1

Structural Plumage Colour as a Signal of Mate Quality in Tree Swallows

Roslyn Dakin

A thesis submitted to the Department of Biology in partial fulfillment of the requirements

for the degree of Bachelor of Science (Honours)

Queen’s University

Kingston, Ontario, Canada

April 2006 2

ABSTRACT

Ornamental plumage colour has long been implicated in sexual selection, and is now thought to function as an honest signal of individual quality in a number of bird species. The purpose of this thesis was to determine whether the bright blue-green structural plumage of tree swallows (Tachycineta bicolor) might function as a signal of mate quality in both males and females. Reflectance spectrometry was used to quantify the plumage colour of breeding adult tree swallows in an Ontario population. Results are discussed with respect to mate choice and sexual selection for honest signaling. First, adult male and female tree swallows differ in plumage colour, with males tending to display bluer, more saturated colour. This sexual dichromatism suggests that structural plumage colour may be sexually selected in tree swallows. Second, birds were found to mate assortatively with respect to plumage colour of the rump region. Tree swallows may therefore use the assessment of plumage colour in mate choice, whether this choice is mutual or unidirectional. Third, relationships were found between plumage colour and body size in males, between plumage colour and ectoparasite load in females, and between reproductive performance and plumage colour in both sexes. Structural plumage colour may therefore be an honest indicator of mate quality in both male and female tree swallows.

Tree swallows may benefit by choosing bluer, more intensely-coloured mates in terms of direct or indirect benefits for their offspring. 3

ACKNOWLEDGEMENTS

I thank Bob Montgomerie, my supervisor, for providing me with the amazing opportunity to work on this project and others, for providing guidance and support along the way, and for trusting me to drive the biology van throughout the spring. I am also grateful to Dr. Steve Lougheed and Briar Howes for teaching me a great deal about field biology and for supporting my failed attempts to catch skinks. Thanks also to Dr. Vicki

Friesen for your input, encouragement and of course, your time in agreeing to be my committee member.

I am very grateful to Dr. Mary Stapleton for taking me along during her work on the tree swallows at QUBS; thank you Mary for teaching me so much about these birds, and for sharing your data. Without your guidance this project would have been impossible! Thank you also to Jason Clarke for your field assistance; I appreciated both your expertise with the spectrometer and your great company.

I am grateful to the rest of the Montgomerie lab, including Christina Cliffe, Kim

MacDonald, Katrina Stockely, Ann McKellar, Kamini Persaud, and Nicole Mideo. Thanks for helping with field and lab work, answering questions, and generally keeping me entertained throughout my time in the lab. Thanks to my wonderful friends for their support and valuable suggestions for my seminar and poster. And finally, I am forever grateful to my parents who suffered through editing this thesis and most of the other things

I have had to write throughout my four years at Queen’s. Even when you fell asleep while reading my essays, I was glad to have your help. 4

TABLE OF CONTENTS

Abstract ……...…………………………………………………………………….. 2

Acknowledgements ….…………………………………………………………….. 3

List of Figures …………….……………………………...... 5

List of Tables ……….…….……………………………………………………….. 6

Introduction and Literature Review ...……………………………………………... 7

Methods …………………………………………………………………………… 18

Results …………………………………………………………………………….. 22

Discussion …………………………………………………………………………. 27

Literature Cited ……………………………………………………………………. 36

Summary …………………………………………………………………………... 41

Appendix …………………………………………………………………………... 53 5

LIST OF FIGURES

Figure 1. Spectral ranges of human and avian colour vision ……………………… 42

Figure 2. Representative reflectance spectra for adult male and female tree

swallows …………………………………………………………………... 43

Figure 3. Relationship between the number of feather mites of males and females

in mated pairs of tree swallows …………….……………………………... 44

Figure 4. Relationships between plumage colour variables of males and females in

mated pairs of tree swallows ….…………………………………… ……... 45

Figure 5. Relationships between plumage colour and morphological variables

in male tree swallows …...…………………………………………………. 46

Figure 6. Relationships between reproductive performance and plumage colour

in both males and females ……...…………………………………………. 47

6

LIST OF TABLES

Table 1. Plumage colour variables in adult male and female tree swallows ……… 48

Table 2. Tests of assortative mating by morphological variables in adult tree

swallow pairs ……………………………………………………………… 49

Table 3. Tests of assortative mating by plumage colour variables in adult tree

swallow pairs ……………………………………………………………… 50

Table 4. Relationships between plumage colour and morphological variables in

male and female tree swallows ……………………………………………. 51

Table 5. Relationships between reproductive performance and plumage colour

in male and female tree swallows …………………………………………. 52

7

INTRODUCTION AND LITERATURE REVIEW

Sexual Selection and Ornamental Plumage

The study of sexual selection has long been associated with questions regarding the evolution of ornamental plumage. Darwin pointed to numerous cases of sexual dimorphism in bird plumage, and noticed that often the male bears more conspicuously ornamented plumage than his mate (Darwin, 1871). He recognized that something other than natural selection must be responsible for maintaining costly traits such as long tails and brightly coloured ornaments. This led Darwin to suggest his theory of sexual selection: ornaments and behaviours that improve the ability to attract or compete for mates are favoured because they improve reproductive success despite costs to survival.

More recently, different theories have attempted to explain how ornamental traits could be produced and maintained by sexual selection. Fisher’s theory of runaway sexual selection proposes that an arbitrary ornamental trait in one sex may become exaggerated over time if it becomes associated with the genes for the preference of the trait in the other sex (Fisher, 1958). Under this model, the sole advantage of mating with an ornamented individual is that offspring will be more attractive and will therefore have greater reproductive success. Alternatively, Zahavi proposed that ornamental traits that reduce survival are favoured only if they are honest signals of mate quality (Zahavi, 1975). Under this model, only individuals in the best condition are able to express exaggerated ornaments; the preference for more ornamented mates is adaptive in terms of indirect benefits of genes inherited by the offspring, and possibly direct benefits of increased parental care as well. Hamilton and Zuk (1982) suggested that ornamental plumage may function as an honest signal of disease resistance, since there is evidence that the extent of 8

bright plumage is positively correlated with the degree of infection by blood parasites across species. Sexual selection for ornamental plumage may thus be mediated by the genetic benefits of choosing a mate that is resistant to parasites.

Although in many species the male is selected to express a higher degree of ornamentation than the female, plumage characteristics may signal female quality as well.

Mutual mate choice by plumage colour is predicted to occur in socially monogamous species where both sexes invest heavily in parental care, because both males and females stand to benefit from the careful selection of mates (Johnstone et al., 1996). For example, experimental work with the blue tit (Parus caeruleus) has shown that both females and males prefer partners that have plumage with greater UV-reflectance (Hunt et al., 1999). In species where such mutual mate choice is beneficial, patterns of assortative mating are predicted. Assortative mating by plumage colour has been observed in blue tits (Andersson et al., 1998) and American goldfinches, Carduelis tristis (MacDougall and Montgomerie,

2003). MacDougall and Montgomerie also suggest that assortative mating by plumage colour could have a role in maintaining the variation in ornamental traits despite strong directional selection.

The Measurement of Plumage Colouration

Many early studies of plumage ornaments assumed that human perception could effectively describe the quality and variation in plumage colour, either by comparison to colour standards or direct scoring by a human observer (such as Hamilton and Zuk, 1982).

However, these methods do not necessarily represent the colour stimulus as it is perceived by birds due to fundamental differences between avian and human visual systems. The 9

human visual system consists of three types of photoreceptors or cones that are maximally sensitive to red, green and blue light respectively. The avian system consists of four types of cones: red, green, blue, and an additional cone sensitive to ultraviolet (UV) wavelengths

(Bennett and Cuthill, 1994). As a result, birds are sensitive to a much larger range of wavelengths (320-700 nm) extending into the short-wave, UVA portion of the spectrum, whereas humans are sensitive to a more limited range (400-700 nm; see Figure 1).

Moreover, due to this additional cone type, birds are able to perceive an additional dimension of colour not available to human perception (Bennet et al., 1994).

The fundamental differences between human and avian visual systems mean that birds can perceive additional colours and patterns that are hidden from human observers

(Bennett et al., 1994). For this reason, early studies of plumage colouration based on human standards may be mistaken in their conclusions, especially with respect to evolutionary hypotheses about the signal function of plumage (Bennett et al., 1994). More recently, technological advances in reflectance spectrometry have allowed the objective measurement of signal quality and variation in plumage colour. This method involves the use of a full-spectrum incident light and a spectrometer to quantify reflectance over the entire bird-visible range, thereby allowing better tests of many hypotheses with respect to plumage signaling. For instance, UV-based signaling, undetectable to early studies of plumage colouration, is now known to be ubiquitous across avian families (Eaton and

Lanyon, 2003). Also, the use of reflectance spectrometry has demonstrated that many species previously though to be sexually monomorphic are in fact dimorphic in plumage colour (Eaton, 2005), such as the blue tit (Hunt et al., 1998) and starling, Sturnus vulgaris

(Cuthill et al., 1999). 10

Nevertheless, data from reflectance spectrometry alone do not indicate what is perceived by the animal. Results from such studies should also be supported by testing behaviour (such as mate preferences) either in the lab or in the field. For example, with blue tits, male colour scores from spectrometry were correlated with the results of mate choice trials in the lab to establish that females perceive and choose males with brighter crests (Hunt et al., 1998).

Mechanisms of Plumage Colour

Plumage colour may be produced by pigments or structural mechanisms. Pigments deposited in the feather include melanins, which usually produce brown, grey and black colours, and carotenoids, which produce red, orange and yellow colours. Structural colours, on the other hand, are produced by the coherent scattering of specific wavelengths of light by the nanostructure of keratin inside the feather, to produce violet, blue, green and iridescent plumage colours.

a) Carotenoid Pigmentation

Ornaments produced by carotenoid pigments have long been implicated in sexual selection (Hill and McGraw, 2006). Carotenoid colours are present in many secondary sexual characters in birds, including both plumage and fleshy ornaments, and carotenoid- based plumage is the target of female mate choice in a number of species (Omland, 1996;

Olson and Owens, 1998; MacDougall and Montgomerie, 2003). Moreover, carotenoid coloration is a good candidate for the honest signaling of mate quality for a number of reasons. First, red, orange and yellow colours may be highly conspicuous and therefore 11

costly in that they may be obvious to predators or aggressive conspecifics. Second, carotenoids cannot be synthesized by animals, but instead must be obtained in the diet. If carotenoids are limited in the environment, displays of bright carotenoid colours may be an honest indication of individual foraging and competitive ability. This is supported by evidence that carotenoid-based plumage colouration is correlated with nutritional condition at the time of molt in both the house finch, Carpodacus mexicanus (Hill and Montgomerie,

1994) and the American goldfinch (McGraw et al., 2002). In addition, infection with gut parasites during molt can suppress the expression of carotenoid colouration in the American goldfinch, most likely because of a reduced ability to absorb dietary carotenoids (McGraw and Hill, 2000).

