SEXUAL SELECTION IN BIRDS

JUSTIN W. MERRY

INTRODUCTION

Buried within The Origin of Species is a two-page section in which Charles Darwin outlined a subset of his evolutionary theory that he coined "sexual selection." He defined it as the "struggle between individuals of one sex, usually males, for the possession of the other sex," (Darwin 1859). Several years later, Darwin (1871) published a second book devoted largely to the subject of sexual selection, titled The Descent of Man, and Selection in Relation to Sex. In it, he elaborated on his theory, presenting many examples and putting forth some preliminary models for the development of sexually selected traits. Over the past century, sexual selection has become one of the most intensively studied fields of evolution (Møller 1994). Its action has affected an enormous range of animals, from insects to "higher" vertebrates. For example, females of two species of stalk-eyed flies, Cyrtodiopsis dalmanni and C. whitei, have been shown to prefer males possessing large eye-stalks (Wilkinson et al. 1998). Studies have also shown that humans have "subconscious" mate preferences (Thornhill and Gangestad 1996). Men generally prefer women with .7 hip-to-waist ratios, small jaws, plush lips, large eyes, and firm, symmetrical breasts. Women, on the other hand, generally have been shown to prefer broad foreheads, large jaws, strong chins, above average upper-body musculature, and a .9 hip-to-waist ratio (Cowley 1996).

Birds have proven to be one of the more interesting groups of animals to study. The secondary sex characters of many avian males have evolved as astonishingly beautiful designs, as exemplified by the enormous fan-like tail of the peacock or the enormous variety of birdsong one hears in the meadows and forests in springtime. In this paper, I summarize some of the important concepts and models of sexual selection, focusing primarily on mate choice, and demonstrate their action using a variety of examples from the ornithological literature. I hope to demonstrate the diversity of characters that have developed through sexual selection, as well as provide some insight into the mechanisms driving their evolution.

TYPES OF SEXUAL SELECTION

Møller (1994) defined sexual selection as occurring "as a result of a non-random association between a (secondary sexual) character and a component of mating success."  This association will cause the character to be selected in a population, regardless of whether it is otherwise useful to an individual. In fact, a character could be potentially detrimental to the survival of an organism and still be selected, provided the costs do not outweigh the benefits of sexual preference.

Darwin (1871) defined two principle types of sexual selection. The first, which he called "male-male contests", involves the competition between two members of the same sex (usually males) for the possession of, or the right to mate with, members of the opposite sex. The penalty for the loser of a contest is usually not death, but rather the inability to mate and pass on his genome in the form of offspring. Mammals such as deer, in which males develop large antlers on their heads, exemplify this type of selection. These secondary sex characters are used in combat with other males to determine dominance hierarchies. However, this type of contest also occurs among birds. Males of many species of galliforms have spurs on their legs or wings that are used in contests with other males (Møller 1994). Birdsong also can be viewed as a male-male contest. Territorial males use bird song as a mechanism to stake out their territory. Strong singers are able to defend their territory from other males with their voice without having to resort to physical combat. A weak vocalist, however, may be challenged by a roaming male and be forced to battle physically for his territory (Gill 1995). Loss of territory will result in the loss of breeding opportunities for the male, and as a result, that male’s genome will not be passed on to the next generation.

The second type of sexual selection is known as "mate choice." Under this model, a member of one sex, usually the female, selects a member of the other sex based on some character or behavior that makes him more appealing than other nearby males. Darwin (1871) felt that female animals had a sense of "aesthetics" or "beauty," and therefore more beautiful males would be selected because they were more attractive. Though such an anthropomorphic mechanism is not justified, females do demonstrate preferences for certain exaggerated characters in males. Male peacocks display their extravagant fan-like tails in the presence of females. The females, in turn, will choose a mate based on the quality of the tail. Therefore, females will select males with the most extravagant tails, and their genomes will be passed on to future generations.

In practice, these two fundamental types of sexual selection are often difficult to separate. For instance, male ring-necked pheasants have large spurs on their legs that are used in male-male contests. However, von Schantz et al. (1989) demonstrated that females of this species prefer males with longer spurs. Therefore, the spurs play a role in both male-male interactions and mate choice. Birdsong can also be an ambiguous character to classify. Though mentioned above as a type of male-male contest, it has also proven to be very important in mate choice. Male quality can be assessed not only by competing territorial males, but also by females as a means of determining the best available mate (Nowicki et al. 1998).

