology 49:57 - 61. Candolin U , Selin M (2012) Density - dependent habitat selection in a growing threespine stickleback population. International Journal of Zoology 2012. Candolin U, Wong BB (2012) Behavioural responses to a changing world: mechanisms and consequences. Oxford Universit y Press. Clarke JM, Schluter D (2011) Colour plasticity and background matching in a threespine stickleback species pair. Biological Journal of the Linnean Society 102:902 - 914. Clutton - Brock T, Huchard E (2013) Social competition and selection in males a nd females. Philosophical Transactions of the Royal Society B: Biological Sciences 368:20130074. Cluttonbrock TH (1984) Reproductive Effort and Terminal Investment in Iteroparous Animals. American Naturalist 123:212 - 229. Dawkins R (1999) The extended phenotype: The long reach of the gene. Oxford University Press. Emlen ST, Oring LW (1977) Ecology, Sexual Selection, and Evolution of Mating Systems. Scie nce 197:215 - 223. Endler JA (1980) Natural selection on color patterns in Poecilia reticulata. Evolution 34:76 - 91. Eshel I (1979). Sexual selection, population density, and availability of mates. Theoretical Population Biology 16: 301 - 314. FitzGerald GJ (1991) The role of cannibalism in the reproductive ecology of the threespine stickleback. Ethology 89:177 - 194. 108 Flamarique IN, Bergstrom C, Cheng CL, Reimchen TE (2013) Role of the iridescent eye in stickleback female mate choice. The Journal of Experimenta l Biology 216:2806 - 2812. Fox J, Weisberg HS (2010) An R companion to applied regression. Sage Publications. Galeotti P, Saino N, Sacchi R, Møller P (1997) Song correlates with social context, testosterone and body condition in male barn swallows. Animal Behaviour 53:687 - 700. Gislason GM, Gudmundsson A , Einarsson A (1998) Population densities of the three - spined stickleback ( Gasterosteus aculeatus L.) in a shallow lake. Proceedings of the International Association of Theoretical and Applied Limnology 26, 2244 - 2250. Gonzalez G, Sorci G, Smith LC, Lope F (2001) Testosterone and sexual signalling in male house sparrows ( Passer domesticus ). Behavioral Ecology and Sociobiology 50:557 - 562. Grant A (1997) Selection pressures on vital rates in density dependent populations. Proceedings of the Royal Society of London B: Biological Sciences 264:303 - 306. Greenfield MD (1994) Cooperation and conflict in the evolution of signal interactions. Annual Review of Ecology and Systematics 97 - 126. Guderley H (2009) Muscle metabolic capacities and plasma cortisol levels of the male three - spine stickleback Gasterosteus aculeatus: are there "femme fatale" or "macho male" effects? Physiological and Biochemical Zoology 82:653 - 61. Guyer C, Johnson VM, Hermann SM (2012) Effects o f Population Density on Patterns of Movement and Behavior of Gopher Tortoises ( Gopherus Polyphemus ). Herpetological Monographs 26:122 - 134. Holveck M - J, Gauthier A - L, Nieberding CM (2015) Dense, small and male - biased cages exacerbate male male competition and reduce female choosiness in Bicyclus anynana. Animal Behaviour 104:229 - 245. Hopwood PE, Moore AJ, Tregenza T, Royle NJ (2015) Male burying beetles extend, not reduce, parental care duration when reproductive competition is high. Journal of Evolutionar y Biology 28:1394 - 1402. Hothorn T, Bretz F, Westfall P (2008) Simultaneous inference in general parametric models. Biometrical Journal 50:346 - 363. Johnsen S (2002) Cryptic and conspicuous coloration in the pelagic environment. Proceedings of the Royal So ciety of London B: Biological Sciences 269:243 - 256 109 Kaiser H, Orford R, Thornycroft R, Naylor M, Field R, Sheppard J, Magellan K (2013) The effects of competitor density on aggressive behaviour and resource defence in a Poeciliid fish. Journal of Applied Ic hthyology 29:1264 - 1268. Kim TW, Christy JH, Choe JC (2007) A preference for a sexual signal keeps females safe. Plos One 2:e422. Kokko H, Rankin DJ (2006) Lonely hearts or sex in the city? Density - dependent effects in mating systems. Philosophical transactions of the Royal Society of London Series B, Biological sciences 361:319 - 34. Kozak GM, Boughman JW (2009) Learned conspecific mate preference in a species pair of sticklebacks. Behavioral Ecology 20:1282 - 1288. Kozak GM, Reisland M, Boughmann JW (2009) Sex differences in mate recognition and conspecific preference in species with mutual mate choice. Evolution 63:353 - 65. Lackey ACR, Boughman JW (2013a) Divergent sexual selection via male competition: ecology is key. Journal of Evolutionary Biology 26:1611 - 24. Lackey ACR, Boughman JW (2013b) Loss of sexual isolation in a hybridizing stickleback species pair. Current Zoology 59:591 - 603. Lehmann GUC (2007) Density - dependent plasticity of sequential mate choice in a bushcricket (Orthoptera : Tettigon iidae). Australian Journal of Zoolology 55:123 - 130. Lewandowski E, Boughman JW (2008) Effects of genetics and light environment on colour expression in threespine sticklebacks. Biological Journal of the Linnean Society 94:663 - 673. Mafli A, Wakamatsu K, Roulin A (2011) Melanin - based coloration predicts Behaviour 81:859 - 863. Maynard - Smith J, Harper D (2003) Animal signals. Oxford University Press. McKinnon JS, Hamele N, Frey N, Ch ou J, McAleavey L, Greene J, Paulson W (2012) Male choice in the stream - anadromous stickleback complex. Plos One 7: e37951. Møller A, Nielsen J (1997) Differential predation cost of a secondary sexual character: sparrowhawk predation on barn swallows. An imal Behaviour 54:1545 - 1551. Møller AP, Martin - Vivaldi M, Merino S, Soler JJ (2006) Density - dependent and geographical variation in bird immune response. Oikos 115:463 - 474. 110 Moura RR, Peixoto PEC (2013) The effect of operational sex ratio on the opportunit y for sexual selection: a meta - analysis. Animal Behaviour 86:675 - 683. Nagel L, Schluter D (1998) Body size, natural selection, and speciation in sticklebacks. Evolution 52:209 - 218. Niemela PT, Vainikka A, Lahdenpera S, Kortet R (2012) Nymphal density, be havioral development, and life history in a field cricket. Behavioral Ecology and Sociobiology 66:645 - 652. Oku K (2009) Effects of density experience on mate guarding behavior by adult male Kanzawa spider mites. Journal of Ethology 27:279 - 283. Ostlund - Ni lsson S (2000) Are nest characters of importance when choosing a male in the fifteen - spined stickleback ( Spinachia spinachia )? Behavioral Ecology and Sociobiology 48:229 - 235. Ostlund - Nilsson S (2001) Fifteen - spined stickleback ( Spinachia spinachia ) female s prefer males with more secretional threads in their nests: an honest - condition display by males. Behavioral Ecology and Sociobiology 50:263 - 269. Ostlund - Nilsson S, Holmlund M (2003) The artistic three - spined stickleback ( Gasterosteous aculeatus ). Behavi oral Ecology and Sociobiology 53:214 - 220. Pelkwijk JT, Tinbergen N (1937) Eine reizbiologische Analyse einiger Verhaltensweisen von Gasterosteus aculeatus L. Zeitschrift für Tierpsychologie 1:193 - 200. Petersson E, Järvi T, Olsen H, Mayer I, Hedenskog M (1999) Male male competition and female choice in brown trout. Animal Behaviour 57:777 - 783. Price AC, Weadick CJ, Shim J, Rodd FH (2008) Pigments, patterns, and fish behavior. Zebrafish 5:297 - 307. Raeymaekers JAM, Delaire L, Hendry AP (2009) Genetically based differences in nest characteristics between lake, inlet, and hybrid threespine stickleback from the Misty system, British Columbia, Canada. Evolutionary Ecology Research 11:905 - 919. Rodenhouse NL, Sherry TW, Holmes RT (1997) Site - dependent regulatio n of population size: a new synthesis. Ecology 78:2025 - 2042, Rowland WJ (1982) The effects of male nuptial coloration on stickleback aggression: a reexamination. Behaviour 80:118 - 126. 111 Rowland WJ (1994) Proximate determinants of stickleback behaviour: An evolutionary perspective. In: Bell MA, Foster SA, editors. The Evolutionary Biology of the Threespine Stickleback. Oxford University Press, pp 297 - 344. Rowland WJ, Bolyard KJ, Halpern AD (1995). The dual effect of stickleback nuptial coloration on rivals: manipulation of a graded signal using video playback. Animal Behaviour 50:267 - 272. Rushbrook BJ, Dingemanse NJ, Barber I (2008) Repeatability in nest construction by male three - spined sticklebacks. Animal Behaviour 75:547 - 553. Rushbrook BJ, Katsiadaki I , Barber I (2007) Spiggin levels are reduced in male sticklebacks infected with Schistocephalus solidus. Journal of Fish Biology 71:298 - 303. Schaedelin FC, Taborsky M (2009) Extended phenotypes as signals. Biological Reviews 84:293 - 313. Shuster SM, Wade MJ (2003) Mating systems and strategies. Princeton University Press. Sprenger D, Lange R, Anthes N (2011) Population density and group size effects on reproductive behavior in a simultaneous hermaphrodite. BMC Evolutionary Biology 11: 107. Stevens M, Mer ilaita S (2009) Animal camouflage: current issues and new perspectives. Philosophical Transactions of the Royal Society B: Biological Sciences 364:423 - 427. Sutherland WJ, Norris K (2002) Behavioural models of population growth rates: implications for cons ervation and prediction. Philosophical Transactions of the Royal Society B: Biological Sciences 357:1273 - 1284. Tinbergen N (1948) Dierkundeles in het meeuwenduin. De Levende Natuur 51:49 - 56. Tinghitella RM, Lehto WR, Minter R (2015) The evolutionary loss of a badge of status alters male competition in three - spine stickleback. Behavioral Ecology 26: 609 - 616. Tinghitella RM, Weigel EG, Head M, Boughman JW (2013) Flexible mate choice when mates are rare and time is short. Ecology and Evolution 3:2820 - 2831. Trail PW (1985) Courtship disruption modifies mate choice in a lek - breeding bird. Science 227:778 - 80. 