Evolution of patterned plumage as a sexual signal in estrildid finches

Evolution of patterned plumage as a sexual signal in estrildid finches Abstract Color patterns, such as bars or dots that cover the body surface of animals are generally thought to play roles in signaling and camouflage. In birds, however, the macroscopic aspects of plumage coloration are less well understood, as past studies typically described plumage colorations by using spectrophotometric analyses. To provide insight into the evolution of plumage patterns as sexual signals, we characterized interspecific and intersexual variations in the plumage patterns of estrildid finches and tested their associations with other courtship signals and life-history traits using a comparative phylogenetic approach. Our results support the idea that plumage patterns in estrildids are favored by sexual selection because large-sized conspicuous plumage patterns are possessed by species with an elaborate courtship dance. These plumage patterns may also play roles in social signaling because patterns are more conspicuous in species with intraspecific brood parasitism. We predict that pattern traits can be favored by mate choice or intrasexual competition when they can serve as honest indicators of individual condition. As our results are consistent between the sexes, we suggest that the same selective force is acting on the evolution of plumage patterns in males and females in parallel. Finally, we also found a trade-off between large size and vivid color patterns, suggesting that too conspicuous patterns are costly, presumably because of the risk of catching the eyes of potential predators. Therefore, plumage patterns are also shaped by natural selection. INTRODUCTION Many animals are not uniformly colored. They can have color patches, spots, or stripes, which are likely to have evolved through predator–prey interactions, or for intraspecific communication (Caro 2005; Stevens and Merilaita 2009). Despite the definition of “color pattern” varying among both individual studies and species, a number of previous studies have described the functions of color patterns in a range of species. In particular, body color patterns are shown to function for camouflage in diverse species, like moths (Bond and Kamil 2002), squirrels (Ancillotto and Mori 2016), and carnivores (Ortolani 1999; Allen et al. 2011), while bold patterns (e.g., red spots on black) that make animals conspicuous serve as aposematic signals in spiders (Brandley et al. 2016) and poison frogs (Darst et al. 2006). Social living has also promoted the evolution of body color patterns that can code individual information in cetaceans (Caro et al. 2011; Krzyszczyk and Mann 2012), or face color patterns in some mammals (squirrels: Ancillotto and Mori 2016; primates: Santana et al. 2012). Lastly, and most importantly for this study, color patterns can affect mating, which has been well documented in fish (e.g., Endler 1983; Puebla et al. 2007). Although color patterns have long been discussed in the context of conspicuousness versus camouflage (Endler 1978; Gluckman and Cardoso 2010), few empirical studies on birds exist on this topic. Considering that birds play an essential role in predator–prey interactions, and can possess colorful plumage that functions for signaling, it would be of great importance to look into plumage color patterns in birds from such a viewpoint. In contrast with conspicuous plumage colors (e.g., carotenoid, UV) that have been extensively studied using pin-point data relying on the spectrometry techniques (Andersson 1994; Hill and McGraw 2006), the signaling functions of plumage color patterns are only beginning to be understood in an evolutionary context (cf. Gluckman and Mundy 2013; Gluckman 2014; Marshall and Gluckman 2015). We suggest that the extensive interspecific variations in the macroscopic aspects of plumage coloration deserve an evolutionary explanation for a complete understanding of how feather colors function as visual stimuli (cf. Endler 2012; Dale et al. 2015). Patterns focused in this study are defined as repetitive and regular appearances of color patches of a certain shape (i.e., dot, bar, and mottle) covering some background color (see also Methods and Figure 1). Considering that many birds have these patterns limited to small areas of the body surface, and that males and females can differ in the presence or conspicuousness of the patterns (Gluckman and Cardoso 2010) (see examples in Figure 1a–g), we hypothesize that in addition to colors, plumage patterns have important signaling roles in birds. Figure 1 View largeDownload slide Examples of interspecific and intersex variations in plumage color patterns (a–g), and the loading plot for the principal components analysis on the geometric features of plumage patterns (h). The patterned area on the body surface is schematically shown in the illustrations (a–g). Data points of the examples (a–g) are shown on the scatter plot, where the data points of males and females overlap for some species (a, e). Note that we performed a single PCA for a mixed sex dataset (see Methods for details), but that we plotted male and female in different colors only for illustrative purpose. Figure 1 View largeDownload slide Examples of interspecific and intersex variations in plumage color patterns (a–g), and the loading plot for the principal components analysis on the geometric features of plumage patterns (h). The patterned area on the body surface is schematically shown in the illustrations (a–g). Data points of the examples (a–g) are shown on the scatter plot, where the data points of males and females overlap for some species (a, e). Note that we performed a single PCA for a mixed sex dataset (see Methods for details), but that we plotted male and female in different colors only for illustrative purpose. So far, several studies have indicated that plumage patterns are used as visual signals among conspecifics. In the zebra finch Taeniopygia guttata, a popular model species of sexual selection with sexually dimorphic plumage patterns, males but not females have black and white stripes on the chest. The melanin coloration of these stripes is shaped by body condition during the juvenile stage (Birkhead et al. 1999), and females prefer males with symmetric chest stripes (Swaddle and Cuthill 1994). Similarly, in common waxbills Estrilda astrild, the regularity of barred plumage on the back is more salient in males than females, and is positively related to body condition and to the expression of colored ornamental traits (Marques et al. 2016). Moreover, patterns can play roles in broader social contexts not strictly limited to male signaling: in the diamond firetail Stagonopleura guttata, the characteristics of flank spots predict social dominance in females (Crowhurst et al. 2012), while in the red-legged partridge, both males and females show off flank stripes during agonistic displays (Bortolotti et al. 2006). Despite the list of within-species evidence for the use of feather pattern traits in signaling, one fundamental question remains: why some species have patterns while others do not, which begs for evolutionary explanation and between-species comparisons. One straightforward approach to this is to provide insight into the interspecific variations in plumage patterns and to identify their relationships with other morphological or behavioral sexual signals. Multiple traits (e.g., plumage patterns and courtship displays) can evolve in the same direction (e.g., general conspicuousness) under the same selective force, such as strong sexual selection, where redundant or multiple ornamentation is favored (Møller and Pomiankowski 1993; Johnson 2000; Candolin 2003). Conversely, it is also possible that both morphological ornamentation and behavioral displays become less favored when organisms are exposed to higher predation risks. Such correlated evolution would also be expected when the traits are under the control of the same pleiotropic mechanisms, e.g., hormones, genes, or neurotransmitters. Moreover, when birds have both morphological ornamentation and behavioral displays, the latter signals may work as an “amplifier” of ornamentation (Zahavi 1978; Hasson 1991). As a textbook example, peacocks Pavo cristatus, which are known for having long trains with eyespots, perform courtship displays facing toward the sun to highlight the iridescent color patterns of the erected feathers (Dakin and Montgomerie 2009). Similarly, visual displays of some birds are performed in a way that enhances the signal efficacy of ornamentation (Candolin 2003; Fusani et al. 2007; Olea et al. 2010; Bortolotti et al. 2011). In addition, the idea of correlated evolution may hold true for the association among different facets of patterns, i.e., coloration and size. Depending on the strength of the selective force, a plumage pattern may become conspicuous or less conspicuous in terms of both color and size. In contrast, if it is too costly to have large-sized vivid color patterns at the same time because of intense predation risks, there might be a trade-off between the 2 aspects of the patterns. Considering that some birds show sexual dichromatism in melanin- and carotenoid-based coloration (Hill and McGraw 2006), such a relationship between size and color could also differ between the sexes. Lastly, comparing the evolution of plumage patterns between sexes would provide insight into their potential roles as sexual signals. Although sexual traits in birds, such as songs and ornamental colors, tend to be male-biased (Dale et al. 2015), it has been reported that plumage patterns are female-biased, suggesting that they function as cryptic sexual signals, which are particularly widespread in females (Gluckman and Cardoso 2010). Estrildid finches (family: Estrildidae) are ideal candidates to investigate the possible role of plumage patterns in sexual selection. First, within-species evidence in some estrildids suggests that patterns can have signaling functions (zebra finch: Swaddle and Cuthill 1994; Birkhead et al. 1999; diamond firetail: Crowhurst et al. 2012; Zanollo et al. 2013, 2014; common waxbill: Marques et al. 2016). Second, a relatively larger number of species has plumage patterns in this taxonomic group (57.5%), with some showing sexual dimorphism in the presence or conspicuousness of patterns (e.g., Marques et al. 2016, Figure 1), suggesting the possibility that plumage patterns may be linked to sexual selection. Third, estrildid finches are characterized by elaborate courtship signals, including acoustic and behavioral displays (Goodwin 1982; Baptista et al. 1999; Soma and Garamszegi 2015). Most importantly, when potential mating partners come close, often perching side by side, they show a ritualized courtship dance (Goodwin 1982; Soma and Garamszegi 2015), which is performed by males of some species (Zanollo et al. 2013; Ullrich et al. 2016) or by both sexes in other species (Ota et al. 2015, 2017; Soma and Iwama 2017). Such behavioral features of estrildid finches often involve display positions in which the plumage color patterns are exposed and might explain why they frequently have flank patterns (e.g., Crowhurst et al. 2012, see also Figure 1a–g). Plumage coloration has been studied from evolutionary perspectives in estrildids, but plumage patterns have not been investigated in a comparative phylogenetic context. In our previous comparative phylogenetic studies of estrildid finches (Soma and Garamszegi 2015; Gomes et al. 2017), we did not find a link between the evolution of dance displays and ornamental coloration, measured based on the coverage and reflectance spectrophotometry. In another comparative phylogenetic study, ornamental coloration in males was higher in species with a gregarious nature, suggesting that social selection affects the evolution of plumage coloration (Gomes et al. 2016). In this study of estrildid finches, we characterized plumage patterns and tested the following set of predictions by applying a phylogenetic comparative approach to provide insight into the evolution of plumage patterns as sexual signals. 1) As visual signals, the conspicuousness of plumage patterns (e.g., size) may have evolved in association with other visual traits, such as pattern color and courtship display. 2) The sexes may undergo different selection regimes, which can be detected by comparing the evolution of plumage patterns between the sexes. 3) Interspecific variations in life history can be potentially linked to the interspecific variations in plumage patterns because it has already been shown to be related to plumage coloration (Gomes et al. 2016). MATERIALS AND METHODS Collecting the plumage pattern data We collected data on plumage patterns by taking pictures of skin specimens at the Natural History Museum at Tring, UK and Yamashina Institute for Ornithology at Chiba, Japan. We sampled 5 individuals of each sex for each estrildid finch species, when possible, and took pictures from 4 angles (ventral, dorsal, and left and right sides) with a scale using a digital camera (α5000; Sony, Tokyo Japan). We also took UV photos (digital camera: D70s, Nikon, Tokyo, Japan; lens: UV-105 mm F4.5, Tochigi Nikon, Tochigi, Japan; filter: U360, Hoya Optronics, Tokyo, Japan; light: Handy UV Lamp 365nm; AsOne, Osaka, Japan), considering that estrildids are likely to have UV vision (Ödeen and Håstad 2003), to verify that there was no pattern in the nonvisible part of the spectrum that was apparent only under UV light. We collected data on both sexes from 126 species, and male data from one species, of the 134 estrildid species. Based on the fact that some estrildid species show intermediate patterns between bars and dots (e.g., overlapped dots forming lines), we applied the definition of pattern trait in a broad sense (i.e., a repetition of the same shape regardless of whether patterns form bars, dots, or mottles) instead of setting small categories dependent on unit shape. We calculated the following 4 geometric features of patterns from the digital images, using ImageJ 1.48v (Schneider et al. 2012). We measured pattern coverage on the dorsal, ventral, and side (average of left and right sides) views, calculated as the percentage of patterned area per total body surface in each picture. As an index of pattern conspicuousness, we measured the size of the unit shape that constituted the pattern, by taking the width of a bar or the diameter of a mottle or a dot (in mm). After checking that each of the 4 variables showed fairly high repeatability within the same sex of each species (r > 0.86, P < 0.0001; Garamszegi 2014), we used sex-specific average values for each species in subsequent analyses. As some species had multiple types of patterns (e.g., zebra finch males showing flank dots, chest stripes, and tail bars), we counted the total number of patterns for each sex of each species. The above 4 measures were treated as 0, respectively, when no plumage pattern was evident. To avoid potential interobserver bias, all measurements were performed by the same observer (M.S.). Using the species for which we measured specimens in the 2 museums, we checked for potential biases originating from the fact that different museums may store specimens differently or may have collections of different ages. However, we found no differences between the museums (linear mixed model including species as a random effect and sex and museum as fixed effects: effect of museum on side coverage, t = 1.01, P = 0.32; effect of museum on ventral coverage, t = 1.11, P = 0.27; effect of museum on dorsal coverage, t = 0.69, P = 0.49; effect of museum on pattern unit size, t = 0.53, P = 0.60). In addition, we recorded whether the patterns that each species/sex had were composed of melanin-based or carotenoid-based coloration or a mixture of both. Melanin and carotenoid are the 2 main pigment types in birds, responsible for red and orange coloration or black, gray, and brown, respectively (Hill and McGraw 2006). Both pigmentations, including pigment-based patterns, are argued to play roles in signaling individual conditions (Griffith et al. 2006; Pérez-Rodríguez et al. 2017). Based on the visual inspection of the feather colors, they were scored as melanin (0); melanin and carotenoid (1); or carotenoid (2). Melanin patterns are colored with combinations of black, gray, brown, and white, while the carotenoid patterns are colored with red and white (see Figure 1a,d for the comparison of melanin and carotenoid patterns expressed in females and males, respectively). As it was found that some birds have white spots on the basis of a black-to-red gradation background, they were categorized as “melanin and carotenoid” patterns. Other sexual signals and life-history traits As potential signals that can coevolve with plumage patterns, we characterized the degree of expression of courtship displays of each sex. Specifically, we quantified the dance repertoires of males and females, under the prediction that the complexity of a visual display (i.e., courtship dance) evolved jointly with plumage conspicuousness to emphasize showiness. Courtship dance is stereotyped within species and is expressed as a combination of several simple actions (e.g., Restall 1996; Zanollo et al. 2013; Ota et al. 2015; Ullrich et al. 2016; Soma and Iwama 2017). We counted the number of dance categories that constituted the courtship displays of each species, separately for males and females (see also Soma and Garamszegi 2015). In addition, we also considered life-history variables that likely affect the cost of reproduction, such as median clutch size, and body size taken as body length. As some estrildid finches in Africa are targets of interspecific brood parasitism by birds in the genus Vidua (Sorenson et al. 2004), the presence or absence of interspecific brood parasitism was scored as 0 or 1 (parasitism absent–present, respectively). Furthermore, some estrildid finches show intraspecific brood parasitism (Yom-Tov 2001), which was scored as 0 or 1 (parasitism absent–present, respectively). Although most species do not defend territories, the degree of gregariousness varies among species and was scored as follows: colonial breeders or species with social systems in which multiple pairs keep contact with each other even during the breeding season were given a score of 3; highly social and gregarious species, in which aggregations mainly occur outside the breeding season were given a score of 2; species that breed usually in pairs or in small parties were given a score of 1; and strictly territorial species were given a score of 0. All of these data were compiled from an earlier study (Soma and Garamszegi 2015), and the number of focal species was limited (n = 85) because of the availability of information. Phylogeny For the comparative phylogenetic analyses described below, we could not obtain an overwhelmingly supported consensus phylogenetic tree with branch lengths, but were able to derive multiple equally likely candidate trees from Jetz et al. (2012). We derived 1000 trees from the dataset for the focal species in each analysis and used their consensus tree without branch length to reconstruct the ancestral state or used each of them in phylogenetic regressions followed by multimodel inference (see below). Statistical analyses Principal component analysis (PCA) The 4 geometric variables that describe plumage patterns (dorsal, ventral, and side coverage and unit size) are mutually dependent and may be related with the number of pattern types of each species/sex. We conducted a PCA to define independent axes that have distinct biological meaning from the raw variables, in which we used square root values for coverage, pattern unit size, and the number of patterns of each sex. We performed a single conventional PCA for a mixed sex dataset, instead of repeating phylogenetic PCA for each sex, in order to obtain PC scores comparable between sexes. Specifically, using data from 127 males and 126 females, we aimed to achieve a dimension reduction by extracting principal component axes (PCs) that account for most of the variance (>70%), have eigenvalues larger than 1 and have biologically interpretable component loadings. Below we show PCA results for PC1 and PC2, however, based on the above criteria, only PC1 was used for subsequent analyses (see Results for details). We confirmed that these PC scored obtained from a conventional PCA were highly correlated with those obtained from phylogenetic PCA (r > 0.997, P < 0.0001). We used R 3.3.1 software (R Core Team 2016) for the analyses described here and below, except otherwise stated. Ancestral state reconstruction To characterize the evolutionary history of plumage patterns, and to compare it between sexes, we reconstructed the ancestral state using the PC scores in Mesquite (Maddison and Maddison 2011). A consensus phylogenetic tree was obtained based on 1000 trees from the dataset (Jetz et al. 2012). Parsimony methods were used as opposed to maximum likelihood, due to the latter being unavailable for our data, i.e., multifurcating tree. The phylogenetic relationship between color and patterns We investigated the association between geometric features and colors of the patterns with the aim of revealing a positive or negative correlation between the 2 different aspects under our predictions. We used species that have a plumage pattern, and tested the effects of pattern color (i.e., melanin, carotenoid, or mixture of both) and sex on PC1. Nonindependence of the data due to the phylogenetic relatedness of species was controlled by Bayesian phylogenetic mixed models (Hadfield and Nakagawa 2010) in R package “MCMCglmm” (Hadfield 2010) because it allows multiple data entries (i.e., male and female PC scores) per species. In the model, we used a Gaussian error distribution, and the priors [G = list(G1= list(V = 1, nu = 0.02), G2 = list(V = 1, nu = 0.02))]. We repeatedly fit the same model using each of the 1,000 phylogenetic trees obtained from the dataset (Jetz et al. 2012), all of which successfully converged, and obtained the mean coefficients for the predictor variables and their 95% confidence interval by model averaging the 1000 outcomes. We weighted parameter estimates based on the DIC of the respective model corresponding to a particular tree. For this particular analysis, we relied on MCMCglmm because we needed to analyze the data with multiple entries (i.e., male and female variables) per species, therefore requiring an approach based on phylogenetic mixed modeling. In the following analyses, however, we applied the phylogenetic generalized least-squares (PGLS) regression technique, available in the package phylolm (Ho et al. 2016), and analyzed males and females separately. The phylogenetic relationship between courtship dance and plumage patterns To examine the potential relationships between plumage patterns and courtship dance, we tested if the PC1 covaried with the dance repertoire at the interspecific level, separately in the 2 sexes. We fitted PGLS regressions to control for phylogenetic dependence by using the 1,000 phylogenetic trees as explained earlier, in which we entered PC1 as predictor variables and dance repertoire as a response variable. As dance repertoires are discrete values (count data), we used regression with Poisson error distribution. For each tree, we fitted models with identical predictor/response structure, and then derived mean and confidence estimates across trees for the estimated parameters (slopes and intercepts). Summary statistics were obtained via model averaging Garamszegi and Mundry 2014). For the PGLS modeling, we used the R package “phylolm” that allows Poisson distribution (Ho et al. 2016) The phylogenetic relationship between plumage patterns and life-history variables We tested for the relationship between plumage pattern and life-history variables that likely affect the cost of reproduction. Specifically, we used PGLS framework available in R package “phylolm” again, and constructed models in which clutch size, body size, presence of interspecific and intraspecific brood parasitism and sociality were entered as predictor variables, and PC1 was used as a response variable (different models for males and females), using Gaussian distribution. As in the above analyses, we used the same set of 1000 phylogenetic trees and applied model averaging to obtain mean and confidence estimates for the estimated parameters (intercept and slopes for each predictor). Parameter estimates from each model were weighted according to their relative fit to the data (Garamszegi and Mundry 2014). RESULTS Principal component analysis PC1 accounted for 72.8% of the total variation in the variables that were taken to describe plumage pattern and had positive loadings (>0.36) for all 5 variables (Table 1, Figure 1h). Therefore, higher PC1 scores can indicate larger and conspicuous patterns. In contrast, PC2 accounted for only 15.2% of the total variation and had a higher positive loading (>0.8) only for dorsal coverage but negative loadings (<−0.30) for ventral coverage and pattern unit size (Table 1, Figure 1h). Given the low eigenvalue of PC2 (Table 1), we mainly focused on PC1 (conspicuousness) hereafter. Table 1 The results of the principal components analysis on mixed sex data of patterned variables PC1 PC2 Side coverage (square root) 0.505 0.141 Ventral coverage (square root) 0.457 −0.300 Dorsal coverage (square root) 0.364 0.803 Pattern unit size 0.425 −0.494 Number of pattern types 0.472 −0.035 Eigenvalue 3.638 0.761 Proportion of variance 0.728 0.152 Cumulative proportion 0.728 0.880 PC1 PC2 Side coverage (square root) 0.505 0.141 Ventral coverage (square root) 0.457 −0.300 Dorsal coverage (square root) 0.364 0.803 Pattern unit size 0.425 −0.494 Number of pattern types 0.472 −0.035 Eigenvalue 3.638 0.761 Proportion of variance 0.728 0.152 Cumulative proportion 0.728 0.880 View Large Table 1 The results of the principal components analysis on mixed sex data of patterned variables PC1 PC2 Side coverage (square root) 0.505 0.141 Ventral coverage (square root) 0.457 −0.300 Dorsal coverage (square root) 0.364 0.803 Pattern unit size 0.425 −0.494 Number of pattern types 0.472 −0.035 Eigenvalue 3.638 0.761 Proportion of variance 0.728 0.152 Cumulative proportion 0.728 0.880 PC1 PC2 Side coverage (square root) 0.505 0.141 Ventral coverage (square root) 0.457 −0.300 Dorsal coverage (square root) 0.364 0.803 Pattern unit size 0.425 −0.494 Number of pattern types 0.472 −0.035 Eigenvalue 3.638 0.761 Proportion of variance 0.728 0.152 Cumulative proportion 0.728 0.880 View Large Comparison of pattern evolution between sexes The reconstruction of ancestral states revealed that the plumage color patterns evolved in a similar fashion in males and females (Figure 2, see also Supplementary Figure S1). Additionally, we also directly compared PC1 between the sexes using the paired t-test and found that PC1 did not differ significantly between the sexes (paired t-test: t = 1.04, P = 0.302, N = 124 species). Figure 2 View largeDownload slide Reconstruction of ancestral states of plumage color patterns (PC1) in males and females (see Supplementary Figure S1 for PC2). Figure 2 View largeDownload slide Reconstruction of ancestral states of plumage color patterns (PC1) in males and females (see Supplementary Figure S1 for PC2). The phylogenetic relationship between color and patterns We examined whether geometric features and the colors of patterns were inter-related by focusing on species that had plumage patterns and for which information on the phylogenetic relationships was available (male: N = 63 species; female: N = 65 species). We found an interspecific relationship between color and conspicuousness of patterns: PC1 scores were lower in species in which patterns were carotenoid-dependent than in species that had melanin-dependent trait expression in both sexes (Figure 3, Table 2). Figure 3 View largeDownload slide Comparison of principal component 1 (PC1) (pattern conspicuousness) of males and females in species with melanin- or carotenoid-based plumage patterns (see Table 2 for statistical outcomes). Square bars at the bottom show a schematic view of the plumage color categories (“Melanin and carotenoid” patterns typically have black-to-red gradation background with white spots, while the other 2 patterns have monotone backgrounds). Figure 3 View largeDownload slide Comparison of principal component 1 (PC1) (pattern conspicuousness) of males and females in species with melanin- or carotenoid-based plumage patterns (see Table 2 for statistical outcomes). Square bars at the bottom show a schematic view of the plumage color categories (“Melanin and carotenoid” patterns typically have black-to-red gradation background with white spots, while the other 2 patterns have monotone backgrounds). Table 2 Effects of color of plumage pattern (0: melanin; 1: melanin and carotenoid; 2: carotenoid) and sex on PC1 Coefficient 95% CI Intercept 1.776 (1.208, 2.297) Color −0.841 (−1.365,−0.282) Sex 0.029 (−0.158, 0.217) Coefficient 95% CI Intercept 1.776 (1.208, 2.297) Color −0.841 (−1.365,−0.282) Sex 0.029 (−0.158, 0.217) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as a significant effect. View Large Table 2 Effects of color of plumage pattern (0: melanin; 1: melanin and carotenoid; 2: carotenoid) and sex on PC1 Coefficient 95% CI Intercept 1.776 (1.208, 2.297) Color −0.841 (−1.365,−0.282) Sex 0.029 (−0.158, 0.217) Coefficient 95% CI Intercept 1.776 (1.208, 2.297) Color −0.841 (−1.365,−0.282) Sex 0.029 (−0.158, 0.217) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as a significant effect. View Large The phylogenetic relationship between courtship dance and plumage patterns Using the species data for which dance and pattern data were available, we examined the potential relationship between PC1 and complexity of the dance repertoire, and found that PC1 covaried positively with dance repertoire in each sex (Table 3, Figure 4). These associations indicate that species with conspicuous plumage patterns have more complex dances. Table 3 Effects of PC1 on the dance repertoire of each sex (a) Male dance repertoire (b) Female dance repertoire Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.532 0.108 (1.317, 1.747) 0.672 0.381 (−0.090, 1.434) PC1 0.029 0.009 (0.011, 0.047) 0.090 0.039 (0.012, 0.167) (a) Male dance repertoire (b) Female dance repertoire Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.532 0.108 (1.317, 1.747) 0.672 0.381 (−0.090, 1.434) PC1 0.029 0.009 (0.011, 0.047) 0.090 0.039 (0.012, 0.167) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as a significant effect. View Large Table 3 Effects of PC1 on the dance repertoire of each sex (a) Male dance repertoire (b) Female dance repertoire Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.532 0.108 (1.317, 1.747) 0.672 0.381 (−0.090, 1.434) PC1 0.029 0.009 (0.011, 0.047) 0.090 0.039 (0.012, 0.167) (a) Male dance repertoire (b) Female dance repertoire Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.532 0.108 (1.317, 1.747) 0.672 0.381 (−0.090, 1.434) PC1 0.029 0.009 (0.011, 0.047) 0.090 0.039 (0.012, 0.167) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as a significant effect. View Large Figure 4 View largeDownload slide Relationship between the complexity of courtship (dance repertoire) and PC1) in males. Species data are shown with semitransparent plots and the regression lines are drawn based on estimated parameters from the phylogenetic analyses summarized in Table 3. Figure 4 View largeDownload slide Relationship between the complexity of courtship (dance repertoire) and PC1) in males. Species data are shown with semitransparent plots and the regression lines are drawn based on estimated parameters from the phylogenetic analyses summarized in Table 3. The phylogenetic relationship between plumage patterns and life-history variables In both males and females, we found a significant association between plumage pattern and intraspecific brood parasitism, as PC1 was positively correlated with this trait (Figure 5, Table 4). The remaining life-history traits were not related significantly to the PC 1 (Table 4). Figure 5 View largeDownload slide Comparison of male and female PC1 between species with and without intraspecific brood parasitism. Figure 5 View largeDownload slide Comparison of male and female PC1 between species with and without intraspecific brood parasitism. Table 4 Effects of life-history traits on PC1 (plumage pattern conspicuousness) in males (a) and females (b), as estimated from the appropriate phylogenetic generalized least-squares models (a) Male PC1 (b) Female PC1 Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.736 2.268 (−2.801, 6.273) 0.952 2.126 (−3.301, 5.205) Clutch size −0.106 0.271 (−0.648, 0.437) −0.203 0.236 (−0.676, 0.270) Body size −0.107 0.146 (−0.399, 0.184) 0.010 0.141 (−0.272, 0.293) Interspecific brood parasitism 0.999 0.508 (−0.017, 2.015) 0.831 0.506 (−0.182, 1.843) Intraspecific brood parasitism 3.286 0.673 (1.940, 4.633) 2.539 0.716 (1.107, 3.872) Sociality −0.125 0.241 (−0.607, 0.358) −0.212 0.233 (−0.678, 0.253) (a) Male PC1 (b) Female PC1 Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.736 2.268 (−2.801, 6.273) 0.952 2.126 (−3.301, 5.205) Clutch size −0.106 0.271 (−0.648, 0.437) −0.203 0.236 (−0.676, 0.270) Body size −0.107 0.146 (−0.399, 0.184) 0.010 0.141 (−0.272, 0.293) Interspecific brood parasitism 0.999 0.508 (−0.017, 2.015) 0.831 0.506 (−0.182, 1.843) Intraspecific brood parasitism 3.286 0.673 (1.940, 4.633) 2.539 0.716 (1.107, 3.872) Sociality −0.125 0.241 (−0.607, 0.358) −0.212 0.233 (−0.678, 0.253) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as significant association. View Large Table 4 Effects of life-history traits on PC1 (plumage pattern conspicuousness) in males (a) and females (b), as estimated from the appropriate phylogenetic generalized least-squares models (a) Male PC1 (b) Female PC1 Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.736 2.268 (−2.801, 6.273) 0.952 2.126 (−3.301, 5.205) Clutch size −0.106 0.271 (−0.648, 0.437) −0.203 0.236 (−0.676, 0.270) Body size −0.107 0.146 (−0.399, 0.184) 0.010 0.141 (−0.272, 0.293) Interspecific brood parasitism 0.999 0.508 (−0.017, 2.015) 0.831 0.506 (−0.182, 1.843) Intraspecific brood parasitism 3.286 0.673 (1.940, 4.633) 2.539 0.716 (1.107, 3.872) Sociality −0.125 0.241 (−0.607, 0.358) −0.212 0.233 (−0.678, 0.253) (a) Male PC1 (b) Female PC1 Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.736 2.268 (−2.801, 6.273) 0.952 2.126 (−3.301, 5.205) Clutch size −0.106 0.271 (−0.648, 0.437) −0.203 0.236 (−0.676, 0.270) Body size −0.107 0.146 (−0.399, 0.184) 0.010 0.141 (−0.272, 0.293) Interspecific brood parasitism 0.999 0.508 (−0.017, 2.015) 0.831 0.506 (−0.182, 1.843) Intraspecific brood parasitism 3.286 0.673 (1.940, 4.633) 2.539 0.716 (1.107, 3.872) Sociality −0.125 0.241 (−0.607, 0.358) −0.212 0.233 (−0.678, 0.253) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as significant association. View Large DISCUSSION The present phylogenetic comparative study focusing on geometric features of plumage patterns in estrildid finches revealed that they might have evolved for signaling function, for at least 2 reasons. First, the PCA showed that interspecific variations in patterns could be described along the spectrum of conspicuousness. The PC1 reflected larger and bold patterns that are possibly composed of multiple shapes (e.g., Figure 1b). Second, through the phylogenetic comparative analyses of PC1, we found evidence that plumage patterns may be coevolutionarily related to another sexual signal (i.e., dance repertoire), and to intraspecific competition (i.e., presence of intraspecific brood parasitism). As these trends were consistent between sexes (Tables 2–4), and males and females showed very similar levels of conspicuousness (Figures 1 and 2) we can infer that basically the same selective force is acting on the evolution of plumage patterns in males and females. Alternatively, selection may favor the trait in one sex only, but its effect is also manifested in the other due to strong genetic correlation between sexes. Lack of sex-biased evolution of plumage patterns Evolution of female ornamentation is a challenging question, and discussed in relation to the presence of intersexual genetic correlation, and/or sexual and social selection (West-Eberhard 1979; Tobias et al. 2012) in estrildid finches (Soma and Garamszegi 2015). Specifically, estrildids are characterized by multiple sexual traits that can be shared between sexes. Females of some species sing courtship songs like males (Goodwin 1982; Geberzahn and Gahr 2011, 2013), whereas females of other species perform courtship dance displays identical to those of males (Ota et al. 2015; Soma and Garamszegi 2015; Soma and Iwama 2017) or have plumage ornamentation comparable to that of males (Zanollo et al. 2014; Marques et al. 2016). However, these sexual signals (song, dance, and plumage coloration) are generally male-biased with regard to presence/absence or degree of elaboration in estrildids (Soma and Garamszegi 2015; see also Dale et al. 2015), which is contrasted with the plumage color pattern (cf. Gluckman and Cardoso 2010). Only a few species (Spermophaga spp.) show female-biased sexual dimorphism with regard to the presence of patterns, and there was no sex difference in PC1. Moreover, the sexes showed similar evolution of PC1 (Figure 2). This is quite different from what was found in peafowls and their related taxa, where eyespot patterns likely have evolved primarily in males for female choice (Sun et al. 2014). Presumably, plumage patterns in female estrildids would function either for sexual signaling to get mates (sexual selection), or for status signaling to repel rivals, even outside the reproductive context (social selection) (e.g., Kabasakal et al. 2017). Although there is limited empirical evidence available showing that selection operates on female estrildid finches, a study on one species supports this idea. Diamond firetail S. guttata, females have more flank spots than males on average (Zanollo et al. 2014), and the number of spots reflects physical condition (Zanollo et al. 2012) and predicts social dominance in females (Crowhurst et al. 2012). Therefore, similar selective forces may mediate the interspecific variation of plumage patterns in both males and females. Hiding versus signaling Diversity of color patterns in animals can be explained through hiding versus signaling strategy (Endler 1978; Stevens and Merilaita 2009; Gluckman and Cardoso 2010). Some animals balance the two by having cryptic patterns all over the body and conspicuous patterns on small areas that are important for communication, like face or tail (Caro 2005; Ancillotto and Mori 2016), while others have dual-purpose color patterns (e.g., squids: Mäthger and Hanlon 2006; Mäthger et al. 2012). In line with these previous insights into nonbird species, our findings indicate that plumage patterns in estrildid finches are also the product of a compromise between hiding and signaling, as suggested by Endler (1978). Estrildid finches need to escape potential predators, but they also need to be showy enough to attract or deter conspecifics. Consequently, they cannot have patterns that are large and red at the same time (Figure 3), or otherwise their appearance would stand out too much; thus, making it difficult to avoid predation. The gregarious nature and colonial breeding of estrildid finches are likely a response to high predation risks, at least in some species (Zann 1996). Although we did not find a direct link between gregariousness and plumage patterns in this study, some estrildid species build nests in a thorny tree or near a wasp nest (Goodwin 1982; Barnard and Markus 1990; Beier et al. 2006), while others add carnivore scat to their nest to prevent predation (Schuetz 2004). These strategies may indirectly suggest how much predation pressure can affect the life of estrildid finches. Phylogenetic comparative studies particularly focusing on this aspect of breeding biology could be designed in the future into this promising direction. Some finches are also known for their showy plumage patterns that play roles in sexual/social contexts (Swaddle and Cuthill 1994; Crowhurst et al. 2012; Zanollo et al. 2013, 2014; Marques et al. 2016). Consistent with such earlier findings, in this comparative study, we gained results that support the hypothesis that plumage patterns evolve alongside their signaling roles in sexual selection. Specifically, we found that the complexity of courtship dance has evolved in the same direction as the conspicuousness of plumage patterns (Table 3, Figure 4). This supports the idea that the 2 visual traits evolved under the same selective force, and our findings thus meet the prediction of the “amplifier” hypothesis (Zahavi 1978; Hasson 1991). However, more detailed investigations are needed to clarify how gestural displays functionally enhance signal efficacy of plumage patterns (or vice versa) in estrildid finches. In general, we expect that motion (i.e., direction, speed) might play an important role in how patterns are perceived by receivers (cf. Dakin and Montgomerie 2009). In the context of predator–prey interactions, it was already reported that particular patterns can make animals less detectable especially while they are moving (Halperin et al. 2017). Such potential roles remain to be explored in the context of sexual signaling. The function of plumage patterns could also include social signaling in addition to a sexual context. Having signaling traits that facilitate individual recognition is particularly favored in both species living in colonies and those with agonistic competitions (Tibbetts and Dale 2007). Here, we uncovered that the presence of intraspecific brood parasitism was associated with conspicuous plumage patterns (Table 4). This is also consistent with our previous study (Soma and Garamszegi 2015), in which we detected associations between dance complexity and intraspecific brood parasitism in both sexes. Apparently, the 2 visual signals, dance and plumage patterns, evolved in the same direction in response to intraspecific parasitism (Table 4). Intraspecific parasitism can cause intense within-species competition, where, for instance, host and parasite females can have a higher chance to directly interact or fight (e.g., Semel and Sherman 2001; Åhlund 2005). In such cases, condition-dependent honest indicators of quality (e.g., Marques et al. 2016) would be effective for avoiding escalated costly aggression. This is comparable to the previous findings from the passerine-wide phylogenetic comparative study showing that females of cooperative breeding species are more likely to have ornamental coloration due to female competition (Dale et al. 2015). To conclude, we have shown that plumage patterns constitute an important visual trait that can be the subject to both natural and sexual selection forces. Such a finding could not have been discovered solely with spectrometric techniques that are more commonly applied for examining the evolution of plumage traits of birds. Our results strongly suggest that the macroscopic aspects of plumage traits may open a new research agenda for understanding the evolution of complex plumage traits in birds, as done in other taxonomic groups (Ortolani 1999; Seehausen et al. 1999; Caro 2005; Caro et al. 2011; Allen et al. 2011; Santana et al. 2012; Allen et al. 2013; Kelley et al. 2013; Ancillotto and Mori 2016). We suggest that color and geometric features should be integrated to reach a comprehensive understanding of the physical appearance of complex plumage traits in animals (Endler 2012). SUPPLEMENTARY MATERIAL Supplementary data are available at Behavioral Ecology online. FUNDING This study was supported by the JSPS Grant-in-Aid for Young Scientists (23680027 and 16H06177) and Hokkaido University WinGS Global Networking Award 2015 to Masayo Soma, and by the Ministry of Economy and Competitiveness (Spain) (CGL2015-70639-P) and the National Research, Development and Innovation Office (Hungary) (K-115970) to L.Z.G. We thank the Natural History Museum and its senior curator Mark Adams, and the Yamashina Institute for Ornithology and its researcher Takeshi Yamasaki for their help assessing their ornithological collection. We also thank Nao Ota for advice on photography, and Honor Scarlett for reviewing the text. Data accessibility: Analyses reported in this article can be reproduced using the data provided by Soma and Garamszegi (2018). REFERENCES Åhlund M . 2005 . Behavioural tactics at nest visits differ between parasites and hosts in a brood-parasitic duck . Anim Behav . 70 : 433 – 440 . Google Scholar CrossRef Search ADS Allen WL , Baddeley R , Scott-Samuel NE , Cuthill IC . 2013 . The evolution and function of pattern diversity in snakes . Behav Ecol . 24 : 1237 – 1250 . Google Scholar CrossRef Search ADS Allen WL , Cuthill IC , Scott-Samuel NE , Baddeley R . 2011 . Why the leopard got its spots: relating pattern development to ecology in felids . Proc Biol Sci . 278 : 1373 – 1380 . Google Scholar CrossRef Search ADS PubMed Ancillotto L , Mori E . 