A third reason for the implication of carotenoid-based ornaments in honest signaling is that carotenoids are involved in various immunological and physiological processes

(Olson and Owens, 1998). It has been suggested that the production of bright, carotenoid- based ornaments may have a direct immunological cost since bodily carotenoid reserves are a finite resource that can be directed towards either immune functioning or plumage display

(Lozano, 1994). Thus, extreme displays of carotenoid colouration may signal immunocompetence. Support for this hypothesis has been found in male mallards (Anas platyrhynchos L.), where the yellow bill colouration varies with immune responsiveness and sperm swimming velocity (Peters et al., 2004). Thus carotenoid colouration may be an important signal of genetic quality as well as individual condition and foraging ability.

12

b) Melanin Pigmentation

Melanin-based ornaments are also sexually dimorphic in many species, with males often displaying a cap on the head or badge on the breast. Unlike carotenoid pigments, melanins can be synthesized by animals from amino acid precursors, and therefore do not have the same obvious costs in production. For this reason, it was previously thought that melanin-based ornaments could not be honest signals of mate quality (Gray, 1996). Indeed, the size and colouration of melanin-based displays are not affected by nutritional condition in house sparrows (Passer domesticus) and brown-headed cowbirds (Molothrus ater;

McGraw et al., 2002), and melanin-based plumage colour is not affected by endoparasitism in either the American goldfinch (McGraw and Hill, 2000) or the house finch (Hill and

Brawner, 1998). Thus, melanin-based ornaments are probably not a direct signal of health.

Nevertheless, it has been suggested that melanin-based ornaments might be honest signals of social dominance. For instance, in the house sparrow, the expression of male melanin-based badges depends on hormonal status (Gonzalez et al., 2001), social rank and aggressive interactions with other males during molt (McGraw et al., 2003). Thus, melanin-based ornaments may signal individual quality not because these ornaments are costly to produce, but because they are costly to express. Only those individuals who can withstand frequent aggressive interactions can afford to express large melanin-based badges of social status. Moreover, melanin-based signals are not necessarily free from intersexual assessments. The colour of melanin-based plumage patches has been found to predict male reproductive success in the black-capped chickadee, Poecile atricapillus (Doucet et al.,

2004), and the size of the black facial mask in males is correlated with the degree of extrapair mating success in the common yellowthroat, Geothlypis frichas (Thusius et al., 13

2001). Melanin-based ornaments may therefore have a role in both intra- and intersexual signaling.

c) Structural Colouration

Structural plumage colours are produced by the coherent scattering of light by the nanostructural elements of the feather. Initially, this type of ornamentation was not thought to be involved in sexually-selected displays (Gray, 1996) because early methods of colour measurement were not able to detect individual variation or instances of sexual dichromatism in structural plumage colour. However, the advent of reflectance spectrometry in recent years has allowed the accumulation of evidence that differences in structural colouration between the sexes and between individuals are often greater than was originally thought (Hunt et al., 1998; Cuthill et al., 1999).

Although the costs of producing structurally-coloured plumage were not initially obvious, recent studies of the mechanisms involved suggest how structural colouration might indicate individual quality. For instance, in eastern bluebirds (Sialis sialis), each barb of a structurally-coloured feather is comprised of an inner layer of dark melanin granules, covered by a spongy layer of keratin rods and air spaces which is in turn covered by an outer keratin cortex (Shawkey et al., 2003). The precision and dimensions of this nanostructual arrangement produce much of the variation in individual plumage hue and saturation (Shawkey et al., 2003; Shawkey et al., 2005). If it is physiologically costly to grow the precise arrangement of feather nanostructural elements that produces the most intense colour, structurally-coloured ornaments may be an honest signal of developmental stability during molt (Hill and McGraw, 2006). In addition, bright structural colours may 14

have a cost of reducing the structural integrity of plumage since the thickness of the cortex is negatively correlated with plumage brightness (Shawkey et al., 2005).

These findings may help to explain the growing body of evidence that structural colour is related to individual condition and quality in a number of avian species. For example, structural colour appears to be an honest signal of male condition in the blue- black grassquit Volatinia jacarina (Doucet, 2002) and the blue grosbeak Guiraca caerulea

(Keyser and Hill, 2000). In satin bowerbirds (Ptilonorhynchus violaceus), the structural colouration of males is related to the intensity of parasite infection (Doucet and

Montgomerie, 2003) and in wild turkeys (Meleagris gallopavo), experimentally increasing parasite infection can suppress the expression of structural plumage ornaments (Hill et al.,

2005). Structural colour also predicts the level of parental care offered by eastern bluebird males (Siefferman and Hill, 2003) and females (Siefferman and Hill, 2005a), and indicates a male’s abilities in nestbox competition (Siefferman and Hill, 2005b).

Given this evidence, it is not surprising that structural plumage colour appears to be involved in mate choice in several species. For example, in bluethroats (Luscinia s. svecica), male attractiveness can be reduced by artificially decreasing the UV reflectance of blue throat feathers (Johnsen et al., 1998). In addition, blue tits have been shown to mate assortatively by the UV reflectance of their blue crown patch (Andersson et al., 1998).

Moreover, it appears that blue tit females adjust parental effort (Limbourg et al., 2004) and offspring sex ratios (Sheldon et al., 1999) in response to the UV attractiveness of their mate.

15

Study Species: The Tree Swallow

The tree swallow (Tachycineta bicolor) is a migratory, cavity-nesting passerine bird with a breeding range south of the tree line throughout North America. The plumage of adult tree swallows is structurally coloured, displaying iridescent blue-green on the upperparts and white below. Males molt directly from their brown juvenile plumage to the adult blue-green plumage before their first year of breeding. Females, however, molt to an intermediate plumage before their first year of breeding, displaying mostly brown upperparts interspersed with greenish blue feathers (Hussell, 1983). By their second year, females attain an adult plumage similar to that of males, although they may retain some brown feathers on the forehead as adults (Stutchbury and Robertson, 1987a).

This sexual dimorphism in plumage maturation suggests that the display of blue- green plumage may be costly to second-year females. It has been established that females in their first year of breeding experience reduced reproductive success, probably due to the slow acquisition of skills that improve foraging abilities (Robertson and Rendell, 2001) and competitive abilities as the birds age (Stutchbury and Robertson, 1985). Females may benefit from delayed plumage maturation if it allows them to avoid the costs associated with adult plumage, such as increased conspicuousness to predators and aggressive conspecifics (Stutchbury and Robertson, 1987b), or the energetic costs of the production of structural colour.

Tree swallows are socially monogamous birds with a high level of parental investment required from both sexes (Lombardo, 1991) but also a high incidence of extrapair paternity (Lifjeld et al., 1993; Kempenaers et al., 1999), with about half of all offspring having extrapair sires. It remains unclear what benefits females may derive from 16

these copulations with other males. Extrapair young do not survive any better than within- pair nest mates before fledging nor do they differ from within-pair nest mates in body size or mass (Kempenaers et al., 1999; Whittingham and Dunn, 2001). There is some evidence that extrapair fertilizations may result in greater hatching success in accordance with the genetic compatibility hypothesis (Kempenaers et al., 1999) but more work is needed on this.

The rate of extrapair paternity in tree swallows is relatively constant across populations despite differences in male breeding density (Conrad et al., 2001), breeding synchrony, or female breeding experience (Dunn et al., 1994), suggesting that female tree swallows may actively select and reject extrapair copulation partners. Yet it remains unclear how female tree swallows might choose particular males to father their extrapair young as would be expected from the good genes hypothesis. For example, female reproductive tactics were not affected by experimental mate replacement with younger males (Barber et al., 1998). In addition, extrapair fathers do not differ from within-pair fathers in any of several measures of body size (Kempenaers et al., 1999), although there is some evidence that extrapair fathers may be in better condition than the males they cuckold

(Kempenaers et al., 2001).

The possibility that plumage colour could be a signal of mate quality used by females when choosing extrapair mates has not yet been investigated in this species. The extent of extrapair paternity found in tree swallows may result in a great deal of variance in male reproductive success, especially if females tend to choose certain males. Such a scenario has the potential to exert strong sexual selection on traits signaling male quality 17

and condition. If plumage colouration is an indication of female quality as well, a system of mutual mate choice by plumage colour would be expected.

For these reasons, I have attempted to quantify the plumage colour of tree swallow breeding pairs, and to compare plumage colour within pairs to test whether there is evidence for assortative mating by plumage colour. I looked for correlations between variables describing individual plumage colour and variables describing morphology and reproductive performance, to determine whether plumage might be an honest signal of mate quality in tree swallows. 18

METHODS

This study was conducted in June 2005 at the Queen’s University Biological Station in southeastern Ontario, Canada (44°34’N, 76°19’W). At this site, several grids of nest boxes are provided for tree swallow nesting. The sample of birds used for this study consisted of breeding pairs where the female displayed adult (blue-green) plumage.

Initially 18 pairs were studied, although one of these pairs was later excluded due to the male having two mates, an unusual situation in this species (Dunn and Robertson, 1992).

Individuals were sexed based on the presence of a brood patch (female) or cloacal protuberance (male). Each bird was banded with a numbered aluminum band to allow identification and a coloured plastic band indicating sex (red for female, blue for male).

Plumage Colour

Birds were captured inside their nest boxes during the nestling period, between 8 and 24 June, for the quantification of plumage colour. Measurements were taken with an

Ocean Optics USB 2000 spectrometer and Pulsed Xenon (PX-2) lamp. All spectral readings were taken perpendicular to the feather surface, with the spectrometer probe encased in a plastic sheath to ensure that the probe was a constant 2 mm from the feather surface and to exclude ambient light. Each reading was the average of 30 reflectance curves, expressed as a proportion of the reflectance from a Spectralon® white standard.