112 Tuomainen U, Candolin U (2011) Behavioural responses to human - induced environmental change. Biological Reviews 86:640 - 657. Ward JL, McLennan DA (2008) Interpopulation differences in courtship and spawning behavior in the brook stickleback ( Culaea inconstans ). The American Midland Naturalist 160:82 - 95. Wong BBM, Tuomainen U, Candolin U (2012) Algal blooms impact the quality of nest construction in three - spined sticklebacks. Animal Behaviour 84:1541 - 1545. Wootton RJ (1976) The biology of the sticklebacks. Academic Press, London. Wootton RJ(1984) A functional biology of sticklebacks. Univ of California Press. Wootton RJ, Smith C (2000). A long - term study of a short - lived fish: the demography of Gasterosteus aculeatus. Behaviour 137(7): 981 - 997. Wright SP (1992) Adjusted p - values for simultaneous inference. Biometrics:1005 - 1013. Zimmerer EJ, Kallman KD (1988) The inheritance of vertical barring (aggressi on and appeasement signals) in the pygmy swordtail, Xiphophorus nigrensis (Poeciliidae, Teleostei). Copeia 299 - 307. 113 CHAPTER FOUR Courtship Experience and Success Affect Male Traits ABSTRACT Mating opportunities and experience are likely to vary for males both among and within populations. Differential experience may lead to variation in traits used in sexual selection and thereby advantage males that are most responsive to courtship experiences. Here we show that, with increasing c ourtship experience, male threespine sticklebacks ( Gasterosteus aculeatus ) alter their body and eye coloration but tend to decrease their latency to court regardless of whether their courtship was previously successful. However, in response to prior courts hip success associated with spawning, males intensify both their nest care and their color displays. Animals are likely to have differential opportunities to acquire experience during their lifetimes. Our study demonstrates how trait values can change as a result of experience, which thus modifies the variation on which sexual selection can act. Furthermore, our study suggests that selection may act on plasticity in response to experience itself. INTRODUCTION Males and females of many species often mate m ore than once in their lifetimes. attention, and the female tends to choose males possessing the traits she prefers (reviewed in Andersson 1994). In addition to fighting off rivals, males exhibit a variety of traits to attract females during courtship. For example, to attract females, males exhibit classic morphological display traits like large tails (Møller and Nielsen 1997) or colorful 114 bodies (Endler 1980). Likewise, male behaviors, such as the mating calls found in many amphibians and insects, can be used by females to locate and evaluate prospective mates (Ryan et al. 1982; Mappes et al. 1996; Hoback and Wagner 1997; Reinh old et al. 1998; Ophir et al. 2010; Voituron et al. 2012). In addition to their morphological traits and emitted behaviors, males may also alter their environment by constructing structures such as burrows, bowers, craters, or nests which may be used in co urtship and/or raising young (Borgia 1985; Hill et al. 2006; Schaedelin and Taborsky 2006; Kim et al. 2007 ). These by females, as they are variable within populations and may indicate male quality (reviewed in Bailey 2012). The traits males exhibit during courtship do not necessarily remain constant as they age, and with each experience, the potential exists for new information to be acquired about courtship. For example, the success or failure of a courtship attempt may shape trait expression and impact future mating attempts. If males can modify their own traits in response to such newly - acquired inf ormation, they may be able to rapidly acclimate to novel conditions and thus enhance their fitness ( Baldwin 1896; Osborn 1896; Robinson and Dukas 1999; Dukas 2013 ). Courtship experience in particular has been shown to affect how males court. Responses to e xperience have been shown to help males reduce their time and effort in pursuing females (Dukas et al. 2012) , target the females most likely to accept them as mates (Dukas 2004; Dukas 2009; Dukas 2008) , and better compete against more - naïve males, thereby increasing their fecundity (Dukas et al. 2006). Experienced males may 115 be better at courting generally, as they can cue onto receptive females ( Siegel and Hall 1979; Ejima et al. 2005 ) and adjust their courtship to suit the females (King and West 1983; Patr icelli et al. 2002) . Most work to date has focused on direct comparisons of behavior between experienced and naïve males. This is understandable, as behavior is thought to change faster than other traits, such as morphology (West - Eberhard 2003), and has wi dely been shown to respond to external simuli (reviewed in Snell - Rood 2013). However, males may show responses to experience in traits other than behavior, and these trait responses may strengthen with additional experience. For example, with experience, m ales may not only increase their behavioral display traits, but also their morphological display traits, such as color, to attract females. Similarly, males may modify their extended phenotypes to increase success with females. Given that such effects may also strengthen with positive feedback from courtship success, studying males in serial courtship can elucidate how population means and trait variation might change in response to experience during the lifetimes of individual animals in ways that may in t urn affect evolutionary processes. Because experience tends to be coupled with increasing age in the wild, it is difficult to determine how male age and experience might interact to shape trait expression. In particular, individuals need not gain experien ce consistently across their lifetime; some experiences may be restricted to specific phases of life, yet they may still impact how behaviors and traits are expressed in the future. Particularly because male mating traits can change at different rates with age (Miller and Brooks 2005), there is a need to decouple experience from age to understand how mating traits change. Here 116 we designed an experiment allowing us to address how increasing courtship experience affects males independent of their age. We con ducted an experiment in which our goal was to determine how courtship experience and success affect both a male's phenotype and his future success in courtship. To investigate the effects of courtship experience and success, we examined threespine stickleb acks ( Gasterosteus aculeatus ) from Cranby Lake using four successive no - courtship experiences across the breeding season to help control for male age (as sticklebacks are generally annual and only survive for a singl e breeding season; Baker, 1994). We then tracked how a suite of male traits change with courtship experience and courtship success and whether past courtship success begets future success. The biology of the threespine stickleback is relatively well - studi ed, particularly with respect to sexual selection (Rowland 1988; Bakker and Milinski 1993; Kraak et al. 1999; Candolin 1999; Tinghitella et al. 2013). Male sticklebacks establish territories for constructing nests to court (Wootton 1976). Larger, redder ma les are generally preferred across populations (Wootton 1976; Milinski and Bakke r 1990; Milinski and Bakker 1992 ; Bakker and Mundwiler 1994; Candolin 1999; Boughman 2007), and because males are responsible for all parental care (Wootton and Wootton 1984), nests have also been suggested to factor into female mate - choice decisions (Barber et al. 2001). Across the breeding season, males may be visited by several females searching for suitable mates, thus providing males with ample opportunities for experience with courtship . Stickleback population density is known to vary temporally and spatially in nature (Tinghitella et al. 2013; Tinghitella, Head and Boughman, unpublished data), 117 potentially creating variation among individuals in the amount of courtship they experience. Finally, given that female sticklebacks are known to alter their mate choice decisions with their own courtship experience (Kozak et al., 2013), it seems reasonable to expect that male sticklebacks might likewise be able to respond. Male sticklebacks are under strong sexual selection through female choice, which should select for males displaying the traits females most prefer. Variation exists within and between populations in male traits, particularly in regard to their morphology, color (Boughman 2001) , behavior (reviewed in Ward and McLennan 2008), and nest characteristics (Wootton 1976). We were therefore interested to learn whether and how the traits that males display during courtship are affected by experience