www.advances.sciencemag.org/cgi/content/full/1/2/e1400155/dc1 Supplementary Materials for Natural and sexual selection act on different axes of variation in avian plumage color The PDF file includes: Peter O. Dunn, Jessica K. Armenta, Linda A. Whittingham Published 27 March 2015, Sci. Adv. 1, e1400155 (2015) DOI: 10.1126/sciadv.1400155 Materials and Methods Fig. S1. Sexual dichromatism in brightness (PC1) and hue (PC2) in relation to mating system and breeding latitude. Fig. S2. Examples of reflectance spectra for males of species with high (splendid fairy-wren, Malurus splendens) and low (American goldfinch, Carduelis tristis) hue (PC2 scores were 35.7 and 9.9, respectively). Fig. S3. An example of duller plumage in males of polygynous species. Table S1. PCA of reflectance data for 977 species of birds. Table S2. Monochromatism in relation to morphological, ecological, and behavioral variables. Table S3. Sexual dichromatism in plumage brightness (PC1 dichro) and hue (PC2 dichro), and variation in brightness (PC1) and hue (PC2) for each sex in relation to life history and ecological variables associated with natural and sexual selection (N = 977 species). Table S4. Sexual dichromatism in plumage brightness (PC1) in relation to life history and ecological variables in PGLS models. Table S5. Sexual dichromatism in plumage brightness (PC1) in relation to life history and ecological variables in the full PGLS model (see table S4). Table S6. Sexual dichromatism in plumage hue (PC2) in relation to life history and ecological variables in PGLS models. Table S7. Sexual dichromatism in plumage hue (PC2) in relation to life history and ecological variables in the full PGLS model (see table S6). Table S8. Female plumage brightness (PC1) in relation to life history and ecological variables in PGLS models. Table S9. Female plumage brightness (PC1) in relation to life history and ecological variables in the full PGLS model (see table S8). Table S10. Male plumage brightness (PC1) in relation to life history and ecological variables in PGLS models.
Table S11. Male plumage brightness (PC1) in relation to life history and ecological variables in the full PGLS model (see table S10). Table S12. Female plumage hue (PC2) in relation to life history and ecological variables in PGLS models. Table S13. Female plumage hue (PC2) in relation to life history and ecological variables in the full PGLS model (see table S12). Table S14. Male plumage hue (PC2) in relation to life history and ecological variables in PGLS models. Table S15. Male plumage hue (PC2) in relation to life history and ecological variables in the full PGLS model (see table S14). References (40 50) Other Supplementary Material for this manuscript includes the following: (available at www.advances.sciencemag.org/cgi/content/full/1/2/e1400155/dc1) Data file S1. Plumage color and ecological data for 977 species (Data977.csv). Data file S2. Nexus file of 100 phylogenetic trees for the 977 species (trees100spp977.tre). Data file S3. Plumage brightness (PC1) in three categories and 100 phylogenetic trees (PC1_LoMedHi_100trees.xml; Simmap file). Data file S4. Plumage hue (PC2) in three categories and 100 phylogenetic trees (PC2_LoMedHi_100trees.xml; Simmap file).
Materials and Methods Data Collection We measured the spectral reflectance of plumage colors from museum specimens of 977 species of birds, representing 87% (20/23) of the orders in Sibley & Monroe (40) and 79% (26/33) of the orders in Clements et al. (41). Species were originally chosen based on two criteria: 1) maximize taxonomic diversity (based on the 23 orders in Sibley and Monroe) and 2) availability of testes mass data (9). We sampled three male and three female specimens for each species. All specimens sampled were adults collected during the breeding season (within the same month or adjacent months) and were not visibly molting. We avoided visibly faded specimens and 90% of specimens were <50 years old, which is approximately when reflectance starts to decrease (42). We sampled specimens from the nominate subspecies, if there were sufficient individuals, otherwise we sampled the subspecies with the greatest number of specimens in the collection. When possible, we chose all six specimens for each species from the same collecting locality, otherwise we chose specimens that were as geographically close as possible. For example, specimens of North American species may have come from different localities within the same state or neighboring states. Color Measurements and Calculations For each specimen, we measured reflectance across the bird-visible spectrum (320 to 700 nm) at six body regions: crown, back, tail, throat, belly and wing coverts. Although species may have more color patches than at these six body locations, they provide homologous sample sites for a large number of species, and several of them are widely used for display purposes (e.g., throat, wing coverts). Each body region was sampled five times, moving the spectrophotometer probe after each reading. All spectrophotometer measurements were collected using an Ocean Optics (Dunedin, Florida) USB2000 spectrophotometer and a PX-2 pulsed Xenon light source with the spectrophotometer probe illuminating and measuring at 90 to the plumage. Measurements were standardized to a white standard (WS-1, Ocean Optics), considered >98% reflective from 250 to 1500 nm wavelengths.