Five readings were taken from each body region displaying blue-green plumage (crown, nape, rump and back), moving the probe haphazardly at least 5 mm between readings.

Spectral curves that did not match the general pattern were considered to be machine errors 19

and were excluded from further analyses. Figure 2 shows representative spectra from the back region of one male and one female.

Tristimulus colour variables were calculated from each spectral curve using the program ColoR 1.7 (available at http://biology.queensu.ca/~color). Total spectral reflectance was calculated as the sum of mean reflectance values for each 1 nm interval over the entire bird-visible range (320-700 nm). The same method was used to calculate total reflectance values for each of the four spectral regions corresponding to the four avian photoreceptors: UV (320-400 nm), blue (400-512 nm), green (512-575 nm) and red (575-

700 nm). Chroma, an index of spectral saturation, was calculated by dividing the total reflectance for a given region by the total spectral reflectance. Another index of saturation was calculated as ‘reflectance amplitude’, or the difference between maximum and minimum percent reflectances. As an index of brightness, I calculated the maximum percent reflectance across the bird-visible range (as in Keyser and Hill, 1999; hereafter referred to as Rmax). An index of hue was calculated as the wavelength of the midpoint between maximum and minimum reflectances (hereafter λRmid). The wavelength of maximum reflectance is often used to indicate hue, but λRmid was preferred here due to anomalous peaks in the spectral readings. For each of these colour variables, a mean value was calculated for each body region of every individual, with the exception of λRmid, which was only calculated for the back and rump regions.

Based on previous work with tree swallows (R. Montgomerie, unpublished data), five colour variables were chosen for analysis (Rmax, reflectance amplitude, blue chroma, green chroma and λRmid) as these variables had previously been found to correlate between the sexes in mated pairs. Colour data were available for all but two birds, for which 20

measurements were incomplete due to their accidental release during data collection (one male and one female from two different pairs).

Morphological Measurements

Several measurements of body size were taken from adult birds. A ruler was used to measure wing chord and the length of the outermost tail feather on the right and left side to the nearest mm. Digital calipers were used to measure right and left tarsus length to the nearest 0.01 mm. The mean of right and left values was used in all analyses.

Birds were weighed using a spring balance to the nearest 0.1 g. An index of individual body condition was calculated as the residuals from the regression of log mass on log tarsus for each sex. There was a significant relationship between log mass and log tarsus for males (r = 0.59, n = 12, p = 0.04) but not for females (r = 0.02, n = 12, p = 0.96).

As a measure of male reproductive condition, the height and diameter of the male cloacal protuberance was measured to the nearest 0.01 mm using digital calipers. Cloacal protuberance volume was calculated as V = π(d/2)2(h). These measurements were taken twice for each male, and an average cloacal protuberance volume was calculated. As measures of ectoparasite load, the number of holes and mites in wing and tail feathers were counted for both males and females.

Morphological measurements were taken within 4 days of spectral quantification for

23 of the 36 individuals in this study. For the remaining 13 individuals, morphological measurements were taken an average of 27.8 (range of 2 to 57) days before spectral readings were taken. Sample sizes vary for these measurements because not all individuals were measured for each morphological variable. 21

Reproductive Performance

Nests were monitored for 11 of the pairs for which plumage data were collected.

Laying date for the first egg and clutch size were recorded earlier in 2005 (M. Stapleton, unpublished data). Nest boxes were also monitored for hatching success and nestling survival. Of the 8 nests for which hatching data were available, 7 had 100% hatching success, and in the remaining nest only one of the 5 eggs failed to hatch. For this reason, hatching success was not included in further analyses. Nestlings were banded with a numbered aluminum band about 14 days after hatching. Any nestlings found dead inside a nest box were recorded, and it was assumed that all nestlings that were not found dead had successfully fledged. The percent of nestlings successfully fledging was calculated for each nest for which these data were available (n = 8).

Statistical Analyses

Statistical analyses were performed using JMP 5.1 (SAS Institute Inc., Cary, North

Carolina, U.S.A.). Distributions often appeared to differ from normal, but due to small sample sizes the fit to normality was difficult to determine. For this reason, nonparametric

Spearman rank correlations are reported for tests of assortative mating and for relationships between plumage colour and measures of individual quality.

22

RESULTS

Plumage Colouration and Sexual Dichromatism

The structural plumage of both male and female adult tree swallows reflects most strongly in the blue and green portions of the spectrum, with a much lower peak in reflectance in the UV region at the lower end of bird-visible light (Figure 2). Within pairs there was a significant difference between the sexes in Rmax of the nape region only (Table

1). Thus, males were not brighter than their mates overall. There were significant differences between the sexes in reflectance amplitude of the crown, nape and rump regions, in blue chroma of all body regions, and in green chroma of the nape and back regions (Table 1). There was also a significant difference between the sexes in λRmid of the rump region indicating that the hue of male plumage is left-shifted compared to that of their mates (Table 1). Thus there is a pattern of sexual dichromatism in plumage hue and colour saturation such that males generally have bluer more intensely-coloured plumage than their mates.

Assortative Mating

There were no significant relationships between male and female morphological traits (tarsus, wing, and tail length) within pairs (Table 2). Correlations between tarsus length and wing chord were relatively high, though far from significant, and may be worth looking at further with larger samples (Table 2). There was no relationship between male and female body mass within pairs, or between male and female body condition as measured by the residuals of log mass on log tarsus (Table 2). With respect to parasites, there was no significant relationship between the number of feather holes within pairs 23

(Table 2), but there was a significant positive relationship between the number of feather mites within pairs (rs = 0.55, p = 0.05, n = 13; Figure 3).

Despite the lack of correlation between morphological traits within pairs, there was a significant positive relationship for both λRmid (rs = 0.52, p = 0.04, n = 16; Figure 4) and blue chroma of the rump region (rs = 0.49, p = 0.05, n = 16; Figure 4). Thus bluer, more intensely coloured males were mated to bluer, more intensely coloured females. Since both members of most pairs were caught and measured on the same day, the observed relationships within pairs may have been the result of feather wear (which would increase with time) influencing plumage colour. Alternatively, the observed pattern of assortative mating could be a byproduct of passive age-assortative mating by arrival date on the breeding grounds and a change in plumage colour with age. I therefore tested the effects of measurement date and first egg date (used as an index of arrival date) on both λRmid and blue chroma measurements.

Rump λRmid was not significantly related to measurement date in either males (rs = -

0.33, p = 0.21, n = 16) or females (rs = -0.21, p = 0.42, n = 17). The fact that these relationships are both negative is interesting and needs more study; it suggests that rump

λRmid measurements became shifted towards shorter wavelengths with time. Rump blue chroma was also not related to measurement date in males (rs = 0.10, p = 0.72, n = 16) or females (rs = -0.12, p = 0.64, n = 17). Similarly, rump λRmid was not related to first egg date in males (rs = -0.25, p = 0.46, n = 11) or females (rs = -0.03, p = 0.94, n = 11); nor was rump blue chroma related to first egg date in males (rs = 0.07, p = 0.83, n = 11) or females

(rs = -0.31, p = 0.36, n = 11). 24

Although plumage colour was not significantly related to either measurement date or first egg date, measurement date was entered into multiple regression models to ensure that this variable was controlled statistically. Male rump λRmid was significantly predicted by partner rump λRmid (F = 5.86, p = 0.03) controlling for male measurement day (F = 1.50, p = 0.24). Female rump λRmid was also significantly predicted by partner rump λRmid (F =

5.36, p = 0.04) controlling for measurement day (F = 0.02, p = 0.89). Similarly, male rump blue chroma was significantly predicted by partner rump blue chroma (F = 7.14, p = 0.02) controlling for measurement day (F = 1.21, p = 0.29), and female rump blue chroma was significantly predicted by partner blue chroma (F = 6.62, p = 0.03) controlling for measurement day (F = 0.67, p = 0.43). Consequently, there is no evidence to suggest that the observed mating pattern by plumage colour is a spurious correlation.

Colour in Relation to Morphology

In males, wing length and body mass were positively correlated with λRmid of the back region, though the correlation with body mass was not quite significant (Figure 5,

Table 4). There were also significant negative relationships between male tarsus length and blue chroma of the nape region, and between male mass and blue chroma of the nape and back regions. Taken together, these results suggest that smaller males tend to display plumage hues shifted towards shorter wavelengths within the blue range (see Figure 1 for a representation of the hues of various spectral regions). There were no significant relationships between measures of parasite load, body condition, or cloacal protuberance volume and plumage colour in males. See the appendix (Table A1) for a complete 25

summary of the relationships between male plumage colour variables and morphological measurements.

In females, there was a significant negative relationship between the number of feather mites and the blue chroma of the crown region (Table 4), a weak positive relationship between number of feather mites and λRmid of the rump region, and a negative relationship between number of feather mites and Rmax of the nape region (Table 4). Thus, more intensely blue-coloured and brighter females tended to have fewer feather mites. In addition, significant negative relationships were found between both female mass and body condition and Rmax of the back region, suggesting that brighter females carried a lower body mass (Table 4). However, the relationships between female plumage brightness and morphological variables were not consistent, so this result may have been a statistical artifact. No relationships were found between female wing, tarsus, or tail lengths and plumage colour. A complete summary of the relationships between female plumage colour variables and morphological measurements can be found in the appendix (Table A2).

Reproductive Performance in Relation to Colour

In males, fledging success was positively related to green chroma for all body regions, though only the correlation with the nape region was significant (Figure 6, Table

5). There were also weak positive relationships between male fledging success and reflectance amplitude of the nape and back regions (Table 5). Clutch size was negatively related to male green chroma of the nape, and this relationship was significant (Figure 6,

Table 5). This result suggests that females mated to more intensely colour-saturated males tended to lay smaller clutches, although it should be interpreted with caution because there 26

was no consistent pattern relating clutch size to male green chroma. There were no significant relationships between first egg date and male plumage colour. A complete summary of the relationships between reproductive performance and male plumage colour can be found in the appendix (Table A3).

In females, there were positive relationships between fledging success and green chroma of the nape, back and rump regions similar to the patterns in males, though none of these was significant (Figure 6, Table 5). In addition, there were significant positive relationships between clutch size and green chroma of the crown region (Figure 6), and between clutch size and both Rmax and reflectance amplitude of the back region (Table 5).