within the male how selection acts on male traits over generations. MATERIALS AND METHODS S tudy population We used minnow traps to collect reproductive sticklebacks ( Gasterosteus aculeatus ) from Cranby Lake, British Columbia in summer 2013. We separated fish by sex using well - established differences in body shape (for males and females), the presence of eggs (females), and nuptial color (males) (McPhail 1984; McPhail 1992; Hatfield 1997), an d we transported them to Michigan State University in East Lansing, MI. Fish were placed into single - sex 75 - gallon holding tanks for 48 hours to acclimate to laboratory conditions. Each tank contained cover (plastic plants and ceramic half - pots) and was ma intained under summer conditions of 14 - hour day lengths and 18°C. Fish 118 were fed defrosted brine shrimp ( Artemia sp.) and bloodworms ( Chironomus spp.) once daily. Encouraging males to create and maintain nests At two - week intervals across the summer breedi ng season, males were introduced to individual 29 - gallon nesting tanks and allowed a maximum of 14 days to complete a nest. Each tank contained nest - building materials (a 900 - mL tray of sand and 12 g wet - weight of Chara spp.), cover (ceramic pot and plasti c plants), and opaque boards along the external perimeter of the tank to visually isolate the males . To encourage nest building, a free - swimming gravid female was introduced to each tank for 10 minutes daily. We identified completed nests by a visible nest opening and/or a male swimming through a nest. Twenty four hours after each male completed a nest, he underwent his first courtship trial, and subsequent trials were conducted at 3 - day intervals. Trial date was used as our proxy for male age, as these annual fish generally complete one breeding season before dying (Wootton and Smith 2000). In order to encourage nest maintenance, on each of the two days between trials, males were presented for 10 minutes with females placed into their nesting tanks in 0. 47 - L glass jars with mesh lids. These jars allowed for visual and olfactory communication between the fish but prevented physical interaction; the use of in - jar females ensured that the males continued to see females and to maintain their own nests once bu ilt. Courtship trials A total of 121 males built complete nests and underwent four no - choice mating trials at 3 - day intervals. No - choice mating trials allow males and females to interact in 119 the absence of male - male competition, in order to focus exclusive ly on male - female interactions (Wagner 1998). No - choice trials have become standard for measuring mating preferences in sticklebacks (Nagel and Schluter 1998). During each courtship trial, a gravid female was placed in an opaque holding container in a nest - gallon tank to acclimate for 5 minutes. After acclimation, the female was released from the holding container. Trials began when the male and female first viewed one another and ended after 20 minutes, or when a female , whichever occurred first. Nesting males were always novel to females. Females that entered nests were prevented from depositing eggs within the eggs. A single observe r (EGW) documented all courtship and nesting behaviors using an event recorder (Observer: Noldus Technologies, Wageningen, The Netherlands). Changes in male behaviors due to experience may be found not only in which behaviors a male emits, but also in the duration of each activity. Therefore we measured the frequency and duration of display - oriented behaviors (zig - nest) and aggressive behaviors (chasing, dorsal - pricking, and biting), as well as male nest - fanning and showing th e nest to the female. Given that females may or may not choose to approach and follow a male and subsequently examine, enter, or deposit eggs within his nest, we also recorded female behaviors in response to males at several points during the courtship seq uence (for further descriptions of these behaviors, see Rowland 1994). Our measures of female choice included responsiveness (the number of times a female followed a male when he led her to the nest; a measure of motivation) 120 and preference score, which mea sured how far courtship progressed (on a scale of 0 - 4, ranging from no response to attempted spawning), and thus how strongly females were attracted to any given male (Kozak and Boughman 2009; Kozak et al. 2009). We calculated per - minute rates for all measured behaviors to correct for trial duration as well as summary metrics to capture nature of courtship, and courtship vigor. We calculated the latency to court as the differ ence between the start of the trial and the first courtship behavior emitted by the male. We also developed a metric to characterize the overall nature of courtship on a 0 - 1 scale describing the relative proportion of courtship activity that was display - or iented in nature, where 0 indicates all courtship behaviors were aggressive, and 1 indicates all behaviors were displays. Finally, we calculated courtship vigor as the total number of all courtship behaviors emitted by a male, corrected for trial duration, as in Kozak et al. (Kozak et al. 2009). We also calculated two variables to investigate how previous mating success, effect of prior courtship experience in the absence of mating. Previous courtship success was assessed in four different ways. First we inquired whether a male had gotten a female to 1) inspect his nest in the previous trial, or 2) enter his nest considered all prior successes (additive success) by adding up across all prior trials whether 3) inspection 121 all other naturally - occuring elements of female reproduction occurred except egg deposition. Male phenotypic data Phenotypic traits, such as body size and male nuptial color, are important for the reproductive success of ma les in some stickleback species (Nagel and Schluter 1998; Boughman 2001; Boughman et al. 2005). To document changes in these traits that might occur with exposure to females, we took several phenotypic measurements before and after each courtship trial, as well as directly prior to initial placement in the 29 - gallon nesting tank. During each individual courtship trial, we observed males in - tank (to avoid excessive handling of the fish) and took color measurements directly before and after each trial. We sc and area of throat coloration), body color intensity, eye color intensity, and body darkness following methods established previously (Boughman 2001; Boughman 2007; Lewandowski and Boughma n 2008; Lackey and Boughman 2013). These measurements were made on a 0 - 5 scale to describe the intensity of color display on Nest measurements Following McKinnon et al. (2012), we recorded all male behavior (fanning, creeping through, boring, etc.) directed toward the nest. Because fanning is crucial to egg survival, and boring and gluing behaviors by the male are crucial to nest construction and maintenance (von Hippel 2000), we divided nest - oriented behavior into two categories: nesting bouts denoted the number of visits th e male made to fan his 122 nest, and nest care was a composite score that included counts of depositing spiggin, boring at the sides of the nest, as well as all movements to swim through the nest (which may alter its shape). To determine whether males changed their nests between trials, we visually observed and recorded the nest location, depth, orientation, amount of plant material used (1 - 5 scale, where 1 is sparse plant use and 5 full plant use throughout the nest), and the visibility/exposure of the nest im mediately before each courtship trial. We noted any and all changes to the nest as described above and above changes had been made. Note that, even though changes were considered in a binary fashion, regardless of the degree of nest change, males were still found to maintain nests 76% of the time across trials. Statistical analysis We asked how courtship experiences and previous courtship success affected male trai ts (display traits, courtship behavior, nests, and nesting behavior). We then asked whether past experience or courtship success led to future courtship success. We conducted mixed - effect, repeated measures model analyses in R (R Core Team, 2013, version 3 .0.2) using the packages car, lme4, lmerTest, optimx, and languageR. We used plots to visually detect deviations from homoscedascity and normality. Multicolinearity of variables was investigated through the Variance Inflation Factor (VIF) of each model. A VIF exceeding 4 (our threshold reference value) was not found, and therefore predictor variables were included in model as independent to search for outliers that might have disproportionately affected estimates. Although 123 rare, when detected, models with and without the outliers were run and showed no differences in the significance of effects. Therefore we included all data points in our analyses. We ran models to inqui re whether and how male traits change in response to age), and the interactions between these two variables as predictors and used each male trait as a response variable. In regard to morphology, we examined possible changes in male throat, eye, and body coloration, as well as body darkness. In regard to behavior, we examined the latency to court, courtship vigor, the overall nature of courtship, and the nature of the firs t action directed by the male toward the female. We also measured how two specific nesting behaviors, nest bout rate and nest care rate, changed between consecutive trials, and whether a male changed or rebuilt his nest (by nest placement, visible amount o f material used, and/or orientation as identified through visual inspection and confirmed through photographs). We next conducted models to inquire whether and how male traits change in response to courtship success (accounting for courtship experience). We used the described above. Our predictors included trial number (to now account for the effects of courtship experience), trial date, previous courtship success, and the intera ction between trial number and date to document responses in male traits to previous courtship success. Previous courtship success was assessed in two different ways: 1) whether a male had gotten a female to inspect his nest in the previous trial, or 2) w hether a male had gotten a female to spawn in the previous trial (previous success). 