CHAPTER 10 Sexual Selection “male gray tre frog (Ha versicolor) singing to attract a mate, J itssane aais ore ner smc av 75 areanance to son. marine iguanas, for example, males weigh twice as much as females. Males become in- tensely territorial during the breeding season, while females remain gregarious throughout the year. In long-tiled widow birds, the adults of the two sexes have plumage so distinct chat it would be easy to mistake them for diferent species. Males arc jet black, carry til feathers several times the length of their own bodies, and have red and yellow shoulder patches. Fe males are colored a cryptic brown, with short tail feathers and no shoulder patches. In gray tree frog, males have dark throats and sing, Females have white throats and are silent, In stalk~ ‘eyed fies, both sexes wear their eyes on the ends of long thin stalks, but males have longer eyestalks than females. In some species of pipetish, females have blue stripes and skin folds on their bellies, Males lack these ornaments.The photos of males and females in Figure 10.1 pro- vide additional examples. In humans, too, females and males are conspicuously different. Our differences exceed the obvious and essential ones in genitalia and reproductive organs. They are found in the ap- pearance of our faces, the sound of our voices, the distribution of our body fat and body bait and our size. The size difference between women and men is documented in Figure 10.2374 PARTI Adaptation (a) Red door Figure 10.1. The differences between males ‘and females (the sexual dimorphism) in red deer (Cervus elaphus), guppies (Poecilia reticulata), and golden toads (Bufo perigienes)_ in (a), the male is (onthe lft in (b) and (0, the male on the top. (Golden toads A difference between the males and females of a species is called a sexi A difference between the rmoephism. In this chapter, we ask why sexual dimorphism occurs ins sexes is called a sexual great variety of organisms. It is a question Charles Darwin (1871) wrotel imorphism. book about, and it has captivated evolutionary biologists ever since 10.1 Sexual Dimorphism and Sex In previous chapters, we have explained the traits of organisms with the th evolution by natural selection, In Chapter 3, for example, we saw how natu tion shapes the beaks of medium ground finches on Daphne Major Whi finches face a drought, and small soft seeds are rate, big beaks confer an ad Big beaks help finches survive by enabling them to open lan 180 170. Figure 10.2 Women and men differ in height For each of more than 200 human societies, the average height of the men is plotted against the average height of the women. The diagonal ine shows where the points would fall if men ‘and women were of equal height. People vary widely in height from society to soc ‘ty: In the shortest society, the average man is about 143 cm tal (about feet, 8 inches) and the average woman about 135 cm (~4’5"); inthe tallest society, the av ‘erage man is about 180 cm tall (~5'11"), and the average woman about 165 cm (-5'5"). But in every society the average man is taller than the average woman, usu ally by about 10%. From Rogers and Mukherjee (1992) sao. 150160 i Female statue amore to eat: When the big-beaked survivors reproduce, they pass genes for big sto their offpring. In a similar way, natural selection can explain a great varie her traits, from the bars on fly wings (ce Figure 9.6, page 338) to the hiding chosen by garter snakes (see Figure 9.10, page 344), But differences between tess of a species, like those shown in Figure 10.1, are often not among them. Tosce why, try to imagine how we might use evolution by natural selection to in the til streamers in male long-tiled widow birds (Figure 10.3). Two ens aise. First, if Tong til feathers can improve the survival or fecundity of ow bied, then why do only males have them? Second, how could enor- y long til feathers improve the survival, or the fecundity, of widow birds in fist place? Long tail feathers probably make male widow birds easier for ors to find and catch. Furthermore, growing long tail feathers requires con- y energy spent on feathers is energy that cannot be spent on ng ofpring. The theory of evolution by natural selection can explain nei why male and female widow birds ate different, nor why the birds’ most ig trt, long tail feathers, exists a all Darwin himself was the first to recognize, sex provides a solution to the ofsexual dimorphism. To see why, consider life withoue sex. For organ- that reproduce without sex (see Chapter 7), getting genes into the next gen- is raightforward,if noc always easy The two big challenges are surviving {enough to reproduce, and then reproducing. Sex complicates life by adding major challenge: finding 2 member of the opposite sex and persuading Jor her ro cooperate. ares Darwin realized that individuals vary not only in their success at su. and reproducing, but also in their success at persuading members of the fe sex to mate. About birds, for example, Darwin wrote, lsmuch as the act of courtship appears to be with many birds a prolonged Witedious affair, so it occasionally happens that certain males and females do succeed during the proper season, in exciting each other’ love, and cons ily do not pair” (1871, page 107) Figure 10.3. The sexual dimorphism in long-tailed widow birds (Euplectes progne) ‘The male is black with long tall feathers and red ‘and yellow shoulder patches: the female is brown and cryptic Chapter 10 Sexual Selection 375 Sexual dimorphism fe often a puzzle, because natural explain it.376 PARTIII Adeptation In its evolutionary consequences, failing to mate is the same as dying young vietim makes no genetic contribution co future generations. Darwin hada applied the label natural selection to differential reproductive success dve 10% ation among individuals in survival and reproduction. Differential reproda success due to variation among individuals in success at getting mates heal sexual selection. We can develop a theory of evolution by sexual selection is logically equivalent to the theory of evolution by natural selection: Ift heritable variation in a trait that affects the ability to obtain mates, then conducive to success will become more common over time, Asymmetries in Sexual Reproduction If sexual selection isto explain differences between the sexes, i¢ will have i ‘on the sexes differently In this section we argue that it often docs. The ogi Figure 10.4 orangutan __-develop to support this claim was clearly articulated by Angus Jobn Bi mothers invest considerably (148) and refined by Robert Trivers (1972). It hinges on a simple obser more time and energy Eggs (or pregnancies) are more expensive than ejacllates in each offspring than In more general terms, mothers typically make a larger parental investi orangutan fathers ically P each offipring than fathers. By parental investment we mean energy andi expended both in constructing an