2016 . Adaptive significance of coat colouration and patterns of Sciuromorpha (Rodentia) . Ethol Ecol Evol . 29 : 241 – 254 . Google Scholar CrossRef Search ADS Andersson M . 1994 . Sexual selection . Princeton (NJ) : Princeton University Press . Baptista LF , Lawson R , Visser E , Bell DA . 1999 . Relationships of some mannikins and waxbills in the estrildidae . J Ornithol . 140 : 179 – 192 . Google Scholar CrossRef Search ADS Barnard P , Markus MB . 1990 . Reproductive failure and nest site selection of two estrildid finches in Acacia woodland . Ostrich . 61 : 117 – 124 . Google Scholar CrossRef Search ADS Beier P , Tungbani AI , Brittingham M . 2006 . Nesting with the wasp Ropalidia cincta increases nest success of red-cheeked cordonbleu (Uraeginthus bengalus) in Ghana . Auk . 123 : 1022 – 1037 . Birkhead TR , Fletcher F , Pellatt EJ . 1999 . Nestling diet, secondary sexual traits and fitness in the zebra finch . Proc R Soc B Biol Sci . 266 : 385 – 390 . Google Scholar CrossRef Search ADS Bond AB , Kamil AC . 2002 . Visual predators select for crypticity and polymorphism in virtual prey . Nature . 415 : 609 – 613 . Google Scholar CrossRef Search ADS PubMed Bortolotti GR , Blas J , Negro JJ , Tella JL . 2006 . A complex plumage pattern as an honest social signal . Anim Behav . 72 : 423 – 430 . Google Scholar CrossRef Search ADS Bortolotti GR , Stoffel MJ , Galván I . 2011 . Wintering snowy owls Bubo scandiacus integrate plumage colour, behaviour and their environment to maximize efficacy of visual displays . Ibis . 153 : 134 – 142 . Google Scholar CrossRef Search ADS Brandley N , Johnson M , Johnsen S . 2016 . Aposematic signals in North American black widows are more conspicuous to predators than to prey . Behav Ecol . 27 : 1104 – 1112 . Google Scholar CrossRef Search ADS Candolin U . 2003 . The use of multiple cues in mate choice . Biol Rev . 78 : 575 – 595 . Google Scholar CrossRef Search ADS PubMed Caro T . 2005 . The adaptive significance of coloration in mammals . BioScience . 55 : 125 – 136 . Google Scholar CrossRef Search ADS Caro T , Beeman K , Stankowich T , Whitehead H . 2011 . The functional significance of colouration in cetaceans . Evol Ecol . 25 : 1231 – 1245 . Google Scholar CrossRef Search ADS Crowhurst CJ , Zanollo V , Griggio M , Robertson J , Kleindorfer S . 2012 . White flank spots signal feeding dominance in female diamond firetails, Stagonopleura guttata . Ethology . 118 : 63 – 75 . Google Scholar CrossRef Search ADS Dakin R , Montgomerie R . 2009 . Peacocks orient their courtship displays towards the sun . Behav Ecol Sociobiol . 63 : 825 – 834 . Google Scholar CrossRef Search ADS Dale J , Dey CJ , Delhey K , Kempenaers B , Valcu M . 2015 . The effects of life history and sexual selection on male and female plumage colouration . Nature . 527 : 367 – 370 . Google Scholar CrossRef Search ADS PubMed Darst CR , Cummings ME , Cannatella DC . 2006 . A mechanism for diversity in warning signals: conspicuousness versus toxicity in poison frogs . Proc Natl Acad Sci USA . 103 : 5852 – 5857 . Google Scholar CrossRef Search ADS PubMed Endler JA . 1978 . A predator’s view of animal color patterns . Evol Biol . 11 : 319 – 364 . Endler JA . 1983 . Natural and sexual selection on color patterns in poeciliid fishes . Environ Biol Fish . 9 : 173 – 190 . Google Scholar CrossRef Search ADS Endler JA . 2012 . A framework for analysing colour pattern geometry: adjacent colours . Biol J Linn Soc . 107 : 233 – 253 . Google Scholar CrossRef Search ADS Fusani L , Giordano M , Day LB , Schlinger BA . 2007 . High-speed video analysis reveals individual variability in the courtship displays of male Golden-collared manakins . Ethology . 113 : 964 – 972 . Google Scholar CrossRef Search ADS Garamszegi LZ . 2014 . Uncertainties due to within-species variation in comparative studies: measurement errors and statistical weights . In: Garamszegi LZ , editor. Modern phylogenetic comparative methods and their application in evolutionary biology . New York (NY) : Springer . p. 157 – 199 . Google Scholar CrossRef Search ADS Garamszegi LZ , Mundry R . 2014 . Multimodel-inference in comparative analyses . In: Garamszegi LZ , editor. Modern phylogenetic comparative methods and their application in evolutionary biology . New York (NY) : Springer . p. 157 – 199 . Google Scholar CrossRef Search ADS Geberzahn N , Gahr M . 2011 . Undirected (solitary) birdsong in female and male blue-capped cordon-bleus (Uraeginthus cyanocephalus) and its endocrine correlates . PLoS One . 6 : e26485 . Google Scholar CrossRef Search ADS PubMed Geberzahn N , Gahr M . 2013 . Song learning in male and female Uraeginthus cyanocephalus, a tropical songbird species . J Comp Psychol . 127 : 352 – 364 . Google Scholar CrossRef Search ADS PubMed Gluckman TL . 2014 . Pathways to elaboration of sexual dimorphism in bird plumage patterns . Biol J Linn Soc . 111 : 262 – 273 . Google Scholar CrossRef Search ADS Gluckman TL , Cardoso GC . 2010 . The dual function of barred plumage in birds: camouflage and communication . J Evol Biol . 23 : 2501 – 2506 . Google Scholar CrossRef Search ADS PubMed Gluckman TL , Mundy NI . 2013 . Cuckoos in raptors’ clothing: barred plumage illuminates a fundamental principle of Batesian mimicry . Anim Behav . 86 : 1165 – 1181 . Google Scholar CrossRef Search ADS Gomes ACR , Funghi C , Soma M , Sorenson MD , Cardoso GC . 2017 . Multimodal signalling in estrildid finches: song, dance and colour are associated with different ecological and life-history traits . J Evol Biol . 30:1336–1346 Gomes ACR , Sorenson MD , Cardoso GC . 2016 . Speciation is associated with changing ornamentation rather than stronger sexual selection . Evolution . 70 : 2823 – 2838 . Google Scholar CrossRef Search ADS PubMed Goodwin D . 1982 . Estrildid finches of the world . New York (NY) : Cornell University Press . Griffith S , Parker TH , Olson VA . 2006 . Melanin- versus carotenoid-based sexual signals: is the difference really so black and red ? Anim Behav . 71 : 749 – 763 . Google Scholar CrossRef Search ADS Hadfield J . 2010 . MCMC methods for multi-response generalized linear mixed models : the MCMCglmm R package . J Stat Softw . 33 : 1 – 22 . Google Scholar CrossRef Search ADS PubMed Hadfield JD , Nakagawa S . 2010 . General quantitative genetic methods for comparative biology: phylogenies, taxonomies and multi-trait models for continuous and categorical characters . J Evol Biol . 23 : 494 – 508 . Google Scholar CrossRef Search ADS PubMed Halperin T , Carmel L , Hawlena D . 2017 . Movement correlates of lizards’ dorsal pigmentation patterns . Funct Ecol . 31 : 370 – 376 . Google Scholar CrossRef Search ADS Hasson O . 1991 . Sexual displays as amplifiers: practical examples with an emphasis on feather decorations . Behav Ecol . 2 : 189 – 197 . Google Scholar CrossRef Search ADS Hill GE , McGraw KJ . 2006 . Bird coloration . Boston (MA) : Harvard University Press . Ho LST , Ane C , Lachlan R , Tarpinian K , Feldman R , Yu Q . 2016 . Package “phylolm.”https://cran.r-project.org/web/packages/ phylolm/index.html (accessed 28 March 2018). Jetz W , Thomas GH , Joy JB , Hartmann K , Mooers AO . 2012 . The global diversity of birds in space and time . Nature . 491 : 444 – 448 . Google Scholar CrossRef Search ADS PubMed Johnson KP . 2000 . The evolution of courtship display repertoire size in the dabbling ducks (Anatini) . J Evol Biol . 13 : 634 – 644 . Google Scholar CrossRef Search ADS Kabasakal B , Poláček M , Aslan A , Hoi H , Erdoğan A , Griggio M . 2017 . Sexual and non-sexual social preferences in male and female white-eyed bulbuls . Sci Rep . 7 : 5847 . Google Scholar CrossRef Search ADS PubMed Kelley JL , Fitzpatrick JL , Merilaita S . 2013 . Spots and stripes: ecology and colour pattern evolution in butterflyfishes . Proc Biol Sci . 280 : 20122730 . Google Scholar CrossRef Search ADS PubMed Krzyszczyk E , Mann J . 2012 . Why become speckled? Ontogeny and function of speckling in Shark Bay bottlenose dolphins (Tursiops sp.) . Mar Mammal Sci . 28 : 295 – 307 . Google Scholar CrossRef Search ADS Maddison WP , Maddison DR . 2011 . Mesquite: a modular system for evolutionary analysis . http://mesquiteproject.wikispaces.com (accessed 28 March 2018). Marques CI , Batalha HR , Cardoso GC . 2016 . Signalling with a cryptic trait: the regularity of barred plumage in common waxbills . R Soc Open Sci . 3 : 160195 . Google Scholar CrossRef Search ADS PubMed Marshall KL , Gluckman TL . 2015 . The evolution of pattern camouflage strategies in waterfowl and game birds . Ecol Evol . 5 : 1981 – 1991 . Google Scholar CrossRef Search ADS PubMed Mäthger LM , Hanlon RT . 2006 . Anatomical basis for camouflaged polarized light communication in squid . Biol Lett . 2 : 494 – 496 . Google Scholar CrossRef Search ADS PubMed Mäthger LM , Shashar N , Hanlon RT . 2012 . Do cephalopods communicate using polarized light reflections from their skin ? J Exp Biol . 212 : 2133 – 2140 . Google Scholar CrossRef Search ADS Møller AP , Pomiankowski A . 1993 . Why have birds got multiple sexual ornaments ? Behav Ecol Sociobiol . 32 : 167 – 176 . Odeen A , Hastad O . 2003 . Complex distribution of avian color vision systems revealed by sequencing the SWS1 opsin from total DNA . Mol Biol Evol . 20 : 855 – 861 . Google Scholar CrossRef Search ADS PubMed Olea PP , Casas F , Redpath S , Viñuela J . 2010 . Bottoms up: great bustards use the sun to maximise signal efficacy . Behav Ecol Sociobiol . 64 : 927 – 937 . Google Scholar CrossRef Search ADS Ortolani A . 1999 . Spots, stripes, tail tips and dark eyes: predicting the function of carnivore colour patterns using the comparative method . Biol J Linn Soc . 67 : 433 – 476 . Google Scholar CrossRef Search ADS Ota N , Gahr M , Soma M . 2015 . Tap dancing birds: the multimodal mutual courtship display of males and females in a socially monogamous songbird . Sci Rep . 5 : 16614 . Google Scholar CrossRef Search ADS PubMed Ota N , Gahr M , Soma M . 2017 . Songbird tap dancing produces non-vocal sounds . Bioacoustics . 26 : 161 – 168 . Google Scholar CrossRef Search ADS Pérez-Rodríguez L , Jovani R , Stevens M . 2017 . Shape matters: animal colour patterns as signals of individual quality . Proc R Soc B Biol Sci . 284 : 20162446 . Google Scholar CrossRef Search ADS Puebla O , Bermingham E , Guichard F , Whiteman E . 2007 . Colour pattern as a single trait driving speciation in Hypoplectrus coral reef fishes ? Proc Biol Sci . 274 : 1265 – 1271 . Google Scholar CrossRef Search ADS PubMed R Core Team . 2016 . R: a language and environment for statistical computing . Vienna (Austria) : R Foundation for Statistical Computing . https://www.R-project.org/ Restall R . 1996 . Munias and mannikins . East Sussex (UK) : Pica Press . Santana SE , Alfaro JL , Alfaro ME . 2012 . Adaptive evolution of facial colour patterns in Neotropical primates . Proc R Soc B Biol Sci . 279 : 2204 – 2211 . Google Scholar CrossRef Search ADS Schneider CA , Rasband WS , Eliceiri KW . 2012 . NIH Image to ImageJ: 25 years of image analysis . Nat Methods . 9 : 671 – 675 . Google Scholar CrossRef Search ADS PubMed Schuetz JG . 2004 . Common waxbills use carnivore scat to reduce the risk of nest predation . Behav Ecol . 16 : 133 – 137 . Google Scholar CrossRef Search ADS Seehausen O , Mayhew PJ , van Alphen JJM . 1999 . Evolution of colour patterns in East African cichlid fish . J Evol Biol . 12 : 514 – 534 . Google Scholar CrossRef Search ADS Semel B , Sherman PW . 2001 . Intraspecific parasitism and nest-site competition in wood ducks . Anim Behav . 61 : 787 – 803 . Google Scholar CrossRef Search ADS Soma M , Garamszegi LZ . 2015 . Evolution of courtship display in Estrildid finches: dance in relation to female song and plumage ornamentation . Front Ecol Evol . 3 : 4 . Google Scholar CrossRef Search ADS Soma M , Garamszegi LZ . 2018 . Data from: evolution of patterned plumage as a sexual signal in estrildid finches . Dryad Digital Repository. http://dx.doi. org/10.5061/dryad.35j39r8 Soma M , Iwama M . 2017 . Mating success follows duet dancing in the Java sparrow . PLoS One . 12 : e0172655 . Google Scholar CrossRef Search ADS PubMed Sorenson MD , Balakrishnan CN , Payne RB . 2004 . Clade-limited colonization in brood parasitic finches (Vidua spp.) . Syst Biol . 53 : 140 – 153 . Google Scholar CrossRef Search ADS PubMed Stevens M , Merilaita S . 2009 . Animal camouflage: current issues and new perspectives . Phil Trans R Soc B . 364 : 423 – 427 . Google Scholar CrossRef Search ADS PubMed Sun K , Meiklejohn KA , Faircloth BC , Glenn TC , Braun EL , Kimball RT . 2014 . The evolution of peafowl and other taxa with ocelli (eyespots): a phylogenomic approach . Proc R Soc B Biol Sci . 281 : 20140823 . Google Scholar CrossRef Search ADS Swaddle JP , Cuthill IC . 1994 . Female zebra finches prefer males with symmetric chest plumage . Proc R Soc B Biol Sci . 258 : 267 – 271 . Google Scholar CrossRef Search ADS Tibbetts EA , Dale J . 2007 . Individual recognition: it is good to be different . Trends Ecol Evol . 22 : 529 – 537 . Google Scholar CrossRef Search ADS PubMed Tobias JA , Montgomerie R , Lyon BE . 2012 . The evolution of female ornaments and weaponry: social selection, sexual selection and ecological competition . Phil Trans R Soc B . 367 : 2274 – 2293 . Google Scholar CrossRef Search ADS PubMed Ullrich R , Norton P , Scharff C . 2016 . Waltzing Taeniopygia: integration of courtship song and dance in the domesticated Australian zebra finch . Anim Behav . 112 : 285 – 300 . Google Scholar CrossRef Search ADS West-Eberhard MJ . 1979 . Sexual selection, social competition, and evolution . Proc Am Philos Soc . 123 : 222 – 234 . Yom-Tov Y . 2001 . An updated list and some comments on the occurrence of intraspecific nest parasitism in birds . Ibis . 143 : 133 – 143 . Google Scholar CrossRef Search ADS Zahavi A . 1978 . Decorative patterns and the evolution of art . New Scientist . 19 : 182 – 184 . Zann RA . 1996 . The zebra finch: a synthesis of field and laboratory studies . Oxford: Oxford University Press . Zanollo V , Griggio M , Robertson J , Kleindorfer S . 2012 . The number and coloration of white flank spots predict the strength of a cutaneous immune response in female Diamond Firetails, Stagonopleura guttata . J Ornithol . 153 : 1233 – 1244 . Google Scholar CrossRef Search ADS Zanollo V , Griggio M , Robertson J , Kleindorfer S . 2013 . Males with a faster courtship display have more white spots and higher pairing success in the diamond firetail, Stagonopleura guttata . Ethology . 119 : 344 – 352 . Google Scholar CrossRef Search ADS Zanollo V , Griggio M , Robertson J , Kleindorfer S . 2014 . Assortative pairings in diamond firetails (Stagonopleura guttata) are not the result of mutual mate choice for an ornament . Ethology . 120 : 951 – 964 . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Behavioral Ecology Oxford University Press

Evolution of patterned plumage as a sexual signal in estrildid finches

Loading next page...
 
/lp/ou_press/evolution-of-patterned-plumage-as-a-sexual-signal-in-estrildid-finches-HUoxgoyQcZ
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
1045-2249
eISSN
1465-7279
D.O.I.