124 We then also considered all (additive) prior successes by adding up across all prior trials whether 3) inspection or 4) spawning was achieved (e.g., 0 - 3 are possible prior success values). We then inquired whether past courtship success predicted future courtship success. These models used trial number, trial date, previous courtship success, and the interaction between trial number and date to predict response rate, inspe ction rate, each of the four measures of courtship success. We examined the residual variation of each significant response variable in the experience model post - hoc to det ermine whether the variation could be explained by other male traits that we measured. The predictors in these models included male condition, size, behavior variables (nature of courtship, vigor, latency to court), and color variables (throat color, eye c olor, body color, and body darkness). In order to reduce the number of variables, Principle Components Analysis (PCA) was done on the color variables for models predicting behavior. Scree plots and loadings showed that the color variables could be combined into the first principle component, which explained 37 percent of the variation in color. Stepwise AIC was used to further reduce these models to find the set of variables that best explained change in male traits after experience. To correct for multiple comparisons across our models, we used False Discovery Rate (FDR), and report both the initial and corrected p - value below; we used stepwise AIC to find the best models where appropriate. All graphs were constructed using the package ggplot2 in R. 125 RESULTS How does courtship experience affect male traits? As the number of prior courtship experiences increased, males increased their throat and eye coloration, increased their vigor, and decreased their latency to court (Figure 4. 1). However, males did not change aspects about their nest or its care, the nature of their courtship, or body coloration in response to courtship experience (all t<2.286, all p>0.13). Courtship latency and courtship vigor also varied, not only with male experience, but also wi th male age, represented by trial date; however this variation was not significant after FDR correction (Table 3.1 ). Several variables that responded to experience showed much residual variation, so we conducted post hoc analyses to understand what contri buted to that variation. We examined the residual variation in each of the four significant variables (latency to court, courtship vigor, throat color, and eye color) in response to the suite of male trait variables shown in Figure 4. 1. The residual variat ion in latency to court was correlated with courtship vigor (slope ± SE = - 0.264 ± 0.067, t = - 3.927, p <<0.0001). The residual variation in vigor was correlated with latency to court (slope ± SE = - 0.0043 ± 0.001, t = - 4.01, p <<0.0001) and the nature of c ourtship (slope ± SE = - 0.0067 ± 0.002, t = - 3.537, p <<0.0001), but male color also trended toward statistical significance (slope ± SE = 0.0023 ± 0.001, t =1.74 , p = 0.083). Males who were more vigorous generally started courting more quickly, courted m ore aggressively, and were often more colorful. The residual variation in throat color was correlated with eye and body color (slope ± SE = 0.1130 ± 0.040, t = 2.831, p = 0.005; and slope ± SE = 0.1090 ± 0.039, t =2.767, p = 0.006, respectively), as well as the latency to court (slope ± SE = - 0.0450 ± 0.022, t = - 126 2.081 p = 0.038). Residual variation in eye color was best explained by darkness and throat color (slope ± SE = - 0.0780 ± 0.034, t = - 2.294, p = 0.020; and slope ± SE = 0.4560 ± 0.040 t =11.516, p <<0.001, respectively). How does courtship success affect male traits? We considered the effects of mating success on male traits in four different ways. We first evaluated the effect of success in previous trials (previ ous success) based on whether a male had gotten a female to 1) inspect his nest in the previous trial, or 2) spawn in the previous trial. We then considered all prior successes (additive success) by adding up across all prior trials whether 3) inspection o r 4) spawning was reached. A summary of the significant findings for eac h stage can be found in Table 3.2 . First, we examined the response of traits to previous inspection success. Here, prior success led to an increase in red throat coloration during the next courtship attempt (Table 3.2 ). However, nest bout rate, body color, and body darkness did not signfica ntly change with previous inspection success (all t<0.179, p>0.054, all adjusted p=0.197). None of the other male traits were affected by previous courtship success (all t <1.584 , all p >0.114, all adjusted p >0.251). We then examined male traits that resp onded to previous spawning success. Between trials, males merely returned their nests to the state which they appeared during the earlier successful trial (Table 3.2 ). In addition, the difference in the rate of nest care between trials merely trended towar d statistical signficance (slope ± SE = 2.593e - 04 ± 1.37e - 04 , df=369.7 t =1.892, p = 0.059, adjusted p= 0.651). None of the other male traits were significantly altered (all t <1.48, all p >0.138, all adjusted p > 0.759). 127 Next, we examined trait responses to additive inspection success. With additive inspection events, males increased their throat and body color (Table 3.2 ). However, none of the following variables changed significantly between trials: nest bout and care ra tes, latency to court, the nature of courtship, courtship vigor, nest changes, eye color, or body darkness (all t<1.247, all p>0.105, all adjusted p=1). Finally, we examined trait responses to additive spawning success. Males increased their nest care rate but made no changes to their nests (Table 3.2 ). No other male traits were significantly affected by additive courtship success (all t< - 1.624, all p>0.105, all adjusted p>0.385). Does past courtship success predict future courtship success? In investiga ting trait changes in response to success, we considered previous courtship success in four different ways: previous inspection success, additive inspection success, previous spawing success, and additive spawing success. We then determined how female beha viors (responsiveness, inspection rate, and preference scores) changed in relation to these four ways of measuring prior courtship success. We found that previous success (measured by inspection and spawning) did not affect future success as measured by fe male responsiveness, inspection, or preference score ( all t < 1.647, all p >0.101, adjusted p >0.302). Nor did additive success affect future success in any measure ( all t <1.564, all p >0.119, all adjusted p >0.357). Past success apparently does not guarantee future success. DISCUSSION The goal of this experiment was to determine whether and how courtship experience and prior success change both a male's phenotype and his current success in courtship. We found that courtship experience altered male color, vigor , and latency 128 to court. Mating success, in contrast, discouraged males to alter their nests. However, past success in courtship did not predict future success. Trait changes in response to e xperience With increasing courtship experience, the throats and ey es of males became more colorful. Even though Cranby males have duller throats than do males from other stickleback populations (Boughman 2001), they nonetheless became redder with courtship experience. This aligns with previous work showing that male thro at coloration intensifies in response to social experience with conspecifics (Guderley 2009). - shifted waters (Albert et al. 2007), should create litating detection female choice in this population (Boughman 2001), it may be that increasing redness ive ability, and nesting status to other males. Indeed, in other populations, redness functions in competition among males, both in territory acquisition (Bakker and Sevenster 1983) and antagonistic interactions (Rowland 1982; Rowland 1994; Rowland et al. 1995; Baube 1997). Because throat color increases along with eye color (an important combination of cues in stickleback signaling; Flamarique et al. 2013), it is likely that these color changes are mediated by the same mechanisms. Although coloration is dy namic in fish and responsive to many cues (Kodric - Brown 1998), a good candidate mechanism here might be melanocyte concentrating hormone, as it has previously been shown to darken both fish skin and eyes (Sköld et al. 2015). 