For each reflectance spectrum, we averaged the reflectance data into bins 20 nm wide across the bird-visible spectrum. We then performed a principal component analysis (PCA) on the 'binned' (mean reflectance) values from each reflectance spectrum (n= 175,860 spectra used in PCA; 5 measurements x 6 body regions x 3 individuals x 2 sexes x 977 species). Multiple methods have been used to analyze spectral data (31, 43, 44), but when different methods have been compared, they appear to yield essentially similar estimates of color (45) or dichromatism (27). We chose PCA analysis for its simplicity and because it yields separate values that represent the brightness of the spectrum and the shape of the spectrum, both of which contribute to differences in color (27). In our analyses, the first two principal components explained more than 97% of the variation in the data (91 and 6% respectively). Similar to previous studies (29, 31, 32), we found that principal component 1 (PC1) loaded evenly across all wavelengths, and principal component 2 (PC2) loaded positively in the shorter (UV/blue/green) wavelengths and negatively in the longer (orange/red) wavelengths (Table S1). Therefore, we interpreted PC1 to represent overall brightness and PC2 to represent hue (see Fig. S2 for examples). Some researchers remove 'brightness" by subtracting the overall average reflectance (based on all specimens) from the reflectance values of each individual before performing PCA. This procedure produces scores for differences in dichromatism between the sexes on the first PCA axis that are highly correlated with the scores from our PC2 axis, which we interpreted as differences in hue (r 2 = 0.916, P<0.0001). Analysis of this PCA axis with brightness removed produced similar results to our PC2. For our initial analyses, we summed these PC scores across all six body regions for each sex and axis (PC1 or 2) to gain an overall index of brightness or hue. For analyses of sexual dichromatism we calculated differences between the sexes in PC1 and PC2. Starting with each axis separately (PC1 or 2), we calculated the average PC score for each body region for males and females of each species. We subtracted the female average from the
male average to obtain a difference for each body region. We then summed these differences from all six body regions in order to produce a dichromatism score for each PC axis and species. We used the sum of the differences of all body regions, rather than the sum of the absolute differences, so that the total score reflected the directionality of the dichromatism across all body regions. Using this system, a positive dichromatism score for PC1 indicates that the species has a relatively brighter male than female, while a negative PC1 score indicates a relatively brighter female than male. Similarly, a positive PC2 score for dichromatism indicates that males of a particular species have relatively longer wavelength hue than females. This longer wavelength hue may be orange or red, or it may be brown, because browns reflected more in longer wavelengths than in shorter wavelengths. A negative PC2 score indicated that females of a particular species have a relatively shorter-wavelength hue than males. A PC1 or PC2 dichromatism score close to zero indicated little or no difference between the sexes. Ecological and life history variables. We analyzed 10 ecological and life history variables thought to be associated with plumage color or dimorphism in birds. Most of these data were compiled in our previous study that analyzed dichromatism using human visual estimates (9). Data came primarily from reference books for North America (46), Europe (47, 48), and Australasia (49) and were supplemented by data in Handbook of the Birds of the World (50). Previous comparative studies have used social mating system as an index of variation in mating success and sexual selection. For example, polygynous and lek mating systems often have greater variation in male mating success than monogamous mating systems. However, dichromatism also occurs in some monogamous groups (6) with low variance in male mating success, so it is important to examine how mating systems interact with ecological variables. Thus, we examined a variety of ecological variables that may influence natural selection on
plumage; e.g., directly through their effects on predation, or indirectly through tradeoffs with reproductive investment (e.g., parental care may reduce the opportunity for gaining mates). These variables included body mass (log-transformed; TotMassLog), breeding latitude (Lat), habitat cover (Habitat), migratory behavior, development (Develop), nest height, cavity nesting and male participation in parental care. Variables and the rationale for their inclusion are as follows: Total body mass (log-transformed). Included to control for overall body size in the analysis of testes mass (see below). Breeding latitude: 0=tropical, 1=tropical to temperate, 2= temperate to polar. Based on whether the center of the breeding range was in the tropics (<24 N or S of equator), subtropics (24-38 N or S of equator) or temperate and polar regions (>38 N or S of equator). Greater seasonality and food may occur at breeding locations farther from the equator, and this may influence opportunities for mating or the need for male parental care. Habitat Cover: 0 = open (grass, marsh, tundra, heath), 1 = semi-open (edge, mixed types, open forest, woodland), 2 = closed forest. Habitat cover may influence the risk of predation. Mating system: coop (cooperative or group-living), lek (lek and promiscuous), mon (monogamous, including species with occasional polygyny; <15%), polyg (polygynous but not lekking), PA (polyandrous). The monogamous and occasionally polygynous categories used by Dunn et al. (9) were combined here because of low sample sizes in some analyses with interactions. Testes mass: We used testis mass (relative to body mass) as a proxy measure of extrapair mating (15). Species with higher levels of extrapair mating experience greater sperm competition, which is correlated with relatively larger testes that produce more sperm. Estimates of testes mass were obtained from Dunn et al. (9) for 964 of 977 species. Results were qualitatively similar when we