offipring and in caring for it. Ul parental investment is measured in fitness. Parental investment increases the productive success of the offipring receiving it. At the same time, it decreas remaining reproductive success that the investing parent may achieve in th ture by way of additional offipring. Consider the parental investments made by male and female orangutans 10.4). Adult orangutans of opposite sex tolerafe each others company fi purpose only (Nowak 1991). After a brief tryst inchuding a copilation til about 15 minutes, the male and female go their separate ways, Ifa pregnany sults, then the mother, who weighs about 40 kg, will carry the feta for $m give birth toa I-ky baby, nurse i for about 3 years, and continue to protect it reaches the age of 7 or 8, For the father, who weighs about 70 kg, the beg and end of parental investment isa few grams of semen, which he can repel matter of hours or days. In their pattern of patental investment, orga typical mammal. In more than 90% of mammal species, females provide sh tial parental care and males provide lite or none (Woodroffe and Vincent 1 Because mammalian mothers provide such intensive parental care present a somewhat extreme example of dispafity in parental investment animal species, neither parent cares for the young. Mated pairs of pare make eggs, fertilize them, and leave them. But in these species, to, female ally make a larger investment in each ofispring than males. E and yolky, with a big supply of stored energy and nutrients. Think of seal some of which are as large as a hen’s eggs. Most sperm, on the othetll arc little more than DNA with a propeller. Even when asi hundreds of millions of sperm, the ejaculate seldom represents more than a tion of the investment contained in a clutch of eggs Shan canal enee Recognizing that eggs are more expensive than ¢jaculates, we can pred mposes differ the factors limiting lifetime reproductive success will often be different fi rales versus males. A female’s potential reproductive success is relative rnd her realized reproductive success is likely to be limited more by theChapter 10 Sexval Selection 377 fags she can make (or pregnancies she can carry) than by the number of males GH convince to mate with her. In contrast, a male's potential reproductive istelatively large, and his realized reproductive success is likely to be lim- more by the number of females he can convince co mate with him than by Hhumber of cjaculates he can make. Access to mates will be a limiting resource files, but not for females. In other words, we predict that sexual selection— tion in mating success—will usually be a more than in the evolution of females Bateman (1948) tested this prediction otent force in the evolution laboratory populations of the fruit fly ils melanogaster. He found, as predicted, that number of mates had a lager ef in the reproductive success of males than on the reproductive success of fe BiThis eul is central to the theory of sexual selection Gee Arnold 1994; Bidand Duvall 1994). great many other researchers have found indirect cor ation of Bateman’ prediction, but few have performed direct tess, Adam Jones ollesques (2000, 2002) have recently done so. It will be useful to compare, in deta their rests for nwo di skinned newts and broad- Binet metric Limits on Reproductive Success sand Pipefish Roman Arguello, and Stevan Arnold (2002) quantified the relative of sexual selection on male and female rough-skinned newts (Tiricha J. Male rough-skinned newts gather in ponds in carly winter to wait for ing emales. Females saunter in during January and February. The females Iho trouble attracting the attention of aniorous males (Figure 10.5a) and jenly finish mating shortly after making their appearance, As a esult, at Bien time there are more males in females prowling the pond, After mat- Males ly their eggs, one at a time, over the course of several weeks or GA female may produce 300 eggs or more. Neither parent cares for th bg The cost of eggs and sperm thus accounts for the entitety of p: fent,The investment per offspring is larger for females than for males. Band colleagues captured all che newts from a single pond in the spring. Ms had finished mating and the fe ales were laying eggs. The biologists the females in individual containers, induced them to lay the rest of their injecting them with hormones, and reared the eggs to hatching. The re fs then took tissue samples from all the adult males, all the ule fernales, But two dozen of cach fernale’s offspring, T cy genotyped each individual Dbfmarker loci, and used the genotypes to identify each hatchling's father. Jones and colleagues sufficient information to determine the number Byand the number of offipring, for all the adults in their sample Imajority of the males, it turned out, failed to mate (Figure 10.5b, upper hose that succeeded, most mated just once or twice. In sharp contrast, females mated at least once, and most m. sd two or three times (Figure ly the males showed a pronounced disparity in number of off FPgure 10.5b, upper right). Those who failed to mate, of course, fathered no iwhile the lucky few fathered as many as 300 or more. All of the females, ing, most between 100 and 300 of them (Figure 10.5b, lower right)Ofgreatest interest is the extent to which access to mates determined repro- dhctive success. Figure 10.5c plots number of offipring versus number of mates for both sexes, along with their statistical best-fit lines. Look first at the data for Females (red dots and line). It appears that mating with more than one male may hive carried some benefit. The slope of the best-fit lin filly distinguishable from zero. F 1r males, however, the association between fhumber of mates and number of offipring was strong and highly significant (blue Hots and line). More mate -ant more offspring, This result is consistent with Bateman’ prediction. In rough-skinned newts, done pote ec the cable nf male sai sco Bsn of females. This means that heritable traits that are associated, in males, Wh flor to mate will end to disappear, while heritable tits associated with Bi saccess will tend to become common. Male rough-skinned newts devel Bp cress on their tails during the breeding season, and Jones and colleagues Hound that males who found mates displayed ificantly taller crests than males st filed. We can infer that wil crests evolved asa result of sexual selection Most evolutionary biologists believe that the pattem we have seen in rou Biber: to mates. And, as we shall see, chs asymmetry hus important conse BBL So before we explore its consequences it will be instructive to © Bison sexual selection in broad-nosed pipeis. Broad-nosed pipefish (Syrgnathus typhle) live in eelgrass beds (Figure 10.54). In Pppefsh familics, as in their kin, the seahorses, the father provides all the parental Bre The male has a brood pouch into which the female lays her eg The male Berries the eges, protects