129 With increasing courtship experience, males courted more quickly. Quicker courting with experience has also been demonstrated in fruit flies (Dukas 2005), where experienced males identified and courted receptive females faster than did naïve males. In our experiment, however, indiv idual males were tracked through a series of courtship experiences, and thus increases in speed detected were detected within males, rather than between groups of males. Experience evidently has compounding effects such that each experience results in ever - faster courting. Furthermore, males saw only gravid females, which can be visually differentiated from nongravid females at a distance (Rowland 1989). Thus, faster identification of appropriate females cannot account for males courting more quickly. Howev er, given the strong competition males face in the wild (Rowland 1988), courting more quickly might be a way for males to beat the competition before it starts. Parsons (1973) suggested that, across many species, quick - courting individuals should have a se lective advantage, which has been subsequently supported by studies conducted in both lab and field (Parsons 1973; Markow 1988; Markow and Sawka 1992; Stoltz et al. 2008). Vigor, a male trait commonly preferred by females (Gibson and Bradbury 1985; Anderss on 1991), also tended to increase with experience. Evidence suggests that female sticklebacks prefer more intense, natural amounts of courtship vigor (Rowland 1995). Thus, males who learn to court more vigorously may have increased mating success with cour tship experience. Vigor may also function to signal parental quality (Knapp and Kovach 1991), or may simply indicate younger mates who are likely to survive to care for young, which is critical to offspring survival in this species (Wootton and Wootton 198 4). 130 Increased vigor and decreased courtship latency appear to be related. Changes in both vigor and latency to court suggest males may become more sexually aroused with experience (Andretic et al. 2005). Because sexual arousal has been shown to reinforce c ues in stimulus - response work (Jenkins and Rowland 2000), arousal may also help males learn how vigorously and quickly court. Because males in other species have been shown to learn from mating experience (Dukas 2004; Dukas et al. 2006; Dukas 2008; Dukas 2009) and alter their behavior mid - courtship - attempt to suit females (Patricelli et al. 2002) , it is important that future studies consider how learning, and potentially also arousal, may change as males age. Interestingly, our results also show multiple elements of stickleback courtship to be plastic. Given that stickleback courtship behaviors have been considered to be a classic example of instinctive behaviors or fixed action patterns (Tinbergen 1951), it is somewhat surprising that we find so much vari ation in response to experience. Although the triggers to emission of courtship behaviors might be fixed in this species, it is clear that experience can affect the intensity or frequency of the ensuing behavioral and morphological displays. In any case, t his classic example of a putative fixed action pattern clearly shows plasticity. Trait changes in response to courtship success Male traits can evidently change in response to courtship success, and these changes depend on the level of success experienced. It appears that reaching the stage of nest inspection by the female tends to intensify male coloration, and that attempted spawning within the nest discourages changes to nests. Achieving nest inspection, but not quite achieving spawning, implying that ma les were almost successful in courtship, 131 the vigor of courtship and courting more quickly. Given that nest inspection is the likely plays a role in mate choice, as suggested by Barber et al. (2001). Having a female enter the nest may indicate to the male that his nest structure was approved. This may explain our finding that, if a male was successful in the past with his nest, he is unlikely to modify it, and why unsuccessful males change their nests. Male traits may also change when courtship success is experienced additively across encoun ters. Multiple nest inspections increase male coloration for advertisement in future courtship attempts, but repeated spawnings encourage the maintenance of nests. The odor and the structural integrity of a nest is likely altered by females repeatedly ente care for nests by depositing glue, smoothing the tunnel, and tidying edges. These behaviors may function either to make the nest more attractive (through nest compactness or added odor) for further courting, or to prepare the nest for fatherhood. help him garner additional clutches before he switches entirely into his parental mode (reviewed in Mayer and Páll 2006). If males are preparing to be parents, then the purpose of the nest shifts from a structure for mating to one for rearing young, such that nest care and tending behaviors are needed to increase offspring survival (Sevenster 1961; Sargent and Ge bler 1980) . Because nest structure is repeatable within individuals (Rushbrook and Barber 2008), explicit tests are needed to disentangle whether males 132 increase nest care to repair nests for further courtship, to correct flaws in nest structure that might affect survival of their young, or both. Does past success predict future success? In many species, repeated encounters with potential mates affect sexual selection, particularly when animals search for mates sequentially. Our results indicate that past s uccess in courtship does not guarantee future success, suggesting that females vary in their mate choice decisions. This is consistent with several studies showing that female preferences vary based on conditions intrinsic and extrinsic to the female (revi ewed in Jennions and Petrie 1997; Cotton et al. 2006 ; Bell et al. 2009) . In particular, a lthough all females in our study were kept in a single - sex tank and were to have affected their mate choice decisions. Females in many species have mate preferences that can be influenced by both potential and prior mates (Shelly and Bailey 1992; Jennions and Petrie 2000; Wong et al. 2004; Kokko and Mappes 2005), and experience with high - acceptance criterion in several of these species (Janetos 1980; Reid and Stamps 1997; Rebar et al. 2011; Fowler - Finn and Rodriguez 2012), including in sticklebacks (Bakker and Milinski 1991 ; Milinski and Bakker 1992; Kozak et al. 2013) . The various experiences of females help to explain variation in acceptance of males with different females across mating mating attempts, particularly for species with complex displays (Coleman et al. 2004). Another potential explanation for why males are not consistently successful is 133 exists to suggest that males tailor their courtship to the individual females with which they interact, not only in in sticklebacks (Kozak et al. 2009), but also in other taxa (Meffert and Regan 2002). In particular, if our gravid females varied in abdominal distention (a sign of fecundity; Fletcher and Wootton 1995), males might have varied i n their courtship behavior directed toward the females (Kraak and Bakker 1998). If males consistently favored more distended females, but these females were presented at random to males and across trials, responses to females could have masked behaviors wh ich could have led to consistent success across trials if our sample size was not large enough to overcome this noise. Nonetheless, modifying courtship behavior is said to be a learned response acquired from prior courtship attempts (Jenkins and Rowland 19 97), which would persist and potentially be adaptive for males even if its positive effects on fitness (here, courtship success) are only occasional (Dukas and Visscher 1994; Dukas 1998) . Furthermore, the traits expressed in our males did not change consistently together across trials. I f male traits change inconsistently, they may experience inconsistent mating success. Our data show that males respond differently due to both the nature of success and experience. Experiences prior to spawning appear to lead to many strong responses in traits for future courtship, and spawning appeared to only discourage changing of nests. These different responses suggest that experiences prior to spawning may generally aid a male in developing and displaying desired traits for future courtship attempts. However, we did not see consistent mating success across trials among our males. This suggests that selection may be weaker on later courtship attempts if a male has been previously successful, compared to fish that h ave yet to 134 succeed. In nature, stickleback males vary strongly in reproductive success, where many males are unable to mate (often due to an inability to secure nesting grounds), and very few males are