Untitled - Université de Sherbrooke

Comments

Transcription

Untitled - Université de Sherbrooke
INFLUENCE DE L’IRIDESCENCE SUR LE SUCCÈS REPRODUCTEUR DES MÂLES
HIRONDELLE BICOLORE (TACHYCINETA BICOLOR)
par
Sonia Van Wijk
mémoire présenté au Département de biologie en vue
de l’obtention du grade de maître ès sciences (M.Sc.)
FACULTÉ DES SCIENCES
UNIVERSITÉ DE SHERBROOKE
Sherbrooke, Québec, Canada, juillet 2015
Le 17 juillet 2015
le jury a accepté le mémoire de madame Sonia Van Wijk
dans sa version finale.
Membres du jury
Professeure Fanie Pelletier
Directrice de recherche
Département de biologie
Professeur Marc Bélisle
Évaluateur interne
Département de biologie
Professeur Dany Garant
Président-rapporteur
Département de biologie
SOMMAIRE
En élaborant la théorie de la sélection sexuelle, Darwin a levé le voile sur les causes de
l’évolution de traits extravagants plus souvent observés chez les mâles, comme la coloration
du plumage chez les oiseaux. Un des mécanismes proposés pour expliquer l’évolution des
couleurs vives chez les mâles est la fonction de signal de condition. Les mâles pourraient donc
être préférés par les femelles puisque celles-ci obtiendraient des bénéfices directs, comme un
meilleur territoire, ou des bénéfices indirects, tels que de bons gènes pour leurs oisillons. La
sélection sexuelle de l’iridescence, un mode de coloration répandu chez les oiseaux, est encore
peu documentée à ce jour. L’iridescence est un mécanisme de coloration directionnel, c’est-àdire qu’il occasionne un changement de brillance avec un mouvement de l’animal. Cette
propriété unique à l’iridescence amène une difficulté de quantification qui a longtemps limité
son étude. Le but de mon projet était donc, dans un premier temps, de fabriquer un appareil de
mesure répétable de l’iridescence déjà publié dans la littérature afin de le tester chez
l’Hirondelle bicolore (Tachycineta bicolor). Dans un deuxième temps, j’ai vérifié si
l’iridescence des mâles Hirondelle bicolore (Tachycineta bicolor) influençait leur succès
reproducteur.
Grâce à l’appareil de mesure, j’ai pu montrer qu’il y avait des différences individuelles dans la
directionnalité de l’iridescence. De plus, j’ai pu établir que ces différences étaient reliées au
nombre de jeunes dans le couple du mâle. Cette variable, pour la première fois associée au
succès reproducteur, pourrait jouer un rôle dans le signalement de la capacité à donner des
soins parentaux. De plus, les mâles brillants avaient un avantage pour les paternités hors
couple, mais seulement dans les sites à faible densité de reproducteurs. Une forte densité de
reproducteurs pourrait limiter le choix de la femelle, possiblement en raison de l’agressivité
des mâles ou du coût énergétique cumulatif de refuser des copulations hors couple.
Globalement, j’ai non seulement montré que l’iridescence est un signal à composantes
iii
multiples pouvant jouer un rôle dans les deux constituantes du succès reproducteur, mais aussi
que les conditions de compétition représentées par la densité de reproducteurs sont
importantes à considérer pour avoir un portrait plus précis des patrons de sélection. Mon projet
de recherche amène ainsi une meilleure compréhension de la sélection sexuelle en milieu
naturel.
Mots-clés : Hirondelle bicolore, Tachycineta bicolor, sélection sexuelle, coloration,
iridescence, aptitude phénotypique, directionnalité, répétabilité.
iv
REMERCIEMENTS
Je remercie tout d’abord ma superviseure Fanie Pelletier qui m’a soutenu tout au long de ma
maîtrise. Merci Fanie, tu m’as permis d’aller beaucoup plus loin que je ne l’aurais cru. Merci à
mes conseillers Marc Bélisle et Dany Garant qui m’ont donné chaque fois de judicieux
conseils.
Merci à toutes ces merveilleuses personnes qui ont travaillé sur le terrain avec moi en 2013 et
2014, et qui ont donc collecté ces précieuses plumes iridescentes! Aux étudiants gradués :
Audrey Bourret, Geneviève Coudé, Paméla Lagrange, Antoine Millet, Clarence Schmitt et
Audrey Turcotte. Aux assistants de terrain : Maxime Carrier, Marie-Pier Laplante, Stéfanie
Larocque-Desroches, Chloé Martineau, Hélène Presseault-Gauvin, Gaëlle Satre et Julie
Tremblay. Nous avons formé une équipe du tonnerre, en ayant pu s’amuser tout en gardant
notre sérieux afin de réaliser cet ambitieux projet de recherche.
C’est un travail colossal d’analyser les plumes au spectrophotomètre tout en étant dans le noir
avec un bruit de lampe à donner mal à la tête!!! Ce travail a été fait avec persévérance et brio
grâce à mes deux assistantes de laboratoire Myriam Cadotte et Sabrina Gignac-Brassard.
Merci à vous deux, vous avez été une partie essentielle au projet. Un merci particulier à Benoit
Couture, technicien en génie mécanique, qui a donné généreusement de son temps pour
m’aider à construire le fameux appareil de mesure. Merci aux étudiants du labo Pelletier :
Jacinthe Gosselin, Simon Guillemette, Martin Leclerc, Alexandre Martin, Gabriel Pigeon,
Clarence Schmitt et Geneviève Turgeon, pour votre solidarité et vos conseils statistiques!
Finalement, merci à mon amour Gabriel Droulers, ma famille et mes amis pour leur
inestimable soutien. Gab, tes conseils de physicien ont apporté une autre dimension à mon
projet!
Mon projet de recherche a bénéficié du support financier du CRSNG, du FRQNT, de la Chaire
de Recherche du Canada en Démographie Évolutive et Conservation, du CSBQ et de la
Faculté des sciences.
v
TABLE DES MATIÈRES
Sommaire .............................................................................................................................. iii
Remerciements ...................................................................................................................... v
Table des matières ................................................................................................................ vi
Liste des abréviations ........................................................................................................... ix
Liste des tableaux .................................................................................................................. x
Liste des figures .................................................................................................................... xi
Chapitre 1 Introduction générale ............................................................................................ 1
1.1 Fondements théoriques de la sélection sexuelle ............................................................ 2
1.1.1 Les reproductions hors couple ................................................................................ 4
1.1.2 La coloration pigmentaire....................................................................................... 6
1.1.3 L’iridescence .......................................................................................................... 6
1.2 Objectifs de recherche ................................................................................................. 10
1.2.1 Modèle d’étude : l’Hirondelle bicolore (Tachycineta bicolor) ............................ 10
1.2.2 Objectifs spécifiques ............................................................................................ 13
1.2.3 Importance du projet............................................................................................. 14
1.2.4 Population à l’étude .............................................................................................. 15
Chapitre 2 Une technique fiable pour quantifier la variabilité individuelle de
l’iridescence chez les oiseaux .................................................................................................. 17
2.1 Introduction de l’article ............................................................................................... 17
2.2 Abstract ....................................................................................................................... 18
2.3 Introduction ................................................................................................................. 19
2.4 Materials and Method.................................................................................................. 21
2.4.1 Data collection ...................................................................................................... 21
2.4.2 Apparatus .............................................................................................................. 22
2.4.3 Iridescence measurements .................................................................................... 22
2.4.4 Color analyses ...................................................................................................... 25
2.4.5 Statistical analyses ................................................................................................ 25
2.4.6 Ethical standards ................................................................................................... 26
vi
2.5 Results ......................................................................................................................... 27
2.6 Discussion.................................................................................................................... 31
2.6.1 Repeatability ......................................................................................................... 32
2.6.2 Potential bias caused by barbule tilt ..................................................................... 33
2.6.3 Limits to the method of measurement .................................................................. 34
2.7 Aknowledgements ....................................................................................................... 35
2.8 References ................................................................................................................... 36
ANNEXE .............................................................................................................................. 40
Chapitre 3 La densité de reproducteurs influence la sélection sexuelle de l’iridescence
chez les mâles Hirondelle bicolore, Tachycineta bicolor....................................................... 47
3.1 Introduction de l’article ............................................................................................... 47
3.2 Abstract ........................................................................................................................ 48
3.3 Introduction ................................................................................................................. 49
3.4 Materials and methods ................................................................................................. 52
3.4.1 Study system and data collection .......................................................................... 52
3.4.2 Iridescence measurements .................................................................................... 53
3.4.3 Genetic analyses ................................................................................................... 56
3.4.4 Statistical analyses ................................................................................................ 56
3.4.5 Ethical standards ................................................................................................... 58
3.5 Results ......................................................................................................................... 58
3.5.1 Patterns of paternity .............................................................................................. 58
3.5.2 Iridescence and reproductive success ................................................................... 58
3.5.3 Iridescence and %WPY in the nest ....................................................................... 63
3.6 Discussion.................................................................................................................... 64
3.6.1 Iridescence and reproductive success ................................................................... 65
3.6.2 Iridescence and %WPY in the nest ....................................................................... 67
3.6.3 Methodological considerations ............................................................................. 68
3.6.4 Conclusion ............................................................................................................ 69
3.7 Aknowledgements ....................................................................................................... 69
3.8 References ................................................................................................................... 70
ANNEXE .............................................................................................................................. 78
vii
Chapitre 4 Discussion générale et conclusion ....................................................................... 83
4.1 Retour sur les résultats ................................................................................................ 83
4.1.1 Mesure et répétabilité de l’iridescence ................................................................. 83
4.1.2 L’iridescence et le succès reproducteur des mâles ............................................... 86
4.2 Perspectives futures ..................................................................................................... 87
4.3 Conclusion ................................................................................................................... 93
BIBLIOGRAPHIE .............................................................................................................. 94
viii
LISTE DES ABRÉVIATIONS
CRSNG
Conseil de recherches en sciences naturelles et en génie du Canada
CSBQ
Centre de la science de la biodiversité du Québec
EPP
Extra-pair paternity
EPY
Extra-pair young
FRQNT
Fonds de recherche du Québec nature et technologies
WPY
Within-pair young
%WPY
Proportion of within-pair young
ix
LISTE DES TABLEAUX
2.1
Effect of body region and year on plumage iridescence of male tree swallows as
quantified by hue, saturation, brightness, and directionality. ....................................... 31
2.2
Parameters from Eqn 1 and 2 and their associated definitions ..................................... 43
2.3
Measurement repeatability of iridescence variables for crown, mantle and rump
feathers of male tree swallows. .................................................................................... 45
2.4
Individual repeatability of iridescence variables for crown, mantle, and rump
feathers, as well as for all feathers from these three regions in male tree swallows. ... 45
3.1
Effect of iridescence and other individual and environmental variables on a) the
annual reproductive success, b) the number of within-pair young (WPY) and c)
the number of extra-pair young (EPY) of male tree swallows (n=202) breeding in
southern Québec, Canada, 2013-2014. ......................................................................... 60
3.2
Effect of iridescence and other individual and environmental variables on the
proportion of within-pair young (%WPY) of male tree swallows (n=202) breeding
in southern Québec, Canada, 2013-2014. ..................................................................... 64
3.3
PC loadings of iridescence variables (hue, brightness, saturation and
directionality) from a principal component analysis performed separately on 2
body regions: crown and back. ..................................................................................... 78
3.4
Mean and standard deviation of iridescence and other individual and
environmental variables in the 202 male tree swallows used to model the
influence of iridescence on reproductive success. ........................................................ 79
x
LISTE DES FIGURES
1.1
Les deux grands types de coloration du plumage, la coloration pigmentaire et la
coloration iridescente. ..................................................................................................... 7
1.2
Changement de brillance chez le Colibri d’Anna (Calypte anna). ................................. 8
1.3
Conformation des barbules iridescentes et non-iridescentes chez le Canard colvert
(Anas platyrhynchos). ..................................................................................................... 9
1.4
Iridescence de l’Hirondelle bicolore. ............................................................................ 12
1.5
Couple d’Hirondelle bicolore. ...................................................................................... 13
1.6
Localisation des 40 fermes où sont installés les nichoirs du suivi de l’Hirondelle
bicolore, dans un gradient d’intensification agricole au Sud du Québec, Canada. ...... 16
2.1
Tree swallow rump feather. .......................................................................................... 23
2.2
Measurement repeatability of iridescence variables for crown (dark grey), mantle
(light grey), and rump feathers (white) of male tree swallows. .................................... 27
2.3
Individual repeatability of iridescence variables for crown (dark grey), mantle
(medium grey), and rump feathers (light grey), as well as for all feathers from
these three regions (white), in male tree swallows. ...................................................... 28
2.4
Histogram of mean rump feathers’ tilt across 271 male tree swallows (left y-axis,
grey bars). ..................................................................................................................... 29
2.5
Variation in iridescence between three body regions of male tree swallows as
quantified by a) hue, b) saturation, c) brightness, and d) directionality. ...................... 30
2.6
Apparatus for quantifying iridescence in bird feathers built according to Meadows
et al. (2011) ................................................................................................................... 41
2.7
Apparatus for quantifying iridescence in bird feathers built according to Meadows
et al. (2011) ................................................................................................................... 42
2.8
Correlation between brightness and three different directionality variables:
angular breadth, directionality as calculated by Madsen et al. (2007), and
directionality calculated similarly to Meadows et al. (2012). ...................................... 44
3.1
Correlation circles representing the association between iridescence variables of
two distinct body regions, namely a) the crown and b) the back of 202 male tree
swallows breeding in southern Québec, Canada, 2013-2014. ...................................... 55
3.2
Effect of back angular breadth on the number of WPY produced by male tree
swallows as predicted by the model shown in Table 3.1b. ........................................... 61
xi
3.3
Effect of the interaction between back brightness and breeding density on the
number of EPY produced by male tree swallows as predicted by the model shown
in Table 3.1c. ................................................................................................................ 62
3.4
Effect of the interaction between crown brightness and breeding density on the
annual reproductive success of male tree swallows as predicted by the model in
Table 3.1a. .................................................................................................................... 62
3.5
Variation in plumage iridescence variables measured on 202 male tree swallows
breeding in southern Québec, Canada, 2013-2014. ...................................................... 80
3.6
Correlation matrix of explanatory variables included in models explaining
reproductive success of 202 male tree swallows breeding in southern Québec,
Canada, 2013-2014. ...................................................................................................... 81
3.7
A social pair of tree swallows. ..................................................................................... 82
4.1
Contrôle de l’angle d’inclinaison des barbules. ........................................................... 85
4.2
Photographie des barbules de deux plumes iridescentes de croupion de mâle par
microscopie en réflexion. ............................................................................................. 90
4.3
Relation entre la couleur et le jour julien de capture. ................................................... 91
xii
Science is not only a disciple of reason but, also, one of romance and passion.