Thus, more intensely green-coloured and brighter females tended to lay larger clutches and fledge more offspring. There were also weak negative relationships between fledging success and Rmax and reflectance amplitude of the crown region in females. However, no other relationships were found between fledging success and Rmax or reflectance amplitude in females, so these results should be interpreted with caution. Finally, there was a significant negative relationship between first egg date and green chroma of the back region in females, and weak negative relationships between first egg date and green chroma of the nape and rump regions (Table 5), suggesting that more intensely-green coloured females were able to begin breeding earlier.

27

DISCUSSION

The results of this study provide evidence that tree swallows mate assortatively with respect to plumage colour and that plumage colour may be an honest signal of mate quality in tree swallows. First, the degree of sexual dichromatism suggests that the expression of plumage colour may be costly to produce and subject to sexual selection. Second, the observed pattern of assortative mating indicates that tree swallows may use plumage colour as a cue when choosing mates. Third, correlations between plumage colour and morphological measurements, and between reproductive performance and plumage colour suggest that structural colour might be a useful signal of body size and reproductive performance in tree swallows.

Sexual Dichromatism and Cost of Structural Colour

Tree swallows have long been known to be sexually dichromatic to some degree

(Stutchbury and Robertson, 1987a), but this study is the first to use reflectance spectrometry to objectively assess sexual dichromatism in the blue-green plumage of adults. I found that male tree swallows have more intensely blue-coloured plumage compared to the greener, less saturated plumage of their mates. This is consistent with the results of structural colour quantification in many other species, in which the hue of male plumage tends to be shifted towards shorter wavelengths compared to that of females (blue tits, Hunt et al., 1998; blue grosbeaks, Keyser and Hill, 1999; satin bowerbirds, Doucet and

Montgomerie 2003).

The intermediate plumage of second-year females (dull brown with some greenish- blue feathers) suggests that there may be costs involved in the expression of blue-green 28

plumage in tree swallows. Indeed, it has been shown that females expressing adult plumage are larger and in better body condition than those with intermediate brown plumage (Lozano and Handford, 1995). Previous studies have suggested that the adaptive significance of delayed plumage maturation for females might be that, as a signal of sex, it reduces aggression from resident males during nest-site exploration (Stutchbury and

Robertson, 1987b). Yet the expression of blue-green plumage may be costly for other reasons, such as the energetic cost of growing the precise nanostructural arrangement that produces structural colour (Hill and McGraw, 2006). Second-year female tree swallows, who may lack foraging skills (Robertson and Rendell, 2001), may not be able to afford the energetic costs of producing blue-green structural colour.

If the expression of saturated blue plumage colour also indicates quality in male tree swallows, it is possible that sexual dichromatism in this species is the result of intense sexual selection on males for success in obtaining mates (within-pair or extrapair).

Interestingly, courtship in tree swallows involves a male display in which he presents his blue-green upperparts in a vertical posture with wings slightly drooped and tail slightly raised while an unfamiliar female approaches from behind (Robertson et al., 1992). If the female is receptive, she descends onto the displaying male in a courtship pounce. Thus, tree swallow courtship display, sexual dichromatism and the potential costs of structural colour all point to the possibility that plumage colouration may be sexually selected in this species.

29

Assortative Mating

Tree swallows did not mate assortatively by any measures of body size or condition, although there was a correlation between partners in the number of feather mites. This correlation could be the result of the shared nest environment where mites might be acquired, or it could indicate similar health or condition between mates.

Despite the absence of other morphological correlations between mates, there was a significant pattern of assortative mating by plumage hue and saturation. More intensely- blue coloured females paired with males who also had intensely-blue coloured plumage, even though the colours of mates were not exactly the same (females were greener). To date, the only other bird shown to mate assortatively by structural plumage colour is the blue tit (Andersson et al., 1998). Experimental manipulations of plumage reflectance in the blue tit indicate that this pattern is most likely the result of mutual mate choice (Hunt et al.,

1999). Selection for mutual mate choice is predicted in species, such as tree swallows, where both parents invest heavily in parental care (Johnstone et al., 1996). If intensely blue-coloured plumage signals individual quality, both males and females may benefit by choosing bluer birds as their social mates. The benefits of choice could be direct (parental care) or indirect (genetic quality of offspring). Thus, assortative mating in tree swallows could be the result of mutual mate choice. However, two other explanations are possible and must be considered.

First, it is possible that the observed pattern was due to tree swallows mating assortatively by some other characteristic that happens to be correlated with plumage colour. For example, it has been suggested that tree swallow plumage colour changes with age (R.J. Robertson, personal communication), and that tree swallows mate assortatively by 30

age with older females tending to pair with older males (Robertson and Rendell, 2001).

The observed pattern of assortative mating by plumage colour may therefore be the result of passive pairing by arrival date on the nesting grounds rather than active mate choice, since older tree swallows initiate nesting earlier in the season (Robertson and Rendell,

2001). A pattern of passive age-assortative mating has been observed in Spanish imperial eagles (Aquila adalberti), where age-assortative pairing is the result of competition over heterogeneous territory quality rather than active mate choice (Ferrer and Penteriani, 2003).

In my study, however, first egg-laying date was not a good predictor of plumage hue or blue chroma in males or females. Assuming that first egg-laying date is correlated with arrival date at the breeding grounds, it is therefore unlikely that the observed pattern of assortative mating by plumage colour was a byproduct of passive age-assortative mating.

Second, the assortative mating observed here could be the result of unidirectional mate choice and intrasexual competition rather than mutual mate choice. Tree swallows arrive at breeding sites early in spring, with males arriving slightly earlier than females to choose and defend a nest site where they will court potential mates. Arriving females may prefer males expressing bluer plumage, but also, if female plumage colour indicates condition, bluer females may be more likely to win competitions for preferred males. The high degree of aggressive intrasexual competition over nest sites in female tree swallows

(Stutchbury and Robertson, 1985) and the fact that structural colouration is a good signal of competitive ability in the eastern bluebird (Siefferman and Hill, 2005b) suggest that unidirectional mate choice is a plausible explanation for assortative mating in this species.

Thus, the evidence here is consistent with both mutual mate choice and unidirectional mate 31

choice by plumage colour. Further study is needed to distinguish between these two possibilities with respect to tree swallow pairing.

Benefits of Mate Choice by Plumage Colour

In males, several trends suggest that plumage colour may be a good signal of overall body size (Figure 5, Table 4). This result is probably not a statistical artifact since correlations of both blue chroma and λRmid and measures of size were consistent in direction

(though not in significance); smaller males tended to display plumage with higher blue chroma and hues shifted towards shorter wavelengths. Under the assumption that more intensely coloured males should be higher quality individuals, this result seems counterintuitive. One possibility is that it represents a shift towards shorter wavelengths within the blue range among smaller, low-quality males (see Figure 1 for a representation of the hues of various spectral regions). Without knowing more about the specific mechanisms of structural colouration in tree swallows, these results are difficult to interpret. Nevertheless, the relative consistency of relationships between male plumage colour and body size suggests that structural colour may be an honest signal of male quality in tree swallows.

In females, relationships between plumage colour variables and the number of feather mites were also interesting. There was a significant negative relationship between the number of feather mites and the blue chroma of the crown region. In addition, the relative consistency of relationships between number of mites and both blue chroma and

λRmid in different plumage regions suggests that the expression of dull green plumage is a 32

signal of mite infestation in female tree swallows. Thus structural plumage colour may be a signal of individual quality in females as well.

If plumage colour is an honest signal of quality in tree swallows, it might also be used during mate choice to predict reproductive performance, since higher quality individuals might provide better care for offspring. In both males and females, fairly consistent positive relationships were found between fledging success and the green chroma of various body regions (Figure 5), suggesting that plumage colour saturation may be a good indication of parental quality in both sexes. This is similar to evidence that structural colouration may signal the quality of parental care in eastern bluebird males

(Siefferman and Hill, 2003) and females (Siefferman and Hill, 2005a).

There was also a significant positive relationship between female green chroma of the crown and clutch size suggesting that females expressing more saturated plumage are able to produce larger clutches (Figure 5). Previous work with tree swallows has shown that clutch size is probably limited by the energy available to females during laying

(Murphy et al., 2000). My results therefore suggest that plumage colouration may be a good indicator of female condition during egg laying. In addition, there were consistent negative relationships between female green chroma of various body regions and first egg- laying date. In tree swallows, females breeding earlier in the season tend to have greater reproductive output (Robertson et al., 1992). Moreover, early arrival on the nesting grounds is probably costly because it exposes females to a high risk of starvation (Murphy et al., 2000). Thus, the relationship between first egg date and female plumage colour supports the hypothesis that structural colour is a good indication of reproductive performance in female tree swallows. 33

Curiously, there was also a significant negative relationship between male green chroma of the nape and clutch size suggesting that females mated to more colour-saturated males tended to produce smaller clutches (Figure 5). There were no other relationships consistent with this trend, so it may be a statistical artifact. The potential signal function of plumage colour for parental abilities of both male and female tree swallows is interesting nonetheless, and needs more investigation.

The significant relationships between plumage colour and both morphology and reproductive performance suggest that the blue-green plumage of tree swallows may signal aspects of individual quality in both males and females. Thus female preference for more ornamented social mates may be adaptive in terms of indirect benefits (good genes) or direct benefits (parental care), or both. Male preference for ornamented mates may be adaptive as well, although further investigation is needed to determine whether mate choice is unidirectional or mutual in tree swallows.

Limitations and Future Directions

There are several considerations that must be taken into account when interpreting the results reported here. This study was based on a small sample size and data on reproductive performance were not available for many of the captured pairs. Further studies on a larger scale may help clarify the potential benefits of mate choice by plumage colour, especially with regards to the numerous weak correlations between plumage colour and aspects of male and female quality. In addition, I did not control for aspects of individual quality that vary over time (mass, parasite load, cloacal protuberance volume), which may have obscured any relationships with plumage colour. There are also several 34

possible indices of individual quality that were not measured here (such as feather growth rates, endoparasite load, degree of parental investment, or male ejaculate quality).