excluded the 13 species without testes data. So, for simplicity, we include all species and impute a mean testes mass for the 13 species with missing data, using the relationship: log testes (g) = - 1.5351 + 0.60257 * log10(total body mass [g]). Migratory: yes or no. Species were considered migratory if they had largely non-overlapping winter and summer ranges (<50% overlap), and resident if there was little seasonal change in distribution (>50% overlap in ranges). Previous studies have suggested a correlation between dichromatism and migratory behavior (6, 22, 23). Development: 1= precocial, 2 = semiprecocial, 3 = semialtricial and altricial. Type of development is often related to male parental care, which can influence the risk of predation or opportunities for mating. Categories were based on Ehrlich et al. (46). Cavity nesting (Cavity): yes or no. Nests in cavities may be less visible to predators. Male cares for young (MaleCareYg): yes (includes tending in precocial species) or no. Male participation in parental care could increase the risk of predation on the male or young. Nest height (NestHt): ground, shrub level, or tree height. This was included because risk of nest predation varies with height (16). Models were examined with the 10 variables above, as well as interactions between the following variables to test for differential effects of ecology on plumage: i) male care for young and nest height - nest predation, which varies with nest height (16), may have a greater influence on plumage in species in which males care for young and are at greater risk of predation because they are near the nest, ii) male care for young and cavity nest - same rationale as above; male plumage may vary based on predation risk, which may be influenced by nest concealment or protection and the presence of the male near young during parental care, and iii) mating system and breeding latitude - the intensity of sexual selection may vary with resource abundance or competition, which may vary with latitude.
Example R code for running PGLS models #color variables are named as follows: #sumdiffpc1- difference between the sexes in PC1 scores across 6 body regions #sumdiffpc2- difference between the sexes in PC2 scores across 6 body regions #FemalePC1sum- sum of PC1 scores for females across 6 body regions # FemalePC2sum - sum of PC2 scores for females across 6 body regions #MalePC1sum- sum of PC1 scores for males across 6 body regions # MalePC2sum - sum of PC2 scores for males across 6 body regions #load libraries library(ape) library(nlme) library(geiger) library(caper) library(mumin) #read nexus file with 100 phylogenetic trees (trees100spp977.tre) trees <-read.nexus("trees100spp977.tre") #read data file with 977 species data <-read.csv("data977.csv", header=t) # assign the values in the data frame to new variables, as follows. See above for #explanation of variables and their names: sumdiffpc1 = data$sumdiffpc1 sumdiffpc2 = data$sumdiffpc2 FemalePC1sum = data$femalepc1sum MalePC1sum = data$malepc1sum FemalePC2sum = data$femalepc2sum MalePC2sum = data$malepc2sum TotMassLog = data$totmasslog Log10Testes = data$log10testes BreedLat3 = data$breedlat3 Cover = data$cover MateSys5 = data$matesys5 Migratory = data$migratory Develop3 = data$develop3 MaleCareYg = data$malecareyg NestHt = data$nestht
Cavity01 = data$cavity01 # assign scientific names (SciName_BL) to the data, which is necessary for #linking the data to the tree. BL= BirdLife taxonomy names(sumdiffpc1) = data$sciname_bl names(sumdiffpc2) = data$sciname_bl names(femalepc1sum) = data$ SciName_BL names(malepc1sum) = data$ SciName_BL names(femalepc2sum) = data$ SciName_BL names(malepc2sum) = data$ SciName_BL names(totmasslog) = data$sciname_bl names(log10testes) = data$sciname_bl names(breedlat3) = data$sciname_bl names(cover) = data$sciname_bl names(matesys5) = data$sciname_bl names(migratory) = data$sciname_bl names(develop3) = data$sciname_bl names(malecareyg) = data$sciname_bl names(nestht) = data$sciname_bl names(cavity01) = data$sciname_bl # create a data frame for the data to provide to the gls function DF = data.frame(sumdiffpc1, sumdiffpc2, FemalePC1sum, MalePC1sum, FemalePC2sum, MalePC2sum,TotMassLog, Log10Testes, BreedLat3, Cover, MateSys5, Migratory, Develop3, MaleCareYg, NestHt, Cavity01) # create an empty list as a container for the 100 models (1 for each phylogeny) res = vector(mode = "list", length = 100) # Run each model in a for() loop and save it to 'res' # example below is for PC1 of males. Modify as needed for other color variables #and predictors. cormartins implements the OU model. for (i in 1:100) { tree = trees[[i]] res[[i]] = gls(malepc1sum ~ TotMassLog + Log10Testes + BreedLat3 + Cover + MateSys5 + Migratory + Develop3 + as.factor(malecareyg) + NestHt + as.factor(cavity01) + BreedLat3*MateSys5 + as.factor(malecareyg)* as.factor(cavity01) + as.factor(malecareyg)*nestht, data= DF, correlation= cormartins(1.0, phy=tree, fixed=true)) } fm = model.avg(res) summary(fm) # P values were averaged over all 100 models to get overall ANOVA table results
#for each variable. for (i in 1:100) { print(anova(res[[i]], type="marginal")) } #figures for categorical variables (eg, mating systems or nestht) were made by #running separate models for each category (eg, monogamy) with a dummy #variable (coded 0, 1) instead of the original variable with all of the categories #(eg, MateSys5).