- Stephen Hawking
xiii
CHAPITRE 1
INTRODUCTION GÉNÉRALE
La sélection sexuelle, soit l’évolution des caractères augmentant l’acquisition de partenaires
sexuels, est proposée pour expliquer la présence de traits extravagants comme les bois des
grands herbivores ou la couleur vive du plumage des oiseaux (Darwin, 1871). Deux
mécanismes principaux peuvent expliquer cette évolution. Le premier se nomme le processus
d’emballement de Fisher, et nécessite le couplage génétique de la préférence de la femelle et
du trait du mâle (Andersson, 1994; Pomiankowski et al., 1991). Le deuxième se nomme le
principe du handicap, et nécessite que le trait soit coûteux à arborer (augmentation du risque
de prédation, coût de fabrication et d’entretien), donc seuls les mâles en bonne condition
peuvent se permettre d’arborer le trait (Isawa et al., 1991; Zahavi, 1975). Il s’ensuivrait une
évolution de la préférence des femelles puisqu’elles en retireraient des bénéfices directs (par
exemple : meilleurs soins parentaux, Hoelzer, 1989) ou indirects (par exemple : gènes
augmentant l’aptitude phénotypique des rejetons; Hamilton et Zuk, 1982; Neff et Pitcher,
2005; Weatherhead et Robertson, 1979).
La coloration produite par la déposition de pigments dans les plumes a été grandement étudiée
chez les oiseaux, et est principalement produite par deux pigments, soit les caroténoïdes (jaune
à rouge) et la mélanine (brun à noir) (McGraw, 2006). Une littérature grandissante suggère
que la coloration pigmentaire peut agir comme un signal de condition du mâle et que ceux
arborant une coloration plus vive seraient préférés par les femelles (Hill, 2006). L’iridescence
est un autre mécanisme de coloration du plumage répandu chez les oiseaux qui a la
particularité de provoquer un changement rapide de brillance avec un mouvement de l’animal
(Osorio et Ham, 2002). Cette propriété unique rend difficile sa quantification précise
(Meadows et al., 2011), ce qui explique que beaucoup moins d’études se sont attardées sur la
1
sélection sexuelle de l’iridescence. Dans ce contexte, le but de mon projet était tout d’abord de
mesurer l’iridescence des mâles Hirondelle bicolore (Tachycineta bicolor) de manière
répétable, et ensuite, de vérifier comment celle-ci influençait leur succès reproducteur.
1.1 Fondements théoriques de la sélection sexuelle
La théorie de l’évolution via le processus de sélection naturelle, développée par Darwin
(1859), explique la présence de plusieurs traits phénotypiques chez les individus d’une
population, par exemple la couleur des plumes d’un oiseau. Pour évoluer, la couleur des
plumes doit être variable au sein d’une population et comporter une base génétique. Si certains
individus subissent moins de prédation que les autres par leur plumage davantage cryptique
par exemple, ils auront éventuellement un succès reproducteur plus élevé (une meilleure
aptitude phénotypique) et les gènes associés à ce plumage augmenteront en fréquence dans la
génération suivante. Après quelques générations, on s‘attend à l’évolution du plumage
cryptique. Des propriétés autres que le camouflage telles que la thermorégulation, la protection
contre les rayons UV, ou la confusion des prédateurs, peuvent aussi avoir permis l’évolution
de la coloration par sélection naturelle (Bortolotti, 2006).
Dans le règne animal, un grand nombre d’espèces ont un dimorphisme sexuel de la coloration,
et de manière générale les mâles arborent une coloration plus vive que les femelles. Ce
phénomène a été expliqué pour la première fois par Darwin (1871) en proposant le principe de
la sélection sexuelle, un dérivé de la sélection naturelle. La sélection naturelle à elle seule ne
semblait pas pouvoir expliquer la présence d’ornements extravagants chez les mâles. Ces
ornements devraient désavantager les mâles en augmentant leur visibilité, donc la probabilité
qu’ils soient capturés par un prédateur par exemple. Un des facteurs à la base de la sélection
sexuelle est l’anisogamie, c’est-à-dire que les mâles produisent beaucoup de petits gamètes
2
peu coûteux et les femelles produisent peu de gros gamètes coûteux (Bateman, 1948; Trivers,
1972). Ainsi, les mâles compétitionnent entre eux pour pouvoir féconder le plus de femelles
puisque chaque fertilisation est peu coûteuse, tandis que les femelles, en ayant investi plus
d’énergie dans chaque gamète, ont tendance à être plus sélective sur leur(s) partenaire(s) de
reproduction. Cela favorise une plus grande variance du succès reproducteur entre les mâles
qu’entre les femelles (Bateman, 1948), et donc une plus grande opportunité pour la sélection
des traits du mâle favorisant l’accès aux femelles. Ces traits peuvent avantager le mâle via la
compétition avec les autres mâles, par exemple sous forme de combats (sélection intrasexuelle), ou via le choix des femelles (sélection inter-sexuelle) (Andersson, 1994).
La coloration vive d’un mâle peut donc avoir évolué sous sélection intra-sexuelle ou intersexuelle, deux mécanismes non mutuellement exclusifs. Par sélection intra-sexuelle, la
coloration du mâle serait ainsi une « arme psychologique », un badge signalant le haut taux de
succès aux combats aux autres mâles (Rohwer, 1982), ce qui permet d’éviter de subir les
conséquences négatives sur la survie des deux individus. La sélection inter-sexuelle, soit le
choix de la femelle pour la coloration vive des mâles, a reçu plus d’attention et d’appui dans la
littérature (Andersson, 1994; Hill, 2006). Les travaux de Fisher (1958) ont établi une base
solide à la théorie du choix des femelles par l’élaboration du processus d’emballement. Dans
celui-ci, certaines femelles ont une préférence pour les mâles colorés, qui peut être simplement
due à un biais sensoriel. L’emballement se fait par le couplage génétique de la coloration du
mâle et de la préférence des femelles pour ce trait, menant à une augmentation du trait sexuel
dans la population. La sélection se stabilise lorsque les coûts en termes de survie de la
coloration du mâle dépassent les bénéfices en succès reproducteur. Un avantage pour la survie
des mâles colorés peut accélérer le phénomène, sans être nécessaire pour que le processus de
sélection s’enclenche. Les théories qui ont suivi celle de Fisher concernent les « mécanismes
indicateurs », c’est-à-dire que la coloration indiquerait la condition du mâle en étant un trait
coûteux à produire ou en agissant comme un handicap (Hamilton et Zuk, 1982; Kodric-Brown
et Brown, 1984; Zahavi, 1975). Si la coloration d’un mâle est un signal de sa qualité en termes
3
de ressources (territoire, soins parentaux), le choix de la femelle peut évoluer étant donné les
bénéfices directs pour celle-ci, soit en termes de survie ou de fécondité (Hoelzer, 1989). Si la
coloration signale la qualité génétique du mâle, le choix d’une femelle peut évoluer étant
donné les bénéfices indirects pour celle-ci, soit l’augmentation de la capacité de survie et/ou
de reproduction de ses jeunes (Hamilton et Zuk, 1982; Neff et Pitcher, 2005, Weatherhead et
Robertson, 1979).
1.1.1 Les reproductions hors couple
Les reproductions hors couple sont une composante du succès reproducteur importante à
considérer pour comprendre la sélection sexuelle chez les oiseaux. Environ 90 % des espèces
d’oiseaux montrent une monogamie sociale (Lack, 1968). Les reproductions hors couple sont
observées chez environ 90 % d’entre elles (Griffith et al., 2002). Obtenir des reproductions
hors couple serait plus bénéfique pour le mâle puisque cela peut grandement augmenter son
succès reproducteur, comparativement à la femelle qui est limitée par le nombre d’œufs
qu’elle peut produire. Ainsi, le comportement de sollicitation des reproductions hors couple
chez la femelle peut être potentiellement expliqué par une corrélation génétique entre les
sexes, puisque ce trait est bénéfique pour le mâle (Forstmeier et al., 2011). Par contre, sachant
que ce comportement peut occasionner des coûts non-négligeables comme le risque de
parasitisme et la perte des soins de la part du mâle social (i.e. : celui donnant les soins aux
jeunes) (Able, 1996; Westneat et al., 1990), le gain de bénéfices via le choix de la femelle
pour l’identité de ses partenaires hors couple est tout à fait possible. En général, les femelles
ne retirent des reproductions hors couple que des bénéfices indirects liés aux caractéristiques
génétiques des mâles. Voici plusieurs de ces bénéfices pouvant être recherchés par la femelle
(voir Griffith et al., (2002) et Jennions et Petrie (2000) pour des revues) :
1. De bons gènes qui augmentent la viabilité des jeunes (Hamilton, 1990; Møller, 1988);
4
2. La compatibilité génétique avec le mâle qui augmente l’hétérozygotie du jeune
(Kempenaers et al., 1999; Stapleton et al., 2007);
3. La diversité génétique à l’intérieur de sa nichée (Kempenaers et al., 1999; Westneat et
al., 1990) ;
4. La prévention de l’infertilité du mâle social (Wetton et Parkin, 1991).
Une des explications de l’existence de la coloration vive des mâles chez les oiseaux est la
sélexion sexuelle par l’entremise des reproductions hors couple. En effet, Møller et Birkhead
(1994) ont montré qu’à l’échelle interspécifique, plus le taux de reproductions hors couple est
élevé, plus les mâles sont colorés. Un mécanisme possible derrière ceci est que les femelles
ont une préférence pour les mâles plus colorés, puisque cela représenterait leur qualité
génétique et amènerait un bénéfice indirect à la femelle via l’augmentation de la viabilité des
jeunes (Hamilton et Zuk, 1982; Neff et Pitcher, 2005, Weatherhead et Robertson, 1979). De
cette façon, les femelles pourraient préférer les mâles colorés à la fois lors du choix du mâle
social et du mâle hors couple. Dans ce cas, cela engendrerait une grande variance du succès
reproducteurs entre les mâles peu et très colorés (Webster et al., 1995). On a ainsi tous les
éléments nécessaires à la sélection sexuelle de la coloration vive du plumage (Andersson,
1994). À ce jour, le mécanisme par lequel la coloration amènerait des bénéfices indirects à la
femelle reste controversé, étant donné le paradoxe du lek (Kotiaho et al., 2008). En effet, le
choix des femelles devrait éroder la variance génétique de coloration entre les mâles, pourtant
cette variance est nécessaire pour que le choix de la femelle lui amène des bénéfices. Un des
mécanismes proposé pour résoudre ce paradoxe est l’existence de variance environnementale
cyclique, qui amène un changement assez lent des meilleurs génotypes pour maintenir le
bénéfice génétique aux oisillons (Fitzpatrick, 1994; Hamilton et Zuk, 1982).
5
1.1.2 La coloration pigmentaire
La coloration des plumes peut être le résultat de deux phénomènes non mutuellement
exclusifs : la coloration pigmentaire et la coloration structurale (Prum, 2006). La coloration
pigmentaire résulte de l’absorption différentielle des longueurs d’onde de la lumière selon la
structure moléculaire du pigment et sa concentration (Andersson et Prager, 2006). Les
pigments les plus fréquents chez les oiseaux sont les caroténoïdes (couleurs jaune à rouge) et
la mélanine (couleurs brun à noir). Les caroténoïdes sont obtenus directement de
l’alimentation, tandis que la mélanine est synthétisée par l’organisme (Andersson et Prager,
2006). La coloration est caractérisée par les trois composantes universelles: la couleur (ou
teinte), la brillance (quantité de lumière réfléchie) et la saturation (pureté de la couleur)
(Hailman, 1977). Une littérature grandissante suggère que la coloration pigmentaire peut agir
comme un signal de condition du mâle et que ceux arborant une coloration plus vive (souvent
plus brillante ou plus saturée) seraient préférés par les femelles (Hill, 2006). Par contre, une
méta-analyse remet en doute la force du lien entre la coloration et la condition du mâle étant
donné un fort biais de publication (Griffith et al., 2006).
1.1.3 L’iridescence
L’iridescence est une coloration dite structurale puisqu’elle est dépendante de l’arrangement
spatial de différentes composantes dans la microstructure des barbules de la plume (les
barbules étant un microfilament d’une plume). Elle est produite par l’interférence des
longueurs d’onde lorsqu’elles traversent des couches uniformes de matériaux (la kératine, la
mélanine et parfois de l’air) ayant différents indices de réfraction (Prum, 2006). Par exemple,
si la couleur observée est bleue, il s’est alors créé une interférence constructive pour la
longueur d’onde bleue. Cela signifie que les multiples réflexions de l’onde par les différents
matériaux restent en phase les unes avec les autres (Figure 1.1). Ce qui détermine la couleur,
6
la brillance et la saturation est une combinaison de l’épaisseur traversée dans chaque structure
par la lumière, des différents indices de réfraction en jeu, ainsi que de la distribution spatiale
des structures (Prum, 2006). Ce type de coloration implique que les structures agissent comme
des miroirs (angle d’incidence = angle de réflexion), occasionnant ainsi un changement de
brillance lorsque l’animal est en mouvement (Figure 1.2; Osorio et Ham, 2002). Ainsi,
l’iridescence peut non seulement être caractérisée par les trois composantes universelles de
couleur, brillance et saturation, mais aussi par le changement de brillance.