10.1093/beheco/ary021
Publisher site
See Article on Publisher Site

Abstract

Abstract Color patterns, such as bars or dots that cover the body surface of animals are generally thought to play roles in signaling and camouflage. In birds, however, the macroscopic aspects of plumage coloration are less well understood, as past studies typically described plumage colorations by using spectrophotometric analyses. To provide insight into the evolution of plumage patterns as sexual signals, we characterized interspecific and intersexual variations in the plumage patterns of estrildid finches and tested their associations with other courtship signals and life-history traits using a comparative phylogenetic approach. Our results support the idea that plumage patterns in estrildids are favored by sexual selection because large-sized conspicuous plumage patterns are possessed by species with an elaborate courtship dance. These plumage patterns may also play roles in social signaling because patterns are more conspicuous in species with intraspecific brood parasitism. We predict that pattern traits can be favored by mate choice or intrasexual competition when they can serve as honest indicators of individual condition. As our results are consistent between the sexes, we suggest that the same selective force is acting on the evolution of plumage patterns in males and females in parallel. Finally, we also found a trade-off between large size and vivid color patterns, suggesting that too conspicuous patterns are costly, presumably because of the risk of catching the eyes of potential predators. Therefore, plumage patterns are also shaped by natural selection. INTRODUCTION Many animals are not uniformly colored. They can have color patches, spots, or stripes, which are likely to have evolved through predator–prey interactions, or for intraspecific communication (Caro 2005; Stevens and Merilaita 2009). Despite the definition of “color pattern” varying among both individual studies and species, a number of previous studies have described the functions of color patterns in a range of species. In particular, body color patterns are shown to function for camouflage in diverse species, like moths (Bond and Kamil 2002), squirrels (Ancillotto and Mori 2016), and carnivores (Ortolani 1999; Allen et al. 2011), while bold patterns (e.g., red spots on black) that make animals conspicuous serve as aposematic signals in spiders (Brandley et al. 2016) and poison frogs (Darst et al. 2006). Social living has also promoted the evolution of body color patterns that can code individual information in cetaceans (Caro et al. 2011; Krzyszczyk and Mann 2012), or face color patterns in some mammals (squirrels: Ancillotto and Mori 2016; primates: Santana et al. 2012). Lastly, and most importantly for this study, color patterns can affect mating, which has been well documented in fish (e.g., Endler 1983; Puebla et al. 2007). Although color patterns have long been discussed in the context of conspicuousness versus camouflage (Endler 1978; Gluckman and Cardoso 2010), few empirical studies on birds exist on this topic. Considering that birds play an essential role in predator–prey interactions, and can possess colorful plumage that functions for signaling, it would be of great importance to look into plumage color patterns in birds from such a viewpoint. In contrast with conspicuous plumage colors (e.g., carotenoid, UV) that have been extensively studied using pin-point data relying on the spectrometry techniques (Andersson 1994; Hill and McGraw 2006), the signaling functions of plumage color patterns are only beginning to be understood in an evolutionary context (cf. Gluckman and Mundy 2013; Gluckman 2014; Marshall and Gluckman 2015). We suggest that the extensive interspecific variations in the macroscopic aspects of plumage coloration deserve an evolutionary explanation for a complete understanding of how feather colors function as visual stimuli (cf. Endler 2012; Dale et al. 2015). Patterns focused in this study are defined as repetitive and regular appearances of color patches of a certain shape (i.e., dot, bar, and mottle) covering some background color (see also Methods and Figure 1). Considering that many birds have these patterns limited to small areas of the body surface, and that males and females can differ in the presence or conspicuousness of the patterns (Gluckman and Cardoso 2010) (see examples in Figure 1a–g), we hypothesize that in addition to colors, plumage patterns have important signaling roles in birds. Figure 1 View largeDownload slide Examples of interspecific and intersex variations in plumage color patterns (a–g), and the loading plot for the principal components analysis on the geometric features of plumage patterns (h). The patterned area on the body surface is schematically shown in the illustrations (a–g). Data points of the examples (a–g) are shown on the scatter plot, where the data points of males and females overlap for some species (a, e). Note that we performed a single PCA for a mixed sex dataset (see Methods for details), but that we plotted male and female in different colors only for illustrative purpose. Figure 1 View largeDownload slide Examples of interspecific and intersex variations in plumage color patterns (a–g), and the loading plot for the principal components analysis on the geometric features of plumage patterns (h). The patterned area on the body surface is schematically shown in the illustrations (a–g). Data points of the examples (a–g) are shown on the scatter plot, where the data points of males and females overlap for some species (a, e). Note that we performed a single PCA for a mixed sex dataset (see Methods for details), but that we plotted male and female in different colors only for illustrative purpose. So far, several studies have indicated that plumage patterns are used as visual signals among conspecifics. In the zebra finch Taeniopygia guttata, a popular model species of sexual selection with sexually dimorphic plumage patterns, males but not females have black and white stripes on the chest. The melanin coloration of these stripes is shaped by body condition during the juvenile stage (Birkhead et al. 1999), and females prefer males with symmetric chest stripes (Swaddle and Cuthill 1994). Similarly, in common waxbills Estrilda astrild, the regularity of barred plumage on the back is more salient in males than females, and is positively related to body condition and to the expression of colored ornamental traits (Marques et al. 2016). Moreover, patterns can play roles in broader social contexts not strictly limited to male signaling: in the diamond firetail Stagonopleura guttata, the characteristics of flank spots predict social dominance in females (Crowhurst et al. 2012), while in the red-legged partridge, both males and females show off flank stripes during agonistic displays (Bortolotti et al. 2006). Despite the list of within-species evidence for the use of feather pattern traits in signaling, one fundamental question remains: why some species have patterns while others do not, which begs for evolutionary explanation and between-species comparisons. One straightforward approach to this is to provide insight into the interspecific variations in plumage patterns and to identify their relationships with other morphological or behavioral sexual signals. Multiple traits (e.g., plumage patterns and courtship displays) can evolve in the same direction (e.g., general conspicuousness) under the same selective force, such as strong sexual selection, where redundant or multiple ornamentation is favored (Møller and Pomiankowski 1993; Johnson 2000; Candolin 2003). Conversely, it is also possible that both morphological ornamentation and behavioral displays become less favored when organisms are exposed to higher predation risks. Such correlated evolution would also be expected when the traits are under the control of the same pleiotropic mechanisms, e.g., hormones, genes, or neurotransmitters. Moreover, when birds have both morphological ornamentation and behavioral displays, the latter signals may work as an “amplifier” of ornamentation (Zahavi 1978; Hasson 1991). As a textbook example, peacocks Pavo cristatus, which are known for having long trains with eyespots, perform courtship displays facing toward the sun to highlight the iridescent color patterns of the erected feathers (Dakin and Montgomerie 2009). Similarly, visual displays of some birds are performed in a way that enhances the signal efficacy of ornamentation (Candolin 2003; Fusani et al. 2007; Olea et al. 2010; Bortolotti et al. 2011). In addition, the idea of correlated evolution may hold true for the association among different facets of patterns, i.e., coloration and size. Depending on the strength of the selective force, a plumage pattern may become conspicuous or less conspicuous in terms of both color and size. In contrast, if it is too costly to have large-sized vivid color patterns at the same time because of intense predation risks, there might be a trade-off between the 2 aspects of the patterns. Considering that some birds show sexual dichromatism in melanin- and carotenoid-based coloration (Hill and McGraw 2006), such a relationship between size and color could also differ between the sexes. Lastly, comparing the evolution of plumage patterns between sexes would provide insight into their potential roles as sexual signals. Although sexual traits in birds, such as songs and ornamental colors, tend to be male-biased (Dale et al. 2015), it has been reported that plumage patterns are female-biased, suggesting that they function as cryptic sexual signals, which are particularly widespread in females (Gluckman and Cardoso 2010). Estrildid finches (family: Estrildidae) are ideal candidates to investigate the possible role of plumage patterns in sexual selection. First, within-species evidence in some estrildids suggests that patterns can have signaling functions (zebra finch: Swaddle and Cuthill 1994; Birkhead et al. 1999; diamond firetail: Crowhurst et al. 2012; Zanollo et al. 2013, 2014; common waxbill: Marques et al. 2016). Second, a relatively larger number of species has plumage patterns in this taxonomic group (57.5%), with some showing sexual dimorphism in the presence or conspicuousness of patterns (e.g., Marques et al. 2016, Figure 1), suggesting the possibility that plumage patterns may be linked to sexual selection. Third, estrildid finches are characterized by elaborate courtship signals, including acoustic and behavioral displays (Goodwin 1982; Baptista et al. 1999; Soma and Garamszegi 2015). Most importantly, when potential mating partners come close, often perching side by side, they show a ritualized courtship dance (Goodwin 1982; Soma and Garamszegi 2015), which is performed by males of some species (Zanollo et al. 2013; Ullrich et al. 2016) or by both sexes in other species (Ota et al. 2015, 2017; Soma and Iwama 2017). Such behavioral features of estrildid finches often involve display positions in which the plumage color patterns are exposed and might explain why they frequently have flank patterns (e.g., Crowhurst et al. 2012, see also Figure 1a–g). Plumage coloration has been studied from evolutionary perspectives in estrildids, but plumage patterns have not been investigated in a comparative phylogenetic context. In our previous comparative phylogenetic studies of estrildid finches (Soma and Garamszegi 2015; Gomes et al. 2017), we did not find a link between the evolution of dance displays and ornamental coloration, measured based on the coverage and reflectance spectrophotometry. In another comparative phylogenetic study, ornamental coloration in males was higher in species with a gregarious nature, suggesting that social selection affects the evolution of plumage coloration (Gomes et al. 2016). In this study of estrildid finches, we characterized plumage patterns and tested the following set of predictions by applying a phylogenetic comparative approach to provide insight into the evolution of plumage patterns as sexual signals. 1) As visual signals, the conspicuousness of plumage patterns (e.g., size) may have evolved in association with other visual traits, such as pattern color and courtship display. 2) The sexes may undergo different selection regimes, which can be detected by comparing the evolution of plumage patterns between the sexes. 3) Interspecific variations in life history can be potentially linked to the interspecific variations in plumage patterns because it has already been shown to be related to plumage coloration (Gomes et al. 2016). MATERIALS AND METHODS Collecting the plumage pattern data We collected data on plumage patterns by taking pictures of skin specimens at the Natural History Museum at Tring, UK and Yamashina Institute for Ornithology at Chiba, Japan. We sampled 5 individuals of each sex for each estrildid finch species, when possible, and took pictures from 4 angles (ventral, dorsal, and left and right sides) with a scale using a digital camera (α5000; Sony, Tokyo Japan). We also took UV photos (digital camera: D70s, Nikon, Tokyo, Japan; lens: UV-105 mm F4.5, Tochigi Nikon, Tochigi, Japan; filter: U360, Hoya Optronics, Tokyo, Japan; light: Handy UV Lamp 365nm; AsOne, Osaka, Japan), considering that estrildids are likely to have UV vision (Ödeen and Håstad 2003), to verify that there was no pattern in the nonvisible part of the spectrum that was apparent only under UV light. We collected data on both sexes from 126 species, and male data from one species, of the 134 estrildid species. Based on the fact that some estrildid species show intermediate patterns between bars and dots (e.g., overlapped dots forming lines), we applied the definition of pattern trait in a broad sense (i.e., a repetition of the same shape regardless of whether patterns form bars, dots, or mottles) instead of setting small categories dependent on unit shape. We calculated the following 4 geometric features of patterns from the digital images, using ImageJ 1.48v (Schneider et al. 2012). We measured pattern coverage on the dorsal, ventral, and side (average of left and right sides) views, calculated as the percentage of patterned area per total body surface in each picture. As an index of pattern conspicuousness, we measured the size of the unit shape that constituted the pattern, by taking the width of a bar or the diameter of a mottle or a dot (in mm). After checking that each of the 4 variables showed fairly high repeatability within the same sex of each species (r > 0.86, P < 0.0001; Garamszegi 2014), we used sex-specific average values for each species in subsequent analyses. As some species had multiple types of patterns (e.g., zebra finch males showing flank dots, chest stripes, and tail bars), we counted the total number of patterns for each sex of each species. The above 4 measures were treated as 0, respectively, when no plumage pattern was evident. To avoid potential interobserver bias, all measurements were performed by the same observer (M.S.). Using