them, and provides them with oxygen and nutrients until ley hatch (Figure 10.6). Jones, Gunilla Rosenqvist, Anders Berglund, and col- Hesgus (2000) caught pipefish off the coast of Sweden before the breeding season Bin and let them mate in barrels in the lab. The data in Figure 10.5e and (9) ine fiom two experiments. In the first, each barrel contained four males and mules. In the second, each barrel contained two mules and six females. The Bon! experiment probably mimics natural conditions more closely than the Be takes a female less time to produce a clutch of eggs than it takes a male to Ber them to hatching (Berglund et al. 1989). As a result, at any given time there Beinor: females with eggs to lay than males with space to accept them. Jones and colleagues took tissue samples from the adults and the offspring, Boryped them at set of marker loci, and used the genotypes to determine each pring’ mother. As in the newt study, this enabled the biologists to determine Benumber of mates and the number of offipring for each adult The results for pipefish are similar to those for newts, except that the roles of s43es are reversed, In pipefish, more females than males failed to mate (Figure se, upper and lower left). Consequently, the reproductive success of females Bmore skewed than the reproductive success of males, with a sharper distinc- Bu berween the winners and the losers (Figure 10.5e, upper and lower right) importantly, it was in females that reproductive success depended most Bialy on mating success (Figure 10.54). In broad-nosed pipefish, sexual selec Bisa more potent force in the evolution of females than it isin the evolution isles. We nced to keep this result in mind as we consider the behavioral con Chapter 10 Sexual Selection 379 Because females typically imeot more in each offepring reproductive success Is mited by the number of However these p Figure 10.6 A pregnant male pipefish, Syngnathus typhle, gives birthPARTI Adaptation ual selection theory compete with each other over 16 to mateo, and that females wil. be choosy, Behavioral Consequences of Asymmetric Limits on Fitness An asymmetry in the factors that limit reproductive success for females males allows us to predict differences in the mating behavior of the two Consider the common pattern seen in Bateman’ fruit flies and rough-skinned newts: For males reproductive success is limited by accedl mates, whereas for females reproductive success is limited by capacity tot and rear ofipring. When this pateern holds, we can predict the following: + Males should be competitive. If the fitness of males is limited by acca females, then we expect that males will compete among themselves ree Females should be choosy. If che fitness of females is not limited by portunities to mate, but any given mating may involve the commitment byl female to a large investment in making and rearing offipring, then we ex that females will be selective about with whom they mate. These predictions have been confirmed in a great variety of animal species will look at some examples shortly At note of caution is in order, however. It is easy to get carried away Bateman and many who followed appear, in hindsight, to have done Knight 2002). Bateman thought that greater sexual selection on males th females is inherent in maleness and femaleness as such, He and others ther assumed that the optimal male strategy, in virtually any species, would be mate with as many females as possible, and that the optimal female str would be to choose one male and mate with him only. These assumptions turned out to be wrong, We will se, later in the chapter, that males of od reasons to stick with one mate and that females often have good sons to be promiscuous. Furthermore, as Jones et al.s pipefish study shows, greater sexual selectioh males than on females is not inherent in the identity of the sexes themsd ‘When access to mates is limiting for females instead of males, we predict thi males will compete with each other aver access to males and that males wl choosy. Toward the end of this chapter, we will return to stich “sex-role reve species to see if they are exceptions that can prove the rules of sexual select For now, however, we will focus on sexual selection involving male-male 6 petition and female choice Male-male competition for mates and female choosiness can play olf ‘owo ways. First, males may fight amongst themselves, head-to-head, cli claw, or antler-to-antler. Sometimes they fight over direct conteol of fe sometimes they fight over control of a resource vital to females, and times they just fight. The females then mate with the winners. This font sexual selection is called intrasexual selection, because the key event i determines reproductive success involves interactions among the members a single sex (the males fight). Second, instead of fighting the males may adil tise for mates by singing, dancing, or showing off bright colors, The fem then choose the male with the best display. This form of sexual selectiog called intersexual selection, because the key event that determines tpl ductive success involves an interaction between members of the two sexes males choose males)Chapter 10 Sexual Selection 381 10.2 Male-Male Competition: Intrasexual Selection aul selection by male~male competition often occurs when individual males fan monopolize access to females. Males may monopolize females through direct 01 of the females themselves or through control of some resource important Diemales, such as feeding territory or nest sites. Male-male competition can also Ector for no apparent reason beyond simply impressing females, In this section, consider examples of research into three forms of male-male competition: right combat, sperm competition, and infanticide. Male-male competition can take the form of combat over al traits including large body size, weaponry, and armor. Male—male cases to females, ut ako selects for tactical cleverness. Our example of male-male combat comes from the marine iguanas hus cristatus) of the Galapagos Islands (Figure 10.7). Marine iguanas Alifestyle unique among the lizards. They make their living grazing on algae fe intertidal zone. Between bouts of grazing, they bask on rocks at the water's f Basking warms the iguanas, which aids digestion and prepares them for next foray into the cold water. Marine iguanas grow to different sizes on erent islands, but, as we mentioned earlier, on any given island the ma than the females (Figure 10.88). The sexual size dimorphism in marine iguanas is an excellent example for the Hy of sexual selection, because we know a great deal about how marine igua~ Mize is affected by natural selection (Wikelki ct al. 1997; Wikelski and mich 1997). Martin Wikelski and Fritz Trillmich documented natural sele fon iguana body size by monitoring the survival of marked individuals on Bilands over one to two years, Natural selection was much harsher on Gen