Figure 1.1 Les deux grands types de coloration du plumage, la coloration pigmentaire et la
coloration iridescente. a) coloration pigmentaire. Un pigment fictif bleu, où la longueur d’onde
bleue est réémise dans toutes les directions b) coloration iridescente bleue, où la longueur
d’onde bleue interfère constructivement et la longueur d’onde rouge interfère destructivement
lors du passage dans une couche de mélanine superposée d’une couche de kératine.
7
Figure 1.2 Changement de brillance chez le Colibri d’Anna (Calypte anna). a) Colibri face à
l’observateur montrant une grande brillance de la zone iridescente, b) le même colibri ayant
tourné la tête montrant une brillance moindre de la zone iridescente. Tiré de (Doucet et
Meadows, 2009).
Les différentes composantes de l’iridescence n’auraient pas toutes le même potentiel de
signalement de la condition du mâle. C’est la brillance et la saturation qui seraient plus
fortement affectées par une diminution de la condition de l’individu (Prum, 2006). Des études
expérimentales ayant infecté des individus à la coccidiose (Hill et al., 2005) ou ayant altéré la
qualité de leur nourriture (McGraw et al., 2002; Meadows et al., 2012) ont constaté qu’après la
mue, ces individus avaient un plumage iridescent plus terne. Des études corrélationnelles ont
aussi montré que les individus ayant une croissance plus faible des plumes durant la mue
avaient un plumage plus terne (Doucet, 2002; Doucet et Montgomerie, 2003). Ainsi, ces
évidences suggèrent qu’une bonne condition lors de la mue serait requise afin de produire les
couches uniformes de la microstructure de la plume nécessaires pour la production de la
brillance de l’iridescence (McGraw, 2006). Les barbules iridescentes sont recourbées par
rapport aux barbules non iridescentes (Figure 1.3). Cette courbure sert probablement à diriger
la lumière vers le destinataire (Doucet et Meadows, 2009; Osorio et Ham, 2002). Toutefois,
elle rend possiblement les plumes iridescentes plus vulnérables à l’abrasion par les bactéries
dégradant la kératine par exemple (Fitzpatrick, 1998; Gunderson et al., 2009; Shawkey et al.,
2007). Cela engendrait aussi un handicap en diminuant l’efficacité d’auto-nettoyage et
l’hydrophobicité des plumes et, par conséquent, pourrait affecter la thermorégulation (Figure
8
1.3; Eliason et Shawkey, 2011). L’iridescence peut ainsi amplifier les différences de condition
entre individus puisqu’elle est sensible au parasitisme et au niveau d’entretien des plumes
(Leclaire et al., 2014), rendant l’évaluation de la condition du mâle plus facile par les femelles
(Fitzpatrick, 1998).
Figure 1.3 Conformation des barbules iridescentes et non-iridescentes chez le Canard colvert
(Anas platyrhynchos). a) Photographie de la plume. Photographie au microscope électronique
à balayage de b) la partie non iridescente, dans laquelle les barbules sont positionnées
perpendiculairement au plan de la plume, et de c) la partie iridescente dans laquelle les
barbules sont positionnées parallèlement au plan de la plume. Dans les encadrés : dessin d’une
goutte d’eau sur les barbules illustrant la différence d’hydrophobicité entre les barbules
iridescentes et non-iridescentes. Échelle : 100 µm. Tiré de (Eliason et Shawkey, 2011).
Les propriétés directionnelles de l’iridescence amènent une difficulté de quantification qui est
à l’origine du faible nombre d’études portant sur ce sujet comparativement à la multitude
d’études sur la coloration pigmentaire (Hill, 2006). De plus, il a été suggéré que les femelles
peuvent faire un choix basé sur la variation naturelle du plumage iridescent. Il a été récemment
découvert que l’iridescence peut influencer le succès copulatoire des mâles. Plus
9
spécifiquement, chez le Paon (Pavo cristatus), les mâles avantagés sont ceux ayant des ocelles
de la queue les plus bleues, les plus brillantes, et démontrant un plus grand changement de
couleur lors d’un mouvement (Dakin et Montgomerie, 2013; Loyau et al., 2007). Chez le
Jardinier satiné (Ptilonorhynchus violaceus), les mâles arborant une coloration bleue plus
saturée sont avantagés (Savard et al., 2011). Aussi, chez l’Hirondelle bicolore, la brillance du
mâle augmente sa probabilité d’obtenir des reproductions hors couple (Bitton et al., 2007).
Enfin, chez l’Étourneau sansonnet (Sturnus vulgaris), la coloration du plumage est reliée à la
polygynie du mâle et à sa condition, par contre il y aurait une forte corrélation avec l’âge
(Komdeur et al., 2005). Ces études montrent que certaines composantes de l’iridescence
peuvent avoir évolué en réponse aux préférences des femelles. Elles n’ont cependant pas
exploré la directionnalité de l’iridescence, qui est pourtant une propriété unique à ce type de
coloration. Il a été suggéré qu’il pourrait y avoir des différences individuelles en
directionnalité : une forte directionnalité (un changement rapide de brillance) pourrait refléter
l’uniformité de la microstructure de la plume (Madsen et al., 2007; Meadows et al., 2012), et
de cette manière refléter la condition. Cette caractéristique pourrait donc véhiculer de
l’information utilisée par la femelle lors de son choix de partenaire.
1.2 Objectifs de recherche
1.2.1 Modèle d’étude : l’Hirondelle bicolore (Tachycineta bicolor)
L’Hirondelle bicolore est un passereau migrateur nichant au Canada et au nord des États-Unis
et hivernant sur les côtes de la Floride et du Mexique. Étant un insectivore aérien, ses habitats
de prédilection sont les champs et les marais (Winkler et al., 2011). C’est un nicheur cavicole
secondaire : il utilise des trous d’arbres morts faits par d’autres oiseaux et s’accommode bien
des nichoirs artificiels (Robertson et al., 1992). La rareté des sites de nidification a modelé le
comportement de reproduction de l’Hirondelle. Par exemple, lors de la migration, le mâle
10
arrive avant la femelle pour défendre un territoire auquel il est très fidèle, davantage que la
femelle (Winkler et al., 2004). Plus souvent, les individus ne forment pas le même couple
année après année (Llambıas et al., 2008). Ils vivent en moyenne 2,7 ans, certains atteignant
les 8 ans (Butler, 1988).
L’hirondelle bicolore, socialement monogame, possède un des plus grands taux de
reproduction hors couple observé chez les oiseaux (Barber et al., 1996; Griffith et al., 2002).
Entre 76 et 89 % des nichées comportent au moins un jeune hors couple, et entre 35 et 52 %
des jeunes sont issus de reproductions hors couple (Kempenaers et al., 2001; Lessard et al.,
2014; O’Brien et Dawson, 2007; Stapleton et al., 2007; Whittingham et Dunn, 2001). Un
grand nombre d’études se sont attardées aux causes de ce taux si élevé de reproductions hors
couple (Dunn et al., 1994, 2009). La femelle semble avoir un grand contrôle de son choix de
partenaire pour ses reproductions hors couple, et ce pour plusieurs raisons. Tout d’abord, la
rareté des sites de nidification naturels amène une défense constante du site par un des
membres du couple y nichant (Robertson et al., 1992). Lorsque la femelle quitte le site pour
s’alimenter par exemple, le mâle ne la suit pas et ne peut donc pas l’empêcher de faire des
reproductions hors couple (Leffelaar et Robertson, 1984). Ensuite, elle peut non seulement
solliciter les reproductions hors couple, mais aussi refuser une tentative de copulation d’un
mâle à faible coût simplement en abaissant la queue, en s’envolant ou en tournant la tête et en
ouvrant son bec, une interaction agonistique qui décourage le mâle (Venier et al., 1993). Elle
semble aussi être libre de choisir le nombre de reproductions hors couple qu’elle désire,
puisque des évidences montrent qu’elle n’a pas de compromis à faire entre le nombre de
reproductions hors couple et la quantité de soins que le mâle procurera aux jeunes de la
nichée. En effet, peu importe si le mâle a accès à sa femelle durant la période fertile ou non, le
mâle social ne déserte pas (Whittingham et al., 1993) et donne autant de soins aux jeunes
(Kempenaers et al., 1998; Lifjeld et al., 1993).
11
Les mâles d’un an et les femelles de deux ans arborent un plumage iridescent bleu-vert
métallique sur le dessus du corps (tête et dos). L’iridescence de l’Hirondelle bicolore résulte
d’une structure parmi les plus simples, soit une couche de mélanosomes superposée d’une
couche de kératine (Figure 1.4 a et b, Eliason et Shawkey, 2010). L’iridescence du plumage
est directionnelle, ainsi on observe un changement de brillance en changeant l’angle de la
plume jusqu’à ce qu’elle devienne noire et matte, résultant de l’absorption presque totale de la
lumière par les mélanosomes sous-jacents (Figure 1.4 c et d). Les femelles d’un an, matures
sexuellement, arborent un plumage brunâtre avec une quantité variable d’iridescence
(généralement moins de 50 % de leur surface dorsale), il s’agit du phénomène de maturation
retardée du plumage (Hussel, 1983; Lozano et Handford, 1995). On note un dimorphisme
sexuel intermédiaire selon un classement issu de 73 espèces d’oiseaux (Owens et Hartley,
1998), la femelle présentant un plumage plus vert et terne que le mâle (Figure 1.5; Robertson
et al. 1992).
Figure 1.4 Iridescence de l’Hirondelle bicolore. a) Plume de croupion b) coupe transversale
d’une barbule vue au microscope électronique à transmission montrant en noir les granules de
mélanine et en gris la matrice de kératine. Échelle = 500 nm. Les lettres représentent d) la
surface dorsale et v) la surface ventrale. Tiré de Eliason et Shawkey (2010). c) plume de
croupion vue perpendiculairement, d) plume de croupion vue à angle rasant. Crédit photo :
Sonia Van Wijk.
12
Figure 1.5 Couple d’ Hirondelle bicolore. Le mâle est à gauche, la femelle à droite. Crédit
photo: Paméla Lagrange.
1.2.2 Objectifs spécifiques
La directionnalité de l’iridescence, soit le changement de brillance avec un mouvement de
l’animal, rend difficile la quantification de l’iridescence de manière répétable. Cela est une des
causes de la rareté des études de sélection sexuelle de ce type de coloration. Les deux objectifs
spécifiques de mon projet étaient donc :
1) de vérifier si un appareil de mesure ayant été démontré répétable chez le colibri
d’Anna (Meadows et al., 2011) l’est aussi chez l’Hirondelle bicolore, en plus de
vérifier l’existence de différences individuelles en directionnalité;
2) de vérifier comment les multiples composantes de l’iridescence influencent le succès
reproducteur des mâles Hirondelle bicolore, via 4 mesures du succès reproducteur : le
succès reproducteur annuel, le nombre de jeunes dans le couple, le nombre de jeunes
hors couple, et la proportion de jeunes dans le couple de la nichée.
13
La littérature indique qu’en général les mâles plus bleus, brillants et saturés ont une meilleure
condition et/ou sont préférés par les femelles (Balenger et al., 2008; Bitton et al., 2007; Hill et
al., 2005; Keyser et Hill, 2000; Zirpoli et al., 2013). De plus, les femelles Hirondelle bicolore
ont un fort contrôle des paternités hors couple (Venier et al., 1993). Ainsi, mon hypothèse est
que les femelles appariées avec des mâles peu ornementés (plus verts, moins brillants et moins
saturés) devraient accepter de subir les coûts possibles de faire des reproductions hors couple
afin d’obtenir les bons gènes via les mâles hors couple très ornementés (Jennions et Petrie,
2000). Ces coûts peuvent inclure les coûts de recherche (Dunn et Whittingham 2007; Milinski
et Bakker, 1992), l’augmentation de la probabilité de parasitisme (Westneat et al., 1990; Able,
1996), ou encore l’augmentation de la probabilité de subir la prédation (Hedrick et Dill, 1993).
Les femelles appariées avec des mâles très ornementés ne risqueraient pas de payer le coût des
reproductions hors couple puisqu’elles ont déjà les bons gènes de leur mâle social. La
première prédiction est donc que les mâles plus ornementés auraient un meilleur succès
reproducteur à la fois dans le couple et hors couple. La deuxième prédiction est que l’effet de
l’iridescence du mâle serait plus prononcé à forte versus faible densité de reproducteurs
puisque la femelle auraient une meilleure capacité ou un coût moindre à effectuer un choix
(Dunn et Whittingham, 2007; Taff et al., 2013).
1.2.3 Importance du projet
L’importance de cette étude est multiple. Tout d’abord, une seule étude a été faite sur le lien
entre l’iridescence et les reproductions hors couple chez l’Hirondelle bicolore (Bitton et al.,
2007). Les auteurs ont mesuré un ensemble de plumes superposées pour imiter le réel
assemblage des plumes sur l’oiseau. Par contre, cette méthode de mesure est peu précise selon
l’étude de Meadows et al., (2011). Ainsi, ma contribution réside dans une quantification
potentiellement plus précise de l’iridescence qui permet une meilleure compréhension de la
sélection sexuelle chez cette espèce. De plus, comme la directionnalité de l’iridescence a le
14
potentiel de véhiculer une information différente à propos de l’individu, explorer les liens
entre la directionnalité et le succès reproducteur est importante afin de mieux comprendre
l’évolution de l’iridescence. Finalement, le système d’étude à large échelle présente des
habitats à densités de reproducteurs variées. Sachant que les conditions de compétition
peuvent influencer la sélection sexuelle (Kokko et Rankin, 2006), étudier l’interaction entre la
densité de reproducteurs et l’iridescence permet d’obtenir un portrait plus exhaustif de la
sélection sexuelle.
1.2.4 Population à l’étude
Afin de répondre aux objectifs fixés, j’ai utilisé un suivi à long terme de la reproduction de
l’Hirondelle bicolore dans le Sud du Québec, Canada. Le système comprend 400 nichoirs
répartis dans 40 fermes, à raison de 10 nichoirs par ferme. L’aire d’étude d’environ 10 200
km2 est située dans un gradient d’intensification agricole, le but initial de ce suivi étant
d’étudier l’impact de l’agriculture sur ce passereau, en déclin d’environ 4 % par année au
Québec depuis 1970 (Sauer et al., 2014). À l’est, le paysage est composé d’une mosaïque de
forêts et de cultures extensives (pâturages, champs de foin) servant pour la production laitière.