able to sire offspring from several clutches. Therefor e repeated success may only be experienced by a select few males. Traits may change to help secure at least one successful siring, after which responses may decrease (as suggested by Whoriskey and FitzGerald 1994) or be redirected to the nest for rearing (to counter selection against offspring), rather than wooing additional females. Population size and the densities of potential mates and competitors are well known to affect sexual selection (Emlen and Oring 1977). Our experimental paradigm permitted mal es to interact with four different females over time, which may represent a competition in the wild. In sparse populations, mating opportunities may be even rarer. However, in populous areas, mating opportunities should vary widely between males because of both competition and encounter rates. If mating opportunities vary between individuals, and these individuals respond to mating experiences, then variation should result from differences in mating experience between males. Demographic parameters may also affect whether males are more likely to modify their behavior based on a single prior experience or a series of such experiences. Learning by males based on their experience sh ould theoretically affect not only the individual trait values expressed by males, but also the overall behavioral variation present in their larger populations. Thus, learning may help to increase and maintain the trait variation upon which selection can act in nature. 135 Modification of traits in response to experience should allow individuals to adjust to unique or changing circumstances in the habitats where they live; these might vary significantly over ecological and evolutionary time. Understanding ind ividuals' responses to experience helps explain the distribution of traits upon which females choose their mates. Furthermore, because suites of traits may not all change together, acclimation through adjustment of certain traits can alter the trait combin ations males display, potentially affecting other aspects of fitness, such as male - male competition and parental care. When male traits change within lifetimes, rather than across generations, the rate at which sexual selection acts can be altered. In part icular, selection may be slowed, as the phenotypes males display are a product of both their genes and differential experience with females (assuming the experience males garner is unrelated to their genotype). Because females may prefer males with good ge nes (which would nullifying this assumption), the effects of plasticity on sexual selection may operate differently depending on the underlying nature of female choice (i.e., whether good genes or runwaway, etc.) across species species. Our study illustrat es the contribution of experience, not merely to variation among males in their behavioral traits, but also in their morphological traits and their nest - oriented behavior, and thus to the extended phenotypes they express. 136 APPENDIX 137 APPENDIX Table 3 .1. Male traits that varied significantly with male age (trial date). Estimate ± SE df t Pr(>t) FDR Corrected p Latency to Court 0.011 ± 0.0047 118.9 2.41 0.017 0.187 Vigor - 2.50e - 04 ± 1.12e - 04 136.6 - 2.25 0.026 0.286 138 Table 3 .2. Male trait changes in response to courtship success, as represented by models are displayed, and non - - df =232). P - values are noted before and after FDR correction using asterisks, where p <0.05*, p <0.01**, and p p <0.10). Inspection Spawning Previous Success Additive Success Previous Success Additive Success Nest Characteristics Nest Bout Rate __ __ __ __ Nest Care Rate __ __ __ 2.261e - 04 ± 9.95e - 05 t = 2.27* / -- Changed Nest __ __ - 0.879 ± 0.43 t = - 2.06* / -- - 0.244 ± 0.0021 t = - 113.60*** / *** Morphology Throat Color Index 0.746 ± 0.28 t = 2.60** / 0.460 ± 0.18 t = 2.52* / -- __ __ Body Color __ 0.126 ± 0.060 t = 2.12* / -- __ __ Body Darkness __ __ __ __ Eye Color __ __ __ __ Courtship Behavior Latency to Court __ __ __ __ Nature of Courtship __ __ __ __ Courtship Vigor __ __ __ __ Nature of First Event __ __ __ __ 139 Figure 4.1. Standardized parameter estimates and confidence intervals for the effect of courtship experience on male traits. Male traits that responded significantly to courtship experience are denoted by asterisks (***, where after FDR adjusted p <0.001). 140 LITERATURE CITED 141 LITERATURE CITED Albert AYK, Millar NP, Schluter D (2007) Character displacement of male nuptial colour in threespine sticklebacks ( Gasterosteus aculeatus ). Biological Journal of the Linnean Society 91:37 - 48 . Andersson M, 1994. Sexual Selection. Princeton, NJ: Princeton Univ. Press. pp. i xx., 1 - 599. Andersson S (1991) Bowers on the savanna: Display courts and mate choice in a lekking widowbird. Behavioral Ecology 2:210 - 218 . Andretic R, van Swinderen B, Greenspan RJ (200 5) Dopaminergic modulation of arousal in Drosophila. Current Biology 15:1165 - 75 . Bailey NW (2012) Evolutionary models of extended phenotypes. Trends in Ecology & Evolution 27:561 - 9 . Baker JA, 1994. Life history variation in female threespine stickleback. In: Bell MA, Foster SA, editors. The Evolutionary Biology of the Threespine Stickleback Oxford, England: Oxford University Press. p. 144 - 187. Bakker TC, Milinski M (1991) Sequential female choice and the previous male effect in sticklebacks. Behavioral E cology and Sociobiology 29:205 - 210 . Bakker TCM, Milinski M (1993) The advantages of being red: Sexual selection in the stickleback. Mar Behavioral Physiology 23:287 - 300 . Bakker TCM, Mundwiler B (1994) Female mate choice and male red coloration in a natural three - spined stickleback ( Gasterosteus aculeatus ) population. Behavioral Ecology 5:74 - 80 . Bakker TCM, Sevenster P (1983) Determinants of dominance in male sticklebacks ( Gasterosteus aculeatus L ). Behaviour 86:55 - 71 . Baldwin JM (1896) A new factor in evolution. American Naturalist 536 - 553 Barber I, Nairn D, Huntingford FA (2001) Nests as ornaments: Revealing construction by male sticklebacks. Behavioral Ecology 12:390 - 396 . Baube CL (1997) Manipulations of signalling environment affect male compet itive success in three - spined sticklebacks. Animal Behaviour 53:819 - 833 . Bell AM, Hankison SJ, Laskowski KL (2009) The repeatability of behaviour: a meta - analysis. Animal Behaviour 77:771 - 783 . 142 Borgia G (1985) Bower quality, number of decorations and mati ng success of male satin bowerbirds ( Ptilonorhynchus violaceus ): an experimental analysis. Animal Behaviour 33:266 - 271 . Boughman JW (2001) Divergent sexual selection enhances reproductive isolation in sticklebacks. Nature 411:944 - 948 . Boughman JW (2007) Condition - dependent expression of red colour differs between s tickleback species. Journal of Evolutionary B iology 20:1577 - 90 . Boughman JW, Rundle HD, Schluter D (2005) Parallel evolution of sexual isolation in sticklebacks. Evolution 59:361 - 373 . Candolin U (1999) Male - male competition facilitates female choice in sticklebacks. Proceedings of the Royal Society B - Biological Sciences 266:785 - 789 . Coleman SW, Patricelli GL, Borgia G (2004). Variable female preferences drive complex male displays. Nature 428(6984): 742 - 745. Cotton S, Small J, Pomiankowski A (2006) Sexual selection and condition - dependent mate preferences. Current Biology 16:R755 - R765 . Dukas R (1998) Cognitive ecology: the evolutionary ecology of information processing and decision making, vol 1. University of Chicago Press . Dukas R (2004) Male fruit flies learn to avoid interspecific courtship. Behavioral Ecology 15:695 - 698 . Dukas R (2005) Experience improves courtship in male fruit flies. Animal Behaviour 69:1203 - 1209 . Dukas R ( 2008) Learning decreases heterospecific courtship and mating in fruit flies. Biology Letters 4:645 - 647 . Dukas R (2009) Dynamics of learning in the context of courtship in Drosophila persimilis and D. pseudoobscura . Animal Behaviour 77:253 - 259 . Dukas R (2 013) Effects of learning on evolution: robustness, innovation and speciation. Animal Behaviour 85:1023 - 1030 . Dukas R, Clark CW, Abbott K (2006) Courtship strategies of male insects: when is learning advantageous? Animal Behaviour 72:1395 - 1404 . Dukas R, D urisko Z, Dukas L (2012) Learning in the context of sexual behaviour and danger in female and male Drosophila pseudoobscura . Animal Behaviour 83:95 - 101 . 