Fig. S1. Sexual dichromatism in brightness (PC1) and hue (PC2) in relation to mating system and breeding latitude. Means and P values are based on full PGLS models (Tables S3-15). P values are for the interaction between mating system and latitude.
Fig. S2. Examples of reflectance spectra for males of species with high (splendid fairy-wren, Malurus splendens) and low (American goldfinch, Carduelis tristis) hue (PC2 scores were 35.7 and 9.9, respectively). Brightness (PC1) scores were 35.2 and 4.9, respectively.
Fig. S3. An example of duller plumage in males of polygynous species. Shown is plumage reflectance (%) from three body regions of male (dashed lines) and female (solid lines) bobolinks (Dolichonyx oryzivorus). Note the low reflectance from the black throat of males (bottom dashed line). Females have a tan throat that has greater reflectance. Males in polygynous species were generally duller (lower reflectance) because of differences on the throat and belly. Among the 44 polygynous species, males were duller (PC1) than females across all six body regions in 25 species (56%), and in 21 (75%) of those species males had darker (black, red or green) feathers with less reflectance on the throat or belly than females (white or tan).
Table S1. PCA of reflectance data for 977 species of birds. PCA was conducted on the correlation matrix. Reflectance bins are the 20 nm bins used for averaging reflectance. Negative PC scores are highlighted in bold. Number Eigenvalue Percent Cumulative percent PCA 1 17.4 91.4 91.4 PCA 2 1.2 6.1 97.5 PCA 3 0.3 1.6 99.0 PC scores Reflectance bin (nm) PC 1 PC 2 PC 3 320 0.212 0.323 0.555 340 0.223 0.275 0.403 360 0.228 0.255 0.230 380 0.230 0.250 0.036 400 0.230 0.240-0.130 420 0.230 0.220-0.244 440 0.231 0.192-0.319 460 0.232 0.152-0.346 480 0.233 0.115-0.353 500 0.237 0.011-0.157 520 0.236-0.069-0.028 540 0.235-0.118 0.007 560 0.234-0.168 0.023 580 0.233-0.215 0.042 600 0.231-0.252 0.060 620 0.228-0.277 0.069 640 0.226-0.294 0.073 660 0.225-0.306 0.078 680 0.223-0.316 0.080
Table S2. Monochromatism in relation to morphological, ecological, and behavioral variables. P values are presented for PGLS models of color variation (PC1 or 2) among 489 monochromatic species (i.e., species with dichromatism scores in the middle 50% [interquartile] of the distribution). Significant relationships (P<0.05) are in bold. The number of categories is indicated in parentheses after each variable. Full PC1 model Reduced PC1 model Full PC2 model Reduced PC2 model Predictor variable Female Male Female Male Female Male Female Male Body Mass (Log) 0.830 0.856 0.000 0.000 0.000 0.000 Testes mass (Log) 0.307 0.360 0.219 0.256 Breeding latitude (3) 0.004 0.006 0.003 0.004 0.483 0.555 Habitat Cover (3) 0.183 0.187 0.217 0.215 0.098 0.109 0.153 0.171 Mating system (5) 0.836 0.823 0.219 0.266 Migratory (2) 0.000 0.000 0.000 0.000 0.018 0.039 0.008 0.012 Development (3) 0.064 0.080 0.003 0.005 0.000 0.000 0.000 0.000 Male Care Young (2) 0.092 0.103 0.021 0.022 0.036 0.055 0.000 0.000 Nest height (3) 0.973 0.951 0.000 0.000 0.000 0.000 Cavity nest (2) 0.429 0.518 0.031 0.035 0.827 0.791 Male Care * Cavity nest 0.613 0.709 0.940 0.873