À l’ouest, les cultures intensives (maïs, soya, céréales) sont davantage présentes avec un
couvert forestier moindre (Figure 1.6). Durant les années 2013 et 2014, nous avons récolté des
plumes iridescentes des mâles nichant dans le système d’étude. Le suivi de la reproduction et
l’attribution d’un statut en couple ou hors couple des oisillons grâce aux analyses génétiques a
permis de vérifier la relation entre l’iridescence des mâles et leur succès reproducteur.
15
Figure 1.6 Localisation des 40 fermes où sont installés les nichoirs du suivi de l’Hirondelle
bicolore, dans un gradient d’intensification agricole au Sud du Québec, Canada. Image issue
de LANDSAT-TM, avec coordonnées UTM zone 18, NAD83. Blanc : cultures intensives, gris
pâle : cultures extensives , gris foncé : forêt, noir : eau.
16
CHAPITRE 2
UNE TECHNIQUE FIABLE POUR QUANTIFIER LA VARIABILITÉ
INDIVIDUELLE DE L’IRIDESCENCE CHEZ LES OISEAUX
2.1 Introduction de l’article
L’objectif principal de cet article était d’améliorer la technique de mesure de l’iridescence
chez les oiseaux. Pour ce faire, j’ai construit un appareil de mesure publié dans la littérature
permettant une mesure répétable de l’iridescence, et j’ai testé celui-ci sur des mâles Hirondelle
bicolore. La contribution scientifique de cette étude comprend d’une part la publication du
plan détaillé pour la construction de l’appareil, qui le rend accessible et permet de répandre
son utilisation. D’autre part, j’ai quantifié avec l’appareil une propriété inhérente à
l’iridescence, soit le changement de brillance occasionné par un mouvement de l’animal. La
découverte de différences individuelles au niveau de cette propriété de l’iridescence a permis
de promouvoir la quantification de cette propriété pouvant aussi être sous sélection sexuelle.
Ma contribution à cet article est majeure puisque j’ai planifié la construction de l’appareil de
mesure et supervisé les deux stagiaires ayant effectué les mesures spectrométriques de plumes.
De plus, j’ai effectué les analyses statistiques, écrit l’article et coordonné le processus de
publication. Ma superviseure, la Pre. Fanie Pelletier, m’a encadrée tout au long du projet,
tandis que mes conseillers, Marc Bélisle et Dany Garant, ont judicieusement apporté des
commentaires et suggestions au projet ainsi qu’aux versions préliminaires de ce manuscrit. Le
manuscrit est en révision majeure au Journal of Avian Biology (13 mai 2015).
17
A RELIABLE TECHNIQUE TO QUANTIFY THE INDIVIDUAL VARIABILITY OF
IRIDESCENCE IN BIRDS
Par
Sonia Van Wijk, Marc Bélisle, Dany Garant et Fanie Pelletier
2.2 Abstract
The study of iridescence in birds emerged only recently, mainly due to the difficulty inherent
in quantifying directionality. Directionality restrains color perception to a limited angle and
thereby causes drastic changes in brightness when an animal is in motion. Although a versatile
technique for quantifying iridescence has been developed recently (Meadows et al. 2011), so
far, it has only been applied to measuring bright red iridescence in one species. Thus, the
reliability of the method for species displaying other iridescent coloration, as well as for
directionality, has yet to be assessed. Additionally, two important methodological aspects
remain to be assessed before this method can be used confidently: 1) whether directionality,
which could be subject to sexual selection, can be quantified in a repeatable way; and 2)
whether individual variability in barbule tilt at the surface of feathers can bias iridescence
measurements. Using feathers collected from 271 male tree swallows (Tachycineta bicolor)
over two years, we found that the method provided repeatable measurements to quantify
variability in directionality across individuals and across three body regions, namely the
crown, mantle and rump. The method was also efficient in controlling individual differences
in barbule tilt and reducing measurement bias associated with this phenomenon. We strongly
encourage researchers to quantify directionality of iridescence in birds to unveil its role in
signaling and sexual selection.
18
2.3 Introduction
Since Endler’s seminal book on natural selection (1986), a vast literature has been established
on the effect of body condition on coloration by carotenoids and melanin and on the
importance of those traits for sexual selection (Griffith et al. 2006, Meunier et al. 2010,
Rahman et al. 2013). Recently, there has also been a growing interest in studying iridescent
coloration, in part because this trait shows potential for condition-dependent expression and
signaling (McGraw et al. 2002, Doucet and Montgomerie 2003, Hill et al. 2005, Kemp and
Rutowski 2007, Lim and Li 2013). There is also evidence that iridescence is a sexuallyselected trait in males because it correlates with both their mating (Papke et al. 2007, Loyau et
al. 2007, Savard et al. 2011, Dakin and Montgomerie 2013) and reproductive success (Bitton
et al. 2007).
In birds, iridescence is produced by the alternation of two or more layers of different materials
(melanin, keratin and, in some cases, air) in the feather microstructure, causing the
constructive interference of certain wavelengths that produces the observed color pattern
(Prum 2006). A unique property of iridescence is directionality, caused by the specularity of
color production. Directionality restrains color perception to a limited angle and thereby
drastic changes in brightness can be observed when an animal is in motion (Osorio and Ham
2002, Meadows et al. 2011). This property could convey information about an individual’s
condition and be used as a sexual signal. In fact, it is suggested that plumage exhibiting greater
directionality has a more ordered microstructure (Madsen et al. 2007, Meadows et al. 2012),
and, therefore, can be indicative of better condition. Moreover, directionality is involved in the
nuptial parade of some species. For instance, in Anna’s hummingbirds (Calypte anna), the
male courtship display involves the movement of erected crown and gorget feathers causing
intense flashes towards the female (Stiles 1982). While studies on iridescent coloration in
birds rarely quantify directionality (For example: Bitton et al. 2007, Leclaire et al. 2014), it is
19
a common practice in butterfly studies (Vukusic et al. 2002, Kemp et al. 2006, 2008).
Recently, Meadows et al. (2011) proposed a new method for quantifying the directionality of
feathers. They designed an apparatus containing illumination and reception arms that can
freely rotate around the feather, mimicking natural conditions in nature for different species.
They obtained repeatable measurements of the bright red iridescence of Anna’s hummingbird
feathers with respect to the main dimensions of color: hue, saturation and brightness (Hailman
1977). However, the reliability of the method for species displaying other iridescent
coloration, as well as for directionality, has not been assessed. Considering that the
directionality of iridescence has the potential to convey important information for sexual
selection, it is imperative to assess if it can be measured in a repeatable way, as well as
quantify its level of individual variation.
Apart from directionality, another distinct feature of iridescence, the barbule tilt, may need to
be considered to reduce measurement biases. Iridescence is produced at the surface of
feathers’ barbules. Barbules are not always oriented parallel to the plane of the vane
depending on species (Osorio and Ham 2002). Within species, the tilt of barbules relative to
the plane of the vane could also vary among individuals, as it is the case with the lamellae (the
equivalent for barbules) of the butterfly Eurema hecabe (Kemp 2008). Controlling for the
angle of this tilt is therefore a standard procedure in butterfly studies to avoid measurement
biases (Kemp 2006, 2008). To our knowledge, no study has assessed the importance of this
potential problem in bird studies.
The main goal of this study is to find a reliable method to quantify iridescence in the tree
swallow (Tachycineta bicolor), which displays a blue-green iridescence from its forehead to
its rump. To do so, we first assessed whether the method described in Meadows et al. (2011)
yields repeatable measurements of tree swallow iridescence. The microstructure causing
iridescence in tree swallows is one of the simplest: a layer of keratin over a layer of melanin
20
(Eliason and Shawkey 2010). It is very different from hummingbird feather microstructure
where multiple alternating layers of melanin and keratin with hollow melanosomes enable
extreme brightness (Greenewalt et al. 1960). Consequently, tree swallow iridescence was
deemed appropriate for assessing whether Meadows' et al. (2011) method can be used in a
variety of species. Then, we quantified the precision of a measure of directionality, as well as
the range of individual differences, a fundamental prerequisite for its use in sexual selection
studies. Finally, we assessed the extent of individual differences in barbule tilt to evaluate
whether controlling for this variation can reduce measurement bias.
2.4 Materials and Method
2.4.1 Data collection
We captured tree swallows in a study system comprising 400 nest boxes located in southern
Québec, Canada (see Ghilain and Bélisle (2008) for a detailed description of the system).
Iridescent feathers were sampled on male tree swallows captured during chick provisioning in
2013 (N=141) and 2014 (N=187). As 57 males were caught in both years, data from one of the
two years was randomly removed for these individuals to avoid pseudoreplication. The sample
size used for our analyses was thus 271 individuals. Forceps were used to collect iridescent
feathers from three body regions: crown, mantle and rump. Care was taken to pluck at the
feather’s base and to avoid contact with the iridescent region. Three feathers were plucked at
the center of each of the three regions, for a total of nine feathers per male. Feathers were
stored at room temperature in manila envelopes until spectral analysis.
21
2.4.2 Apparatus
Iridescence was assessed for each feather using an apparatus built following Meadows et al.
(2011). Briefly, the apparatus consisted in two arms and a translational stage (PT1, Thorlabs,
Newton, NJ, USA) that could all rotate around a single axis. A collimating lens (COL-UV/VIS
from Avantes) was fixed on each arm with an optical fiber that either joined a lamp source
(200 μm fiber) or a spectrometer (100 μm fiber), respectively. Spectral measurements were
performed with an Avaspec-2048 spectrometer coupled with an Avalight-XE Pulsed Xenon
Lamp relative to a WS-2 standard (Avantes Inc., Broomfield, CO, USA). A detailed plan of
the apparatus and specifications related to the machining are available in the Supplementary
Material Figure 2.6 andFigure 2.7. We used a 20° angle between the light arm and the
spectrometer arm. Using a small angle such as this avoids producing large light ellipses on the
stage that might reduce repeatability due to the difficulty of perfectly overlapping the light and
spectrometer beams at all points when rotating the stage to quantify directionality (see below).
The alignment of the two beams was confirmed before and after the analysis of all feathers
from a region by shining a laser pointer down the spectrometer fiber. Feathers were mounted
on the translational stage which was then adjusted in height so that the surface of the feather
reached the center of rotation.
2.4.3 Iridescence measurements
All measurements were performed in a dark room. The light source was turned on 15 minutes
before starting measurements to stabilize it and a blank was performed every 15 minutes
during measurements to control for lamp drift. For each spectrum, the average of five scans at
25 pulses per 250 ms integration time, with a boxcar smoothing function of 10 pixels, was
taken using the software Avasoft 8 (Avantes). Spectra were expressed as percent reflectance of
light relative to the white standard. Measurements were made blind to the identity of the
22
individuals and feather order was randomized. Individual feathers were put on the matte black
art-quality cardboard covering the stage so that the 2 mm diameter beam from the light source
was located at the center of the iridescent region (Figure 2.1). The feather was held flat by a
piece of black cardboard with a 7 mm aperture to control for variation in the feather’s
curvature. The stage was then rotated until maximal reflectance was reached, as determined by
evaluating the spectra on the software, in order to control for barbule tilt. The degree of tilt
was taken from the stage’s protractor. The degree of tilt was 0° when the stage was not rotated
(parallel to the base), positive if rotated clockwise, and negative if rotated counter-clockwise,
as in Osorio and Ham (2002). Reflectance spectra were binned in 1 nm increments from 300
to 700 nm which corresponds to the visual range of birds (Hart 2001). The quantification
method was modified for crown feathers, considering their smaller size: 1 mm diameter beam
and 15 pulses per 150 ms integration time.
Figure 2.1 a) Tree swallow rump feather. The black spot corresponds to the area measured. b)
Typical spectrum from a male Tree swallow rump feather (Bird ID 2591-96205) describing
three color variables used in the study. H1: Hue, wavelength of maximal reflectance. B2:
Brightness, mean reflectance from 300 to 700 nm (grey region), S8: Saturation, ratio of
maximal reflectance – minimum reflectance on B2.
23
Directionality can be quantified using several methods. The angular span in which iridescence
is emitted is used in butterfly studies (Kemp et al. 2006, 2008), whereas the rate of decrease in
brightness when taken 5° away from maximal reflectance (Madsen et al. 2007), and the
difference in brightness when taken 10° away from maximal reflectance (Meadows et al.
2012) have both been used in bird studies. To choose which method to adopt here, we assessed
the strength of the correlation between brightness and directionality, measured using each of
the three methods, for a subsample of 39 crown feathers (see the Supplementary Material
Figure 2.8 for more details). We decided to proceed using the angular span (hereafter referred
to as directionality), as it was the variable least correlated with brightness and, consequently,
had more potential to carry different biological information. Specifically, directionality was
measured by noting the angles at which the signal was lost, rotating the stage both clockwise
and counter-clockwise. The angular span was the maximum minus the minimum angle. The
signal was considered lost when the spectra fell below 3% of maximal reflectance. For tree
swallow feathers, the minimum level of reflectance measured during rotation varied among
feathers, ranging between 1 and 2% reflectance. Thus, we chose to use a threshold of 3%
because it is above the minimum level of all measured feathers in our sample.
In order to evaluate the amplitude of bias that could be caused by not controlling barbule tilt,
we quantified brightness loss when deviating from the angle producing maximal reflectance.
To do this, we took a spectrum for each degree from -3 to 3° around the angle of maximal
reflectance using a subsample of one rump feather from 20 males. The degree range of -3 to 3°
corresponds to the range of variability in barbule tilt for tree swallow males.
24
2.4.4 Color analyses
Spectra were summarized with the R package PAVO v. 0.5-1 (Maia et al. 2013) by calculating
the three main dimensions of color: hue, saturation and brightness (Hailman 1977,
Montgomerie 2006). Hue corresponds to the wavelength of maximal reflectance, saturation is
the ratio of the maximum difference in percent reflectance from 300 to 700 nm on brightness,
and brightness is the average percent reflectance in 1 nm intervals from 300 to 700 nm (Figure
2.1). Saturation is usually measured in the ultraviolet (UV) region of the spectrum (Doucet et
al. 2006, Bitton et al. 2007, Balenger et al. 2008), in part because it has been shown that UVreflectance influences mate choice in species for which the UV region represent an important
part of the spectrum (Bennett et al. 1997, Hunt et al. 1999). In this study, we decided to
calculate the ratio of the maximum difference in percent reflectance on brightness over the
whole range of wavelengths because the tree swallows’ spectrum has a low UVcontribution (Figure 2.1). This measure was independent from any specific region of the
spectrum, thus avoiding high correlations between saturation and hue.