143 Dukas R, Visscher PK (1994) Lifetime learning by foraging honey bees. Animal behaviour 48:1007 - 1012 . Ejima A, Smith BP, Lucas C, Levine JD, Griffith LC (2005) Sequential learning of pheromonal cues modulates memory consolidation in trainer - specific associative courtship conditioning. Current Biology 15:194 - 206 . Emlen ST, Oring LW (1977) Ecology, Sexual S election, and Evolution of Mating Systems. Science 197:215 - 223 . Endler JA (1980) Natural selection on color patterns in Poecilia reticulata . Evolution 34:76 - 91 . Flamarique IN, Bergstrom C, Cheng CL, Reimchen TE (2013) Role of the iridescent eye in stickl eback female mate choice. The Journal of Experimental Biology 216:2806 - 2812 . Fletcher D, Wootton R (1995) A hierarchical response to differences in ration size in the reproductive performance of female three spined sticklebacks. Journal of Fish Biology 46:657 - 668 . Fowler - Finn KD, Rodriguez RL (2012) Experience - Mediated Plasticity in Mate Preferences: Mating Assurance in a Variable Environment. Evolution 66:459 - 468 . Gibson RM, Bradbury JW (1985) Sexual selection in lekking sage grouse: Phenotypic correl ates of male mating success. Behavioral Ecology and Sociobiology 18:117 - 123 . Guderley H (2009) Muscle metabolic capacities and plasma cortisol levels of the male three - spine stickleback Gasterosteus aculeatus: are there "femme fatale" or "macho male" effe cts? Physiological and Biochemical Zoology 82:653 - 61 . Hatfield T (1997) Genetic divergence in adaptive characters between sympatric species of stickleback. The American Naturalist 149:1009 - 1029 . Hill PS, Wells H, Shadley JR (2006) Singing from a construc ted burrow: why vary the shape of the burrow mouth? Journal of Orthoptera Research 15:23 - 29 . Hoback WW, E Wagner W (1997) The energetic cost of calling in the variable field cricket, Gryllus lineaticeps . Physiological Entomology 22:286 - 290 . Janetos AC (1980) Strategies of female mate choice: a theoretical analysis. Behavioral Ecology and Sociobiology 7:107 - 112 . 144 Jenkins JR, Rowland WJ (1997) Learning Influences Courtship Preferences of Male Threespine Sticklebacks ( Gasterosteus aculeatus ). Ethology 103: 954 - 965 . Jenkins JR, Rowland WJ (2000) Stimulus - specific and response - specific habituation in courting stickleback: Developmental and functional considerations. Behaviour 137:933 - 945 . Jennions MD, Petrie M (1997) Variation in mate choice and mating prefe rences: A review of cause s and consequences. Biological R eviews of the Cambridge Philosophical Society 72:283 - 327 . Jennions MD, Petrie M (2000) Why do females mate multiply? A review of the genetic benefits. Biological reviews of the Cambridge Philosophi cal Society 75:21 - 64 . Kim TW, Christy JH, Choe JC (2007) A preference for a sexual signal keeps females safe. Plos One 2:e422 . King AP, West MJ (1983) Epigenesis of cowbird song a joint endeavour of males and females. Nature 305:704 - 706 . Knapp RA, Kovach JT (1991) Courtship as an honest indicator of male parental quality in the bicolor damselfish, Stegastes partitus . Behavioral Ecology 2:295 - 300 . Kodric - Brown A (1998) Sexual dichromatism and temporary color changes in the reproduction of fishes. American Zoologist 38:70 - 81 . Kokko H, Mappes J (2005) Sexual selection when fertilization is not guaranteed. Evolution 59:1876 - 85 . Kozak GM, Boughman JW (2009) Learned conspecific mate preference in a species pair of sticklebacks. Behavioral Ecology 20:1282 - 1288 . Kozak GM, Head ML, Lackey AC, Boughman JW (2013) Sequential mate choice and sexual isolation in threespine s tickleback species. Journal of Evolutionary B iology 26:130 - 40 . Kozak GM, Reisland M, Boughmann JW (2009) Sex differences in mate recognition and conspecific preference in species with mutual mate choice. Evolution 63:353 - 65 . Kraak SBM, Bakker TCM (1998) Mutual mate choice in sticklebacks: attractive males choose big females, which lay big eggs. Animal Behaviour 56:859 - 866 . Kr aak SBM, Bakker TCM, Mundwiler B (1999) Sexual selection in sticklebacks in the field: correlates of reproductive, mating, and paternal success. Behavioral Ecology 10:696 - 706 . 145 Lackey ACR, Boughman JW (2013) Divergent sexual selection via male competition: ecology is key. Journal of Evolutionary Biology 26:1611 - 24 . Lewandowski E, Boughman J (2008) Effects of genetics and light environment on colour expression in threespine sticklebacks. Biological Journal of the Linnean Society 94:663 - 673 . Mappes J, Alata lo RV, Kotiaho J, Parri S (1996) Viability costs of condition - dependent sexual male display in a drumming wolf spider. Proceedings of the Royal Society of London Series B: Biological Sciences 263:785 - 789 . Markow TA (1988) Reproductive behavior of Drosophi la melanogaster and D. nigrospiracula in the field and in the laboratory. Journal of Comparative Psychology 102:169 . Markow TA, Sawka S (1992) Dynamics of mating success in experimental groups of Drosophila melanogaster (Diptera: Drosophilidae ). Journal o f Insect Behavior 5:375 - 383 . Mayer I, Páll M (2006) Hormonal control of reproductive behaviour in the stickleback. In: Östlund - Nilsson S, Mayer I, Hunt ingford F (eds) Biology of the Three - spined S tickleback. CRC Press, Boca Raton, FL, pp 249 - 269 . McKinnon JS, Hamele N, Frey N, Chou J, McAleavey L, Greene J, Paulson W (2012) Male choice in the stream - anadromous stickleback complex. Plos One 7 . McPhail JD (1984) Ecology and evolution of sympatric sticklebacks ( Gasterosteus ): morphological and geneti c evidence for a species pair in Enos Lake, British Columbia. Canadian Journal of Zoology 62:1402 - 1408 . McPhail JD (1992) Ecology and evolution of sympatric sticklebacks ( Gasterosteus ): evidence for a species - pair in Paxton Lake, British Columbia. Canadia n Journal of Zoology 70:361 - 369 . Meffert LM, Regan JL (2002) A test of speciation via sexual selection on female preferences. Animal Behaviour 64:955 - 965 . Milinski M, Bakker TCM (1990) Female sticklebacks use male coloration in mate choice and hence avoi d parasitized males Nature 344:330 - 333 . Milinski M, Bakker TCM (1992) Costs Influence Sequential Mate Choice in Sticklebacks, Gasterosteus a culeatus . Proceedings of the Royal Society B - Biological Sciences 250:229 - 233 . Miller LK, Brooks R (2005) The effects of genotype, age, and social environment on male ornamentation, mating behavior, and attractiveness. Evolution 59:2414 - 25 . 146 Møller A, Nielsen J (1997) Differential predation cost of a secondary sexual character: sparrowhawk predation on barn swallow s. Animal Behaviour 54:1545 - 1551 . Nagel L, Schluter D (1998) Body size, natural selection, and speciation in sticklebacks. Evolution 52:209 - 218 , Qual ity and Quant ity 41: 673 - 690 . Ophir AG, Schrader SB, Gillooly JF (2010) Energetic cost of calling: General constraints and species - specific differences. Journal of Evolutionary Biology 23:1564 - 1569 . Osborn HF (1896) A mode of evolution requiring neither natural selection nor the inheritance of acquired characters . Transactions of the New York Academy of S ciences 15:141 - 142 . Parsons PA (1973) Behavioural and ecological genetics. Clarendon Press, Oxford . Patricelli GL, Uy JAC, Walsh G, Borgia G (2002) Male displays adjusted to female's response - Macho courtship by the satin bowerbird is tempered to avoid frightening the female. Nature 415:279 - 280 . Rebar D, Zuk M, Bailey NW (2011) Mating experience in field crickets modifies pre - and postcopulatory female choice in parallel. Behavioral Ecology 22:303 - 309. Reid ML, Stamps JA (1997) Female mate choice tactics in a resource - based mating system: field tests of alternative models. The American Naturalist 150:98 - 121 . Reinhold K, Greenfield MD, Jang Y, Broce A (1998) Energetic cost of sexual attractiveness: ultrasonic advertisement in wax moths. Animal Behaviour 55:905 - 913 . Robinson BW, Dukas R (1999) The influence of phenotypic modifications on evolution: the Baldwin effect and modern perspectives. Oikos 582 - 589 . Rowland WJ (1982 ) The effects of male nuptial coloration on stickleback aggression: a reexamination. Behaviour 80:118 - 126 . Rowland WJ (1988) Aggression versus courtship in threespine sticklebacks and the role of habituation to neighbours. Animal Behaviour 36:348 - 357 . Rowland WJ (1989) Mate choice and the supernormality effect in female sticklebacks ( Gasterosteus aculeatus ). Behavioral Ecology and Sociobiology 24:433 - 438 . 147 Rowland WJ (1994) Proximate determinants of stickleback behaviour: An evolutionary perspective. In : Bell MA, Foster SA (eds) The evolutionary biology of the threespine stickleback. Oxford University Press, pp 297 - 344 . Rowland WJ (1995) Do female stickleback care about male courtship vigour? Manipulation of display tempo using video playback. Behaviour 132:951 - 961 . Rowland WJ, Bolyard KJ, Halpern AD (1995) The dual effect of stickleback nuptial coloration on rivals: manipulation of a graded signal using video playback. Animal Behaviour 50:267 - 272 . Rushbrook BJ, Barber I (2008) A comparison of nest bui lding by three - spined sticklebacks Gasterosteus aculeatus from still and flowing waters. Journal of Fish Biology 73:746 - 752 . Ryan MJ, Tuttle MD, Rand AS (1982) Bat predation and sexual advertisement in a neotropical anuran. American Naturalist 119:136 - 139 . Sargent RC, Gebler JB (1980) Effects of nest site concealment on hatching success, reproductive success, and paternal behavior of the threespine stickleback, Gasterosteus aculeatus . Behavioral Ecology and Sociobiology 7:137 - 142 . Schaedelin FC, Taborsky M (2006) Mating craters of Cyathopharynx furcifer (Cichlidae) are individually specific, extended phenotypes. Animal Behaviour 72:753 - 761 . Sevenster P (1961) A causal analysis of a displacement activity (fanning in Gasterosteus acu leatus L.). Behaviour Supplement I - 170 . Shelly TE, Bailey WJ (1992) Experimental manipulation of mate choice by male katydids: the effect of female encounter rate. Behavioral Ecology and Sociobiology 30:277 - 282 . Siegel RW, Hall JC (1979) Conditioned resp onses in courtship behavior of normal and mutant Drosophila . Proceedings of the National Academy of Sciences 76:3430 - 3434 . Sköld HN, Yngsell D, Mubashishir M, Wallin M (2015) Hormonal regulation of colour change in eyes of a cryptic fish. Biology Open BIO20159993 . Snell - Rood EC (2013) An overview of the evolutionary causes and consequences of behavioural plasticity. Animal Behaviour 85:1004 - 1011 . Stoltz J, Elias D, Andrade M (2008) Females reward courtship by competing males in a cannibalistic spider. Behavioral Ecology and Sociobiology 62:689 - 697 . Tinbergen N (1951) The Study of Instinct. Oxford University Press . 