THEORETICAL MODELS OF MATE CHOICE

In the great majority of species displaying sexually selected characters, it is the male who must compete with rivals for the opportunity to mate. Why aren’t males as selective in the choice of mates as females tend to be? Bateman (1948) demonstrated the probable answer to this by observation of Drosophila reproduction. The number of offspring fathered by a male increase steadily with the number of females he inseminates. However, female fecundity remains constant regardless of the number of mating events she experiences. Therefore, a promiscuous male will have a large number of offspring carrying his promiscuity trait, whereas a promiscuous female gains nothing when compared with a monogamous female. In fact, the extra time spent mating may actually be a costly expenditure of time and energy that makes the promiscuous female less fit. Because of the fecundity constraints placed on a female, a female that selects high "quality" mates will have more fit offspring and will consequently out compete her less selective conspecifics. Males, on the other hand, have virtually nothing to lose from mating with every receptive female they encounter (Andersson and Iwasa 1996).

The selection of the "highest quality," or fittest mates as alluded to above can have many meanings. Fitness was defined by Mills and Beatty (1979) as the "propensity" of an organism "to survive and reproduce" viable, fit offspring in that organism’s particular niche. The fitness of a female, therefore, will be enhanced if her male offspring are more likely to survive and reproduce than other conspecifics. In polygynous species, only a few males will achieve copulations with a particular population of females. In some cases, females will occur a fecundity cost in order to mate with a dominant male. For example, in red-winged blackbirds, Agelaius phoeniceus, the preferred males (males achieving the largest number of copulations) provide little parental care, while males with fewer mates provide significantly more parental care. Why do females consistently continue to choose the "preferred" male? Weatherhead and Robertson (1979) proposed the "sexy-son hypothesis," in explanation. In their model, females choose the preferred red-winged blackbirds because their male offspring will later be preferred, either because they out compete other males in a dominance hierarchy or are simply more attractive to females. This benefit outweighs the cost of decreased fecundity caused by single parenting. Workers' attempts to genetically model this type of system have experienced mixed results. Some have confirmed the plausibility of the system (Curtsinger and Heisler 1988), while others have found it unworkable (Kirkpatrick 1985).

The "good genes" model describes another possible motivation for female preference. A high quality mate, that is, a strong, dominant mate in good health, is strong and dominant because he has a "good" genome capable of defending itself against predators, disease, and competition with other males. Male-male interactions fit very nicely into this hypothesis. Males that win the most contests will be the most dominant males, and consequently will have greater exposure to mates. A secondary sex character that allows a male to gain an advantage either in deterrence of competitors or in the male-male fights themselves will allow that male to gain greater access to mates, and consequently his genes will out compete those of other males. Over time, exaggerated traits such as antlers, spurs, or territorial song will develop in the population.

Mate choice can be based on the search for "good genes" in a mate as well, but the mechanism of choice is less clear. How is it that females choose one male over another? What roles do the exaggerated secondary sex characters displayed by many males play in this choice? Numerous models have been proposed to explain the mechanisms underlying mate choice. I will discuss two below.

The handicap model, popularized by Zahavi (1975, 1977), attempts to explain how the exaggerated characters exemplified in such animals as male peacocks and the long-tailed widowbird are used in mate choice (Pycraft 1914; Cronin 1991; Andersson 1982). In his model, these exaggerated traits are costly to the organism. It is clear that male peacocks have limited locomotor abilities, which could impair their ability to forage or escape from predators. A male who is able to sustain himself with food, defend himself from rivals, avoid predation, and still carry around a large tail indicates that he has "good genes." The more substantial the handicap, in this case a tail, the higher quality the male's genome, and consequently the more attractive he is to females. Examples of handicaps are numerous among birds. Bright, conspicuous coloration, exaggerated feather ornaments, and long, easy-to-locate songs all can be viewed as handicaps. Though costly to their possessors in terms of viability, the advantages of increased progeny keep the handicaps, as well as female preferences for handicaps, prominent in the population.

Clearly, enormous tails were not present in ancestral peacocks. How did these monstrosities develop in the first place? Møller (1994) suggested that many "hidden preferences" exist in populations that have not yet had the occasion to become important in mate choice. Individual females in a population may be attracted to a wide variety of characters. If an ancestral female peacock happened to be attracted to long tails, and that character happened to be an indicator of high quality males, than her offspring will be composed of high quality males with long tails and high quality females with a preference for long tails. Such a combination results in sexual selection for not only long tails in males, but also preference for long tails in females. This process, first proposed by Ronald Fisher (1915, 1930) and coined "runaway selection," would cause rapid coevolution of both a trait and the preference for it to ever more extreme expressions. As long tails became increasingly costly for male peacocks, only the fittest males would be able to maintain the trait and still survive to reproduce. At this point, the trait can be interpreted as a Zahavi handicap. This handicap eventually limits the expression of the trait, as an overly exaggerated trait would be fatal to even the strongest of males (Getty 1998).