2.4.5 Statistical analyses
Repeatability is the proportion of total variation due to among-group variation when taking
into account within-group variation (Lessells and Boag 1987, Wolak et al. 2012). To
determine the measurement precision of the apparatus, repeatability was calculated with a
subsample of 30 males from 2013. One feather from each of the three regions was randomly
selected and measured three times. We refer to this measure as “measurement repeatability”,
in which within-group variation occurs among measurements of the same feather. To
determine if there are individual differences in iridescence variables, repeatability was also
calculated based on the whole feather sample (three feathers from each of the three regions)
collected on males captured in 2013 and 2014 (N=271). We refer to this measure as
25
“individual repeatability”, in which within-group variation occurs among measurements of the
three different feathers of each region. Again, the sample order was randomized between each
series of measurements. Repeatability estimates and confidence intervals for hue, saturation,
brightness, directionality, and barbule tilt were calculated using intra-class correlation
coefficients with the library ICC v. 2.2.1 (Wolak et al. 2012). The alpha level used to estimate
confidence intervals was 0.05.
To assess differences in iridescence between regions, linear mixed-models with restricted
maximum likelihood (REML) were used with the library lme4 (Bates et al. 2014). Response
variables were the four iridescence variables (hue, saturation, brightness, directionality) and
explanatory variables were regions (crown, mantle, rump) and year. The identity of
individuals was added as a random variable and explained a significant amount of variation in
each case (Likelihood Ratio Test (LRT), p<0.0001). All statistical analyses were performed
with R 3.1.2 (R Core Team 2014).
2.4.6 Ethical standards
Animals were captured and handled in compliance with the Canadian Council on Animal
Care, under the approval of the Université de Sherbrooke’s Animal Ethics Committee
(protocol number FP2014-01-Université de Sherbrooke).
26
2.5 Results
Measurement repeatability of the five variables measured (hue, saturation, brightness,
directionality, and barbule tilt) was significantly different from zero for the three regions
sampled (range = 0.48-0.89, Figure 2.2). No difference in repeatability was found between
regions for either hue, directionality or tilt. However, repeatability of saturation was higher for
rump than for crown feathers and repeatability of brightness was higher for rump than for
mantle feathers. No region was systematically more repeatable for all variables measured.
Figure 2.2 Measurement repeatability of iridescence variables for crown (dark grey), mantle
(light grey), and rump feathers (white) of male tree swallows. Error bars are 95% confidence
intervals. One feather from 30 males was measured three times. For exact values, see
Supplementary Material Table 2.3.
27
Individual repeatability of the three main dimensions of color (hue, saturation, brightness), as
well as of directionality, was significantly different from zero for the three regions sampled
(Figure 2.3) suggesting significant individual differences in iridescence. The repeatability of
directionality was comparable to that of the three main dimensions of color. When pooling all
nine feathers from each individual together, repeatability was still significant for all variables
(Figure 2.3). However, it was significantly lower than the repeatability of the three regions
considered separately, likely due to differences between regions.
Figure 2.3 Individual repeatability of iridescence variables for crown (dark grey), mantle
(medium grey), and rump feathers (light grey), as well as for all feathers from these three
regions (white), in male tree swallows. Error bars are 95% confidence intervals. Three feathers
from each of the 271 males from 2013 and 2014 were measured. For exact values, see
Supplementary Material Table 2.4.
28
Individual repeatability of barbule tilt was significant for the three regions sampled, i.e. the
confidence intervals never included zero, but was lower overall than for the other variables
(Figure 2.3). In a case where measurements would be obtained with a stage fixed parallel to
the base, the brightness readings of males with a barbule tilt that did not correspond to 0°
(55% of all males) would be between 2 to 16% lower than if the stage was adjusted to account
for individual barbule tilt (Figure 2.4). This corresponds to a loss of 13 to 104% when reported
on the 15% variation range of brightness across all males. The more the tilt deviates from 0°,
the more the brightness loss increases in a non-linear fashion (Figure 2.4).
Figure 2.4 Histogram of mean rump feathers’ tilt across 271 male tree swallows (left y-axis,
grey bars). Relationship between percent brightness loss and tilt of the feather (right y-axis,
black dots depict mean ± standard error). The brightness loss is the difference between
measurement at maximal reflectance (when the tilt is controlled for) and measurement with
the stage fixed parallel to the base (when the tilt is not controlled for). One rump feather from
20 males was used to estimate brightness loss.
29
Linear mixed-model analyses indicated that iridescence variables were all significantly
different between feathers sampled from the rump and crown regions, the rump having a lower
hue, higher saturation and brightness, and lower directionality (Table 2.1; Figure 2.5). The
mantle did not differ significantly from rump for brightness and directionality, when
Bonferroni correction was applied (α = 0.0125, for each response variable). However, mantle
feathers were greener (+4.7 nm) and had lower saturation (-0.02) than rump feathers. The hue
and directionality of feathers sampled in 2014 increased by 5.2 nm and by 2.6°, respectively,
compared to those sampled in 2013.
Figure 2.5 Variation in iridescence between three body regions of male tree swallows as
quantified by a) hue, b) saturation, c) brightness, and d) directionality. One datum corresponds
to one male (i.e. mean from three feathers). Boxplots indicate 1st and 3rd quartiles (edges),
median (line) and 1.5 times the interquartile range (whiskers).
30
Table 2.1 Effect of body region and year on plumage iridescence of male tree swallows as
quantified by hue, saturation, brightness, and directionality. Linear mixed effect models
included male identity as a random effect as it was significant (LRT, P < 0.0001) in each
model. PRV: Percent random variance explained by male identity.
Response variable Fixed effect
Estimate
Hue (nm)
467.794 1.258 371.95 <0.001 *
PRV= 49%
Saturation
PRV= 28%
Brightness (%)
PRV= 25%
Directionality (°)
PRV= 21%
Intercept (Rump, 2013)
SE
t value
P
Mantle
4.742 0.618
7.68 <0.001 *
Crown
26.293 0.619
42.50 <0.001 *
Year 2014
5.196 1.564
3.32 <0.001 *
Intercept (Rump, 2013)
0.934 0.011
83.28 <0.001 *
Mantle
-0.020 0.008
2.70 0.0070 *
Crown
-0.081 0.008
10.60 <0.001 *
Year 2014
-0.011 0.013
Intercept (Rump, 2013)
18.121 0.203
Mantle
0.320 0.145
Crown
-1.703 0.145
Year 2014
0.136 0.240
Intercept (Rump, 2013)
0.84
0.40
89.06 <0.001 *
2.21
0.027
11.76 <0.001 *
0.57
0.57
28.478 0.169 168.48 <0.001 *
Mantle
0.292 0.129
Crown
5.977 0.129
46.41 <0.001 *
Year 2014
2.560 0.196
13.05 <0.001 *
2.27
0.023
* Significant after Bonferroni correction (α = 0.0125 for each response variable)
2.6 Discussion
Our study contributes to the improvement of bird iridescence quantification in three ways.
First, we showed that the apparatus described in Meadows et al. (2011) is a reliable method for
31
quantifying individual variability in iridescence using a species, the tree swallow, whose
iridescence is derived from a different feather structure. Second, by showing that the apparatus
quantifies individual variability in directionality, we validated an important first step in
studying the sexual selection of this unique property of iridescence. Third, we showed that
controlling barbule tilt, which is easily done with the apparatus, reduces measurement bias and
should therefore be considered systematically when analyzing iridescent coloration in birds.
2.6.1 Repeatability
Overall, we found that measurement repeatability ranges from 0.48 to 0.89 for all five
components of iridescence in three body regions, namely the crown, mantle and rump. Our
measurement repeatability estimates regarding hue and brightness were similar to the ones
reported in Meadows et al. (2011) for gorget feathers of Anna’s hummingbirds. This result
suggests that our apparatus performed adequately to quantify blue iridescence for tree swallow
feathers, which have a much simpler iridescent structure than that of Anna’s hummingbirds.
Comparing our repeatability values for iridescence with those reported elsewhere is difficult
given the differences in method of measurement, repeatability calculation and species studied
(Perrier et al. 2002, Legagneux et al. 2010). We outline the importance of systematically
reporting measurement repeatability as an intra-class correlation coefficient (Lessells and
Boag 1987, Wolak et al. 2012) for better future comparisons. Although feathers from the three
body regions we studied had different morphologies, none suffered from an overall lower
repeatability. Yet, rump feathers have the potential to be highly repeatable given that their
barbs are more hooked to one another and therefore less mobile. However, if we had solely
focused on rump feathers to assess the iridescence of the entire blue plumage of male tree
swallows, we might have missed information provided by the crown that appears to contrast
with that of the mantle and rump.
32
We found individual repeatability values ranging between 0.37 and 0.71 for the main
dimensions of color (hue, saturation and brightness) in the three body regions, which
reinforces their potential for being sexual signals. These three dimensions have recently been
linked to mating success in males from species displaying iridescence, namely the peacock
(Pavo cristatus), the tree swallow and the satin bowerbird (Ptilonorhynchus violaceus) (Loyau
et al. 2007, Bitton et al. 2007, Savard et al. 2011, Dakin and Montgomerie 2013). The rather
low repeatability observed in some iridescence variables highlights the importance of
sampling more than one feather in order to have a precise representation of the iridescence of a
particular body region. Such a high variation in iridescence among feathers of a single region
is indeed commonly reported in the literature (Komdeur et al. 2005, Leclaire et al. 2014). The
method also captured variation in directionality among both individuals and body regions.
Crown feathers appeared to have the greatest variability in directionality, from 29 to 44°
(Figure 5). This suggests that it may also be used by females for mate choice and thus be
subjected to sexual selection, as it has been reported in the butterfly Colias Eurytheme (Kemp
et al. 2008). This is the first report of the repeatability of the directionality of iridescence in
birds measured as an angular span.
2.6.2 Potential bias caused by barbule tilt
We detected individual differences in feather barbule tilt of tree swallows. Indeed, brightness
measurements were increasingly underestimated for individuals with a large barbule tilt when
no adjustment was performed by rotating the feather until maximal reflectance. Without this
adjustment, males with the largest barbule tilt could be unjustly classified among the dullest
males. We therefore emphasize the importance of controlling for this source of bias when
measuring iridescence, particularly as it is easily done with the apparatus built following
Meadows et al. (2011). One should at least verify the degree of individual variation in barbule
tilt, as well as how large the bias can be, in the species under study. In the few studies where it
33
is reported, barbule tilt can be nonexistent (Toomey et al. 2010) or vary up to 7° (Brink and
Van der Berg 2004, Madsen et al. 2007).
2.6.3 Limits to the method of measurement
Here, we measured single feathers because Meadows et al. (2011) reported that measuring
multiple feathers so as to mimic the natural arrangement of feathers on birds resulted in lower
repeatability estimates. However, measuring single feathers may result in artificial differences
between regions caused by different barb arrangements. First, crown feathers were duller than
rump feathers, possibly because they have sparser barbs than rump feathers, thus causing a
decrease in relative iridescent coverage of the illuminated region. On a live bird, however, the
spatial arrangement of feathers can vary between body parts and, as a result, the crown could
appear as bright as the rump. Second, the black cardboard on which feathers were placed
during measurement reflects a significant amount of light (10% in our study). As crown and
mantle feathers appear to have sparser barbs than rump ones, cardboard reflectance possibly
contributed more to the spectrum of these two upper body regions by increasing their spectrum
baseline and lowering their saturation level. That could partly explain why we observed crown
and mantle feathers having lower saturation than rump feathers. The assessment of differences
between regions is important to determine which regions should be pooled for further
biological analyses. For instance, hue should not be affected by differences in barb
arrangement because it is independent from signal intensity. Therefore, when making the
decision to average measurements across regions using reflectance spectrometry of single
feathers, one should rely on hue differences instead of brightness or saturation. Using digital
photography of entire patches would be a useful tool to the decision to average or not, as it
avoids the problems mentioned above by taking the spatial arrangement of feathers into
account. However, this method is generally put aside given that it does not capture UV
reflectance. Albeit rarely used, UV-photography does exist (Stevens et al. 2007) and its use
34
should be encouraged as it can be very powerful and less time-consuming. Despite the
potential drawbacks discussed, the method presented here is still an effective way to quantify
individual differences in iridescence for a given region.
In conclusion, we urge researchers quantifying iridescence in birds to control for barbule tilt
when variation in this characteristic is observed to limit measurement error and bias. Because
directionality is inherent to iridescent coloration, we also encourage researchers to quantify it
in order to unveil its role in signaling and, ultimately, have a more complete understanding of
iridescence evolution.
2.7 Aknowledgements
We thank Benoit Couture, a Mechanical Engineering technician in the Université de
Sherbrooke’s Physics department, for helping us build the apparatus. We are grateful to
Myriam Cadotte and Sabrina Gignac-Brassard for helping analyze the feathers. Last but not
least, we thank all the graduate students and field assistants who helped collect feathers, as
well as the farm owners who gave us access to their land. This work was supported by the
Natural Sciences and Engineering Research Council of Canada (NSERC) discovery grants to
D.G., M.B., and F.P., the Fonds de Recherche du Québec – Nature et Technologies (FRQNT)
“Projet de recherche en équipe” grant awarded to D.G., M.B., and F.P., and by the Canada
Research Chair held by F.P. and M.B. S.V.W was supported by NSERC and FRQNT
scholarships.
35
2.8 References
Balenger, S. L., Johnson, L. S. and Masters, B. S. 2008. Sexual selection in a socially
monogamous bird: male color predicts paternity success in the mountain bluebird, Sialia
currucoides. - Behav. Ecol. Sociobiol. 63: 403–411.
Bates, D., Maechler, M. and Bolker, B. M. 2014. lme4: Linear mixed-effects models using
Eigen and S4.