Table S3. Sexual dichromatism in plumage brightness (PC1 dichro) and hue (PC2 dichro), and variation in brightness (PC1) and hue (PC2) for each sex in relation to life history and ecological variables associated with natural and sexual selection (N = 977 species). Shown are P values for each variable in the full PGLS model for each color variable (see Tables S4-15). Breeding latitude may result in either sexual or natural selection on plumage, but we included it under sexual selection here because of its interaction with mating system. All results are based on Ornstein-Uhlenbeck models of evolutionary change averaged over 100 trees. Numbers in parentheses refer to the number of categories for a variable. Significant relationships (P<0.05) are in bold. PC1 PC2 Predictor variable Dichro Female Male Dichro Female Male Sexual selection Mating system (5) 0.000 0.574 0.021 0.006 0.459 0.011 Testes mass (Log) 0.234 0.250 0.117 0.009 0.217 0.228 Breeding latitude (3) 0.038 0.676 0.305 0.004 0.823 0.033 Breeding lat * Mating system 0.000 0.810 0.322 0.017 0.257 0.027 Natural selection Body Mass (Log) 0.602 0.001 0.001 0.208 0.000 0.053 Habitat Cover (3) 0.415 0.005 0.009 0.031 0.198 0.211 Migratory (2) 0.645 0.000 0.000 0.489 0.123 0.606 Development (3) 0.071 0.006 0.001 0.564 0.000 0.000 Male Care Young (2) 0.014 0.404 0.344 0.332 0.372 0.147 Nest height (3) 0.039 0.660 0.737 0.039 0.237 0.103 Cavity nest (2) 0.102 0.848 0.549 0.532 0.777 0.681 Male Care * Cavity nest 0.181 0.806 0.394 0.553 0.868 0.644 Male Care * Nest height 0.056 0.529 0.640 0.017 0.038 0.733
Table S4. Sexual dichromatism in plumage brightness (PC1) in relation to life history and ecological variables in PGLS models. Results are based on Ornstein-Uhlenbeck models of evolutionary change. Shown are P values for each variable in the full and reduced (1-3) models. Predictor variables Full model 1 2 3 Body Mass (Log) 0.602 0.602 Testes mass (Log) 0.234 0.217 Breeding latitude (3) 0.038 0.030 0.023 0.021 Habitat Cover (3) 0.415 Mating system (5) 0.000 <0.0001 <0.0001 <0.0001 Migratory (2) 0.645 Development (3) 0.071 0.068 0.055 0.106 Male Care Young (2) 0.014 0.012 0.015 0.028 Nest height (3) 0.039 0.079 0.091 Cavity nest (2) 0.102 Breeding lat * Mating system 0.000 0.001 0.000 0.002 Male Care * Cavity nest 0.181 Male Care * Nest height 0.056 0.085 0.096 AIC 6158.9 6161.0 6158.8 6164.7 Δ AIC 0.18 2.25 0.00 5.96
Table S5. Sexual dichromatism in plumage brightness (PC1) in relation to life history and ecological variables in the full PGLS model (see table S4). Coefficients Estimat SE Z P (Intercept) 6.10 1.792 3.40 0.001 Body mass (Log) 0.17 0.438 0.38 0.704 Testes mass (Log) -0.43 0.419 1.01 0.311 BreedLat -Tropical -3.39 1.536 2.21 0.027 BreedLat- Subtropical -3.32 1.202 2.76 0.006 Habitat - Open -0.45 0.499 0.90 0.369 Habitat - Closed -0.65 0.490 1.32 0.187 MateSys- Lekking -5.81 2.055 2.82 0.005 MateSys - Monogamous -3.80 1.179 3.22 0.001 MateSys- Polyandrous -1.72 2.303 0.75 0.456 MateSys- Polygynous -7.40 1.709 4.33 0.000 Migratory - yes 0.09 0.530 0.17 0.862 Development - Precocial -1.14 0.825 1.38 0.167 Development - Semiprecocial 0.97 1.208 0.80 0.422 MaleCareYg- yes -2.19 1.087 2.01 0.044 Cavity -yes 4.49 2.601 1.72 0.085 NestHt- shrub -4.02 1.903 2.11 0.035 NestHt - tree -2.50 1.644 1.52 0.129 BreedLat*MateSys: Tropical x Lek 9.82 2.664 3.68 0.000 SubTropical x Lek -2.24 3.507 0.64 0.523 Tropical x Monogamous 3.03 1.652 1.83 0.067 SubTropical x Monogamous 3.31 1.379 2.40 0.016 Tropical x Polyandrous 0.87 3.789 0.23 0.818 SubTropical x Polyandrous 2.69 3.667 0.73 0.464 Tropical x Polygynous 3.57 2.634 1.36 0.175 SubTropical x Polygynous 1.62 2.673 0.61 0.545 MaleCare (yes) x Cavity nest -3.78 2.668 1.42 0.157 MaleCare (yes) x Shrub level nest 3.91 1.915 2.04 0.042 MaleCare (yes) x Tree level nest 1.82 1.653 1.10 0.271