the species for which we measured specimens in the 2 museums, we checked for potential biases originating from the fact that different museums may store specimens differently or may have collections of different ages. However, we found no differences between the museums (linear mixed model including species as a random effect and sex and museum as fixed effects: effect of museum on side coverage, t = 1.01, P = 0.32; effect of museum on ventral coverage, t = 1.11, P = 0.27; effect of museum on dorsal coverage, t = 0.69, P = 0.49; effect of museum on pattern unit size, t = 0.53, P = 0.60). In addition, we recorded whether the patterns that each species/sex had were composed of melanin-based or carotenoid-based coloration or a mixture of both. Melanin and carotenoid are the 2 main pigment types in birds, responsible for red and orange coloration or black, gray, and brown, respectively (Hill and McGraw 2006). Both pigmentations, including pigment-based patterns, are argued to play roles in signaling individual conditions (Griffith et al. 2006; Pérez-Rodríguez et al. 2017). Based on the visual inspection of the feather colors, they were scored as melanin (0); melanin and carotenoid (1); or carotenoid (2). Melanin patterns are colored with combinations of black, gray, brown, and white, while the carotenoid patterns are colored with red and white (see Figure 1a,d for the comparison of melanin and carotenoid patterns expressed in females and males, respectively). As it was found that some birds have white spots on the basis of a black-to-red gradation background, they were categorized as “melanin and carotenoid” patterns. Other sexual signals and life-history traits As potential signals that can coevolve with plumage patterns, we characterized the degree of expression of courtship displays of each sex. Specifically, we quantified the dance repertoires of males and females, under the prediction that the complexity of a visual display (i.e., courtship dance) evolved jointly with plumage conspicuousness to emphasize showiness. Courtship dance is stereotyped within species and is expressed as a combination of several simple actions (e.g., Restall 1996; Zanollo et al. 2013; Ota et al. 2015; Ullrich et al. 2016; Soma and Iwama 2017). We counted the number of dance categories that constituted the courtship displays of each species, separately for males and females (see also Soma and Garamszegi 2015). In addition, we also considered life-history variables that likely affect the cost of reproduction, such as median clutch size, and body size taken as body length. As some estrildid finches in Africa are targets of interspecific brood parasitism by birds in the genus Vidua (Sorenson et al. 2004), the presence or absence of interspecific brood parasitism was scored as 0 or 1 (parasitism absent–present, respectively). Furthermore, some estrildid finches show intraspecific brood parasitism (Yom-Tov 2001), which was scored as 0 or 1 (parasitism absent–present, respectively). Although most species do not defend territories, the degree of gregariousness varies among species and was scored as follows: colonial breeders or species with social systems in which multiple pairs keep contact with each other even during the breeding season were given a score of 3; highly social and gregarious species, in which aggregations mainly occur outside the breeding season were given a score of 2; species that breed usually in pairs or in small parties were given a score of 1; and strictly territorial species were given a score of 0. All of these data were compiled from an earlier study (Soma and Garamszegi 2015), and the number of focal species was limited (n = 85) because of the availability of information. Phylogeny For the comparative phylogenetic analyses described below, we could not obtain an overwhelmingly supported consensus phylogenetic tree with branch lengths, but were able to derive multiple equally likely candidate trees from Jetz et al. (2012). We derived 1000 trees from the dataset for the focal species in each analysis and used their consensus tree without branch length to reconstruct the ancestral state or used each of them in phylogenetic regressions followed by multimodel inference (see below). Statistical analyses Principal component analysis (PCA) The 4 geometric variables that describe plumage patterns (dorsal, ventral, and side coverage and unit size) are mutually dependent and may be related with the number of pattern types of each species/sex. We conducted a PCA to define independent axes that have distinct biological meaning from the raw variables, in which we used square root values for coverage, pattern unit size, and the number of patterns of each sex. We performed a single conventional PCA for a mixed sex dataset, instead of repeating phylogenetic PCA for each sex, in order to obtain PC scores comparable between sexes. Specifically, using data from 127 males and 126 females, we aimed to achieve a dimension reduction by extracting principal component axes (PCs) that account for most of the variance (>70%), have eigenvalues larger than 1 and have biologically interpretable component loadings. Below we show PCA results for PC1 and PC2, however, based on the above criteria, only PC1 was used for subsequent analyses (see Results for details). We confirmed that these PC scored obtained from a conventional PCA were highly correlated with those obtained from phylogenetic PCA (r > 0.997, P < 0.0001). We used R 3.3.1 software (R Core Team 2016) for the analyses described here and below, except otherwise stated. Ancestral state reconstruction To characterize the evolutionary history of plumage patterns, and to compare it between sexes, we reconstructed the ancestral state using the PC scores in Mesquite (Maddison and Maddison 2011). A consensus phylogenetic tree was obtained based on 1000 trees from the dataset (Jetz et al. 2012). Parsimony methods were used as opposed to maximum likelihood, due to the latter being unavailable for our data, i.e., multifurcating tree. The phylogenetic relationship between color and patterns We investigated the association between geometric features and colors of the patterns with the aim of revealing a positive or negative correlation between the 2 different aspects under our predictions. We used species that have a plumage pattern, and tested the effects of pattern color (i.e., melanin, carotenoid, or mixture of both) and sex on PC1. Nonindependence of the data due to the phylogenetic relatedness of species was controlled by Bayesian phylogenetic mixed models (Hadfield and Nakagawa 2010) in R package “MCMCglmm” (Hadfield 2010) because it allows multiple data entries (i.e., male and female PC scores) per species. In the model, we used a Gaussian error distribution, and the priors [G = list(G1= list(V = 1, nu = 0.02), G2 = list(V = 1, nu = 0.02))]. We repeatedly fit the same model using each of the 1,000 phylogenetic trees obtained from the dataset (Jetz et al. 2012), all of which successfully converged, and obtained the mean coefficients for the predictor variables and their 95% confidence interval by model averaging the 1000 outcomes. We weighted parameter estimates based on the DIC of the respective model corresponding to a particular tree. For this particular analysis, we relied on MCMCglmm because we needed to analyze the data with multiple entries (i.e., male and female variables) per species, therefore requiring an approach based on phylogenetic mixed modeling. In the following analyses, however, we applied the phylogenetic generalized least-squares (PGLS) regression technique, available in the package phylolm (Ho et al. 2016), and analyzed males and females separately. The phylogenetic relationship between courtship dance and plumage patterns To examine the potential relationships between plumage patterns and courtship dance, we tested if the PC1 covaried with the dance repertoire at the interspecific level, separately in the 2 sexes. We fitted PGLS regressions to control for phylogenetic dependence by using the 1,000 phylogenetic trees as explained earlier, in which we entered PC1 as predictor variables and dance repertoire as a response variable. As dance repertoires are discrete values (count data), we used regression with Poisson error distribution. For each tree, we fitted models with identical predictor/response structure, and then derived mean and confidence estimates across trees for the estimated parameters (slopes and intercepts). Summary statistics were obtained via model averaging Garamszegi and Mundry 2014). For the PGLS modeling, we used the R package “phylolm” that allows Poisson distribution (Ho et al. 2016) The phylogenetic relationship between plumage patterns and life-history variables We tested for the relationship between plumage pattern and life-history variables that likely affect the cost of reproduction. Specifically, we used PGLS framework available in R package “phylolm” again, and constructed models in which clutch size, body size, presence of interspecific and intraspecific brood parasitism and sociality were entered as predictor variables, and PC1 was used as a response variable (different models for males and females), using Gaussian distribution. As in the above analyses, we used the same set of 1000 phylogenetic trees and applied model averaging to obtain mean and confidence estimates for the estimated parameters (intercept and slopes for each predictor). Parameter estimates from each model were weighted according to their relative fit to the data (Garamszegi and Mundry 2014). RESULTS Principal component analysis PC1 accounted for 72.8% of the total variation in the variables that were taken to describe plumage pattern and had positive loadings (>0.36) for all 5 variables (Table 1, Figure 1h). Therefore, higher PC1 scores can indicate larger and conspicuous patterns. In contrast, PC2 accounted for only 15.2% of the total variation and had a higher positive loading (>0.8) only for dorsal coverage but negative loadings (<−0.30) for ventral coverage and pattern unit size (Table 1, Figure 1h). Given the low eigenvalue of PC2 (Table 1), we mainly focused on PC1 (conspicuousness) hereafter. Table 1 The results of the principal components analysis on mixed sex data of patterned variables PC1 PC2 Side coverage (square root) 0.505 0.141 Ventral coverage (square root) 0.457 −0.300 Dorsal coverage (square root) 0.364 0.803 Pattern unit size 0.425 −0.494 Number of pattern types 0.472 −0.035 Eigenvalue 3.638 0.761 Proportion of variance 0.728 0.152 Cumulative proportion 0.728 0.880 PC1 PC2 Side coverage (square root) 0.505 0.141 Ventral coverage (square root) 0.457 −0.300 Dorsal coverage (square root) 0.364 0.803 Pattern unit size 0.425 −0.494 Number of pattern types 0.472 −0.035 Eigenvalue 3.638 0.761 Proportion of variance 0.728 0.152 Cumulative proportion 0.728 0.880 View Large Table 1 The results of the principal components analysis on mixed sex data of patterned variables PC1 PC2 Side coverage (square root) 0.505 0.141 Ventral coverage (square root) 0.457 −0.300 Dorsal coverage (square root) 0.364 0.803 Pattern unit size 0.425 −0.494 Number of pattern types 0.472 −0.035 Eigenvalue 3.638 0.761 Proportion of variance 0.728 0.152 Cumulative proportion 0.728 0.880 PC1 PC2 Side coverage (square root) 0.505 0.141 Ventral coverage (square root) 0.457 −0.300 Dorsal coverage (square root) 0.364 0.803 Pattern unit size 0.425 −0.494 Number of pattern types 0.472 −0.035 Eigenvalue 3.638 0.761 Proportion of variance 0.728 0.152 Cumulative proportion 0.728 0.880 View Large Comparison of pattern evolution between sexes The reconstruction of ancestral states revealed that the plumage color patterns evolved in a similar fashion in males and females (Figure 2, see also Supplementary Figure S1). Additionally, we also directly compared PC1 between the sexes using the paired t-test and found that PC1 did not differ significantly between the sexes (paired t-test: t = 1.04, P = 0.302, N = 124 species). Figure 2 View largeDownload slide Reconstruction of ancestral states of plumage color patterns (PC1) in males and females (see Supplementary Figure S1 for PC2). Figure 2 View largeDownload slide Reconstruction of ancestral states of plumage color patterns (PC1) in males and females (see Supplementary Figure S1 for PC2). The phylogenetic relationship between color and patterns We examined whether geometric features and the colors of patterns were inter-related by focusing on species that had plumage patterns and for which information on the phylogenetic relationships was available (male: N = 63 species; female: N = 65 species). We found an interspecific relationship between color and conspicuousness of patterns: PC1 scores were lower in species in which patterns were carotenoid-dependent than in species that had melanin-dependent trait expression in both sexes (Figure 3, Table 2). Figure 3 View largeDownload slide Comparison of principal component 1 (PC1) (pattern conspicuousness) of males and females in species with melanin- or carotenoid-based plumage patterns (see Table 2 for statistical outcomes). Square bars at the bottom show a schematic view of the plumage color categories (“Melanin and carotenoid” patterns typically have black-to-red gradation background with white spots, while the other 2 patterns have monotone backgrounds). Figure 3 View largeDownload slide Comparison of principal component 1 (PC1) (pattern conspicuousness) of males and females in species with melanin- or carotenoid-based plumage patterns (see Table 2 for statistical outcomes). Square bars at the bottom show a schematic view of the plumage color categories (“Melanin and carotenoid” patterns typically have black-to-red gradation background with white spots, while the other 2 patterns have monotone backgrounds). Table 2 Effects of color of plumage pattern (0: melanin; 1: melanin and carotenoid; 2: carotenoid) and sex on PC1 Coefficient 95% CI Intercept 1.776 (1.208, 2.297) Color −0.841 (−1.365,−0.282) Sex 0.029 (−0.158, 0.217) Coefficient 95% CI Intercept 1.776 (1.208, 2.297) Color −0.841 (−1.365,−0.282) Sex 0.029 (−0.158, 0.217) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as a significant effect. View Large Table 2 Effects of color of plumage pattern (0: melanin; 1: melanin and carotenoid; 2: carotenoid) and sex on PC1 Coefficient 95% CI Intercept 1.776 (1.208, 2.297) Color −0.841 (−1.365,−0.282) Sex 0.029 (−0.158, 0.217) Coefficient 95% CI Intercept 1.776 (1.208, 2.297) Color −0.841 (−1.365,−0.282) Sex 0.029 (−0.158, 0.217) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as a significant effect. View Large The phylogenetic relationship between courtship dance and plumage patterns Using the species data for which dance and pattern data were available, we examined the potential relationship between PC1 and complexity of the dance repertoire, and found that PC1 covaried positively with dance repertoire in each sex (Table 3, Figure 4). These associations indicate that species with conspicuous plumage patterns have more complex dances. Table 3 Effects of