Bennett, A. T. D., Cuthill, I. C., Partridge, J. C. and Lunau, K. 1997. Ultraviolet plumage
colors predict mate preferences in starlings. - Proc. Natl. Acad. Sci. U. S. A. 94: 8618–
8621.
Bitton, P.-P., O’Brien, E. L. and Dawson, R. D. 2007. Plumage brightness and age predict
extrapair fertilization success of male tree swallows, Tachycineta bicolor. - Anim. Behav.
74: 1777–1784.
Brink, D. J. and Van der Berg, N. G. 2004. Structural colours from the feathers of the bird
Bostrychia hagedash. - J. Phys. D. Appl. Phys. 37: 813–818.
Dakin, R. and Montgomerie, R. 2013. Eye for an eyespot: how iridescent plumage ocelli
influence peacock mating success. - Behav. Ecol. 24: 1048–1057.
Doucet, M. and Montgomerie, R. 2003. Multiple sexual ornaments in satin bowerbirds :
ultraviolet plumage and bowers signal different aspects of male quality. - Behav. Ecol.
14: 503–509.
Doucet, S. M., Shawkey, M. D., Hill, G. E. and Montgomerie, R. 2006. Iridescent plumage in
satin bowerbirds: structure, mechanisms and nanostructural predictors of individual
variation in colour. - J. Exp. Biol. 209: 380–90.
Eliason, C. M. and Shawkey, M. D. 2010. Rapid, reversible response of iridescent feather
color to ambient humidity. - Opt. Express 18: 21284–92.
Endler, J. A. 1986. Natural selection in the wild. - Princeton University Press.
Ghilain, A. and Bélisle, M. 2008. Breeding success of tree swallows along a gradient of
agricultural intensification. - Ecol. Appl. 18: 1140–1154.
Greenewalt, C., Brandt, W. and Friel, D. 1960. The iridescent colors of hummingbird feathers.
- Proc. Am. Philos. Soc. 104: 249–253.
36
Griffith, S. C., Parker, T. H. and Olson, V. A. 2006. Melanin- versus carotenoid-based sexual
signals: is the difference really so black and red? - Anim. Behav. 71: 749–763.
Hailman, J. 1977. Optical signals: animal communication and light. - Indiana University Press.
Hart, N. S. 2001. The visual ecology of avian photoreceptors. - Prog. Retin. Eye Res. 20: 675–
703.
Hill, G. E., Doucet, S. M. and Buchholz, R. 2005. The effect of coccidial infection on
iridescent plumage coloration in wild turkeys. - Anim. Behav. 69: 387–394.
Hunt, S., Cuthill, I. C., Bennett, A. T. D. and Griffiths, R. 1999. Preferences for ultraviolet
partners in the blue tit. - Anim. Behav. 58: 809–815.
Kemp, D. J. 2006. Heightened phenotypic variation and age-based fading of ultraviolet
butterfly wing coloration. - Evol. Ecol. Res. 8: 515–527.
Kemp, D. J. 2008. Resource-mediated condition dependence in sexually dichromatic butterfly
wing coloration. - Evolution 62: 2346–58.
Kemp, D. J. and Rutowski, R. L. 2007. Condition dependence, quantitative genetics, and the
potential signal content of iridescent ultraviolet butterfly coloration. - Evolution 61: 168–
83.
Kemp, D. J., Vukusic, P. and Rutowski, R. L. 2006. Stress-mediated covariance between
nano-structural architecture and ultraviolet butterfly coloration. - Funct. Ecol. 20: 282–
289.
Kemp, D. J., Macedonia, J. M., Ball, T. S. and Rutowski, R. L. 2008. Potential direct fitness
consequences of ornament-based mate choice in a butterfly. - Behav. Ecol. Sociobiol. 62:
1017–1026.
Komdeur, J., Oorebeek, M., van Overveld, T. and Cuthill, I. C. 2005. Mutual ornamentation,
age, and reproductive performance in the European starling. - Behav. Ecol. 16: 805–817.
Leclaire, S., Pierret, P., Chatelain, M. and Gasparini, J. 2014. Feather bacterial load affects
plumage condition, iridescent color, and investment in preening in pigeons. - Behav.
Ecol. 25: 1192–1198.
Legagneux, P., Théry, M., Guillemain, M., Gomez, D. and Bretagnolle, V. 2010. Condition
dependence of iridescent wing flash-marks in two species of dabbling ducks. - Behav.
Processes 83: 324–30.
37
Lessells, C. M. and Boag, P. T. 1987. Unrepeatable repeatabilities : a common mistake. - Auk
104: 116–121.
Lim, M. L. M. and Li, D. 2013. UV-green iridescence predicts male quality during jumping
spider contests. - PLoS One 8: e59774.
Loyau, A., Gomez, D., Moureau, B., Théry, M., Hart, N. S., Jalme, M. S., Bennett, A. T. D.
and Sorci, G. 2007. Iridescent structurally based coloration of eyespots correlates with
mating success in the peacock. - Behav. Ecol. 18: 1123–1131.
Madsen, V., Dabelsteen, T., Osorio, D. and Osorno, J. L. 2007. Morphology and
ornamentation in male magnificent frigatebirds: variation with age class and mating
status. - Am. Nat. 169 Suppl: S93–S111.
Maia, R., Eliason, C. M., Bitton, P.-P., Doucet, S. M. and Shawkey, M. D. 2013. PAVO : an R
package for the analysis, visualization and organization of spectral data. - Methods Ecol.
Evol. 4: 906–913.
McGraw, K. J., Mackillop, E. a, Dale, J. and Hauber, M. E. 2002. Different colors reveal
different information: how nutritional stress affects the expression of melanin- and
structurally based ornamental plumage. - J. Exp. Biol. 205: 3747–55.
Meadows, M. G., Morehouse, N. I., Rutowski, R. L., Douglas, J. M. and McGraw, K. J. 2011.
Quantifying iridescent coloration in animals: a method for improving repeatability. Behav. Ecol. Sociobiol. 65: 1317–1327.
Meadows, M. G., Roudybush, T. E. and McGraw, K. J. 2012. Dietary protein level affects
iridescent coloration in Anna’s hummingbirds, Calypte anna. - J. Exp. Biol. 215: 2742–
2750.
Meunier, J., Figueiredo Pinto, S., Burri, R. and Roulin, A. 2010. Eumelanin-based coloration
and fitness parameters in birds: a meta-analysis. - Behav. Ecol. Sociobiol. 65: 559–567.
Montgomerie, R. 2006. Analyzing Colors. - In: Hill, G. E. and McGraw, K. J. (eds), Bird
Coloration: Vol. 1. Mechanisms and Measurements. Harvard University Press, pp. 90–
147.
Osorio, D. and Ham, A. D. 2002. Spectral reflectance and directional properties of structural
coloration in bird plumage. - J. Exp. Biol. 205: 2017–27.
Papke, R., Kemp, D. J. and Rutowski, R. L. 2007. Multimodal signalling: structural ultraviolet
reflectance predicts male mating success better than pheromones in the butterfly Colias
eurytheme L. (Pieridae). - Anim. Behav. 73: 47–54.
38
Perrier, C., de Lope, F., Møller, A. P. and Ninni, P. 2002. Structural coloration and sexual
selection in the barn swallow Hirundo rustica. - Behav. Ecol. 13: 728–736.
Prum, R. 2006. Anatomy, Physics, and Evolution of Structural Colors. - In: Hill, G. and
McGraw, K. (eds), Bird Coloration: Vol. 1. Mechanisms and Measurements. Harvard
University Press, pp. 295–353.
R Development Core Team 2014. R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.
Rahman, M. M., Kelley, J. L. and Evans, J. P. 2013. Condition-dependent expression of preand postcopulatory sexual traits in guppies. - Ecol. Evol. 3: 2197–213.
Savard, J.-F., Keagy, J. and Borgia, G. 2011. Blue, not UV, plumage color is important in
satin bowerbird Ptilonorhynchus violaceus display. - J. Avian Biol. 42: 80–84.
Stevens, M., Parraga, C. a, Cuthill, I. C., Partridge, J. C. and Troscianko, T. S. 2007. Using
digital photography to study animal coloration. - Biol. J. Linn. Soc. 90: 211–237.
Stiles, F. G. 1982. Aggressive and courtship displays of the male Anna’s Hummingbird. Condor 84: 208–225.
Toomey, M. B., Butler, M. W., Meadows, M. G., Taylor, L. A., Fokidis, H. B. and McGraw,
K. J. 2010. A novel method for quantifying the glossiness of animals. - Behav. Ecol.
Sociobiol. 64: 1047–1055.
Vukusic, P., Sambles, J. R., Lawrence, C. R. and Wootton, R. J. 2002. Limited-view
iridescence in the butterfly Ancyluris meliboeus. - Proc. Biol. Sci. 269: 7–14.
Wolak, M. E., Fairbairn, D. J. and Paulsen, Y. R. 2012. Guidelines for estimating
repeatability. - Methods Ecol. Evol. 3: 129–137.
39
ANNEXE
Electronic Supplementary Material
A reliable technique to quantify the individual variability of iridescence in birds
Sonia Van Wijk, Marc Bélisle, Dany Garant and Fanie Pelletier
40
Figure 2.6 Apparatus for quantifying iridescence in bird feathers built according to Meadows
et al. (2011)
Specifications for building the apparatus:
-
Material: Aluminium
-
PT1 translational Stage from http://www.thorlabs.com/
-
The critical parameter regarding the apparatus machining is the axis of rotation of the
two arms and stage (here: 6 inches from the base). It is important that the two arms and
stage rotate along the exact same axis of rotation and that the rotation is perfectly
circular. This prevents problems related to the precise alignment of the two beams and
the feather.
41
Figure 2.7 Apparatus for quantifying iridescence in bird feathers built according to Meadows
et al. (2011)
42
In birds, directionality has been calculated in two studies. Madsen et al. (2007) measured
directionality from Eqn 1 and Meadows et al. (2012) measured it from Eqn 2 (parameters
described in table A1). We calculated both measures in a subsample of 39 crown feathers (1
feather per male). For Eqn 2, we used an angle of 5° instead of 10°.
[1]
2 − 2( ± 5°)
2 + 2( ± 5°)
[2] 2 − 2( ± 10°)
Table 2.2 Parameters from Eqn 1 and 2 and their associated definitions.
Parameter
Definition
B2
Brightness
B2( ± 5°)
Mean brightness when the feather is rotated 5°
away from maximal reflectance
B2( ± 10°)
Mean brightness when the feather is rotated 10°
away from maximal reflectance
43
Figure 2.8 Correlation between brightness and three different directionality variables: angular
breadth, directionality as calculated by Madsen et al. (2007), and directionality calculated
similarly to Meadows et al. (2012). The difference in directionality measurement for the
second study is the use of an angle of 5° instead of 10°. Black line represents the regression
between two variables (upper panel). Pearson correlations are shown with increasing font size
with increasing correlation strength (lower panel).
44
Table 2.3 Measurement repeatability of iridescence variables for crown, mantle and rump
feathers of male tree swallows. Brackets are 95% confidence intervals. One feather from 30
males was measured three times.
Region
Iridescence variable
Crown
Mantle
Rump
Hue
0.89 [0.82 - 0.94] 0.81 [0.69 - 0.90] 0.88 [0.79 - 0.93]
Saturation
0.59 [0.39 - 0.76] 0.85 [0.74 - 0.92] 0.88 [0.79 - 0.94]
Brightness
0.72 [0.56 - 0.84] 0.62 [0.42 - 0.78] 0.87 [0.78 - 0.93]
Directionality
0.48 [0.26 - 0.68] 0.76 [0.61 - 0.87] 0.70 [0.53 - 0.83]
Tilt
0.71 [0.55 - 0.84] 0.50 [0.28 - 0.69] 0.49 [0.28 - 0.69]
Table 2.4 Individual repeatability of iridescence variables for crown, mantle, and rump
feathers, as well as for all feathers from these three regions in male tree swallows. Brackets are
95% confidence intervals. Three feathers from the 271 males from 2013 and 2014 were
measured.
Region
Iridescence
variable
Hue
Crown
Mantle
Rump
All feathers
0.71 [0.66 - 0.76]
0.58 [0.52 - 0.64]
0.61 [0.55 - 0.67]
0.31 [0.26 - 0.36]
Saturation
0.39 [0.32 - 0.47]
0.58 [0.52 - 0.64]
0.56 [0.50 - 0.63]
0.27 [0.23 - 0.32]
Brightness
0.37 [0.30 - 0.45]
0.41 [0.33 - 0.48]
0.56 [0.50 - 0.63]
0.23 [0.19 - 0.27]
Directionality
0.50 [0.43 - 0.57]
0.46 [0.38 - 0.53]
0.65 [0.59 - 0.70]
0.13 [0.10 - 0.17]
Tilt
0.30 [0.23 - 0.38]
0.23 [0.15 - 0.31]
0.34 [0.27 - 0.42]
0.14 [0.11 - 0.18]
45
References
Madsen, V., Dabelsteen, T., Osorio, D. and Osorno, J. L. 2007. Morphology and
ornamentation in male magnificent frigatebirds: variation with age class and mating
status. - Am. Nat. 169 Suppl: S93–S111.
Meadows, M. G., Morehouse, N. I., Rutowski, R. L., Douglas, J. M. and McGraw, K. J. 2011.
Quantifying iridescent coloration in animals: a method for improving repeatability. Behav. Ecol. Sociobiol. 65: 1317–1327.
Meadows, M. G., Roudybush, T. E. and McGraw, K. J. 2012. Dietary protein level affects
iridescent coloration in Anna’s hummingbirds, Calypte anna. - J. Exp. Biol. 215: 2742–
2750.