Andersson (1982) conducted one of the best experiments supporting the handicap model. Male long-tailed widowbirds, Euplectes progne, have enormous tails that can reach over a half-meter. In his study, Andersson experimentally altered the tails of male widowbirds, either shortening or lengthening them. Females dramatically favored the males with longer tails over both the control males and the shortened-tailed males, as would be expected if females prefer larger handicaps. It should be noted that an alternative hypothesis was offered to explain the preference: perhaps long tails are chosen instead simply because they make the males more easily detectable. A more visible male would be better able to advertise his location to females, and thus he would achieve more copulations.

The handicap model is intuitively appealing and straightforward. However, it relies on "honest signaling" by males of their fitness levels. This entails that animals possessing a handicap have marginal fitness, and therefore can just sustain their handicap and maintain viability. A more taxing handicap would prove fatal to the male, while a less extravagant handicap would also lower fitness by reducing the animal’s appeal to females. Though it is generally the case that "better males do better by advertising more," one must be cautious (Getty 1998). Frequently, secondary sex characters are indirect measurements of quality, and may scale allometrically with a true indicator of fitness. For example, in widowbirds, tail length is a selected trait indicative of male quality (Gill, 1995). However, in truth, body size may be the actual determinant of viability. A larger male should be able to support a longer tail. Before one states that tail length is a handicap indicating male quality, however, it must be demonstrated that larger males actually do have relatively larger tails and that these tails contribute an added cost to the organism. Though it seems very likely that this is indeed the case, a worker must experimentally determine the validity of his assumptions before any other tests can be done.

For example, intensity of coloration is frequently considered a handicap for male birds. Presumably, more intense reds, yellows, and blues will make a bird more conspicuous to predators, and thus will result in a handicapping cost. However, recent studies have demonstrated that this might not necessarily be the case. The "unprofitable prey hypothesis" suggests that predators will avoid the most brightly colored prey and instead attack more drab individuals because brightly colored prey are generally more fit and better able to escape predation (Baker and Parker 1979). This hypothesis has proven difficult to test, however. Götmark (1992, 1996) has conducted several tests with mounted specimens of pied flycatchers (Ficedula hypoleuca) and European blackbirds (Turdus merula). In the flycatcher study, Götmark placed two mounted specimens, one male and one female, several meters apart in a field and found that sparrowhawks (Accipiter gentilis) avoided the brightly colored male flycatcher, and instead attacked the drab, dull colored female. It was hypothesized that the brightly colored male was avoided because its bright coloration advertised its high fitness, and consequently, a high ability to escape attack (Götmark 1992).

In a second study, Götmark painted red spots on a mounted specimen of one European blackbird and placed it in a field along with an unpainted (normal) mount. Once again, sparrowhawks avoided the brightly colored specimen and instead attacked the unpainted mount (Götmark 1996). These results could also be a function of the typical hesitancy that many avian predators demonstrate when presented with a novel prey type (Schlenoff 1984). An ideal test of the unprofitable prey hypothesis would not rely on sexual dimorphism or the generation of new conspicuous traits, but a dampening of already-present traits. For example, it could be interesting to test how often a male American Goldfinch, Carduelis tristis, in its breeding plumage is attacked compared to a drab male in winter plumage.

Another model for mate choice, referred to as "fluctuating asymmetry", involves the overall symmetry of an organism. In this model, a strong, healthy animal will develop symmetrically, meaning that characters on one side of an animal will be of equal size and shape as those on another organism. Disease, malnutrition, and injuries will impair an animal’s ability to develop symmetrically. Random deviations from perfect bilateral symmetry caused by these stresses are defined as fluctuating asymmetries. Organisms with better genomes would hypothetically be better able to develop symmetrically, and thus have lower degrees of fluctuating asymmetry (Watson and Thornhill 1994). Over the past decade, the scientific community has developed great interest in symmetry, as it seems to be a universal determinant of attractiveness around the animal kingdom, from invertebrates to humans (Thornhill and Gangestad 1996).