S than on Santa Fé, but it was clearly at work on both islands. Moreover, se fon was stabilizing, Medium-sized iguanas survived at higher rates than either iguanas or large iguanas (Figure 10.86). tential agents of this natural selection on body size are few. Marine iguanas ¢ with other species for food and have virtually no predators Figure 10.7 A Galapagos marine iguana foraging on algae in the intertidal zone382 PARTI Adaptation Figure 10.8 Natural selection Santa Fe ‘on body size in marine iguanas {@) Histograms showing the size di ‘tributions of male and female marine iguanas on two diferent Galipages Islands, Genovesa and Santa Fé. The asterisks mark the maximum sizes at which iguanas were able to maintain their weight in two different years (1991-1992 and 1992-1993). From Wikelsi eta (1997) (0) Survival rates of marked individual of citfr. ent sizes snout-vent length, mm) ‘rom March 1991 to March 1992 on Genovesa and from February 1980 to February 1992 on Santa Fé. The sam: ple sizes, or number of inaivicuasin each group, ae given by n. From \wrkelsk and Tilmich (1997). Genovesa ateceltpiigcl Fema ‘oo — SSnout-vent length (mm Size ciasses (ren) | « =] 3 Other than reproduction, about all the iguanas have to contend with is comp tion for food among themselves Larger iguanas can harvest more algae, and thus gather more energy, butt also expend more energy on metabolism, Wikelski and colleagues (1997) fo during two different years that small iguanas ran a net energy strplus, but jguanas ran a net energy deficit. Consistent with the hypothesis that the aa ity of food limits body size, the ws on Santa Fé and Genoves i weight during both 1991-1992, a bad year for algae, and 199: good year (sce also Wikelski and Thom 2000). The largest sizes at which il were able to maintain their weight are indicated by the asterisks in Figure 108 Now compare Figure 10.8a with Figure 10,8b. The maximum sizes at iguanas could sustain their weight are close to the optimal sizes for survival thermore, the largest females in each population are near the optimal size fo vival, but the langest males are much larger than the optimal size. The lan ‘of male marine iguanas is thus an evolutionary puzzle: We cannot explain it by ural selection, because Wikelski and Trillmich have shown that natural selection against it. Iv is exactly the kind of puzele for which Darwin invoked sexual set As we discussed eatlier, a crucial issue in sexual selection is the relat investment per offipring made by females versus males. In marine ig parental investment by females is much larger. Each female di awway from the basking and feeding areas, buries her eggs, guards the nest fora days, and then abandons it (Rauch 1988). Males provide no parental care all parental investment by females consists mostly of producing eggs, and paren vestment by males consists entirely of producing ejaculates. Females lay a clutch of one to six eggs each year, into which they put about 20% of th mass (Rauch 1985; Rauch 1988; Wikelski and Trillmich 1997), C female investment, the cost of the single ejaculate needed to fertilize all the eg a clutch is paltry. This difference in investment suggests that the maximum pol tial reproductive success of males is much higher than that of females. Numbet ‘mates will limit the lifetime reproductive success of males, but not females, The iguanas’ mating behavior is consistent with these inferences. Females ulate only once each reproductive season. Martin Wikelski, Silke Biurle, andi ficld assistants followed several dozen marked females on Genovesa. TheChapter 10 Sexual Selection 383 Figure 10.9 Male marine iguanas in combat Buatched the females from dawn to dusk every day during the entire fg mating season in 1992-1993 and 1993-1994 (Wikelski and Biurle ako watched the marked females from dawn to dusk every day dur- jaent nesting seasons. Every marked female that dug a nest and laid en seen copulating, but no marked female had been seen copulating once. Male iguanas, in conerast, attempt to copulate many times with Females. But the opportunity to copulate with females is a privi Juana has to fight for (Figure 10.9), Bithe mating season each year, male iguanas stake out territories on the Fe females bask between feeding bouts. In these small, densely packed WFigure 10.102), males attempt to claim and hold ground by ousting Confiontations begin with head-bobbing threats and escalate to ead pushing. If neither male backs down, fights can end with bites pus injuries on the head, neck, lanks,and legs (Trillmich 1983). Males Fettiories are more attractive to females than males that do not Number of males O12348678010 45 ‘Number of copulations JO Mating success in male marine iguanas (a) A cluster of iguana fon Camaato Islet, Galapagos. Lines show boundaries of mating Hanus 16, 1978: numbers identity territory owners. As the scale bar Imating territories are only a few square meters in size. The dark ndieates where Krisztina Tillmich sat to watch the iguanas. (Camaanio BOM of shoreline and supports a population of nearly 2,000 iguanas.) (883). (b) Histogram showing variation in number of copulations ob- Hsiguanas on mating territories shown in (a) Note the break inthe the most successful male, iguana 59, got more than four times as 25 any of his rivals. The histogram includes only males that claimed fat lest a short time during the mating season. From Tillmch (1983)384 PARTI Adaptation females congrogate, Large iguanas win more fights, thus get to copulate with more females. This pattern of sexual selection has led to the evolution of large body in males. (Trillmich 1983; Rauch 1985; Partecke et al. 2002). Because only some ‘manage to claim territories, because some males manage to maintain thei for a longer period than others, and because some territories and territory crs are more attractive to females than others, there is extreme vatiation males in the number of copulations obtained (Figure 10.10b). Because claiming and holding a territory involves combat with other bigger males tend to win. In the iguana colony that Krisztina Trillmich (1M studied on Camaafio Islet, the male that got 45 copulations (Figure 10.108 more than any other male, was iguana 59 (his territory is shown in 10.10a). His neighbor, iguana 65, was the second most succesfal with 10 tions. Both of their territories were females’ favorite early-morning and bt ternoon basking places. Trillmich reported that iguana i the colony; that to claim his territory; he had to eject four other males who to take it; and thae during his tenure, he lost parts oft to four neighboring who were pushing their territories in from the sides. Wikelski and co-Wo studied iguana colonies on Genovesa and Santa Fé (Wikelski tal. 1996; Wi and Trillmich 1997). Consistent with Krisztina Trillmich’ observations, thee searchers found that the mean size of males that actually got to copulate vali nificantly larger than the