46
CHAPITRE 3
LA DENSITÉ DE REPRODUCTEURS INFLUENCE LA SÉLECTION SEXUELLE
DE L’IRIDESCENCE CHEZ LES MÂLES HIRONDELLE BICOLORE,
TACHYCINETA BICOLOR
3.1 Introduction de l’article
L’objectif de ce deuxième article était de vérifier si l’iridescence est un trait soumis à la
sélection sexuelle chez les mâles Hirondelle bicolore. Pour ce faire, nous avons suivi la
reproduction de cette espèce dans un système de 400 nichoirs au Sud du Québec. Nous avons
par la suite déterminé la relation entre l’iridescence des mâles et les deux composantes de leur
succès reproducteur, soit leur nombre de jeunes dans le couple et hors couple. La contribution
scientifique de cette étude comprend d’une part l’inclusion d’une propriété unique à
l’iridescence, soit le changement de brillance occasionné par un mouvement de l’animal. Cette
propriété a été pour la première fois reliée au succès reproducteur chez les oiseaux. D’autre
part, ce système à large échelle nous a permis de vérifier si la sélection sexuelle changeait
selon la densité en reproducteurs. Les conditions environnementales peuvent grandement
influencer la sélection sexuelle, pourtant celles-ci sont souvent ignorées. Ma contribution à cet
article est majeure puisque j’ai supervisé les stagiaires pour les mesures de coloration, effectué
les analyses statistiques, écrit l’article et coordonné le processus de publication. Ma
superviseure, la Pre. Fanie Pelletier, m’a encadrée tout au long du projet, tandis que mes
conseillers, Marc Bélisle et Dany Garant, ont judicieusement apporté des commentaires et
suggestions au projet et aux versions préliminaires de ce manuscrit. Audrey Bourret a
coordonné les analyses génétiques en plus d’apporter des commentaires aux versions
préliminaires du manuscrit. Le manuscrit est en révision au journal Behavioral Ecology and
Sociobiology (7 juillet 2015).
47
BREEDING DENSITY INFLUENCES THE SEXUAL SELECTION OF
IRIDESCENCE IN MALE TREE SWALLOWS, TACHYCINETA BICOLOR
Par
Sonia Van Wijk, Audrey Bourret, Marc Bélisle, Dany Garant et Fanie Pelletier
3.2 Abstract
Sexual selection is an important driver of species evolution and secondary sexual traits have
been widely studied throughout the animal kingdom, especially in birds. However, iridescent
coloration is a complex trait that has received little attention despite its potential importance as
a signal of individual condition. Using 202 male tree swallows (Tachycineta bicolor) from a
two-year study, we investigated how iridescence was related to four measures of fitness
(annual reproductive success, the number of within-pair young and extra-pair young as well as
the proportion of within-pair young produced) under variable breeding density. A property
inherent to iridescence was included in its quantification, namely the angular breadth of
iridescence emission. Our results showed that iridescence influenced all four measures of
fitness. A positive association was detected between the angular breadth of iridescence and
male number of within-pair young, suggesting that it could signal information about the
quality of paternal care. The advantage of bright males for extra-pair reproduction dampened
at high breeding density, potentially due to a higher cost of female choice induced by male
behavior. Altogether, our results showed that different selective pressures can act on the
multiple components of iridescence via both within-pair and extra-pair components of
reproductive success. Also, environmental conditions like breeding density are important
factors influencing sexual selection processes.
48
3.3 Introduction
Sexual selection acts on secondary sexual traits enhancing the access to sexual partners and is
thus an important driver of species evolution (Darwin 1871; Andersson 1994). Given the
anisogamy found between sexes, sexual selection is more common in males, as they show a
higher variance in reproductive success than females (Bateman 1948). The correlation
between a male trait and reproductive success can result from two non-exclusive mechanisms:
male-male competition (intra-sexual selection) and female choice (inter-sexual selection)
(Andersson 1994). Traits that signal individual’s condition and therefore its potential fitness
(often referred to as “quality” (Bergeron et al. 2011)) are good candidates for evolution by
either intra or inter-sexual selection. The handicap principle (Zahavi 1975) states that
sexually-selected traits can act as reliable signals only if they induce a cost to the individual
bearing them, such that only individuals in good condition can display the full traits, thereby
inhibiting the possibility to cheat. Female preference for these traits can thus evolve by
bringing either direct benefits (e.g. territory quality, paternal care (Hoelzer 1989)) or indirect
benefits to her (e.g. genes increasing the fitness of offspring (Weatherhead and Robertson
1979; Hamilton and Zuk 1982; Neff and Pitcher 2005)).
The skew of conspicuous coloration towards males of numerous animal species is thought to
have evolved by sexual selection. In fishes, the guppy (Poecilia reticulata) and the threespined stickleback (Gasterosteus aculeatus) have been extensively studied. More colorful
males of these two species show a better condition and are preferred by females (Milinski and
Bakker 1990; Houde and Endler 1990; Rahman et al. 2013). Sexual selection has also been
suggested in some species of butterflies, spiders and reptiles, but the relative importance of
intra and inter-sexual selection have yet to be clarified (Andersson 1994). On the other hand,
evidence of female choice for color displays in birds is substantial and is now recognized as an
important selective agent (Hill 2006).
49
In birds, iridescence is a structural coloration in which the alternation of two or more layers of
different materials (melanin, keratin and, in some cases, air) in the feather microstructure
causes the constructive interference of certain wavelengths that produces the observed color
pattern (Prum 2006). Iridescence has a potential for condition-dependence, because its
structural origin could make it highly sensitive to abrasion by feather-degrading bacteria and
dependent upon the levels of preening that can be provided by the bird (Leclaire et al. 2014),
therefore amplifying among-male differences in condition (Fitzpatrick 1998). Indeed, many
studies reported a link between iridescence and various proxies of condition (McGraw et al.
2002; Doucet and Montgomerie 2003; Hill et al. 2005; Legagneux et al. 2010; Meadows et al.
2012). The first few studies on the sexual selection of iridescence observed relationships with
female preference (Bennett et al. 1997), mating success (Loyau et al. 2007; Savard et al. 2011;
Dakin and Montgomerie 2013) and reproductive success (Bitton et al. 2007), suggesting its
importance for sexual selection. Still, studies directly linking iridescence and reproductive
success remain scarce.
Inherent to iridescent coloration is directionality, the high brightness change with movements
of the animal (Osorio and Ham 2002). This property makes the quantification of iridescence
difficult, which is one of the reasons why its study is much less developed than that of
plumage colored by pigments (carotenoids or melanin) (Hill 2006). Individual differences in
directionality have recently been found using a recent repeatable technique (Meadows et al.
2011; Van Wijk et al. in review). It has been suggested that a plumage exhibiting a high
brightness change would have a more ordered microstructure (Madsen et al. 2007, Meadows
et al. 2012), a feature that could be indicative of a better condition. This property may be a
distinct channel of information about the individual (Van Wijk et al. in review), and thus
including this variable when studying sexual selection of iridescence in birds has the potential
to reveal great insights.
50
With the accumulating evidence of widespread extra-pair paternity (EPP) in socially
monogamous birds (Griffith et al. 2002), EPP was suggested to drive the sexual selection of
male coloration by increasing the variance in male reproductive success (Møller and Birkhead
1994). However, taking into account the number of within-pair young (WPY) besides the
number of extra-pair young (EPY) appear crucial to the study of sexual selection. Indeed, both
fitness components can cause variation in reproductive output (Webster et al. 1995;
Whittingham and Dunn 2005) and alternative tactics maximizing either WPY or EPY
component may coexist (Delhey et al. 2003). Sexual selection studies should also address how
ecological conditions influence the opportunity for selection and the strength of selection
itself. Indeed, there is growing support on the strong influence of ecological factors on sexual
selection, such as weather conditions, habitat quality and breeding density (Robinson et al.
2012; Taff et al. 2013; Lessard et al. 2014). For instance, mate accessibility increases with
breeding density and may thereby affect the cost of mate choice (Kokko and Rankin 2006).
Such an effect could explain why, for example, Taff et al. (2013) observed that sexual
selection on plumage coloration occurred only at high density in the common yellowthroat
(Geothlypis trichas). The inherent bias towards studying high-density population in behavioral
ecology may thus prevent to appreciate the full array of selection patterns that occur under
natural conditions (Stamps 2011).
The goal of this study was to assess how iridescence and its directionality influence the
reproductive success of male tree swallows. To do this, we evaluated the influence of the
different components of iridescence on four measures of fitness (annual reproductive success,
the number of WPY and EPY as well as the proportion of WPY) and their effects under
variable breeding density. Male tree swallows display blue-green iridescent upperparts with
white underparts. Females display a greener and duller iridescence than males, although an
overlap in coloration exists between sexes (Cohen 1984). Between 35 and 52% of tree
swallow young are EPY (Kempenaers et al. 2001; O’Brien and Dawson 2007; Lessard et al.
2014). Females have a high control over EPPs given that no mate guarding occurs (Leffelaar
51
and Robertson 1984) and because the cost of doing extra-pair copulations seems low (Lifjeld
et al. 1993; Venier et al. 1993; Whittingham et al. 1993). More ornamented males of species in
which coloration depends upon feather structure are generally brighter, more saturated or have
a more UV-shifted hue (Keyser and Hill 2000; Hill et al. 2005; Bitton et al. 2007; Balenger et
al. 2008; Zirpoli et al. 2013). Given that these plumage features likely act as honest signals of
male condition, females paired with poorly ornamented males would be better off to pay the
cost of obtaining EPPs (Westneat et al. 1990; Able 1996) and get good genes from highly
ornamented extra-pair males (Jennions and Petrie 2000). On the other hand, females paired
with highly ornamented males would not be prone to engage in extra-pair copulations because
they would already have the good genes via their social partner. As a result, we predicted that
more ornamented males would be more successful at siring both WPY and EPY. Female
preference would also result in a higher proportion of WPY in the nest of more ornamented
males. We also predicted that the effect of male ornamentation would be more pronounced at
high versus low density of breeders because of the greater ability and/or lower cost of finding
and selecting a good mate (Dunn and Whittingham 2007; Taff et al. 2013).
3.4 Materials and methods
3.4.1 Study system and data collection
Covering 10 200 km2, the study system was located in southern Québec, Canada, and included
an East-West agricultural intensification gradient. Large-scale intensive cultures were present
on the western side of the gradient whereas small-scale extensive cultures occurred on its
eastern side. Since 2004, 400 nest boxes equally distributed across 40 farms were monitored
once every two days during the breeding season (see Ghilain and Bélisle (2008) for a detailed
description of the study system). Females were captured during incubation and males during
food provisioning. Captured adults and 12 days-old nestlings were marked with a US Fish and
52
Wildlife Service aluminum band and were bled from the brachial vein. Blood samples were
collected on a qualitative P8 grade filter paper (Fisher Scientific), then dried and stored at
room temperature until DNA extraction (see below). Measurements of weight (±0.01g), nonflattened wing length (±0.5mm) and tarsus length (±0.02mm) were taken on every adult. The
presence of feather parasites on the external rectrices (feather mites and holes from chewing
lice) was noted. We collected iridescent feathers from males in 2013 (N=141) and 2014
(N=187). Three feathers were plucked at the center of each of three regions: crown, mantle
and rump, for a total of nine feathers per male. Feathers were stored at room temperature in
manila envelopes until spectral analyses (see below).
3.4.2 Iridescence measurements
Spectral analyses were performed with a spectrometer Avaspec-2048 coupled with AvalightXE Pulsed Xenon Lamp relative to a WS-2 standard (Avantes Inc., Broomfield, CO, USA).
We used the apparatus described in Van Wijk et al. (in review) and in Meadows et al. (2011)
allowing repeatable measurement of tree swallow iridescence. Briefly, the apparatus consisted
in two arms and a translational stage (PT1, Thorlabs, Newton, NJ, USA) that can rotate around
a single axis. A collimating lens (COL-UV/VIS from Avantes) is fixed on each arm with an
optical fiber that either join the lamp source (200 μm fiber) or the spectrometer (100 μm fiber),
respectively. The light and spectrometer arms were set at a 20° angle from one another. For
each spectrum of mantle and rump feathers, the average of five scans at 25 pulses per 250-ms
integration time with a boxcar smoothing function of 10 pixels was taken using the software
Avasoft 8 (Avantes), and a 2-mm diameter light beam was used. The quantification method
was modified for crown feathers considering their smaller size: we used a 1-mm diameter
beam and 15 pulses per 150-ms integration time. Individual feathers were put on a black
cardboard covering the stage so that the light beam was located at the center of the iridescent
region. The feather was held flat by putting a piece of black cardboard with a 7-mm aperture
53
over it. The stage was then rotated until reaching maximal reflectance in order to control for
individual differences in barbule tilt (Van Wijk et al. in review). After spectrum recording, we
measured directionality adapted from Kemp et al. (2006) where they used a camera able to
detect exclusively UV to measure the angular breadth of UV emission of the butterfly Colias
eurytheme wing scales. In our case, angular breadth was taken by noting the angles at which
the signal was lost, rotating the stage both clockwise and counter-clockwise. The signal was
considered lost when spectra fell below 3% of maximal reflectance (Van Wijk et al. in
review). Directionality will therefore be referred to as angular breadth throughout the text. A
feather with a wider angular breadth thus likely shows a less directional iridescence because
the amount of reflectance is emitted over a larger angle. Reflectance spectra were binned in 1nm increment from 300 to 700 nm that correspond to the visual range of birds (Hart 2001) and
summarized with the package
PAVO
v. 0.5-1 (Maia et al. 2013). We calculated the three
universal dimensions of color: hue, brightness and saturation (Hailman 1977; Montgomerie
2006). Hue corresponds to the wavelength of maximal reflectance, brightness is the average
percent reflectance in 1-nm intervals from 300 to 700 nm, and saturation is the ratio of the
maximum difference in percent reflectance on brightness. More details of iridescence analyses
are described in (Van Wijk et al. in review.).
Based on previous analyses (Van Wijk et al. in review), we considered that male tree swallows
had two different iridescent regions, crown and back (i.e. rump and mantle). Individual
variables of iridescence consisted in the average of the three crown feathers for crown, and the
average of the three rump and three mantle feathers for back. As iridescence variables were
correlated, a Principal Component Analysis (PCA) was performed for each of the two regions
(Figure 3.1). Variables were log-transformed or squared when necessary in order to follow
multinormality. The two PCA showed different correlation structures among iridescence
metrics. We kept PC1 and PC2 in each case since they explained a total of 79 and 72% of
overall variation for crown and back iridescence, respectively. Thus, for crown measurements,
a male with a high PC1 score displayed brighter and more saturated crown, and a male with a
54
high PC2 score displayed a greener plumage (hue shifted to longer wavelengths; Figure 3.1).