Fluctuating asymmetry has also been found to be important in mate choice in birds. It has been postulated that paired sexual ornaments are costly to produce, and therefore are an "honest" indicator of developmental stress (Møller 1990). It would therefore be expected that bilaterally expressed ornaments in male sexually dimorphic birds would exhibit, on average, greater asymmetry than homologous structures found in female conspecifics. Møller and Hoglund (1991) investigated skins of 16 different species of birds from 16 different families in attempt to validate this theory. Of the birds, 15 had ornamental tail feathers, while one had ornamental wing feathers. As expected, they found that in all cases males exhibited greater degrees of asymmetry in their sexual ornaments than those found in homologous structures in females. Though this raises fluctuating asymmetry as a possible determinant of female-choice, behavioral choice tests would be necessary to determine the importance of its actual role in these species.

Møller has since undertaken study of the barn swallow, Hirundo rustica. Males of this species exhibit long, paired tail streamers that average lengths of 107 mm. Females also have tail streamers, however they are slightly shorter, averaging 90 mm (Møller 1994). To determine whether fluctuating asymmetries played a role in female choice in this species, he experimentally manipulated male tail streamers to create males with varying degrees of asymmetry. Symmetric males were created by experimentally clipping and re-gluing both tail feathers to be the exact same length. Asymmetrical males were generated by clipping and re-gluing the tail feathers so that they varied in length by an average of 23 mm. Møller found that males with symmetric tail feathers formed pair bonds faster than those with asymmetric streamers.

Some studies have found that even "unnatural" variations in symmetry can influence female preference. One such study involves the zebra finch, Taeniopygia guttata, which involved the placement of colored bands on the legs of males. Males that had symmetric leg band colorations (Swaddle and Cuthill 1994) were preferred over asymmetrical males in which the relative position of each color was switched top-to-bottom.

Not all studies of such characters have demonstrated preference for symmetry, however. Jennions (1998) tested female choice for male zebra finches with symmetric (2 bands per leg) and asymmetric (3 bands on one leg, 1 on the other) leg-band combinations. Surprisingly, no preference was detected for either leg band combination, despite the strong leg-band preferences found in other studies (e.g. Swaddle and Cuthill 1994). Male paradise whydahs, Vidua paradisaea, have very long, elaborate tail feathers that can reach lengths of a half-meter. Oakes and Barnard (1994) conducted an experiment on the retrices of these African birds similar to Møller’s (1992) experiment on barn swallows. They found that instead of preference for symmetry, females preferred males with asymmetric tail feathers. The authors attributed this result to a longer maximum tail length in asymmetric males, though they did not follow-up this hypothesis with experimentation.

The female’s search for "good genes" in their mates represents the most widely acknowledged theory for the basis of mate choice. Preference for handicaps and assessment of fluctuating asymmetries are two of the mechanisms that females employ to assess the quality of their mates. These are by no means the only methods of mate choice, but are the most heavily studied areas. One interesting alternative is called "sensory bias." Under this hypothesis, females of a species are predisposed to a particular preference by constraints imposed by unrelated traits (Ryan 1998). This idea is not necessarily new: it is the basis for Fisher’s runaway selection model. However, this hypothesis predicts that preferences for a trait will exist without any sort of selective advantage to the development of these preferences. This can occur because females of a species would already be predisposed to exhibit a favorable response to a particular stimulus, and males that develop the desired trait would be selected regardless of the quality of their genome.

A study indicating the validity of this theory was conducted recently by Jones and Hunter (1998). Male least auklets, Aethia pusilla, unlike many of their close relatives such as the crested auklet, do not possess crests.  Male pygmy auklets, A. pygmaea, however, possess elaborate crests that are involved in mate choice. The workers demonstrated that female least auklets reacted more strongly to model males with crests than those without. Thus, it appears preferences for crests evolved prior to the divergence of these two species, though the crest trait did not evolve in A. pusilla.

BIRD-SPECIFIC ISSUES INVOLVING SEXUAL SELECTION

Many species of birds have evolved monogamous behaviors, in which males and females form pair bonds that last throughout the entire breeding season. These behaviors may have evolved out of necessity. Theoretically, if resources are so limited that both males and females have to rear the offspring, rather than only one parent, the advantage a male would receive from multiple copulation events would be nullified (Johnson and Burley 1997). If an unbiased sex ratio exists in the population, all males should be able to acquire a mate under a monogamous system, eliminating the non-biased variation in mating success required by many models for the development of sexual selection (Møller 1994). Therefore, males would not be expected to develop exaggerated characters, but rather would exhibit cryptic plumage and characters much like those that are typically present in females.