Table S6. Sexual dichromatism in plumage hue (PC2) in relation to life history and ecological variables in PGLS models. Shown are P values for each variable in the full and reduced (1-3) models. Predictors Full model 1 2 3 Body Mass (Log) 0.208 0.330 0.290 0.323 Testes mass (Log) 0.008 0.010 0.010 0.009 Breeding latitude (3) 0.004 0.004 0.004 0.004 Habitat Cover (3) 0.031 0.035 0.039 0.032 Mating system (5) 0.006 0.007 0.007 0.007 Migratory (2) 0.489 0.477 Development (3) 0.564 Male Care Young (2) 0.332 0.413 0.439 0.521 Cavity nest (2) 0.532 Nest height (3) 0.039 0.060 0.061 0.218 Breeding lat * Mating system 0.017 0.018 0.020 0.010 Male Care * Cavity nest 0.553 Male Care * Nest height 0.017 0.020 0.019 AIC 4937.0 4932.8 4930.5 4937.6 Δ AIC 6.4 2.3 0.0 7.1
Table S7. Sexual dichromatism in plumage hue (PC2) in relation to life history and ecological variables in the full PGLS model (see table S6). Coefficients Estimate SE Z P (Intercept) 4.050 0.982 4.12 0.000 Body mass (Log) -0.285 0.231 1.23 0.219 Testes mass (Log) 0.582 0.222 2.62 0.009 BreedLat -Tropical -3.002 0.855 3.51 0.000 BreedLat- Subtropical -2.455 0.788 3.11 0.002 Habitat - Open -0.646 0.264 2.45 0.014 Habitat - Closed -0.472 0.260 1.81 0.070 MateSys- Lekking -3.155 1.151 2.74 0.006 MateSys - Monogamous -3.277 0.687 4.76 0.000 MateSys- Polyandrous -3.669 1.253 2.93 0.003 MateSys- Polygynous -2.006 0.952 2.11 0.035 Migratory - yes 0.210 0.278 0.76 0.450 Development - Precocial 0.422 0.439 0.96 0.338 Development - Semiprecocial 0.155 0.638 0.24 0.808 MaleCareYg- yes 0.423 0.604 0.70 0.484 Cavity -yes -0.895 1.381 0.65 0.517 NestHt- shrub 0.273 1.013 0.27 0.788 NestHt - tree 1.933 0.879 2.20 0.028 BreedLat*MateSys: Tropical x Lek 4.021 1.444 2.78 0.005 SubTropical x Lek 5.348 1.921 2.78 0.005 Tropical x Monogamous 3.517 0.912 3.85 0.000 SubTropical x Monogamous 2.489 0.848 2.93 0.003 Tropical x Polyandrous 2.484 2.018 1.23 0.219 SubTropical x Polyandrous 2.311 1.989 1.16 0.246 Tropical x Polygynous 2.140 1.418 1.51 0.132 SubTropical x Polygynous 0.825 1.512 0.55 0.586 MaleCare (yes) x Cavity nest 0.869 1.416 0.61 0.540 MaleCare (yes) x Shrub level nest 0.218 1.021 0.21 0.831 MaleCare (yes) x Tree level nest -2.002 0.885 2.26 0.024
Table S8. Female plumage brightness (PC1) in relation to life history and ecological variables in PGLS models. Shown are P values for each variable in the full and reduced (1-4) models. Predictors Body Mass (Log) Testes mass (Log) Breeding latitude (3) Habitat Cover (3) Mating system (5) Migratory (2) Development (3) Full Model 1 2 3 4 0.001 0.002 0.000 0.002 0.001 0.250 0.332 0.676 0.001 0.655 0.001 0.000 0.005 0.004 0.009 0.007 0.010 0.574 0.298 0.345 0.000 0.000 0.000 0.000 0.000 0.006 0.009 0.009 0.004 0.003 Male Care Young (2) 0.404 0.417 0.210 Nest height (3) 0.660 0.770 0.739 Cavity nest (2) 0.848 Breeding lat * Mating system 0.810 0.821 Male Care * Cavity nest Male Care * Nest height 0.806 0.529 0.566 0.448 AIC 7389.6 7422.9 7405.8 7433.5 7450.48 Δ AIC 0.0 33.3 16.2 44.0 60.9
Table S9. Female plumage brightness (PC1) in relation to life history and ecological variables in the full PGLS model (see table S8). Coefficients Estimate SE Z P (Intercept) -8.78 3.645 2.41 0.016 Body mass (Log) 2.78 0.838 3.31 0.001 Testes mass (Log) -0.92 0.804 1.15 0.252 BreedLat -Tropical -1.74 3.088 0.56 0.574 BreedLat- Subtropical 0.50 2.828 0.18 0.859 Habitat - Open 1.71 0.958 1.79 0.074 Habitat - Closed -1.88 0.938 2.00 0.046 MateSys- Lekking -3.72 4.225 0.88 0.379 MateSys - Monogamous -1.29 2.485 0.52 0.605 MateSys- Polyandrous -3.98 4.535 0.88 0.380 MateSys- Polygynous -4.15 3.450 1.20 0.229 Migratory - yes 5.76 1.015 5.67 0.000 Development - Precocial 0.53 1.605 0.33 0.741 