In addition, although I was able to exclude the explanation that assortative mating by plumage colour was a byproduct of passive age-assortative mating, I was not able to directly test whether plumage colour varies with age. It is possible that the observed sex differences in hue and saturation are a byproduct of blue-green males being, on average, younger than blue-green females. Even if plumage colour was found to vary with age, it is still likely an important signal of individual quality. For instance, as tree swallows age, they are able to maintain better condition during breeding (Robertson and Rendell, 2001).

An understanding of whether plumage colour varies with age in tree swallows might clarify the kind of information conveyed by plumage signals.

Finally, since the results of this study are purely correlational, experimental corroboration would be valuable. For example, controlled tests should be undertaken to determine the effects of various stressors during molt (disease or nutritional) on the expression of tree swallow plumage colours. Experimental work would be especially useful in distinguishing between the various explanations for the assortative mating observed here (mutual mate choice versus unidirectional mate choice). For instance, tests of mate preference in a controlled environment could determine whether both male and female tree swallows prefer mates based on plumage colouration, as would be predicted by the mutual mate choice hypothesis. Further observations in the field could also indicate whether bluer females are more likely to win aggressive competition for nest sites. If so, assortative mating by plumage colour might be the product of unidirectional mate choice and female competition rather than mutual mate choice. 35

The results of this study also point to a possible explanation for extrapair paternity in tree swallows. If females choose extrapair mates based on plumage characteristics, they may obtain a genetic benefit from these copulations in terms of increased viability or increased attractiveness in extrapair offspring. Tree swallow extrapair young are no more likely to survive until fledging, or to be male (i.e. no sex ratio bias), than their within-pair siblings (Whittingham and Dunn, 2001). Nevertheless, it may be beneficial for females to raise more attractive offspring if mutual mate choice operates in this species. Tests of plumage colour and the good genes hypothesis for extrapair paternity in tree swallows are warranted. Further investigations of tree swallow plumage colour have the potential to help us understand a great deal of the reproductive behaviour of this species. 36

LITERATURE CITED

Andersson, S., J. Ornborg, M. Andersson. 1998. Ultraviolet sexual dimorphism and assortative mating in blue tits. Proceedings of the Royal Society of London, Series B, 265, 445-450.

Barber, C.A., R.J. Robertson and P.T. Boag. 1998. Experimental mate replacement does not increase extra-pair paternity in tree swallows. Proceedings of the Royal Society of London, Series B, 265, 2187-2190.

Bennett, A.T.D. and I.C. Cuthill. 1994. Ultraviolet vision in birds: what is its function? Vision Research, 34, 1471-1478.

Bennett, A.T.D., I.C. Cuthill, K.J. Norris. 1994. Sexual selection and the mismeasure of color. The American Naturalist, 144, 848-860.

Conrad, K.F., P.V. Johnston, C. Crossman, B. Kempenaers, R.J. Robertson, N.T. Wheelwright and P.T. Boag. 2001. High levels of extra-pair paternity in an isolated, low-density, island population of tree swallows (Tachycineta bicolor). Molecular Ecology, 10, 1301-1308.

Cuthill, I.C., A.T.D. Bennett, J.C. Partridge, E.J. Maier. 1999. Plumage reflectance and the objective assessment of avian sexual dichromatism. The American Naturalist, 153, 183-200.

Darwin, C. 1871. The Descent of Man and Selection in Relation to Sex. London, Murray.

Delhey, K., A. Johnsen, A. Peters, S. Andersson and B. Kempenaers. 2003. Paternity analysis reveals opposing selection pressures on crown coloration in the blue tit (Parus caeruleus), Proceedings of the Royal Society of London, Series B, 270, 2057-2063.

Doucet, S.M. 2002. Structural plumage coloration, male body size, and condition in the blue-black grassquit. The Condor, 104, 30-38.

Doucet, S.M., D.J. Mennill, R. Montgomerie, P.T. Boag and L.M. Ratcliffe. 2005. Achromatic plumage reflectance predicts reproductive success in male black-capped chickadees. Behavioral Ecology, 16, 218-222.

Doucet, S.M. and R. Montgomerie. 2003. Structural plumage colour and parasites in satin bowerbirds Ptilonorhynchus violaceus: implications for sexual selection. Journal of Avian Biology, 34, 237-242.

37

Dunn, P.O., L.A. Whittingham, J.T. Lifjeld, R.J. Robertson and P.T. Boag. 1994. Effects of breeding density, synchrony and experience on extrapair paternity in tree swallows. Behavioral Ecology, 5, 123-129.

Dunn, P.O. and R.J. Robertson. 1992. Geographic variation in the importance of male parental care and mating systems in tree swallows. Behavioral Ecology, 3, 291-299.

Eaton, M.D. 2005. Human vision fails to distinguish widespread sexual dichromatism among sexually “monochromatic” birds. Proceedings of the National Academy of Sciences, 102, 10942-10946.

Eaton, M.D. and S.M. Lanyon. 2003. The ubiquity of avian ultraviolet plumage reflectance. Proceedings of the Royal Society of London, Series B, 270, 1721-1726.

Ferrer, M., and V. Penteriani. 2003. A process of pair formation leading to assortative mating: passive age assortative mating by habitat heterogeneity. Animal Behaviour, 66, 137-143.

Fisher, R.A. 1958. The Genetical Theory of Natural Selection. New York, Dover Publications.

Gonzalez, G., G. Sorci, L.C. Smith and F. de Lope. 2001. Testosterone and sexual signalling in male house sparrows (Passer domesticus). Behavioral Ecology and Sociobiology, 50, 557-562.

Gray, D.A. 1996. Carotenoids and sexual dichromatism in North American passerine birds. The American Naturalist, 148, 453-480.

Hamilton, W.D. and M. Zuk. 1982. Heritable true fitness and bright birds: a role for parasites? Science, 218, 384-387.

Hill, G.E. and K.J. McGraw. 2006. Bird Coloration, Vol. I: Measurements and Mechanisms. Cambridge, Harvard University Press.

Hill, G.E. and R. Montgomerie. 1994. Plumage colour signals nutritional condition in the house finch. Proceedings of the Royal Society of London, Series B, 258, 47-52.

Hill, G.E., S.M. Doucet and R. Buchholz. 2005. The effect of coccidial infection on iridescent plumage coloration in wild turkeys. Animal Behaviour, 69, 387-394.

Hill, G.E. and W.R. Brawner. 1998. Melanin-based plumage coloration in the house finch is unaffected by coccidial infection. Proceedings of the Royal Society of London, Series B, 265, 1105-1109.

38

Hunt, S., A.T.D. Bennett, I.C. Cuthill, R. Griffiths. 1998. Blue tits are ultraviolet tits. Proceedings of the Royal Society of London, Series B, 265, 451-455.

Hunt, S., I.C. Cuthill, A.T.D. Bennett, R. Griffiths. 1999. Preferences for ultraviolet partners in the blue tit. Animal Behaviour, 58, 809-815.

Hussell, D.J.T. 1983. Age and plumage colour in female Tree Swallows. Journal of Field Ornithology, 54, 312-318.

Johnsen, A., S. Andersson, J. Ornborg and J.T. Lifjeld. 1998. Ultraviolet plumage ornamentation affects social mate choice and sperm competition in bluethroats (Aves: Luscinia s. svecica): a field experiment. Proceedings of the Royal Society of London, Series B, 265, 1313-1318.

Johnstone, R.A., J.D. Reynolds and J.C. Deutsch. 1996. Mutual mate choice and sex differences in choosiness. Evolution, 50, 1382-1392.

Kempenaers, B., B. Congdon, P. Boag and R.J. Robertson. 1999. Extrapair paternity and egg hatchability in tree swallows: evidence for the genetic compatibility hypothesis? Behavioral Ecology, 10, 304-311.

Kempenaers, B., S. Everding, C. Bishop, P. Boag and R.J. Robertson. 2001. Extra-pair paternity and the reproductive role of male floaters in the tree swallow (Tachycineta bicolor). Behavioral Ecology and Sociobiology, 49, 251-259.

Keyser, A.J. and G.E. Hill. 1999. Condition-dependent variation in the blue-ultraviolet coloration of a structurally based plumage ornament. Proceedings of the Royal Society of London, Series B, 266, 771-777.

Keyser, A.J. and G.E. Hill. 2000. Structurally based plumage coloration is an honest signal of quality in male blue grosbeaks. Behavioral Ecology, 11, 202-209.

Lifjeld, J.T., P.O. Dunn, R.J. Robertson and P.T. Boag. 1993. Extra-pair paternity in monogamous tree swallows. Animal Behaviour, 45, 213-229.

Limbourg, T., A.C. Mateman, S. Andersson and C.M. Lessells. 2004. Female blue tits adjust parental effort to manipulated male UV attractiveness. Proceedings of the Royal Society of London, Series B, 271, 1903-1908.

Lombardo, M.P. 1991. Sexual differences in parental effort during the nestling period in tree swallows (Tachycineta bicolor). The Auk, 108, 393-404.

Lozano, G.A. 1994. Carotenoids, parasites and sexual selection. Oikos, 70, 309-311.

39

Lozano, G.A. 1995. A test of an assumption of delayed plumage maturation hypotheses using female tree swallows. The Wilson Bulletin, 107, 153-164.

MacDougall, A.K. and R. Montgomerie. 2003. Assortative mating by carotenoid-based plumage colour: a quality indicator in American goldfinches, Carduelis tristis. Naturwissenschaften, 90, 464-467.

McGraw, K.J., E.A. Mackillop, J. Dale and M.E. Hauber. 2002. Different colors reveal different information: how nutritional stress affects the expression of melanin- and structurally based ornamental plumage. Journal of Experimental Biology, 205, 3747-3755.

McGraw, K.J. and G.E. Hill. 2000. Differential effects of endoparasitism on the expression of carotenoid- and melanin-based ornamental coloration. Proceedings of the Royal Society of London, Series B, 267, 1525-1531.

McGraw, K.J., G.E. Hill, R. Stradi, R.S. Parker. 2002. The effect of dietary carotenoid access on sexual dichromatism and plumage pigment composition in the American goldfinch. Comparative Biochemistry and Physiology, 131, 261-269.

McGraw, K.J., J. Dale and E.A. Mackillop. 2003. Social environment during molt and the expression of melanin-based plumage pigmentation in male house sparrows (Passer domesticus). Behavioral Ecology and Sociobiology, 53, 116-122.