148 Tinghitella RM, Weigel EG, Head M, Boughman JW (2013) Flexible mate choice when mates are rare and time is short. Ecol ogy and Evol ution 3:2 820 - 2831 . Voituron Y, Brepson L, Richardson C, Joly P, Lengagne T (2012) Energetics of calling in the male treefrog Hyla arborea : When being large means being sexy at low cost. Behaviour 149:775 - 793 . von Hippel FA (2000) Vigorously courting male stickleb acks are poor fathers. Acta Ethologica 2:83 - 89 . Wagner WE (1998) Measuring female mating preferences. Animal Behaviour 55:1029 - 1042 . Ward JL, McLennan DA (2008) Interpopulation differences in courtship and spawning behavior in the brook stickleback ( Culaea inconstans ). The American Midland Naturalist 160:82 - 95 . West - Eberhard MJ (2003) Developmental plasticity and evolution. Oxford University Press . Whoriskey FG, FitzGerald, GJ 1994. Ecology of the threespine stickleback on the breeding grounds. In: Bell MA, Foster SA, editors. The Evolutionary Biology of the Threespine Stickleback Oxford, England: Oxford University Press. p. 188 - 206 . Wong BB, Jennions MD, Keogh JS (2004) Sequential male mate choice in a fish, the Pacific blue - eye Pseudomugil signife r . Behavioral Ecology and Sociobiology 56:253 - 256 . Wootton RJ (1976) The biology of the sticklebacks. Academic Press, London . Wootton RJ(1984) A functional biology of sticklebacks. Univ of California Press Wootton RJ, Smith C (2000). A long - term study of a short - lived fish: the demography of Gasterosteus aculeatus . Behaviour 137(7): 981 - 997 . 149 CONCLUSION Through this series of studies, we have demonstrated the role of social experience in altering sexual selection. In particular, we have examined how different forms of experience modify traits in males and females, and how experience ultimately influences mating success. This work strengthens our understanding of how within - lifetime responses to soc ial conditions can affect how sexual selection shapes traits. In Chapter 1, we provide evidence that female reproductive investment need not change in concert with reproductive behavior, or at all. Our work reveals that limited plasticity in reproductive investment may keep females from mitigating the effects of relaxed choosiness when mates are rare . When female choice relaxes, but their reproductive investment remains constant, less - preferred males may profit. The fixed nature of investments can combine with plastic behaviors to lead to the persistence of less - preferred male traits and ultimately slow the rate of selection on them. This work underscores the importance of investigating responses to male availability in both reproductive investment and beha vior by females to determine the ultimate outcomes of sexual selection. It also raises the question of whether male trait investments act in parallel with those of females (e.g., decrease in female preference for red when male expression of red also decrea ses) to mitigate conflicts between female investment and behavior. If so, to which traits does this apply, and how often does this occur? In Chapter 2, we show that males are evaluated by females based on several different traits and to different degrees as courtship progresses. Although males allocate resources into several morphological, behavioral, and nesting traits, only a subset of these are indicative of male condition and are ever preferred by a female. In contrast to 150 several populations of stickl eback, Cranby Lake sticklebacks lack both a condition - dependent relationship for red throat coloration (the preferred sexual signal in many populations) and a female preference for it in mating. This highlights the extent of population divergence in condit ion - dependent sexual signals within a rapidly - evolving species and helps to clarify that red is not only difficult to detect given the ecology of Cranby Lake, but it is also an uninformative signal of male condition to females in that population. Interesti ngly, we show that preferences in this population depend more on male behavior, which may be both a more visible and more reliable indicator of male quality than nesting or morphological traits. This work therefore suggests that the relative importance of traits upon which females choose is based both on baseline female preference for the trait (quality - indicating or not) and the time at which the traits are assessed. If some traits act as initial advertisements, and others to secure mates, future work is needed to investigate how ecology may disrupt the detection of signals used at all stages of mating. This will further our understanding of how sexual signals may be altered, lost, or exchanged for alternatives which may be more reliable in the habitat occ upied by that particular organism. In Chapter 3, we test how male traits might change in response to mate competition. We find that males tended to express traits that correspond with the perceived male density to which they are temporarily exposed. Males under conditions of high competition have risk - mitigating darker coloration, which helps to avoid negative interactions with other males, and they build smaller nests, which require less investment to build and are less costly if destroyed. Conversely, mal es under conditions of low competition build larger nests and have far more colorful bodies, suggesting that 151 these males optimize their potential for detection by females rather than preparing to face rivals. The rapid and lingering manner in which males r espond to the short - term shifts in demography is particularly telling, as demography is likely to vary spatially and temporally for these and many other organisms. This work therefore illuminates how mate competition and female choice may shape traits, and how acclimation to social conditions may influence rates of adaptation to female preferences. Future work should determine the permanence of these effects as well as whether or not a critical window of exposure is required. In Chapter 4, we address how the traits he displays to females. We find that experience in mating is not predictive of future success and that the male traits that change in response to experience are not the same as those that respond to mating success. In particular, the phenotypic traits that allow for advertisement to females (such as coloration) increase with courtship trait assessed latest in cour tship, is very likely to remain unchanged when a male was previously successful in courtship. This response in nests can indicate, as we learned in Chapter 3, that females indeed assess nests, and males receive the cue to keep suitable nests structurally t he same when they enjoy mating success (i.e., females enter the nest). Additionally, we find that mating success operates additively to alter the expression of multiple traits as experience accrues. However, a single, immediately - prior, successful mating a ttempt can still change some male traits (coloration and nest care). These responses indicate that male traits respond to experience and success, and expression of these traits may differ between sequential mating attempts. This has 152 important implications in regard to determining not only male trait values in the population, but also how females formulate relative preferences among mates. If experiences with the opposite sex change male traits and female preferences, then selection can fluctuate in strength over time and space, particularly as opportunities to mate vary and experiences (and potential successes) accumulate. Although some work has been done to determine how female preferences and searching strategies change with exposure to males, future studi es should examine whether demography - induced changes in female preferences keep pace with or lag behind commensurate changes in male traits (i.e., whether a within - lifetime arms race exists between the male traits that change with experience and female pre ferences for them). Sexual selection is undergoing a renaissance as we discover more about how within - lifetime variation, including modifications in physical or social environments, can influence rapid responses in animals. This is particularly timely and important, as rapid environmental changes induced by human activity are increasing the frequency and urgency with which organisms need to respond. The studies here contribute to our understanding of how variation in demographic parameters and social exper ience affect the physical and behavioral traits displayed by males and females. They also help predict the changes to sexual selection that should result when demography and social experience vary.