A great number of monogamous species, however, exhibit sexual dimorphism as great as in polygynous species. For example, male mallards, Anas platyrhynchos, have a stunning green head, a red neck, and a mostly-gray body. They completely lack the brown mottled plumage of females. However, mallards are monogamous (though males will seek forced copulations when the opportunity presents itself), forming pair bonds before winter which lasts all the way into the spring breeding season and finally break during the first week of incubation (Williams 1984). What fitness advantage does a monogamous male possessing exaggerated characters receive?

Darwin (1871) addressed this problem by noting arrival times at breeding grounds. He found that males typically arrive first, followed by a steady flux of females. As females arrive, males compete through song, plumage, and courtship dance for mates. Presumably, the healthiest females would arrive at the breeding grounds first. Therefore, the males who are preferred by either male-male contests or mate choice will form pair bonds with the highest quality females. Their offspring, subsequently, would possess higher fitness, and consequently would out compete rivals from lower quality parents. Genetic models have been developed that demonstrate that these sorts of conditions will result in sexual dimorphism through sexual selection (Kirkpatrick et al. 1990). However, the extent to which exaggerated secondary sex characters can evolve is severely limited by other selective pressures on males. Males involved in protection of a female and the rearing of young must not be hindered to such a degree that they are unable to perform these tasks. Such a handicap would reduce the fitness of both parents.  Therefore, in regions with low resource availability males typically have less-exaggerated secondary sex characteristics than do males from regions of abundant resources (Futuyma 1998)

If males are to mate only once, it follows that they too may also evolve mate choice behaviors (Darwin 1871). This has been confirmed by Kirkpatrick et al.’s (1990) genetic monogamy model, and has been experimentally demonstrated in at least one species of bird, the Crested Auklet, Aethia cristatella. In this seabird, both sexes display crest ornaments. Jones and Hunter (1993) conducted field choice tests on both sexes with "realistic models" that could be manipulated to display varying crest lengths. They determined that both males and females reacted more energetically to opposite sex models with more accentuated crests, demonstrating "mutual sexual selection."

Birdsong has been widely attributed as an indicator of quality. As discussed above, high-quality vocalists will not only deter territorial disputes with competing males, but also are selected by females as mates. How is it that high-quality vocalists are determined? Many studies have shown that individual males that exhibit extraordinarily complex songs or have an extremely large repertoire will be  selected (Gill 1995). For example, Mountjoy and Lemon (1996) demonstrated that female European Starlings, Sturnus vulgaris, preferred male starlings with larger vocal repertoires, regardless of the territory the males possessed. In addition, they found strong correlations between the health and age of male starlings and their vocal repertoires.

It is difficult to classify birdsong as a handicap, as costs seem relatively small. The seemingly most significant include locator signals to predators and lost time devoted to learning new songs. Time can be an expensive commodity, however, and it could be that birdsong, or rather the learning of new song, is therefore handicapping. It is also possible that birdsong, is unique and should be classified separately. Instead of conveying significant costs on a bird, it could be that birdsong is a handicap-free indicator of mate quality. Males that have better genomes would have more time to devote to learning songs as they would have to spend less time foraging and/or maintaining territory.

CONCLUSIONS

The theory of sexual selection began with Darwin’s (1859, 1871) writings, in which he proposed a model explaining the sexual dimorphism seen in many animals. Through most of the following 100 years, sexual selection was rarely discussed, and when it was, the discussion was very theory-laden with very little empirical backing (Møller 1994). Over the past twenty to thirty years, however, sexual selection has become a very exciting field, with many quantitative studies being published using new technologies, mathematical models, and statistical analyses. The aves have proved to be one of the most intensely studied taxonomic groups, as they contain some of the most extreme and fascinating examples of sexual selection on the planet. There are many competing theories and models that attempt to explain how sexual selection operates and evolves, most of which have empirical backing in numerous taxa. No one theory or explanation is universal. Instead, different theories, or in some cases combinations of theories, are accurate in different cases. Rather than siding exclusively with one theory or explanation, scientists must keep in mind all competing explanations as we attempt to describe the complex system of interactions involved in mate choice.

LITERATURE CITED

Copyright © 1998-1999 Justin W. Merry All Rights Reserved

Back to Thoughts on Biology