Development - Semiprecocial 6.67 2.352 2.83 0.005 MaleCareYg- yes 2.18 2.298 0.95 0.343 Cavity -yes -0.84 5.075 0.17 0.869 NestHt- shrub 3.61 3.718 0.97 0.333 NestHt - tree 1.66 3.342 0.50 0.620 BreedLat*MateSys: Tropical x Lek -1.43 5.260 0.27 0.786 SubTropical x Lek 2.31 6.980 0.33 0.741 Tropical x Monogamous 0.61 3.301 0.18 0.854 SubTropical x Monogamous 1.99 3.063 0.65 0.516 Tropical x Polyandrous 2.49 7.315 0.34 0.734 SubTropical x Polyandrous 4.19 7.211 0.58 0.562 Tropical x Polygynous -5.70 5.142 1.11 0.269 SubTropical x Polygynous 3.60 5.410 0.66 0.506 MaleCare (yes) x Cavity nest -1.36 5.198 0.26 0.793 MaleCare (yes) x Shrub level nest -3.71 3.753 0.99 0.324 MaleCare (yes) x Tree level nest -0.13 3.375 0.04 0.969
Table S10. Male plumage brightness (PC1) in relation to life history and ecological variables in PGLS models. Shown are P values for each variable in the full and reduced (1-3) models. Predictors Full model 1 2 3 Body Mass (Log) 0.001 0.001 0.001 0.001 Testes mass (Log) 0.117 0.107 0.134 0.115 Breeding latitude (3) 0.305 0.311 0.345 0.003 Habitat Cover (3) 0.009 0.012 0.010 0.015 Mating system (5) 0.021 0.012 0.015 0.000 Migratory (2) 0.000 0.000 0.000 0.000 Development (3) 0.001 0.001 0.002 0.002 Male Care Young (2) 0.344 0.355 0.365 Nest height (3) 0.737 Cavity nest (2) 0.549 0.712 Breeding lat * Mating system 0.322 0.327 0.340 Male Care * Cavity nest 0.394 0.573 Male Care * Nest height 0.640 AIC 7584.7 7590.5 7595.6 7635.6 Δ AIC 0.0 5.8 10.9 50.9
Table S11. Male plumage brightness (PC1) in relation to life history and ecological variables in the full PGLS model (see table S10). Coefficients Estimate SE Z P (Intercept) -3.59 3.901 0.92 0.359 Body mass (Log) 2.84 0.926 3.06 0.002 Testes mass (Log) -1.17 0.889 1.31 0.190 BreedLat -Tropical -4.99 3.396 1.47 0.142 BreedLat- Subtropical -2.95 3.076 0.96 0.339 Habitat - Open 1.20 1.058 1.13 0.257 Habitat - Closed -2.50 1.039 2.40 0.016 MateSys- Lekking -8.42 4.570 1.84 0.066 MateSys - Monogamous -5.04 2.710 1.86 0.063 MateSys- Polyandrous -5.83 4.990 1.17 0.243 MateSys- Polygynous -11.56 3.786 3.05 0.002 Migratory - yes 5.79 1.118 5.17 0.000 Development - Precocial -0.42 1.760 0.24 0.811 Development - Semiprecocial 7.42 2.597 2.85 0.004 MaleCareYg- yes 1.34 2.332 0.57 0.566 Cavity -yes 0.25 4.063 0.06 0.951 NestHt- shrub -0.09 3.640 0.03 0.980 NestHt - tree 3.68 5.602 0.66 0.512 BreedLat*MateSys: Tropical x Lek 7.65 5.783 1.32 0.186 SubTropical x Lek -0.40 7.672 0.05 0.959 Tropical x Monogamous 3.32 3.624 0.92 0.360 SubTropical x Monogamous 5.34 3.331 1.60 0.109 Tropical x Polyandrous 3.35 8.061 0.42 0.678 SubTropical x Polyandrous 6.82 7.929 0.86 0.390 Tropical x Polygynous -1.75 5.660 0.31 0.757 SubTropical x Polygynous 6.23 5.916 1.05 0.293 MaleCare (yes) x Cavity nest -5.12 5.739 0.89 0.373 MaleCare (yes) x Shrub level nest -0.64 4.100 0.16 0.877 MaleCare (yes) x Tree level nest 0.79 3.677 0.21 0.831
Table S12. Female plumage hue (PC2) in relation to life history and ecological variables in PGLS models. Shown are P values for each variable in the full and reduced (1-3) models. Predictors Full model 1 2 3 Body Mass (Log) 0.000 0.000 0.000 0.000 Testes mass (Log) 0.217 0.199 0.190 Breeding latitude (3) 0.823 0.829 Habitat Cover (3) 0.198 0.189 0.260 Mating system (5) 0.459 0.430 Migratory (2) 0.123 0.121 0.021 0.008 Development (3) 0.000 0.000 0.000 0.000 Male Care Young (2) 0.372 0.374 0.185 0.172 Nest height (3) 0.237 0.226 0.000 0.000 Cavity nest (2) 0.777 Breeding lat * Mating system 0.257 0.256 Male Care * Cavity nest 0.868 Male Care * Nest height 0.038 0.034 0.017 0.015 AIC 4680.9 4678.5 4689.3 4685.0 Δ AIC 2.5 0.0 10.9 6.5