mean size ofall males that tied to copulate (Table il Ife assume that body size is heritable in marine iguanas, then we have tion, heritability, and differential mating success. These ar tion by sexual selection. We thus have an explanation for why male marine get so much bigger than the optimal size for survival. Male iguanas get big bigger males get more mates and pass on more of their big-male genes Male-male combat, analogous to that in marine iguanas, happens in variety of species, including the red deer shown in Figure 10.1 (ce Cla Brock 1985). In addition to large body size, this kind of sexual selection lel the evolution of other traits that are assets in combat, such as weaponry armor. Male-male combat can also lead to the evolution of alternative nal ing strategies (see Box 10.1) Table 10.1 Sexual selection differentials for male body size in marine iguana colonies on Santa Fé and Genovesa Body size is given as snout-vent length (SVL).The standardized selection di ferential (see Chapter 8) i the difference between the average body size of all ‘males that copulated atleast once and the average body size ofall males that ied (0 copulate, expressed in standard deviations of the distribution of body. sizes of all males that tried to copulate. (The standard deviation is the square root of the variance.) Both standardized selection differentials are positive (P< 005), indicating that males that got to copulate were larger on average than males that tried (o copulate. From Wikelski and Tillmich (1997). Average Standard Standardized size (SVL) deviation selection differen Sana Fé Males that copulated 4o1 13 All males that tried to 30 26 Genovesa Mates that copulated All males that tried to386 PARTI Adaptation Box 10.1 (Continued) Figure 10.12. Aherative mating strategies in coho salmon This gue shows a coho mating group. The large sh atthe rah (ustreo) i female that has ‘built 3 nest and is ready to lay her eggs. Downstream b from the female are four males that have opted forthe ti o> fighting strategy. These males, three hooknoses and 2 joc hove sorted themssvescutby soe Two theriacs FORBES tov opnd te aay sing. They hand ’ hiding places near the female, one behind a rock, the wv we other in a shallow. After Gross (191) (Figure 10.12). Some males fight for a position relative fitness of hooknoses versus jacks depend close to the female. These fighters quickly sort part, on the frequency of each type of male in th themselves out by size. When the female lays her breeding population. ‘eggs, the males spawn over them in order. The fist There is an important distinction becween i male to spawn fertilizes the most eggs. Other males iguana example and the coho example, In mati do not fight for position, but instead look for 3 hid- iguanas, the small nonterritorial males appear tol ing place near the female. When the female lays her making the best of a bad situation while they eggs, these sneakers attempt to dart out and spawn to a large enough size to successfully fight for ritory. In coho, a male irreversibly becomes eithet Among hooknoses, those that adopt the fighting hooknose or a jack. Which strategy a male od strategy are more successful. Among jacks, those that pursues depends on a mixture of environmental a adopt the sneaky strategy are more successful. The genetic factors. Sperm Compe! Male~male competition can Male~male competition does not necessarily stop when copulation is ove take the form of sperm real determinant of a male's mating success is not whether he copulate competition, Whether his sperm fertilize eggs. fan animal has internal fertilization, and if male mates with two or more different males within a short period, then sperm from the males will bein a race to the eggs. Indeed, females may pol liters or clutches in which differene ofipring are fathered by different Batches of offipring with multiple fathers have been documented in a vate animals, inchuding squirrels (Boellstorff et al. 1994), bears (Schenk and Ka 1995), birds (Gibbs et al. 1990), lizards (Olsson et al. 1994), and spiders (Wi 1991). It happens in humans too; Smith (1984) reviews several repors of with different fathers. Given sperm competition, what tits contribute to victory? One us right simply be the production of large ejaculates containing many sp sperm competition is something ofa lottery, then the more tickets a male buy beter his chances of winning. This hypothesis has been tested by Matthew (1991) with the Mediterranean fruit fy, Certs cpiteta (Figure 10.13). Gage periment was based on the observation that, although ejaculates are cheaper eggs, they are not five (se, for example, Nakatsuru and Kramer 1982). Gage Figure 10.13 AMeaiter. sone that if male Mediterranean fruit fies are subject to any constants on faneal ral Corebies production, they might benefie from conserving their sperm, using during capitata copulation only the minimum number necessary to ensure complete fertifemale's eggs. But if larger ejaculates contribute to victory in sperm compe mules whose sperm are a risk of competition should release more sperm copulation than males whose sperm are not at risk, Ifthe number of sperm is unimportant to the outcome of competition, then males should release Bane number of sperm regardless ofthe rsk of competition ge nised and mated male medilies under two sets of conditions. One group Pie nsed by themselves and allowed to mate in private; the other group of Be ased inthe company of another male and allowed to mate in the presence iar second male. Immediately after each mating, Gage dissected the females ounted the number of sperm the males had released. Males raised and din the presence of a potential rival ejaculated more than 2 1/2 times as By sperm (average + standard error = 3,520 + 417) as males ised and dinisolation (1,379 + 241), a highly significant difference (P < 0.0001) Bes inerpretation was that large ejaculates do contribute to victory in sperm ton and that male medflies dispense their sperm to balance the twin pri- wsecessfil fertilization and conserving sperm. Tn sdition to large ejaculates, sperm competition has apparently led to various F adaptations. Males may guard their mates, prolong copulation, deposit a cop- plug, or apply pheromones that reduce the female's attractiveness (Gilbert ches illen-Tallberg 1981;Thornhill nd Alcock 1983: fil and Taborsky 2002). During copulation in many species of damselflies, the ses special structures on his penis to scoop out sperm left by the female’ pre Beate (Figure 10.14; Waage 1984, 1986). RE, Hooper and M.. Siva-Jothy 6) used genetic paternity tests to show that this strategy is highly effective. In Mansell species they studied, the second male to mate with a female fertilized Hyal ofthe eggs produced during her fine postcopulatory bout of oviposition. inticide species of mammals, competition between males continues even beyond Reption. One example, discovered by B.C. R. Bertram (1975) and also studied (Ghig Packer and Anne Pusey (reviewed in Packer etal. 1988), happens in lions bic social unit of lions isthe pride. The core of a pride is a group of closely Bai females—mothers, daughters, sisters, nieces, aunts, and so on—and their BsAbo in the pride is a small group of adule males; two or three is a typical The males are usually related to each other, but not to the adult females. Hstem is maintained because females reaching sexual maturity stay in the they were born into, whereas newly mature males move to another pride ‘move for young adule males from one pride to another is no stroll in the Behe adult males already resident in the new pride resist the invaders. That is ales stay with their other male kin: Each group, the residents and the forms a Coalition, The residents fight the newcomers, sometimes violent- Ber the right to live in the pride. Ifthe residents win, they stay in the pride the newcomers search for a different pride to take over. Ifthe residents lose, are evicted, and the newcomers have exclusive access to the pride’s fe- exclusive, that is, until another coalition of younger, stronger, or more Bemus males comes along and kicks them out. Pusey and Packer found that Brerage time a coalition of males holds a pride isa little over two years. Be- fesidence in a pride is the key to reproductive sucess in lions, males inva Chapter 10 Sexual Selection 387 Spematheca | Bursa ‘copulatrix Figure 10.14 Sperm com: petition in damselfies uring copulation (top, the male uses the barbed horns on his penis bottom) to remove sperm left by the female's pre- vious mates. Redrawn from Waage (1984)388° PARTI Adaptation Male-male competition can take the form of infanticide By kiling other males’ cubs, When males cannot discriminate among the Figure 10.15 Lion infanticide in this photo by George B. Schaller, a male lion has just kiled anther male's cub, Which it now caries in its mouth victorious coalition quickly begin trying to father cubs. One impa quick fatherhood, however, isthe presence of sill-nursing cubs futheed of the previous coalition. That is because females do not return to dition nti afer their cubs are weaned How can the males overcome this problem? They frequently empl vious, if grisly, solution: They kill any cubs in the pride that have weaned (Figure 10.15). Packer and Pusey have shown that this strategy cubs’ mothers to return to breeding condition an average of eight than they otherwise would. Infinticide by males isthe cub deaths in the frst year of life and over 10% of all ion mortality Infanticide improves the males’ reproductive prospects, but is obvi mental to the reproductive success of the females. The females have for making the best of their own interests in this bad situation (Packet 1983). One is to defend their cubs from infanticidal males, which fe do, occasionally atthe cost of their own lives. Nonetheless, Packerand port that young cubs rarely survive more than 2 months in the preseng coalition of males. The females’ other tactic isto spontaneously abot nancies in progress when a new coalition gains residence in the pride the females’ losses: They do not waste energy and time on cubs that killed anyway shortly after bireh, With this shife in focus to female strategy, we leave the subject of male-male conflict and move tothe oh sexual selection: female choice 10.3 Female Choice There isa great variety of species in which male reproductive succes by opportunitics to mate, bat in which males are unable to monopole tnales themselves or any resource vital to females. In many. saci advertise for mates, Females typically inspect advertiscments of seen fore they choose a mate. Sexual selection by female choice leads to they of elaborate courtship displays by males. Charles Darwin first asserted that female choice i an importa selection in 1871,in The Deen of Man, and Selection in Relation t Si ‘widely accepted todiy the notion that females actively discriminate ami vidal males was controversial for several decades. Most evolutionagy thought that female discrimination was limited to choosing a male of species (ee Trivers 1985). Beyond allowing females to identify a mae species, male courtship displays were thought to function primanlyingChapter 10 Sexual Selection 389 j general female reluctance to mate. Once ready to mate, female would ac~ any male at hand. Be begin ths section by describing ewo sets of experiments that demonstrate females are in fact highly selective, actively choosing particular males from png the many available. We then consider the functions of female choosines. tal benefits to a choosy female include the acquisition of good genes for iipring-and the acquisition of resources offered by males. Alternatively, fe imay prefer male displays that exploit preexisting sensory biases bul into females’ nervous systems le Choice in Barn Swallows nt demonstrating active female choice comes from the work 5 Moller on barn swallows. Barn swallows, Hinindo rustica, are small in- geting birds that breed in colonies of up to 80 individuals. The swallows (1988) studied breed in Denmark during spring and summer, after spend- fie European winter in Africa. Upon arriving in a Danish breeding colony, imal swallow sets up a territory a few square meters in size, He then tries to mate by displaying his tail while perching and flying. Each female visits pil males, then chooses one to pair with. Once paired, the male and female ther build a mud nest in the male's territory. In this nest the pair raises one y c before summer's end, two clutches of young. The female in the eggs by herself, but both parents feed the chicks. fist glance, barn swallows may not seem promising subjects for a study of selection, The fact that the males help care for the young should tend to fe parental investment by the two sexes, and the fact that the swallows appear Bate monogamously suggests that neither sex should be in short supply for the Ban swallows are, however, sexually dimorphic. The males are mote brigh sad they tend to be slightly larger (Figure 10.16a). The difference is that the outermost tal feathers, which are elongated in both additional factors suggest that sexual selection may be at work in barn pws. First, even in a monogamous species in which both sexes care for the males and females may vary in quality, both as parents and as donors of {0 offipring. The members of both sexes should thus benefit, in higher a8 a ‘lle Ah "a nth) Figure 10.16 Sexual di- ‘morphism in barn swal- ows. (a) A female (eft) and a male (ight). Note that the male has richer col (rs, and longer outer tal feathers. (b) The distribu: tion of tal streamer length in male and female barn swallows, The distributions overlap, but the average male has longer tail feath- fers than the average fe- male. From Moller (1991).390 PARTI Adaptation to mate with malee whose tail feathers are long} reproductive success by trying to identify