Murphy, M.T., B. Armbrecth, E. Vlamis and A. Pierce. 2000. Is reproduction in tree swallows cost free? The Auk, 117, 902-912.

Olson, V.A. and I.P.F. Owens. 1998. Costly sexual signals: are carotenoids rare, risky or required? Trends in Ecology and Evolution, 13, 510-514.

Omland, K.E. 1996. Female mallard mating preferences for multiple male ornaments. Behavioral Ecology and Sociobiology, 39, 353-360.

Peters, A., A.G. Denk, K. Delhey and B. Kempenaers. 2004. Carotenoid-based bill colour as an indicator of immunocompetence and sperm performance in male mallards. Journal of Evolutionary Biology, 17, 1111-1120.

Robertson, R.J., Stutchbury, B.J. and R.R. Cohen. 1992. Tree Swallow. In The Birds of North America, No. 11 (A. Poole, P. Stettenheim and F. Gill, Eds.). Washington, DC, The American Ornithologists’ Union.

Robertson, R.J. and W.B. Rendell. 2001. A long-term study of reproductive performance in tree swallows: the influence of age and senescence on output. Journal of Animal Ecology, 70, 1014-1031.

40

Shawkey, M.D., A.M. Estes, L.M. Siefferman and G.E. Hill. 2003. Nanostructure predicts intraspecific variation in ultraviolet-blue plumage colour. Proceedings of the Royal Society of London, Series B, 270, 1455-1460.

Shawkey, M.D., A.M. Estes, L. Siefferman and G.E. Hill. 2005. The anatomical basis of sexual dichromatism in non-iridescent ultraviolet-blue structural coloration of feathers. Biological Journal of the Linnaean Society, 84, 259-271.

Sheldon, B.C., S. Andersson, S.C. Griffith, J. Ornborg and J. Sendecka. 1999. Ultraviolet colour variation influences blue tit sex ratios. Nature, 402, 874-877.

Siefferman, L. and G.E. Hill. 2003. Structural and melanin coloration indicate parental effort and reproductive success in male eastern bluebirds. Behavioral Ecology, 14, 855-861.

Siefferman, L. and G.E. Hill. 2005a. Evidence for sexual selection on structural plumage coloration in female eastern bluebirds (Sialia sialis). Evolution, 59, 1819-1828.

Siefferman, L. and G.E. Hill. 2005b. UV-blue structural coloration and competition for nestboxes in male eastern bluebirds. Animal Behaviour, 69, 67-72.

Stutchbury, B.J. and R.J. Robertson. 1985. Floating populations of female tree swallows. The Auk, 102, 651-654.

Stutchbury, B.J. and R.J. Robertson. 1987a. Two methods of sexing adult tree swallows before they begin breeding. Journal of Field Ornithology, 58, 236-242.

Stutchbury, B.J. and R.J. Robertson. 1987b. Signaling subordinate and female status: two hypotheses for the adaptive significance of subadult plumage in female tree swallows. The Auk, 104, 717-723.

Thusius, K.J., K.A. Peterson, P.O. Dunn and L.A. Whittingham. 2001. Male mask size is correlated with mating success in the common yellowthroat. Animal Behaviour, 62, 435-446.

Whittingham, L.A. and P.O. Dunn. 2001. Survival of extrapair and within-pair young in tree swallows. Behavioral Ecology, 12, 496-500.

Zahavi, A. 1975. Mate selection – a selection for handicap. Journal of Theoretical Biology, 53, 205-214.

41

SUMMARY

1. The structural plumage colour of adult tree swallows reflects most strongly in the blue and green regions of the bird-visible spectrum. There is significant sexual dimorphism such that males display bluer, more saturated plumage colour than their female counterparts. There is also considerable variation between individual tree swallow plumage scores within sexes.

2. Tree swallows did not mate assortatively by any measures of body size or condition.

However, tree swallow pairs did mate assortatively by plumage hue and saturation such that bluer males were paired with bluer females. This suggests that there may be mutual mate choice by plumage colour in this species, although unidirectional mate choice and female- female competition could also result in the observed pattern of assortative mating.

3. In males, there were consistent relationships between plumage colour and measures of body size such that smaller males tended to display plumage hues shifted towards shorter wavelengths within the blue range. In females, there were relationships between plumage colour and ectoparasite load such that females with more feather mites tended to display less saturated plumage colour. Structural plumage colour may therefore be an honest indicator of mate quality in both male and female tree swallows.

4. There were consistent positive relationships between percent of clutch successfully fledged and plumage colour saturation for both males and females, suggesting that plumage colour may be a good indication of reproductive performance and parental abilities in tree swallows.

42

Figure 1. Spectral ranges of human and avian colour vision. Birds are able to perceive wavelengths in the UV portion of the spectrum (320 – 400 nm) in addition to the range of wavelengths perceived by humans (400 – 700 nm). 43

Figure 2. Representative reflectance spectra for a mated pair of tree swallows. Each curve is a single reading taken from the blue-green plumage of the back region. The spectral range from 320 to 700 nm encompasses the known sensitivity of avian photoreceptors. The male reading has a blue-shifted hue and greater saturation (blue chroma) than that of the female. 44

Figure 3. Within-pair relationship in the number of feather mites of monogamously mated adult male and female tree swallows (rs = 0.55, p = 0.05, n = 13), with Model II regression line shown to illustrate the linear trend. 45

a) b)

Rmid λ Female

Male λRmid

Figure 4. Within-pair relationships in plumage colour variables of monogamously mated

adult male and female tree swallows. Relationships between (a) λRmid (rs = 0.52, p = 0.04)

and (b) blue chroma of the rump region (rs = 0.49, p = 0.05) are shown for 16 mated pairs,

with Model II regression lines shown to illustrate linear trends.

46

a) b)

Rmid λ ale M

Figure 5. Relationships between plumage colour of the back region and morphological

variables in male tree swallows. The relationships shown are between (a) λRmid of the back

and wing chord (rs = 0.51, n = 16, p = 0.04) and (b) blue chroma of the back and body mass

(rs = -0.65, n = 12, p = 0.02), with Model II regression lines.

47

a) b)

c) d)

Figure 6. Relationships between reproductive performance and plumage colour in adult male and female tree swallows. The relationships shown are between (a) clutch size and male green chroma of the nape region (rs = -0.69, n = 11, p = 0.02), (b) clutch size and female green chroma of the crown region (rs = 0.63, n = 11, p = 0.04), (c) percent of clutch successfully fledged and male green chroma of the nape region (rs = 0.81, n = 8, p = 0.01) and (d) percent of clutch successfully fledged and female green chroma of the nape region

(rs = 0.63, n = 8, p = 0.10). Model II regression lines are shown to illustrate linear trends. 48

Table 1. A comparison of plumage colour variables in mated adult male and female tree swallows (means ± SE) using two-tailed, paired t-tests (n = 16 pairs). Significant differences are highlighted in bold.

Comparison Colour Variable Body Region Male Female t p

Rmax crown 20.26 ± 1.56 16.77 ± 1.28 1.44 0.17

nape 14.08 ± 1.05 11.80 ± 0.71 2.31 0.04

back 18.58 ± 1.09 17.25 ± 1.25 0.28 0.78

rump 15.95 ± 0.96 13.00 ± 1.35 1.39 0.19

Reflectance crown 13.67 ± 1.14 10.31 ± 0.93 2.03 0.06 amplitude nape 8.87 ± 0.77 6.11 ± 0.46 3.64 0.003

back 12.11 ± 0.84 10.09 ± 0.87 1.37 0.19

rump 10.48 ± 0.74 7.48 ± 0.97 2.23 0.04

Blue chroma crown 0.352 ± 0.006 0.304 ± 0.005 6.88 < 0.0001

nape 0.358 ± 0.007 0.322 ± 0.005 3.97 0.001

back 0.370 ± 0.005 0.349 ± 0.005 4.16 0.001

rump 0.383 ± 0.004 0.349 ± 0.006 5.77 < 0.0001

Green chroma crown 0.230 ± 0.003 0.225 ± 0.002 1.50 0.16

nape 0.219 ± 0.003 0.209 ± 0.003 2.40 0.03

back 0.223 ± 0.003 0.215 ± 0.003 2.31 0.04

rump 0.217 ± 0.003 0.211 ± 0.003 1.45 0.17

λRmid back 420.7 ± 2.1 423.7 ± 2.8 1.35 0.20

rump 412.8 ± 2.2 420.9 ± 2.3 4.31 0.006

49

Table 2. Within-pair correlations for morphological variables in monogamously mated adult tree swallow pairs. Body condition is measured as the residuals from a regression of log mass on log tarsus. Spearman-rank correlations and p values are shown; significant within-pair relationships are highlighted in bold.

Morphological Variable rs n p Tarsus Length (mm) 0.32 16 0.22

Wing Chord (mm) 0.33 17 0.20

Tail Length (mm) -0.20 16 0.44

Mass (g) 0.005 11 0.99

Body Condition (g) -0.04 11 0.92

No. of Holes -0.21 13 0.49

No. of Mites 0.55 13 0.05

50

Table 3. Within-pair correlations for plumage colour variables in monogamously mated adult tree swallow pairs. Spearman-rank correlations and p values are shown; significant within-pair relationships are highlighted in bold.

Colour Variable Body Region rs n p

Rmax crown 0.21 17 0.42 nape 0.24 16 0.37 back 0.26 16 0.33 rump 0.29 16 0.28 Reflectance amplitude crown 0.15 17 0.56 nape 0.27 16 0.31 back 0.21 16 0.44 rump 0.26 16 0.34 Blue chroma crown 0.18 17 0.49 nape -0.06 16 0.81 back 0.37 16 0.16 rump 0.49 17 0.05 Green chroma crown 0.03 17 0.91 nape -0.02 16 0.95 back 0.36 16 0.18 rump 0.22 16 0.42

λRmid back 0.28 16 0.29 rump 0.52 16 0.04 51

Table 4. Relationships between plumage colour and morphological variables in male and

female tree swallows. Spearman-rank correlations (rs, n, p) are shown; significant

relationships are highlighted in bold.