PC1 on the dance repertoire of each sex (a) Male dance repertoire (b) Female dance repertoire Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.532 0.108 (1.317, 1.747) 0.672 0.381 (−0.090, 1.434) PC1 0.029 0.009 (0.011, 0.047) 0.090 0.039 (0.012, 0.167) (a) Male dance repertoire (b) Female dance repertoire Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.532 0.108 (1.317, 1.747) 0.672 0.381 (−0.090, 1.434) PC1 0.029 0.009 (0.011, 0.047) 0.090 0.039 (0.012, 0.167) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as a significant effect. View Large Table 3 Effects of PC1 on the dance repertoire of each sex (a) Male dance repertoire (b) Female dance repertoire Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.532 0.108 (1.317, 1.747) 0.672 0.381 (−0.090, 1.434) PC1 0.029 0.009 (0.011, 0.047) 0.090 0.039 (0.012, 0.167) (a) Male dance repertoire (b) Female dance repertoire Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.532 0.108 (1.317, 1.747) 0.672 0.381 (−0.090, 1.434) PC1 0.029 0.009 (0.011, 0.047) 0.090 0.039 (0.012, 0.167) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as a significant effect. View Large Figure 4 View largeDownload slide Relationship between the complexity of courtship (dance repertoire) and PC1) in males. Species data are shown with semitransparent plots and the regression lines are drawn based on estimated parameters from the phylogenetic analyses summarized in Table 3. Figure 4 View largeDownload slide Relationship between the complexity of courtship (dance repertoire) and PC1) in males. Species data are shown with semitransparent plots and the regression lines are drawn based on estimated parameters from the phylogenetic analyses summarized in Table 3. The phylogenetic relationship between plumage patterns and life-history variables In both males and females, we found a significant association between plumage pattern and intraspecific brood parasitism, as PC1 was positively correlated with this trait (Figure 5, Table 4). The remaining life-history traits were not related significantly to the PC 1 (Table 4). Figure 5 View largeDownload slide Comparison of male and female PC1 between species with and without intraspecific brood parasitism. Figure 5 View largeDownload slide Comparison of male and female PC1 between species with and without intraspecific brood parasitism. Table 4 Effects of life-history traits on PC1 (plumage pattern conspicuousness) in males (a) and females (b), as estimated from the appropriate phylogenetic generalized least-squares models (a) Male PC1 (b) Female PC1 Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.736 2.268 (−2.801, 6.273) 0.952 2.126 (−3.301, 5.205) Clutch size −0.106 0.271 (−0.648, 0.437) −0.203 0.236 (−0.676, 0.270) Body size −0.107 0.146 (−0.399, 0.184) 0.010 0.141 (−0.272, 0.293) Interspecific brood parasitism 0.999 0.508 (−0.017, 2.015) 0.831 0.506 (−0.182, 1.843) Intraspecific brood parasitism 3.286 0.673 (1.940, 4.633) 2.539 0.716 (1.107, 3.872) Sociality −0.125 0.241 (−0.607, 0.358) −0.212 0.233 (−0.678, 0.253) (a) Male PC1 (b) Female PC1 Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.736 2.268 (−2.801, 6.273) 0.952 2.126 (−3.301, 5.205) Clutch size −0.106 0.271 (−0.648, 0.437) −0.203 0.236 (−0.676, 0.270) Body size −0.107 0.146 (−0.399, 0.184) 0.010 0.141 (−0.272, 0.293) Interspecific brood parasitism 0.999 0.508 (−0.017, 2.015) 0.831 0.506 (−0.182, 1.843) Intraspecific brood parasitism 3.286 0.673 (1.940, 4.633) 2.539 0.716 (1.107, 3.872) Sociality −0.125 0.241 (−0.607, 0.358) −0.212 0.233 (−0.678, 0.253) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as significant association. View Large Table 4 Effects of life-history traits on PC1 (plumage pattern conspicuousness) in males (a) and females (b), as estimated from the appropriate phylogenetic generalized least-squares models (a) Male PC1 (b) Female PC1 Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.736 2.268 (−2.801, 6.273) 0.952 2.126 (−3.301, 5.205) Clutch size −0.106 0.271 (−0.648, 0.437) −0.203 0.236 (−0.676, 0.270) Body size −0.107 0.146 (−0.399, 0.184) 0.010 0.141 (−0.272, 0.293) Interspecific brood parasitism 0.999 0.508 (−0.017, 2.015) 0.831 0.506 (−0.182, 1.843) Intraspecific brood parasitism 3.286 0.673 (1.940, 4.633) 2.539 0.716 (1.107, 3.872) Sociality −0.125 0.241 (−0.607, 0.358) −0.212 0.233 (−0.678, 0.253) (a) Male PC1 (b) Female PC1 Coefficient SE 95% CI Coefficient SE 95% CI Intercept 1.736 2.268 (−2.801, 6.273) 0.952 2.126 (−3.301, 5.205) Clutch size −0.106 0.271 (−0.648, 0.437) −0.203 0.236 (−0.676, 0.270) Body size −0.107 0.146 (−0.399, 0.184) 0.010 0.141 (−0.272, 0.293) Interspecific brood parasitism 0.999 0.508 (−0.017, 2.015) 0.831 0.506 (−0.182, 1.843) Intraspecific brood parasitism 3.286 0.673 (1.940, 4.633) 2.539 0.716 (1.107, 3.872) Sociality −0.125 0.241 (−0.607, 0.358) −0.212 0.233 (−0.678, 0.253) Bold typeface is used when 95% confidence interval (CI) does not contain zero; thus, it can be interpreted as significant association. View Large DISCUSSION The present phylogenetic comparative study focusing on geometric features of plumage patterns in estrildid finches revealed that they might have evolved for signaling function, for at least 2 reasons. First, the PCA showed that interspecific variations in patterns could be described along the spectrum of conspicuousness. The PC1 reflected larger and bold patterns that are possibly composed of multiple shapes (e.g., Figure 1b). Second, through the phylogenetic comparative analyses of PC1, we found evidence that plumage patterns may be coevolutionarily related to another sexual signal (i.e., dance repertoire), and to intraspecific competition (i.e., presence of intraspecific brood parasitism). As these trends were consistent between sexes (Tables 2–4), and males and females showed very similar levels of conspicuousness (Figures 1 and 2) we can infer that basically the same selective force is acting on the evolution of plumage patterns in males and females. Alternatively, selection may favor the trait in one sex only, but its effect is also manifested in the other due to strong genetic correlation between sexes. Lack of sex-biased evolution of plumage patterns Evolution of female ornamentation is a challenging question, and discussed in relation to the presence of intersexual genetic correlation, and/or sexual and social selection (West-Eberhard 1979; Tobias et al. 2012) in estrildid finches (Soma and Garamszegi 2015). Specifically, estrildids are characterized by multiple sexual traits that can be shared between sexes. Females of some species sing courtship songs like males (Goodwin 1982; Geberzahn and Gahr 2011, 2013), whereas females of other species perform courtship dance displays identical to those of males (Ota et al. 2015; Soma and Garamszegi 2015; Soma and Iwama 2017) or have plumage ornamentation comparable to that of males (Zanollo et al. 2014; Marques et al. 2016). However, these sexual signals (song, dance, and plumage coloration) are generally male-biased with regard to presence/absence or degree of elaboration in estrildids (Soma and Garamszegi 2015; see also Dale et al. 2015), which is contrasted with the plumage color pattern (cf. Gluckman and Cardoso 2010). Only a few species (Spermophaga spp.) show female-biased sexual dimorphism with regard to the presence of patterns, and there was no sex difference in PC1. Moreover, the sexes showed similar evolution of PC1 (Figure 2). This is quite different from what was found in peafowls and their related taxa, where eyespot patterns likely have evolved primarily in males for female choice (Sun et al. 2014). Presumably, plumage patterns in female estrildids would function either for sexual signaling to get mates (sexual selection), or for status signaling to repel rivals, even outside the reproductive context (social selection) (e.g., Kabasakal et al. 2017). Although there is limited empirical evidence available showing that selection operates on female estrildid finches, a study on one species supports this idea. Diamond firetail S. guttata, females have more flank spots than males on average (Zanollo et al. 2014), and the number of spots reflects physical condition (Zanollo et al. 2012) and predicts social dominance in females (Crowhurst et al. 2012). Therefore, similar selective forces may mediate the interspecific variation of plumage patterns in both males and females. Hiding versus signaling Diversity of color patterns in animals can be explained through hiding versus signaling strategy (Endler 1978; Stevens and Merilaita 2009; Gluckman and Cardoso 2010). Some animals balance the two by having cryptic patterns all over the body and conspicuous patterns on small areas that are important for communication, like face or tail (Caro 2005; Ancillotto and Mori 2016), while others have dual-purpose color patterns (e.g., squids: Mäthger and Hanlon 2006; Mäthger et al. 2012). In line with these previous insights into nonbird species, our findings indicate that plumage patterns in estrildid finches are also the product of a compromise between hiding and signaling, as suggested by Endler (1978). Estrildid finches need to escape potential predators, but they also need to be showy enough to attract or deter conspecifics. Consequently, they cannot have patterns that are large and red at the same time (Figure 3), or otherwise their appearance would stand out too much; thus, making it difficult to avoid predation. The gregarious nature and colonial breeding of estrildid finches are likely a response to high predation risks, at least in some species (Zann 1996). Although we did not find a direct link between gregariousness and plumage patterns in this study, some estrildid species build nests in a thorny tree or near a wasp nest (Goodwin 1982; Barnard and Markus 1990; Beier et al. 2006), while others add carnivore scat to their nest to prevent predation (Schuetz 2004). These strategies may indirectly suggest how much predation pressure can affect the life of estrildid finches. Phylogenetic comparative studies particularly focusing on this aspect of breeding biology could be designed in the future into this promising direction. Some finches are also known for their showy plumage patterns that play roles in sexual/social contexts (Swaddle and Cuthill 1994; Crowhurst et al. 2012; Zanollo et al. 2013, 2014; Marques et al. 2016). Consistent with such earlier findings, in this comparative study, we gained results that support the hypothesis that plumage patterns evolve alongside their signaling roles in sexual selection. Specifically, we found that the complexity of courtship dance has evolved in the same direction as the conspicuousness of plumage patterns (Table 3, Figure 4). This supports the idea that the 2 visual traits evolved under the same selective force, and our findings thus meet the prediction of the “amplifier” hypothesis (Zahavi 1978; Hasson 1991). However, more detailed investigations are needed to clarify how gestural displays functionally enhance signal efficacy of plumage patterns (or vice versa) in estrildid finches. In general, we expect that motion (i.e., direction, speed) might play an important role in how patterns are perceived by receivers (cf. Dakin and Montgomerie 2009). In the context of predator–prey interactions, it was already reported that particular patterns can make animals less detectable especially while they are moving (Halperin et al. 2017). Such potential roles remain to be explored in the context of sexual signaling. The function of plumage patterns could also include social signaling in addition to a sexual context. Having signaling traits that facilitate individual recognition is particularly favored in both species living in colonies and those with agonistic competitions (Tibbetts and Dale 2007). Here, we uncovered that the presence of intraspecific brood parasitism was associated with conspicuous plumage patterns (Table 4). This is also consistent with our previous study (Soma and Garamszegi 2015), in which we detected associations between dance complexity and intraspecific brood parasitism in both sexes. Apparently, the 2 visual signals, dance and plumage patterns, evolved in the same direction in response to intraspecific parasitism (Table 4). Intraspecific parasitism can cause intense within-species competition, where, for instance, host and parasite females can have a higher chance to directly interact or fight (e.g., Semel and Sherman 2001; Åhlund 2005). In such cases, condition-dependent honest indicators of quality (e.g., Marques et al. 2016) would be effective for avoiding escalated costly aggression. This is comparable to the previous findings from the passerine-wide phylogenetic comparative study showing that females of cooperative breeding species are more likely to have ornamental coloration due to female competition (Dale et al. 2015). To conclude, we have shown that plumage patterns constitute an important visual trait that can be the subject to both natural and sexual selection forces. Such a finding could not have been discovered solely with spectrometric techniques that are more commonly applied for examining the evolution of plumage traits of birds. Our results strongly suggest that the macroscopic aspects of plumage traits may open a new research agenda for understanding the evolution of complex plumage traits in birds, as done in other taxonomic groups (Ortolani 1999; Seehausen et al. 1999; Caro 2005; Caro et al. 2011; Allen et al. 2011; Santana et al. 2012; Allen et al. 2013; Kelley et al. 2013; Ancillotto and Mori 2016). We suggest that color and geometric features should be integrated to reach a comprehensive understanding of the physical appearance of complex plumage traits in animals (Endler 2012). SUPPLEMENTARY MATERIAL Supplementary data are available at Behavioral Ecology online. FUNDING This study was supported by the JSPS Grant-in-Aid for Young Scientists (23680027 and 16H06177) and Hokkaido University WinGS Global Networking Award 2015 to Masayo Soma, and by the Ministry of Economy and Competitiveness (Spain) (CGL2015-70639-P) and the National Research, Development and Innovation Office (Hungary) (K-115970) to L.Z.G. We thank the Natural History Museum and its senior curator Mark Adams, and the Yamashina Institute for Ornithology and its researcher Takeshi Yamasaki for their help assessing their ornithological collection. We also thank Nao Ota for advice on photography, and Honor Scarlett for reviewing the text. Data accessibility: Analyses reported in this article can be reproduced using the data provided by Soma and Garamszegi (2018). REFERENCES Åhlund M . 2005 . Behavioural tactics at nest visits differ between parasites and hosts in a brood-parasitic duck . Anim Behav . 70 : 433 – 440 . Google Scholar CrossRef Search ADS Allen WL , Baddeley R , Scott-Samuel NE , Cuthill IC . 2013 . The evolution and function of pattern diversity in snakes . Behav Ecol . 24 : 1237 – 1250 . Google Scholar CrossRef Search ADS Allen WL , Cuthill IC , Scott-Samuel NE , Baddeley R . 