and attract che best mite Second, although barn swallows appear to be monogamous, may 4160 sometimes solicit copulations with females other than their pain these offers of extra-pair copulation are accepted by the females, a 0% they are, the successful male may benefit by fathering offpring that 0m male will help to raise. Third, the trait that differs most prominently b sexes—the length of the long outer tail feathers—is precisely the tat show off when advertising for mates. Moller hypothesized that sexta does indeed occur in barn swallows, through the mechanism of femal and that females prefer to mate with males displaying longer tail athe Moller captured and color-banded 44 males that had established bur not yet attracted mates. He divided them at random into four gr and altered the tail feathers of each group as follows Shortened tail feathers. Moller clipped about 2 centimeter oul middle of each outer tail feather, then reattached the feather tips ti with superglue, Mock-altered (control I). Moller clipped the tail feathers, This did not change the length of the feathers, but ei subjected the birds to the same handlin, and lengthened groups. Unaltered (control 11). Moller captured and banded these bint nothing to their tal feathers. Elongated tail feathers. Moller added, by clipping and gluing, the feather removed from the shortened group into the middles of thet ers of these birds Moller then released the birds back into their colony: He predicted males prefer males with longer tail feathers, chen the males with el would attract mates sooner, and fledge more young, than either contol control II males. The control males should, in turn, be more succes shortened males. If, on the other hand, the females have no preference tail feathers, then there should be no differences in success among the ‘The results appear in Figure 10.17. The elongated males, on aveag mates more quickly than the control males, and control males atta more quickly than shortened males (Figure 10.172). Among the adv tracting a mate quickly is that the male and his parmer can get an eae rearing a clutch of chicks. An earlier start means that the parents atm to have time to muse a second clutch before the summer ends (Figare nally, the advantage of raising two clutches is that the parents ledge ma over the course of the summer (Figure 10.176). The female barn swallows bad additional opportunities to be choo cr they were solicited by males for extra-pair copulations. Moller 4 watch the birds closely enough to estimate the rates at which the mal attempted to copulate with females other than their pai-mate, the the females they solicited accepted these copulations, and the rates af study males'pair-mates copulated with other males. The males in the il groups showed no differences in the rates at which they attempted t0 pair copulations. Nor did their pair-mates differ in the rates at which tChapter 10 Sexual Selection 391 PrP P Pe” Pe Pr PIA 2 100: 10 g : : 375 $ 6 3 so ot 2 4 8 2s- | 3 | 2 panes ue a o | ° Jes Bees can Govan Soranes —'Srranes Corea Conran ener Go Corian nosed Tal vearment Tal treatment Tal weatment 10.17 Anders Moller data on reproductive success and tail length in male barn swallows (a) Length of time by males to attract a mate. Red dots represent the average for each group; the vertical lines represent the standard l(a atta measure of the amount of variation within each group). There were significant differences among {P< 00). () Differences in time required to attract a mate carried over into which males had time to raise a second (chicks, Bar height indicates the percentage of males in each group who, with thelr mates, raised two dutches. Again, significant variation among groups (P< 0.02). (c) Differences in second-clutch success carried over into reproductive renumber of hicks each male had fledged by summer's end. The meanings of dots and lines are the same a in (a. significant variation among groups (P < 0,001). Elongated males fledged more chicks than control males, who in asd moce chicks than shortened males. From Molle (1988) ‘The males in the groups did differ, however, in the rates at which the Par females they propositioned actually accepted them for copulation 12) Furthermore, the males’ pait-mates differed in the rates at which pied extr-pair copulations with other males (Table 10.2). Apparently Ivho had to settle for less desirable short-tailed males attempted to com Uy copulating out-of-pair with more desirable long-tiled males. Thanks females, the long-tailed males won again, at che expense of the short-tal Bote on extra-pair copulations in bieds and other animals, see Box 10.2.) ws experiment demonstrates that female barn swallows are choosy. As predicted, females prefer pait-mates with longer tal feathers. And for males, fone ofthe more desirable mates results in higher reproductive succes. Choice in Gray Tree Frogs Bob set of experiments demonstrating active female choice comes from SEH, Carl Gerhardt and colleagues on gray tee frogs. Gray tee frogs, is live in woodlands inthe eastern United States, During the breeding Extra-pair copulations in Moller's barn swallow experiment Pers reported are rates, measured as extra-pair copulations per ales give statistical significance of variation among groups. rom 988), Male Tail Treatment Shortened Control Lengthened P tails 1 ils 0 0 0 0.090 <0001 003% = (ood om <001PART III Adaptation vs. Shor call ## Females choosing: 30 Long ‘Shor call ‘Females choosing: 38 1 Figure 10.20 Gerhardt et a's data ‘on the preferences of female gray tree frogs (2) Most females prefer long calls toshort cals, even when the short calls ace initially louder (P < 0.001). After Ger hard el al. (1996). (b) Most females wal pass a loudspeaker playing short cals to approach a loudspeaker playing long cals P< 0001), After Gecharat el al. (1996). (© Females discriminate most strongly agent short calls. female's “phonotax sis score" for a particular test call isthe time ittook her to approach a control call with 18 pulses per second divided by the ‘ime it took her to approach the test call Higher scares indicate a stronger prefer lence for the test call relative to the con trol. Each data point isthe average score ‘of 10 females; the whiskers show 11 stan dard error. From Bush eta (2002). ° s 5 é hers found that 38 of 53 females ( speaker to approach the long-calling speaker. Female gray tre frogs prefer Sarah Bush, Gerhardt, and Johannes Schul (2002) quantified female i to matewith males that he more precisely: By placing females in front ofa single loudspeaker, plying @ Tonger call and mateo that ad noting how enthusiastically each female responded to experimental ll erick caterncpan, $8.2 Standard control call, the researchers learned that females ate pati ss : rly _disdainfal of unusually short cals (Figure 10.20¢).As calls got longer, femal less choosy—although they do still prefer songs with more tris. Ths