Morphological Variable Colour Body CP Vol. Variable Region Wing (mm) Tarsus (mm) Mass (g) Condition (g) No. of Mites (mm3) 0.34, 17, 0.18 -0.003, 16, 0.99 0.07, 12, 0.82 0.00, 12, 1.00 0.03, 15, 0.90 0.02, 17, 0.93 Male Rmax crown nape 0.15, 17, 0.57 0.14, 16, 0.61 -0.22, 12, 0.49 -0.34, 12, 0.29 0.08, 15, 0.78 0.16, 17, 0.55 back 0.33, 16, 0.22 -0.09, 15, 0.76 -0.27, 12, 0.39 -0.34, 12, 0.28 0.23, 15, 0.40 0.38, 16, 0.15 rump -0.21, 16, 0.44 0.08, 15, 0.79 0.10, 12, 0.76 -0.03, 12, 0.93 0.09, 15, 0.74 0.46, 16, 0.07 Blue chroma crown -0.37, 17, 0.14 -0.15, 16, 0.58 -0.40, 12, 0.20 -0.37, 12, 0.24 -0.002, 15, 0.995 -0.14, 17, 0.58 nape -0.37, 17, 0.15 -0.52, 16, 0.04 -0.66, 12, 0.02 -0.16, 12, 0.62 -0.12, 15, 0.67 -0.04, 17, 0.89 back -0.30, 16, 0.26 -0.18, 15, 0.53 -0.65, 12, 0.02 -0.47, 12, 0.12 -0.03, 15, 0.91 -0.01, 16, 0.97 rump -0.22, 16, 0.40 0.17, 15, 0.54 0.07, 12, 0.82 -0.11, 12, 0.73 -0.12, 15, 0.66 0.39, 16, 0.14 0.51, 16, 0.04 0.26, 15, 0.19 0.54, 12, 0.07 0.19, 12, 0.56 0.16, 15, 0.57 0.17, 16, 0.53 λRmid back rump 0.39, 16, 0.14 0.13, 15, 0.64 0.29, 12, 0.37 0.13, 12, 0.69 0.09, 15, 0.76 -0.10, 16, 0.71 -0.32, 17, 0.20 -0.10, 17, 0.72 -0.27, 12, 0.41 -0.27, 12, 0.39 0.26, 15, 0.36 Female Rmax crown nape 0.31, 16, 0.25 0.38, 16, 0.14 0.31, 12, 0.33 0.31, 12, 0.33 -0.51, 14, 0.06 back 0.13, 17, 0.62 -0.11, 17, 0.68 -0.57, 12, 0.05 -0.59, 12, 0.04 0.08, 15, 0.77 rump -0.09, 17, 0.72 0.04, 17, 0.87 0.26, 12, 0.41 0.26, 12, 0.42 -0.10, 15, 0.73 Blue chroma crown -0.05, 17, 0.85 0.08, 17, 0.75 -0.11, 12, 0.72 -0.17, 12, 0.59 -0.70, 15, 0.004 nape 0.05, 16, 0.86 0.28, 16, 0.30 0.36, 12, 0.25 0.33, 12, 0.29 -0.42, 14, 0.14 back 0.13, 17, 0.62 0.02, 17, 0.94 0.03, 12, 0.93 0.06, 12, 0.86 -0.35, 15, 0.21 rump 0.02, 17, 0.94 0.13, 17, 0.62 0.19, 12, 0.55 0.22, 12, 0.50 -0.24, 15, 0.39 -0.10, 17, 0.70 -0.29, 17, 0.26 -0.17, 12, 0.60 -0.15, 12, 0.65 0.34, 15, 0.22 λRmid back rump -0.05, 17, 0.86 -0.19, 17, 0.46 -0.19, 12, 0.55 -0.24, 12, 0.46 0.46, 15, 0.08

52

Table 5. Relationships between reproductive performance and plumage colour in male and female tree swallows. Spearman-rank correlations (rs, n, p) are shown; significant relationships are highlighted in bold.

Colour Body Reproductive Performance Variable Region 1st Egg Date Clutch Size % Fledged

Male Rmax crown -0.05, 11, 0.87 0.32, 11, 0.33 -0.10, 8, 0.81 nape 0.12, 11, 0.73 -0.33, 11, 0.32 0.28, 8, 0.50

back 0.04, 11, 0.92 -0.25, 11, 0.46 0.50, 8, 0.21 rump -0.38, 11, 0.25 0.43, 11, 0.19 0.15, 8, 0.72

Reflectance crown -0.08, 11, 0.81 0.12, 11, 0.72 0.03, 8, 0.95 amplitude nape 0.10, 11, 0.77 -0.40, 11, 0.23 0.50, 8, 0.07 back 0.05, 11, 0.87 -0.23, 11, 0.49 0.46, 8, 0.25 rump -0.45, 11, 0.17 0.42, 11, 0.19 0.15, 8, 0.72 Blue chroma crown 0.45, 11, 0.17 -0.18, 11, 0.59 -0.27, 8, 0.52 nape 0.26, 11, 0.43 -0.23, 11, 0.49 -0.04, 8, 0.93 back -0.04, 11, 0.92 -0.22, 11, 0.51 0.37, 8, 0.37 rump 0.07, 11, 0.83 0.45, 11, 0.16 -0.38, 8, 0.35 Green chroma crown -0.21, 11, 0.54 0.05, 11, 0.89 0.57, 8, 0.14 nape -0.09, 11, 0.79 -0.69, 11, 0.02 0.87, 8, 0.01 back 0.11, 11, 0.75 -0.01, 11, 0.97 0.60, 8, 0.12 rump -0.35, 11, 0.28 0.22, 11, 0.51 0.41, 8, 0.31

Female Rmax crown 0.20, 11, 0.56 0.39, 11, 0.24 -0.65, 8, 0.08 nape -0.19, 11, 0.57 0.13, 11, 0.69 0.17, 8, 0.69 back -0.29, 11, 0.39 0.70, 11, 0.02 -0.14, 8, 0.74 rump -0.48, 11, 0.14 0.22, 11, 0.61 0.18, 8, 0.49 Reflectance crown 0.21, 11, 0.54 0.38, 11, 0.25 -0.65, 8, 0.08 amplitude nape -0.26, 11, 0.43 0.23, 11, 0.50 0.41, 8, 0.31 back -0.35, 11, 0.28 0.60, 11, 0.05 0.03, 8, 0.95 rump -0.44, 11, 0.18 0.28, 11, 0.40 0.22, 8, 0.61 Blue chroma crown 0.23, 11, 0.50 0.09, 11, 0.80 -0.28, 8, 0.27 nape 0.14, 11, 0.69 -0.17, 11, 0.61 0.22, 8, 0.61 back -0.10, 11, 0.77 0.23, 11, 0.49 -0.28, 8, 0.50 rump -0.31, 11, 0.36 0.26, 11, 0.44 0.26, 8, 0.54 Green chroma crown -0.19, 11, 0.57 0.63, 11, 0.04 -0.11, 8, 0.79 nape -0.59, 11, 0.06 0.23, 11, 0.49 0.63, 8, 0.10 back -0.65, 11, 0.03 0.34, 11, 0.31 0.55, 8, 0.16 rump -0.46, 11, 0.15 0.17, 11, 0.61 0.43, 8, 0.28 53

APPENDIX

Table A1. Relationships between plumage colour and morphological variables in male tree

swallows. Spearman-rank correlations (rs, n, p) are shown; significant relationships are

highlighted in bold.

Colour Body Morphological Variable Variable Region Wing (mm) Tarsus (mm) Tail (mm) Mass (g) Condition (g) No. of Holes No. of Mites CP Vol. (mm3)

Rmax crown 0.34, 17, 0.18 -0.003, 16, 0.99 -0.04, 16, 0.89 0.07, 12, 0.82 0.00, 12, 1.00 -0.11, 15, 0.71 0.03, 15, 0.90 0.02, 17, 0.93 nape 0.15, 17, 0.57 0.14, 16, 0.61 0.16, 16, 0.56 -0.22, 12, 0.49 -0.34, 12, 0.29 0.08, 15, 0.77 0.08, 15, 0.78 0.16, 17, 0.55 back 0.33, 16, 0.22 -0.09, 15, 0.76 -0.05, 15, 0.87 -0.27, 12, 0.39 -0.34, 12, 0.28 0.01, 15, 0.97 0.23, 15, 0.40 0.38, 16, 0.15 rump -0.21, 16, 0.44 0.08, 15, 0.79 -0.31, 15, 0.26 0.10, 12, 0.76 -0.03, 12, 0.93 -0.14, 15, 0.62 0.09, 15, 0.74 0.46, 16, 0.07 Reflectance crown 0.33, 17, 0.20 0.14, 16, 0.61 0.07, 16, 0.80 0.09, 12, 0.77 -0.06, 12, 0.86 -0.16, 15, 0.58 0.04, 15, 0.87 0.01, 17, 0.97 amplitude nape 0.20, 17, 0.44 -0.04, 16, 0.87 0.16, 16, 0.55 -0.34, 12, 0.28 -0.41, 12, 0.19 0.22, 15, 0.43 0.20, 15, 0.48 0.18, 17, 0.49 back 0.31, 16, 0.24 -0.05, 15, 0.85 -0.05, 15, 0.85 -0.21, 12, 0.50 -0.42, 12, 0.17 0.00, 15, 1.00 0.40, 15, 0.14 0.35, 16, 0.18 rump -0.14, 16, 0.60 0.25, 15, 0.38 -0.21, 15, 0.46 0.16, 12, 0.62 0.00, 12, 1.00 -0.16, 15, 0.57 0.10, 15, 0.72 0.36, 16, 0.17 Blue chroma crown -0.37, 17, 0.14 -0.15, 16, 0.58 -0.03, 16, 0.91 -0.40, 12, 0.20 -0.37, 12, 0.24 0.36, 15, 0.19 -0.002, 15, 0.995 -0.14, 17, 0.58 nape -0.37, 17, 0.15 -0.52, 16, 0.04 -0.17, 16, 0.54 -0.66, 12, 0.02 -0.16, 12, 0.62 0.40, 15, 0.14 -0.12, 15, 0.67 -0.04, 17, 0.89 back -0.30, 16, 0.26 -0.18, 15, 0.53 -0.05, 15, 0.86 -0.65, 12, 0.02 -0.47, 12, 0.12 0.30, 15, 0.27 -0.03, 15, 0.91 -0.01, 16, 0.97 rump -0.22, 16, 0.40 0.17, 15, 0.54 -0.15, 15, 0.60 0.07, 12, 0.82 -0.11, 12, 0.73 -0.39, 15, 0.15 -0.12, 15, 0.66 0.39, 16, 0.14 Green chroma crown 0.33, 17, 0.19 0.32, 16, 0.23 0.33, 16, 0.21 -0.007, 12, 0.98 -0.16, 12, 0.62 -0.17, 15, 0.55 -0.05, 15, 0.84 0.16, 17, 0.53 nape 0.21, 17, 0.43 0.09, 16, 0.75 0.21, 16, 0.43 -0.30, 12, 0.35 -0.39, 12, 0.21 0.32, 15, 0.25 0.12, 15, 0.67 -0.07, 17, 0.79 back 0.26, 16, 0.33 0.28, 15, 0.32 0.22, 15, 0.44 0.08, 12, 0.81 -0.35, 12, 0.27 -0.11, 15, 0.71 0.30, 15, 0.27 0.24, 16, 0.37 rump 0.37, 16, 0.16 0.11, 15, 0.69 0.08, 15, 0.78 0.34, 12, 0.27 0.15, 12, 0.65 -0.14, 15, 0.62 0.13, 15, 0.66 0.17, 16, 0.53