2011 . Why the leopard got its spots: relating pattern development to ecology in felids . Proc Biol Sci . 278 : 1373 – 1380 . Google Scholar CrossRef Search ADS PubMed Ancillotto L , Mori E . 2016 . Adaptive significance of coat colouration and patterns of Sciuromorpha (Rodentia) . Ethol Ecol Evol . 29 : 241 – 254 . Google Scholar CrossRef Search ADS Andersson M . 1994 . Sexual selection . Princeton (NJ) : Princeton University Press . Baptista LF , Lawson R , Visser E , Bell DA . 1999 . Relationships of some mannikins and waxbills in the estrildidae . J Ornithol . 140 : 179 – 192 . Google Scholar CrossRef Search ADS Barnard P , Markus MB . 1990 . Reproductive failure and nest site selection of two estrildid finches in Acacia woodland . Ostrich . 61 : 117 – 124 . Google Scholar CrossRef Search ADS Beier P , Tungbani AI , Brittingham M . 2006 . Nesting with the wasp Ropalidia cincta increases nest success of red-cheeked cordonbleu (Uraeginthus bengalus) in Ghana . Auk . 123 : 1022 – 1037 . Birkhead TR , Fletcher F , Pellatt EJ . 1999 . Nestling diet, secondary sexual traits and fitness in the zebra finch . Proc R Soc B Biol Sci . 266 : 385 – 390 . Google Scholar CrossRef Search ADS Bond AB , Kamil AC . 2002 . Visual predators select for crypticity and polymorphism in virtual prey . Nature . 415 : 609 – 613 . Google Scholar CrossRef Search ADS PubMed Bortolotti GR , Blas J , Negro JJ , Tella JL . 2006 . A complex plumage pattern as an honest social signal . Anim Behav . 72 : 423 – 430 . Google Scholar CrossRef Search ADS Bortolotti GR , Stoffel MJ , Galván I . 2011 . Wintering snowy owls Bubo scandiacus integrate plumage colour, behaviour and their environment to maximize efficacy of visual displays . Ibis . 153 : 134 – 142 . Google Scholar CrossRef Search ADS Brandley N , Johnson M , Johnsen S . 2016 . Aposematic signals in North American black widows are more conspicuous to predators than to prey . Behav Ecol . 27 : 1104 – 1112 . Google Scholar CrossRef Search ADS Candolin U . 2003 . The use of multiple cues in mate choice . Biol Rev . 78 : 575 – 595 . Google Scholar CrossRef Search ADS PubMed Caro T . 2005 . The adaptive significance of coloration in mammals . BioScience . 55 : 125 – 136 . Google Scholar CrossRef Search ADS Caro T , Beeman K , Stankowich T , Whitehead H . 2011 . The functional significance of colouration in cetaceans . Evol Ecol . 25 : 1231 – 1245 . Google Scholar CrossRef Search ADS Crowhurst CJ , Zanollo V , Griggio M , Robertson J , Kleindorfer S . 2012 . White flank spots signal feeding dominance in female diamond firetails, Stagonopleura guttata . Ethology . 118 : 63 – 75 . Google Scholar CrossRef Search ADS Dakin R , Montgomerie R . 2009 . Peacocks orient their courtship displays towards the sun . Behav Ecol Sociobiol . 63 : 825 – 834 . Google Scholar CrossRef Search ADS Dale J , Dey CJ , Delhey K , Kempenaers B , Valcu M . 2015 . The effects of life history and sexual selection on male and female plumage colouration . Nature . 527 : 367 – 370 . Google Scholar CrossRef Search ADS PubMed Darst CR , Cummings ME , Cannatella DC . 2006 . A mechanism for diversity in warning signals: conspicuousness versus toxicity in poison frogs . Proc Natl Acad Sci USA . 103 : 5852 – 5857 . Google Scholar CrossRef Search ADS PubMed Endler JA . 1978 . A predator’s view of animal color patterns . Evol Biol . 11 : 319 – 364 . Endler JA . 1983 . Natural and sexual selection on color patterns in poeciliid fishes . Environ Biol Fish . 9 : 173 – 190 . Google Scholar CrossRef Search ADS Endler JA . 2012 . A framework for analysing colour pattern geometry: adjacent colours . Biol J Linn Soc . 107 : 233 – 253 . Google Scholar CrossRef Search ADS Fusani L , Giordano M , Day LB , Schlinger BA . 2007 . High-speed video analysis reveals individual variability in the courtship displays of male Golden-collared manakins . Ethology . 113 : 964 – 972 . Google Scholar CrossRef Search ADS Garamszegi LZ . 2014 . Uncertainties due to within-species variation in comparative studies: measurement errors and statistical weights . In: Garamszegi LZ , editor. Modern phylogenetic comparative methods and their application in evolutionary biology . New York (NY) : Springer . p. 157 – 199 . Google Scholar CrossRef Search ADS Garamszegi LZ , Mundry R . 2014 . Multimodel-inference in comparative analyses . In: Garamszegi LZ , editor. Modern phylogenetic comparative methods and their application in evolutionary biology . New York (NY) : Springer . p. 157 – 199 . Google Scholar CrossRef Search ADS Geberzahn N , Gahr M . 2011 . Undirected (solitary) birdsong in female and male blue-capped cordon-bleus (Uraeginthus cyanocephalus) and its endocrine correlates . PLoS One . 6 : e26485 . Google Scholar CrossRef Search ADS PubMed Geberzahn N , Gahr M . 2013 . Song learning in male and female Uraeginthus cyanocephalus, a tropical songbird species . J Comp Psychol . 127 : 352 – 364 . Google Scholar CrossRef Search ADS PubMed Gluckman TL . 2014 . Pathways to elaboration of sexual dimorphism in bird plumage patterns . Biol J Linn Soc . 111 : 262 – 273 . Google Scholar CrossRef Search ADS Gluckman TL , Cardoso GC . 2010 . The dual function of barred plumage in birds: camouflage and communication . J Evol Biol . 23 : 2501 – 2506 . Google Scholar CrossRef Search ADS PubMed Gluckman TL , Mundy NI . 2013 . Cuckoos in raptors’ clothing: barred plumage illuminates a fundamental principle of Batesian mimicry . Anim Behav . 86 : 1165 – 1181 . Google Scholar CrossRef Search ADS Gomes ACR , Funghi C , Soma M , Sorenson MD , Cardoso GC . 2017 . Multimodal signalling in estrildid finches: song, dance and colour are associated with different ecological and life-history traits . J Evol Biol . 30:1336–1346 Gomes ACR , Sorenson MD , Cardoso GC . 2016 . Speciation is associated with changing ornamentation rather than stronger sexual selection . Evolution . 70 : 2823 – 2838 . Google Scholar CrossRef Search ADS PubMed Goodwin D . 1982 . Estrildid finches of the world . New York (NY) : Cornell University Press . Griffith S , Parker TH , Olson VA . 2006 . Melanin- versus carotenoid-based sexual signals: is the difference really so black and red ? Anim Behav . 71 : 749 – 763 . Google Scholar CrossRef Search ADS Hadfield J . 2010 . MCMC methods for multi-response generalized linear mixed models : the MCMCglmm R package . J Stat Softw . 33 : 1 – 22 . Google Scholar CrossRef Search ADS PubMed Hadfield JD , Nakagawa S . 2010 . General quantitative genetic methods for comparative biology: phylogenies, taxonomies and multi-trait models for continuous and categorical characters . J Evol Biol . 23 : 494 – 508 . Google Scholar CrossRef Search ADS PubMed Halperin T , Carmel L , Hawlena D . 2017 . Movement correlates of lizards’ dorsal pigmentation patterns . Funct Ecol . 31 : 370 – 376 . Google Scholar CrossRef Search ADS Hasson O . 1991 . Sexual displays as amplifiers: practical examples with an emphasis on feather decorations . Behav Ecol . 2 : 189 – 197 . Google Scholar CrossRef Search ADS Hill GE , McGraw KJ . 2006 . Bird coloration . Boston (MA) : Harvard University Press . Ho LST , Ane C , Lachlan R , Tarpinian K , Feldman R , Yu Q . 2016 . Package “phylolm.”https://cran.r-project.org/web/packages/ phylolm/index.html (accessed 28 March 2018). Jetz W , Thomas GH , Joy JB , Hartmann K , Mooers AO . 2012 . The global diversity of birds in space and time . Nature . 491 : 444 – 448 . Google Scholar CrossRef Search ADS PubMed Johnson KP . 2000 . The evolution of courtship display repertoire size in the dabbling ducks (Anatini) . J Evol Biol . 13 : 634 – 644 . Google Scholar CrossRef Search ADS Kabasakal B , Poláček M , Aslan A , Hoi H , Erdoğan A , Griggio M . 2017 . Sexual and non-sexual social preferences in male and female white-eyed bulbuls . Sci Rep . 7 : 5847 . Google Scholar CrossRef Search ADS PubMed Kelley JL , Fitzpatrick JL , Merilaita S . 2013 . Spots and stripes: ecology and colour pattern evolution in butterflyfishes . Proc Biol Sci . 280 : 20122730 . Google Scholar CrossRef Search ADS PubMed Krzyszczyk E , Mann J . 2012 . Why become speckled? Ontogeny and function of speckling in Shark Bay bottlenose dolphins (Tursiops sp.) . Mar Mammal Sci . 28 : 295 – 307 . Google Scholar CrossRef Search ADS Maddison WP , Maddison DR . 2011 . Mesquite: a modular system for evolutionary analysis . http://mesquiteproject.wikispaces.com (accessed 28 March 2018). Marques CI , Batalha HR , Cardoso GC . 2016 . Signalling with a cryptic trait: the regularity of barred plumage in common waxbills . R Soc Open Sci . 3 : 160195 . Google Scholar CrossRef Search ADS PubMed Marshall KL , Gluckman TL . 2015 . The evolution of pattern camouflage strategies in waterfowl and game birds . Ecol Evol . 5 : 1981 – 1991 . Google Scholar CrossRef Search ADS PubMed Mäthger LM , Hanlon RT . 2006 . Anatomical basis for camouflaged polarized light communication in squid . Biol Lett . 2 : 494 – 496 . Google Scholar CrossRef Search ADS PubMed Mäthger LM , Shashar N , Hanlon RT . 2012 . Do cephalopods communicate using polarized light reflections from their skin ? J Exp Biol . 212 : 2133 – 2140 . Google Scholar CrossRef Search ADS Møller AP , Pomiankowski A . 1993 . Why have birds got multiple sexual ornaments ? Behav Ecol Sociobiol . 32 : 167 – 176 . Odeen A , Hastad O . 2003 . Complex distribution of avian color vision systems revealed by sequencing the SWS1 opsin from total DNA . Mol Biol Evol . 20 : 855 – 861 . Google Scholar CrossRef Search ADS PubMed Olea PP , Casas F , Redpath S , Viñuela J . 2010 . Bottoms up: great bustards use the sun to maximise signal efficacy . Behav Ecol Sociobiol . 64 : 927 – 937 . Google Scholar CrossRef Search ADS Ortolani A . 1999 . Spots, stripes, tail tips and dark eyes: predicting the function of carnivore colour patterns using the comparative method . Biol J Linn Soc . 67 : 433 – 476 . Google Scholar CrossRef Search ADS Ota N , Gahr M , Soma M . 2015 . Tap dancing birds: the multimodal mutual courtship display of males and females in a socially monogamous songbird . Sci Rep . 5 : 16614 . Google Scholar CrossRef Search ADS PubMed Ota N , Gahr M , Soma M . 2017 . Songbird tap dancing produces non-vocal sounds . Bioacoustics . 26 : 161 – 168 . Google Scholar CrossRef Search ADS Pérez-Rodríguez L , Jovani R , Stevens M . 2017 . Shape matters: animal colour patterns as signals of individual quality . Proc R Soc B Biol Sci . 284 : 20162446 . Google Scholar CrossRef Search ADS Puebla O , Bermingham E , Guichard F , Whiteman E . 2007 . Colour pattern as a single trait driving speciation in Hypoplectrus coral reef fishes ? Proc Biol Sci . 274 : 1265 – 1271 . Google Scholar CrossRef Search ADS PubMed R Core Team . 2016 . R: a language and environment for statistical computing . Vienna (Austria) : R Foundation for Statistical Computing . https://www.R-project.org/ Restall R . 1996 . Munias and mannikins . East Sussex (UK) : Pica Press . Santana SE , Alfaro JL , Alfaro ME . 2012 . Adaptive evolution of facial colour patterns in Neotropical primates . Proc R Soc B Biol Sci . 279 : 2204 – 2211 . Google Scholar CrossRef Search ADS Schneider CA , Rasband WS , Eliceiri KW . 2012 . NIH Image to ImageJ: 25 years of image analysis . Nat Methods . 9 : 671 – 675 . Google Scholar CrossRef Search ADS PubMed Schuetz JG . 2004 . Common waxbills use carnivore scat to reduce the risk of nest predation . Behav Ecol . 16 : 133 – 137 . Google Scholar CrossRef Search ADS Seehausen O , Mayhew PJ , van Alphen JJM . 1999 . Evolution of colour patterns in East African cichlid fish . J Evol Biol . 12 : 514 – 534 . Google Scholar CrossRef Search ADS Semel B , Sherman PW . 2001 . Intraspecific parasitism and nest-site competition in wood ducks . Anim Behav . 61 : 787 – 803 . Google Scholar CrossRef Search ADS Soma M , Garamszegi LZ . 2015 . Evolution of courtship display in Estrildid finches: dance in relation to female song and plumage ornamentation . Front Ecol Evol . 3 : 4 . Google Scholar CrossRef Search ADS Soma M , Garamszegi LZ . 2018 . Data from: evolution of patterned plumage as a sexual signal in estrildid finches . Dryad Digital Repository. http://dx.doi. org/10.5061/dryad.35j39r8 Soma M , Iwama M . 2017 . Mating success follows duet dancing in the Java sparrow . PLoS One . 12 : e0172655 . Google Scholar CrossRef Search ADS PubMed Sorenson MD , Balakrishnan CN , Payne RB . 2004 . Clade-limited colonization in brood parasitic finches (Vidua spp.) . Syst Biol . 53 : 140 – 153 . Google Scholar CrossRef Search ADS PubMed Stevens M , Merilaita S . 2009 . Animal camouflage: current issues and new perspectives . Phil Trans R Soc B . 364 : 423 – 427 . Google Scholar CrossRef Search ADS PubMed Sun K , Meiklejohn KA , Faircloth BC , Glenn TC , Braun EL , Kimball RT . 2014 . The evolution of peafowl and other taxa with ocelli (eyespots): a phylogenomic approach . Proc R Soc B Biol Sci . 281 : 20140823 . Google Scholar CrossRef Search ADS Swaddle JP , Cuthill IC . 1994 . Female zebra finches prefer males with symmetric chest plumage . Proc R Soc B Biol Sci . 258 : 267 – 271 . Google Scholar CrossRef Search ADS Tibbetts EA , Dale J . 2007 . Individual recognition: it is good to be different . Trends Ecol Evol . 22 : 529 – 537 . Google Scholar CrossRef Search ADS PubMed Tobias JA , Montgomerie R , Lyon BE . 2012 . The evolution of female ornaments and weaponry: social selection, sexual selection and ecological competition . Phil Trans R Soc B . 367 : 2274 – 2293 . Google Scholar CrossRef Search ADS PubMed Ullrich R , Norton P , Scharff C . 2016 . Waltzing Taeniopygia: integration of courtship song and dance in the domesticated Australian zebra finch . Anim Behav . 112 : 285 – 300 . Google Scholar CrossRef Search ADS West-Eberhard MJ . 1979 . Sexual selection, social competition, and evolution . Proc Am Philos Soc . 123 : 222 – 234 . Yom-Tov Y . 2001 . An updated list and some comments on the occurrence of intraspecific nest parasitism in birds . Ibis . 143 : 133 – 143 . Google Scholar CrossRef Search ADS Zahavi A . 1978 . Decorative patterns and the evolution of art . New Scientist . 19 : 182 – 184 . Zann RA . 1996 . The zebra finch: a synthesis of field and laboratory studies . Oxford: Oxford University Press . Zanollo V , Griggio M , Robertson J , Kleindorfer S . 2012 . The number and coloration of white flank spots predict the strength of a cutaneous immune response in female Diamond Firetails, Stagonopleura guttata . J Ornithol . 153 : 1233 – 1244 . Google Scholar CrossRef Search ADS Zanollo V , Griggio M , Robertson J , Kleindorfer S . 2013 . Males with a faster courtship display have more white spots and higher pairing success in the diamond firetail, Stagonopleura guttata . Ethology . 119 : 344 – 352 . Google Scholar CrossRef Search ADS Zanollo V , Griggio M , Robertson J , Kleindorfer S . 2014 . Assortative pairings in diamond firetails (Stagonopleura guttata) are not the result of mutual mate choice for an ornament . Ethology . 120 : 951 – 964 . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Behavioral EcologyOxford University Press

Published: Apr 4, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off