Table S13. Female plumage hue (PC2) in relation to life history and ecological variables in the full PGLS model (see table S12). Coefficients Estimate SE Z P (Intercept) -3.03 0.880 3.45 0.001 Body mass (Log) 0.88 0.203 4.32 0.000 Testes mass (Log) -0.23 0.196 1.15 0.251 BreedLat -Tropical 0.05 0.742 0.07 0.943 BreedLat- Subtropical 0.34 0.682 0.50 0.619 Habitat - Open 0.04 0.231 0.18 0.857 Habitat - Closed 0.41 0.228 1.79 0.074 MateSys- Lekking -1.13 1.017 1.11 0.266 MateSys - Monogamous -0.71 0.599 1.18 0.238 MateSys- Polyandrous -0.07 1.089 0.06 0.951 MateSys- Polygynous -1.22 0.832 1.47 0.142 Migratory - yes -0.40 0.244 1.63 0.102 Development - Precocial 0.49 0.387 1.27 0.205 Development - Semiprecocial 3.50 0.560 6.24 0.000 MaleCareYg- yes 0.65 0.572 1.14 0.256 Cavity -yes 0.37 1.220 0.30 0.762 NestHt- shrub 1.47 0.895 1.65 0.100 NestHt - tree 0.33 0.808 0.41 0.684 BreedLat*MateSys: Tropical x Lek 0.07 1.264 0.05 0.959 SubTropical x Lek 0.37 1.678 0.22 0.824 Tropical x Monogamous -0.03 0.794 0.04 0.970 SubTropical x Monogamous -0.12 0.739 0.17 0.866 Tropical x Polyandrous -1.90 1.758 1.08 0.280 SubTropical x Polyandrous -4.60 1.735 2.65 0.008 Tropical x Polygynous -0.23 1.237 0.18 0.856 SubTropical x Polygynous -1.47 1.306 1.12 0.261 MaleCare (yes) x Cavity nest -0.23 1.249 0.18 0.857 MaleCare (yes) x Shrub level nest -0.59 0.905 0.65 0.513 MaleCare (yes) x Tree level nest 1.33 0.817 1.62 0.105
Table S14. Male plumage hue (PC2) in relation to life history and ecological variables in PGLS models. Shown are P values for each variable in the full and reduced (1-2) models. Predictors Full model 1 2 Body Mass (Log) 0.053 0.050 0.000 Testes mass (Log) 0.228 0.244 Breeding latitude (3) 0.033 0.031 0.037 Habitat Cover (3) 0.211 0.209 0.212 Mating system (5) 0.011 0.011 0.009 Migratory (2) 0.606 0.589 Development (3) 0.000 0.000 0.000 Male Care Young (2) 0.147 0.183 Nest height (3) 0.103 0.000 0.000 Cavity nest (2) 0.681 Breeding lat * Mating system 0.027 0.023 0.020 Male Care * Cavity nest 0.644 Male Care * Nest height 0.733 AIC 5481.0 5480.5 5478.0 Δ AIC 3.0 2.5 0.0
Table S15. Male plumage hue (PC2) in relation to life history and ecological variables in the full PGLS model (see table S14). Coefficients Estimate SE Z P (Intercept) 1.11 1.234 0.90 0.370 Body mass (Log) 0.60 0.307 1.95 0.051 Testes mass (Log) 0.35 0.292 1.21 0.226 BreedLat -Tropical -2.96 1.143 2.59 0.010 BreedLat- Subtropical -2.14 1.051 2.04 0.042 Habitat - Open -0.59 0.354 1.67 0.096 Habitat - Closed -0.03 0.355 0.08 0.939 MateSys- Lekking -4.40 1.492 2.95 0.003 MateSys - Monogamous -4.00 0.918 4.36 0.000 MateSys- Polyandrous -3.73 1.672 2.23 0.026 MateSys- Polygynous -3.31 1.267 2.61 0.009 Migratory - yes -0.19 0.366 0.53 0.598 Development - Precocial 0.88 0.577 1.53 0.127 Development - Semiprecocial 3.64 0.854 4.26 0.000 MaleCareYg- yes 0.97 0.607 1.60 0.109 Cavity -yes -0.78 1.859 0.42 0.674 NestHt- shrub 1.88 1.327 1.41 0.158 NestHt - tree 2.55 1.180 2.16 0.031 BreedLat*MateSys: Tropical x Lek 3.93 1.931 2.03 0.042 SubTropical x Lek 5.62 2.560 2.19 0.028 Tropical x Monogamous 3.50 1.221 2.86 0.004 SubTropical x Monogamous 2.39 1.126 2.12 0.034 Tropical x Polyandrous 0.56 2.692 0.21 0.835 SubTropical x Polyandrous -2.26 2.651 0.85 0.394 Tropical x Polygynous 1.80 1.894 0.95 0.343 SubTropical x Polygynous -0.69 1.978 0.35 0.726 MaleCare (yes) x Cavity nest 0.91 1.904 0.48 0.635 MaleCare (yes) x Shrub level nest -0.49 1.335 0.37 0.713 MaleCare (yes) x Tree level nest -0.98 1.191 0.82 0.413