λRmid back 0.51, 16, 0.04 0.26, 15, 0.19 0.23, 15, 0.41 0.54, 12, 0.07 0.19, 12, 0.56 -0.33, 15, 0.23 0.16, 15, 0.57 0.17, 16, 0.53 rump 0.39, 16, 0.14 0.13, 15, 0.64 0.18, 15, 0.52 0.29, 12, 0.37 0.13, 12, 0.69 0.05, 15, 0.87 0.09, 15, 0.76 -0.10, 16, 0.71

54

Table A2. Relationships between plumage colour and morphological variables in female

tree swallows. Spearman-rank correlations (rs, n, p) are shown; significant relationships are

highlighted in bold.

Colour Body Morphological Variable Variable Region Wing (mm) Tarsus (mm) Tail (mm) Mass (g) Condition (g) No. of Holes No. of Mites

Rmax crown -0.32, 17, 0.20 -0.10, 17, 0.72 -0.04, 17, 0.88 -0.27, 12, 0.41 -0.27, 12, 0.39 0.15, 15, 0.60 0.26, 15, 0.36 nape 0.31, 16, 0.25 0.38, 16, 0.14 0.21, 16, 0.44 0.31, 12, 0.33 0.31, 12, 0.33 -0.43, 14, 0.12 -0.51, 14, 0.06 back 0.13, 17, 0.62 -0.11, 17, 0.68 -0.05, 17, 0.85 -0.57, 12, 0.05 -0.59, 12, 0.04 0.17, 15, 0.54 0.08, 15, 0.77 rump -0.09, 17, 0.72 0.04, 17, 0.87 0.14, 17, 0.60 0.26, 12, 0.41 0.26, 12, 0.42 -0.23, 15, 0.42 -0.10, 15, 0.73 Reflectance crown -0.37, 17, 0.15 -0.14, 17, 0.59 -0.02, 17, 0.94 -0.22, 12, 0.49 -0.22, 12, 0.48 0.15, 15, 0.60 0.32, 15, 0.25 amplitude nape 0.38, 16, 0.14 0.34, 16, 0.20 0.39, 16, 0.12 -0.11, 12, 0.73 -0.12, 12, 0.71 -0.28, 14, 0.34 -0.34, 14, 0.23 back 0.21, 17, 0.41 -0.06, 17, 0.82 0.05, 17, 0.84 -0.51, 12, 0.09 -0.51, 12, 0.09 0.24, 15, 0.40 0.18, 15, 0.51 rump -0.05, 17, 0.84 0.03, 17, 0.92 0.15, 17, 0.57 0.28, 12, 0.39 0.26, 12, 0.42 -0.14, 15, 0.60 -0.14, 15, 0.61 Blue chroma crown -0.05, 17, 0.85 0.08, 17, 0.75 -0.14, 17, 0.59 -0.11, 12, 0.72 -0.17, 12, 0.59 0.06, 15, 0.84 -0.70, 15, 0.004 nape 0.05, 16, 0.86 0.28, 16, 0.30 0.18, 16, 0.50 0.36, 12, 0.25 0.33, 12, 0.29 -0.35, 14, 0.22 -0.42, 14, 0.14 back 0.13, 17, 0.62 0.02, 17, 0.94 -0.03, 17, 0.90 0.03, 12, 0.93 0.06, 12, 0.86 -0.004, 15, 0.99 -0.35, 15, 0.21 rump 0.02, 17, 0.94 0.13, 17, 0.62 0.24, 17, 0.36 0.19, 12, 0.55 0.22, 12, 0.50 -0.08, 15, 0.79 -0.24, 15, 0.39 Green chroma crown -0.29, 17, 0.25 -0.28, 17, 0.28 0.18, 17, 0.49 -0.04, 12, 0.90 -0.03, 12, 0.93 0.27, 15, 0.32 0.23, 15, 0.41 nape 0.18, 16, 0.49 -0.45, 16, 0.08 0.19, 16, 0.48 -0.34, 12, 0.28 -0.29, 12, 0.35 0.22, 14, 0.44 0.32, 14, 0.26 back 0.16, 17, 0.53 -0.22, 17, 0.40 0.15, 17, 0.57 -0.29, 12, 0.37 -0.25, 12, 0.43 0.29, 15, 0.29 0.36, 15, 0.19 rump 0.08, 17, 0.76 0.06, 17, 0.83 0.36, 17, 0.16 0.42, 12, 0.17 0.39, 12, 0.21 -0.02, 15, 0.95 0.15, 15, 0.58

λRmid back -0.10, 17, 0.70 -0.29, 17, 0.26 0.04, 17, 0.87 -0.17, 12, 0.60 -0.15, 12, 0.65 0.09, 15, 0.74 0.34, 15, 0.22 rump -0.05, 17, 0.86 -0.19, 17, 0.46 0.06, 17, 0.81 -0.19, 12, 0.55 -0.24, 12, 0.46 0.05, 15, 0.86 0.46, 15, 0.08

55

Table A3. Relationships between reproductive performance and plumage colour in male

tree swallows. Spearman-rank correlations (rs, n, p) are shown; significant relationships are

highlighted in bold.

Reproductive Performance Colour Variable Body Region 1st Egg Date Clutch Size % Fledged

Rmax crown -0.05, 11, 0.87 0.32, 11, 0.33 -0.10, 8, 0.81 nape 0.12, 11, 0.73 -0.33, 11, 0.32 0.28, 8, 0.50 back 0.04, 11, 0.92 -0.25, 11, 0.46 0.50, 8, 0.21

rump -0.38, 11, 0.25 0.43, 11, 0.19 0.15, 8, 0.72 Reflectance amplitude crown -0.08, 11, 0.81 0.12, 11, 0.72 0.03, 8, 0.95 nape 0.10, 11, 0.77 -0.40, 11, 0.23 0.50, 8, 0.07 back 0.05, 11, 0.87 -0.23, 11, 0.49 0.46, 8, 0.25 rump -0.45, 11, 0.17 0.42, 11, 0.19 0.15, 8, 0.72 Blue chroma crown 0.45, 11, 0.17 -0.18, 11, 0.59 -0.27, 8, 0.52 nape 0.26, 11, 0.43 -0.23, 11, 0.49 -0.04, 8, 0.93 back -0.04, 11, 0.92 -0.22, 11, 0.51 0.37, 8, 0.37 rump 0.07, 11, 0.83 0.45, 11, 0.16 -0.38, 8, 0.35 Green chroma crown -0.21, 11, 0.54 0.05, 11, 0.89 0.57, 8, 0.14 nape -0.09, 11, 0.79 -0.69, 11, 0.02 0.87, 8, 0.01 back 0.11, 11, 0.75 -0.01, 11, 0.97 0.60, 8, 0.12 rump -0.35, 11, 0.28 0.22, 11, 0.51 0.41, 8, 0.31

λRmid back -0.05, 11, 0.89 0.17, 11, 0.61 -0.22, 8, 0.61 rump -0.25, 11, 0.46 -0.15, 11, 0.66 0.47, 8, 0.24

56

Table A4. Relationships between reproductive performance and plumage colour in female

tree swallows. Spearman-rank correlations (rs, n, p) are shown; significant relationships are

highlighted in bold.

Reproductive Performance Colour Variable Body Region 1st Egg Date Clutch Size % Fledged

Rmax crown 0.20, 11, 0.56 0.39, 11, 0.24 -0.65, 8, 0.08 nape -0.19, 11, 0.57 0.13, 11, 0.69 0.17, 8, 0.69 back -0.29, 11, 0.39 0.70, 11, 0.02 -0.14, 8, 0.74 rump -0.48, 11, 0.14 0.22, 11, 0.61 0.18, 8, 0.49 Reflectance amplitude crown 0.21, 11, 0.54 0.38, 11, 0.25 -0.65, 8, 0.08 nape -0.26, 11, 0.43 0.23, 11, 0.50 0.41, 8, 0.31 back -0.35, 11, 0.28 0.60, 11, 0.05 0.03, 8, 0.95 rump -0.44, 11, 0.18 0.28, 11, 0.40 0.22, 8, 0.61 Blue chroma crown 0.23, 11, 0.50 0.09, 11, 0.80 -0.28, 8, 0.27 nape 0.14, 11, 0.69 -0.17, 11, 0.61 0.22, 8, 0.61 back -0.10, 11, 0.77 0.23, 11, 0.49 -0.28, 8, 0.50 rump -0.31, 11, 0.36 0.26, 11, 0.44 0.26, 8, 0.54 Green chroma crown -0.19, 11, 0.57 0.63, 11, 0.04 -0.11, 8, 0.79 nape -0.59, 11, 0.06 0.23, 11, 0.49 0.63, 8, 0.10 back -0.65, 11, 0.03 0.34, 11, 0.31 0.55, 8, 0.16 rump -0.46, 11, 0.15 0.17, 11, 0.61 0.43, 8, 0.28

λRmid back -0.55, 11, 0.08 0.37, 11, 0.26 0.22, 8, 0.61 rump 0.03, 11, 0.94 0.08, 11, 0.82 -0.06, 8, 0.88