consistent with observations made under more natural conditions by J Schwartz, Bryant Buchanan, and Gerhardt (2001). In the midst of a chon males, and other background noise, females are less able to make fi ‘inctions among males. But they do continue to discriminate, especialyChapter 10 Sexual Selection The experiments of Gerhardt and colleagues show that female gray tree frogs hoon, preferring, as predicted, males giving longer call. And they show that s-alling males are more likely to find mates Female choice, as illustrated by barn swallows and gray tree frogs, is thought be the sclectve force responsible for the evolution of a great variety of male sement displays—fiom the gaudy tail feathers of the peacock, to the ng of crickets, to the leg cuts of wolf spiders. Some male displays, like Of peacocks, are loud and clear; others, like those of barn swallows, a sublle. I is curious that an extra 2 em added to two tail feathers should male barn swallow so much more attractive to females 28 to dramatical~ gove his reproductive success. Why should the females care about such a diference? And for that matter, why should females care about any of the ements, cven the loud ones, that males use to attract mates? We will con: three explanations. sy Females May Get Better Genes for Their Offspring Posibilty is that the displays given by males are indicators of genetic qual inales giving more attractive displays are genetically superior to males Biractve displays, hen choosy females will secure better genes for their off (Fisher 1915, Williams 1966, Zahavi 1975). son Welch and colleagues (1998) used an elegant experiment to investigate mule gray tree frogs giving long calls are genetically superior to males short calls (Figure 10,21). During two breeding seasons, the researchers fed unfertilized eggs from wild females. They divided each female's clutch Separate batches of eggs, then fertilized one batch of eggs with sperm from a alling male and the other batch of eggs with sperm from a short-calling Mhey reared some of the tadpoles from each batch of eggs on a generous iid the others on a restricted dict. experimental design allowed Welch and colleagues to compare the fitness les that were maternal half-siblings—that is, tadpoles with the same abut different fathers. When comparing tadpoles fathered by long-41 BUnokais Figure 10.33 Mating and reproductive suc- cess of unokais (killers) versus non-unokais among Yenomamd men These graphs show the average number of wives (top) and children a — | (bottom) for adult men of various ages. Taken to- =} gether the data show that unokais are more suc: is (P ~ 0.00001). Plotted from data in Chagnon (1988) 2024 25-90 8140 > Cast than norruna Chapter 10 Sexual Selection 411412 PARTI Adaptation 20 Kids: None wt east 208 30s 40s 50s ‘Age years) Figure 10.34 Men with chi- dren are taller, on average, than childless men. The heawy horizontal bars show, fora sam ple of Polish men, the average height of individuals in each age dass, by whether oF not they have children, The colored boxes show #1 standard deviation around the mear: the whiskers show the range about the mean that includes 95% of the men. |s the number of men in each category. Though small, the dt ferences between men with versus without children are sta Aistically significant for men in their twenties (P = 0,005), thir ties (P = 0.001), and forties (P = 0.002). The difference is ‘ot significant for men in their fifties (P = 0.863). From Pawlowski eta. (2000) sexual dimorphiem in body compete for mates, but larg. males do not necessarrly win limited data sugges he Summary colleagues then compared bachelors to married men. The married men taller by a slight bue statistically significant margin. In addition, men with one more children were significantly taller than childless men (Figure 10.34) exception to this pattern was the group of men in their fifties, within whig there was no difference in height between fathers versus childless ml Pawlowski and colleagues note that the men in their fifties reached mary ge shortly after World War II, when the ratio of women to men in Poland unusually high. The researchers speculate that the men in their fifties had exp rienced less intense sexual selection, through female choice, than is the nor Additional evidence suggesting that female choice favors tall men comes fi a study by Ulrich Mueller and Allan Mazur (2001). Mueller and Mazur sy members of the class of 1950 from the United States Military Academy a Point. Among these career officers, unlike in many other more diverse popt tions, height was not associated with social status or socioeconomic se Height was, however, associated with reproductive success. The tallest men ‘over their lifetimes, more wives, and younger second wives, than other men. result, che tallest men had more children Daniel Nettle (2002) examined the relationship between height and repro success in women. Analyzing data from a lige national health survey in Bil Nettle found a weak but significant effect. Unlike in men, selection on wom stabilizing: Women of slighty less than aver shorter or taller women. The cause appears to be that women of moderate hi were healthier, on average, than extremely short or extremely tall women. Thal the higher fitness of lightly shorter-than-a is due to natural sli (On che data we have reviewed, then, there is some evidence that sexual tion, primarily through female choice, is responsible for the fact that men ae than women, It is best, however, to consider the evolutionary significance off al size dimorphism in humans unresolved. The studies we vational, and the associations they documented are small. As a result, che evil they provide about causation is only suggestive. It is possible that we humans i ply inherited our sexual size dimorphism from our ancestors, who were mone ually dimorphic in size than we are (McHenry 1992). What is really needall settle the issue is data from a larger number of cultures on the relationship bet body size, number of mates, survival, and reproductive success for both Wome ‘men, Preferably, the data would come from hunter-gatherer cultures, whose mi bers live the lifestyle ancestral for our species. The most technically challenging& tor to measure accurately is the reproductive success of men. Modern techi for genetic analysis have made it feasible, in principle, o collect such dita (5 10.18, page 392). However, the research remains to be done. Sexual dimorphism, difference in form or behav- win invoked sexual selection. Sexual selection eg ior between females and males, is common. The differ-_ferential reproductive success resulting from ‘ari ence often involves. traits, like the enormous tail in mating succes. feathers of the peacock, that appear to be opposed by Mating success is often a more important del natural selection. To explain these puzzling traits, Dar-_ nant of fitness for one sex than for the other, Us