Evolution and Significance of Sexual Display in Ptilonorhynchidae
Publication or External Link
There is spectacular diversity of sexual displays among taxa. Charles Darwin was first to suggest that these traits have arisen through sexual selection (Darwin 1871). The theory of sexual selection addresses competition for mates that occurs within and between sexes (i.e. mate choice) and leads to variation in individual reproductive success. Darwin's insights into display evolution and mate choice were largely rejected by Alfred Russell Wallace (1878), who independently conceived of Natural Selection, and were controversial for decades but are now a cornerstone of evolutionary theory. The important role of sexual displays in mate choice has been demonstrated by extensive research in many species (e.g. Andersson 1994).
Although the importance of sexual displays is well established, display evolution remains a controversial topic and active area of research. Multiple hypotheses have been offered to explain the evolution of sexual displays. The "good genes" hypothesis suggests displays communicate reliable information about male quality that allows females to make informed comparisons between potential mates and choose the best one (e.g. Fisher, 1915; Hamilton and Zuk,1982; Maynard Smith, 1976). These indicators of male quality may become costly to produce and maintain and expose the bearer to increased risk. The "handicap" hypothesis predicts that the inherent cost of certain displays and the variation among males in their ability to bear those costs ensure that the traits are honest indicators of relative male quality (Zahavi 1975, 1977). "Runaway selection" is a competing hypothesis that suggests displays evolve through a non-adaptive process that, in most models, is sustained by genetic correlation between display and preference (Fisher, 1930; Lande, 1981, 1987). Display elaboration via run-away is driven by open-ended preference and limited by natural selection. Sensory drive is a theory of signal evolution that addresses display design or form, including characteristics such as color, pattern, and frequency (e.g. Lythgoe, 1979; Endler, 1992a,b; Endler and Basolo, 1998). Sensory drive shapes displays to maximize communication efficiency within the particular constraints of local habitat parameters and sensitivities of the sensory system that receive the display. Thus, differences in display design are largely attributed to variation in habitats and sensory abilities among taxa. For example, display color differences among bird species may be due to the variation in habitat light spectral composition and differences in their visual spectral sensitivity (see Hill and McGraw, 2006). A variation of sensory drive called sensory exploitation suggests that males may evolve displays that take advantage of sensory biases which have evolved in other behavioral contexts, such as foraging, to manipulate females into mating (Ryan and Rand, 1990; Kirkpatrick and Ryan, 1991; Endler and Basolo, 1995). Sensory exploitation may be mal-adaptive for females if the display is not somehow correlated with male quality and may therefore lead to selection on females to eliminate or avoid susceptibility to the display.
In this dissertation, I explore the evolution and development of sexual displays in species of bowerbirds, family Ptilonorhynchidae. Most bowerbirds are polygynous species with lek-like mating systems. They have long been an important model for studying sexual selection (e.g. Darwin, 1971; Marshall, 1954; Gilliard, 1969; Borgia, 1985; Frith et al, 2004). Females are highly selective in choosing mates and mating success is strongly skewed among males (e.g. Borgia, 1985; Borgia 1992). Males display in various modalities and express morphological, behavioral and off-body elements. They build a courtship structure called a bower that provides protection to females from forced copulation as they attend to behavioral display. Males decorate their bower and the surrounding display court with objects of various type and color that they collect from their environment and arrange in specific patterns (e.g. Borgia, 1985; Diamond, 1987; Lenz, 1994; Madden et al. 2004; Endler et al., 2005; Endler and Day, 2006; Borgia, 2008). They also destroy bowers of other males and plunder decorations (e.g. Borgia, 1985b; Borgia and Mueller, 1992). Courtship consists of vocalizations, behavioral elements and displays of their colorful plumage. All display types are highly differentiated throughout the bowerbird family and studies in multiple species reveal a strong correlation between mating success and many display elements (e.g. Borgia, 1985; Borgia, 1992; Madden, 2003; Frith et al. 2004).
Display development in bowerbirds is complex and is thought to involve an extensive learning component (e.g. Marshall, 1954; Borgia, 1986; Loffredo and Borgia, 1986; Diamond, 1988; Frith et al., 2004; Madden, 2008; Endler et al., 2010). Bowerbirds are long lived species and males are not fully mature until more than five years old. During their juvenile period they observe the displays of adult males and practice display with other juveniles. Many species are also excellent vocal mimics demonstrating that learning is involved at least in vocal display acquisition (e.g. Loffredo and Borgia, 1986; Coleman et al., 2007; Kelly and Healy, 2010).
In chapter one, I test the sensory drive hypothesis in bowerbirds by investigating the relationship between visual spectral sensitivity and the color of plumage and decoration displays. Bowerbirds have strong and specific color preferences and dislikes that differ between species. Sensory drive suggests that these differences may be driven by variation in spectral sensitivity that has arisen as a result of local adaptation to habitat (Endler 1992b). Spectrophotometry and microspectrophotometry were used to measure spectral transmission through ocular media (lens, cornea, and aqueous humor) and measure sensitivity of retinal photoreceptors that include four spectrally distinct cone receptors involved in color vision. I also used fluorescent microscopy to quantify the relative number of cone types from mounted retinal tissue. Furthermore, I sequenced the genes that code for opsin pigments to estimate photoreceptor spectral sensitivity. I found no differences in spectral sensitivity among 13 species that would explain the large interspecific variation in display coloration. However, I did find that bowerbird's lenses are relatively transparent to ultraviolet (UV) wavelengths which could influence the evolution of UV reflective displays.
In chapter two, I investigated the occurrence of natural hybridization between two bowerbird species, Chlamydera nuchalis and C. maculata. In other sympatric species of bowerbirds hybridization is rare or absent which may be attributable to prezygotic reproductive isolation that is due to large differences in appearance and display (see Coyne and Orr, 2004). In this study I detected over 20% hybrid individuals based on phenotype analysis. I also provide evidence of mitochondrial introgression and show that females of both species hybridize. Given that females show a high degree of effort in intraspecific mate choice (Borgia, 1995a; Madden, 2003) it is not clear why they would mate with males of another species. Based on the distribution of bowers throughout the contact zone, it does not appear that females lack access to males of their own species, which is one of the more common reasons that bird species hybridize. Studies of other avian hybrid systems show that display learning between species drives hybridization. I provide behavioral evidence that male bowerbirds in this contact zone may learn displays from the other species.
In chapter three I investigate hypotheses about the design of decoration displays. Recent work suggests that great bowerbirds, Chlamydera nuchalis, arrange objects to create a visual illusion called Forced Perspective that attracts females and increases male mating success (Endler et al. 2010; Kelly and Endler, 2012a,b). According to this hypothesis males arrange objects from small to large with distance from the bower to create an even size gradient such that all objects subtend the same visual angle to the female eye. The supposed effect is to make all objects appear the same size and make displays in the foreground appear closer. However, the data from these studies does not support the claim that males do create even size gradients nor does it show a link between the supposed illusion and mating success (Anderson, 2012; Borgia et al., 2012). I hypothesize that males are simply placing smaller objects closer to the bower for the practical purpose of avoiding interference from obstacles during courtship display. I conducted a survey of human subjects who assessed patterns of size-related bower decoration design. Survey participants were instructed to evaluate the size related pattern of object distribution in decoration displays based on visual examination of digital images. I included 40 bowers from great (C. nuchalis) spotted (C. maculata) and western (C. gutatta) bowerbirds, which share similar decoration schemes. Results of this study do not support either hypothesis, however I suggest that the obstacle avoidance hypothesis may be challenging for naive observers to detect.