For ease of reading, hereafter we will refer to PC1 as ‘crown brightness’, and PC2 as ‘crown
hue’. For back measurements, a male with a high PC1 score displayed brighter and more
saturated back as for crown, but a male with a high PC2 score displayed a wider angular
breadth (less directional plumage; Figure 3.1). Therefore, PC1 will be referred to as ‘back
brightness’, and PC2 as ‘back angular breath’. Values of variable loadings on the principal
components are shown in Supplementary Table 3.3.
Figure 3.1 Correlation circles representing the association between iridescence variables of
two distinct body regions, namely a) the crown and b) the back of 202 male tree swallows
breeding in southern Québec, Canada, 2013-2014. Horizontal and vertical axes represent PC1
and PC2 from a Principal Component Analysis (PCA) performed for each region separately.
The two PC combined explained a total 79% and 72% of variance for crown and back,
respectively. The size of the angle between two vectors represents the strength and direction of
the correlation between them (< 90°: positive, 90 °: null, > 90°: negative). PC1 and PC2 from
the two PCA were the iridescence variables retained for modeling (Table 3.1 and Table 3.2).
For ease of reading, crown PC1 and PC2 were referred to as crown brightness and hue,
respectively. Back PC1 and PC2 were referred to as back brightness and angular breadth,
respectively.
55
3.4.3 Genetic analyses
A detailed description of DNA extraction, genotyping, molecular sexing and paternity
assignment methods used in this study can be found in Lessard et al. (2014). Briefly, birds
were genotyped at 6 microsatellites markers: HRU7, Hir 19, Hir 22, IBI MS5-29, TBI 81, TBI
104. Polymerase chain reactions (PCR) were performed to amplify each locus and PCR
products were visualized using an AB-3130xl automated DNA sequencer, while alleles were
scored using GENEMAPPER v. 4.1 (Applied Biosystems). Paternity assignments were performed
with
CERVUS
v. 3.0.3 (Kalinowski et al. 2007) and young origin (WPY or EPY) was
determined based on comparisons between social and genetic father identities. Candidate
fathers of each nestling included all the males caught within a 15-km radius around the nest
box of origin (Lessard et al. 2014).
3.4.4 Statistical analyses
We used a total of 202 males in the analyses. We considered only first breeding attempts to
avoid confounding factors associated with second breeding attempts. We removed males with
missing information about the status (WPY or EPY) of nestlings in their nests (N=28). As 57
males (21%) were sampled in both years, we randomly selected one observation for those
individuals to avoid potential pseudoreplication problems. To model the effect of iridescence
on the annual reproductive success (quantified by the sum of WPY and EPY produced by a
male), the number of WPY and the number of EPY, we used three separate generalized linear
mixed-models with a Poisson structure and log link function, fitted with the library lme4
(Bates et al. 2014). To model the effect of iridescence on the proportion of within-pair young
in the nest (%WPY), we used generalized linear mixed-models with a binomial error
distribution and a logit link function. The same structure of explanatory variables was used for
the four different models. The identity of the farm where the male was caught was added as a
56
random variable since it was significant (Likelihood Ratio Test (LRT), P<0.001). In the full
model, we included the four iridescence variables (PC representing crown brightness, crown
hue, back brightness and back angular breadth) and the interaction between these variables
with tree swallow density (number of nests occupied by tree swallows in a 15-km radius,
ranging from 3 to 63). Because there were differences in iridescence variables among years
(see Supplementary Figure 3.5), we removed year effect by using the residuals from a linear
regression. In all models, we included variables that were previously found to be important
determinants of male reproductive success in our study system (Lessard et al. 2014): presence
of parasites (feather mites and holes made by chewing lice, factor: 0 or 1), wing and tarsus
length, laying date, and the two-way interactions between wing and tarsus length with density.
We also included minimal age (number of years since the first observation in the system) as
age is recurrently related to the number of EPY obtained by a male (Akçay and Roughgarden
2007; Cleasby and Nakagawa 2012), and mass corrected for Julian day and time of day (Rioux
Paquette et al. 2014; Whittingham and Dunn 2014). All continuous variables were centered
and scaled by their standard deviation in order to obtain comparable regression coefficients
and limit collinearity between interactions and main effects (Schielzeth 2010). Unscaled mean
and standard deviation of variables are available in Supplementary Table 3.4. No problematic
levels of collinearity were detected since the highest variance inflation factor (VIF) in the full
models was 1.77 for crown hue (likely caused by collinearity with back angular breadth,
r=0.5). The correlation matrix of explanatory variables is available in Supplementary Figure
3.6. The final model was selected with backward deletion of the least significant term using
P=0.05 as the threshold value for elimination (Crawley 2007). We calculated conditional R2 to
report the amount of variance explained by models (Nakagawa and Schielzeth 2013). All
statistical analyses were performed with R 3.1.2 (R Core Team 2014).
57
3.4.5 Ethical standards
Animals were captured and handled in compliance with the Canadian Council on Animal
Care, under the approval of the Université de Sherbrooke Animal Ethics Committee (protocol
number FP2014-01-Université de Sherbrooke).
3.5 Results
3.5.1 Patterns of paternity
In 2013-2014, 58% of all EPY were associated with an extra-pair father that was a social male
in a nest within the 15-km radius. The level of extra-pair paternity was high: 84% of all nests
contained at least one EPY. Also 25% of nests consisted only in EPY and 40% of all males
were not associated with any EPY. The average ±SD number of EPY in a nest was 2.6±1.8
(52±36% of young in a nest). A male had on average 2.2±1.9 WPY and 1.5±2.0 EPY assigned
to him. No relationship was detected between the number of WPY and EPY assigned to a
male (Pearson correlation: -0.05, P=0.48).
3.5.2 Iridescence and reproductive success
In 2013-2014, iridescence was associated with the annual reproductive success (Table 3.1a).
Back brightness and angular breadth were both positively related to reproductive success, but
their effect was driven by either WPY or EPY component. Back angular breadth was
positively related to the number of WPY with a difference up to 1.4 WPY between the male
58
with the smallest angular breadth to the one with the largest angular breadth (Table 3.1b,
Figure 3.2). Back brightness was positively related to the number of EPY at low density (i.e.:
1st quartile: 22 nests in a 15-km radius) with an increase of 1.8 EPY from the dullest to the
brightest male. The positive relationship between back brightness and the number of EPY was
reduced with increasing density, with no relationship being detected at high density (i.e.: 3rd
quartile: 45 nests in a 15-km radius; Table 3.1c, Figure 3.3). Crown brightness was also
associated with annual reproductive success, and the direction of the relationship depended on
density as well. Crown brightness was positively related to annual reproductive success at low
density, whereas a negative relationship was present at high density, where there was a
difference of 1.8 young between the dullest and the brightest male (Table 3.1a, Figure 3.4).
The relationship between crown brightness and annual reproductive success was mainly
driven by the number of EPY, as a similar effect was detected in the EPY model and none in
the WPY model (Table 3.1b and c). We replaced back angular breadth for crown hue in the
annual reproductive success model, as moderate collinearity existed between both variables
(r=0.5, see Supplementary Figure 3.6). As a result, crown hue was positively related to annual
reproductive success (βcrown
hue=0.12±0.04,
P=0.007). Crown hue had also a marginally
nonsignificant
the
of
effect
on
number
59
EPY
(β
crown
hue=0.11±0.07,
P=0.10).
Table 3.1 Effect of iridescence and other individual and environmental variables on a) the annual reproductive success, b) the
number of within-pair young (WPY) and c) the number of extra-pair young (EPY) of male tree swallows (n=202) breeding in
southern Québec, Canada, 2013-2014. Final generalized linear mixed effect models (Poisson structure and log link function)
included farm identity as a random effect. All variables were centered and scaled by their standard deviation. Reference
categories included (†) absence of parasites and/or (‡) year 2013. Refer to PCA in Figure 3.1 when interpreting effects of
iridescence. R2: Conditional R2.
Other variables
Iridescence
a)
Predictor
Intercept
Crown brightness
Back brightness
Back angular breadth
Crown brightness*density
Back brightness*density
Presence of parasites †
Wing length
Tarsus length
Density
Minimal age
Mass
Laying date
Year (2014) ‡
Wing length*density
Tarsus length*density
Est.
1.31
-0.03
0.15
0.15
-0.10
Annual RS
R2= 0.25
SE z value
0.11
12.41
0.04
0.72
0.04
3.51
0.04
3.55
0.04
2.37
-0.35
0.03
-0.02
-0.13
0.11
0.04
0.04
0.06
3.24
0.66
0.58
2.04
0.001
0.51
0.56
0.04
0.10
0.27
0.12
-0.12
0.04
0.09
0.04
0.04
2.20
3.05
2.62
2.89
0.03
0.002
0.01
0.004
P value
< 0.001
0.47
<0.001
<0.001
0.02
b) WPY
R2= 0.22
Est.
SE z value
0.68 0.07
9.44
0.12
-0.15
-0.26
-0.12
60
0.05
0.05
0.07
0.05
2.43
2.85
3.60
2.26
c)
P value
< 0.001
EPY
R2=0.23
Est.
SE z value
0.45 0.16
2.84
0.00 0.06
0.03
0.18 0.07
2.81
P value
0.004
0.98
0.005
-0.17
-0.16
-0.70
0.12
0.08
0.07
0.14
0.19
0.06
0.06
0.17
0.07
0.07
0.10
0.06
0.07
2.61
2.52
4.18
1.77
1.03
0.71
2.47
2.74
0.009
0.01
<0.001
0.08
0.30
0.48
0.01
0.006
0.40
0.20
-0.13
0.14
0.07
0.06
2.76
2.74
2.11
0.006
0.006
0.04
0.02
0.004
<0.001
0.02
Figure 3.2 Effect of back angular breadth on the number of WPY produced by male tree
swallows as predicted by the model shown in Table 3.1b. The solid line represents the model
prediction and dashed lines, 95% confidence intervals. For data representation, males were
grouped in five categories of back PC2 values (presented as mean ± SE).The solid line
represents the model prediction and dashed lines, 95% confidence intervals. For data
representation, males were grouped in five categories of back PC2 values (presented as mean
± SE).
61
Figure 3.3 Effect of the interaction between back brightness and breeding density on the
number of EPY produced by male tree swallows as predicted by the model shown in Table
3.1c. The predicted relationship between back brightness and number of EPY is shown at low
(1st quartile: 22 nests in a 15-km radius, dotted), medium (median 2nd quartile: 27 nests, solid
line) and high density (3rd quartile: 45 nests, dashed line).
Figure 3.4 Effect of the interaction between crown brightness and breeding density on the
annual reproductive success of male tree swallows as predicted by the model in Table 3.1a.
The predicted relationship between crown brightness and annual reproductive success is
shown at low (1st quartile: 22 nests in a 15-km radius, dotted), medium (median 2nd quartile:
27 nests, solid line) and high density (3rd quartile: 45 nests, dashed line).
62
The other variables that were associated with reproductive success matched for the most part
the effects found in Lessard et al. (2014). Presence of parasites decreased the number of EPY,
with the same effect found for the annual reproductive success. We also detected a positive
interaction between wing length and density on the number of EPY, whereas a negative
interaction between tarsus length and density on the same fitness parameter. Minimal age and
mass, variables not included in the analysis of Lessard et al. (2014), had positive effect on the
number of EPY, albeit not revealed in the annual reproductive success model (Table 3.1a and
c).
3.5.3 Iridescence and %WPY in the nest
All iridescence variables were related to male %WPY in the nest. Back brightness and angular
breadth were positively related to %WPY (Table 3.2). Moving from male with the dullest to
the brightest back and to the male with the smallest angular breadth to the largest angular
breadth corresponded to an increase of 24% and 42% in %WPY, respectively. Crown hue had
a positive effect on %WPY: moving from the male with the bluest to the greenest crown
resulted in an increase of 32%. Crown brightness was negatively related to %WPY at high
density: moving from the male with the dullest to the brightest crown resulted in a decrease of
42%. This effect lessened with decreasing density and was absent at low density. Besides
increasing the number of EPY (Table 3.1c), minimal age was negatively related to %WPY
(Table 3.2). Both laying date and number of nestlings in the nest were positively related to
%WPY. As for the number of WPY, tarsus length and density were negatively related to
%WPY.
63
Table 3.2 Effect of iridescence and other individual and environmental variables on the
proportion of within-pair young (%WPY) of male tree swallows (n=202) breeding in southern
Québec, Canada, 2013-2014. Final generalized linear mixed effect models (binomial error
distribution and logit link function) included farm identity as a random effect. All variables
were centered and scaled by their standard deviation. Refer to PCA in Figure 3.1 when
interpreting effects of iridescence. R2: Conditional R2
Other
variables
Iridescence
% WPY
R2=0.28
Predictor
Intercept
Crown brightness
Crown hue
Back brightness
Back angular breadth
Crown brightness*density
Minimal age
Tarsus length
Density
Laying date
Number of nestlings
in the nest
Est.
-1.58
-0.22
0.22
0.18
0.32
-0.23
-0.17
-0.20
-0.49
0.31
SE z value P value
0.42
3.73 <0.001
0.08
2.56
0.01
0.10
2.24
0.02
0.09
2.06
0.04
0.09
3.47 <0.001
0.09
2.65
0.01
0.08
2.07
0.04
0.08
2.52
0.01
0.15
3.18
0.002
0.09
3.37 <0.001
0.26 0.08
3.32
<0.001
3.6 Discussion
In this study, we found that different iridescence components of male tree swallows affected
reproductive success, through their effect on either the number of WPY or EPY. Notably, we
determined that the change in brightness with the movements of the animal as measured by the
angular breadth of iridescence affected reproductive success. This is the first report of a link
between this unique property of iridescence and reproductive success, suggesting a role in
paternal care signaling. We further detected interactions between breeding density and
brightness. Generally, higher density dampened the fitness benefits of bright birds, suggesting
64
that female preferences were constrained when the level of intra-sexual competition was
higher.
3.6.1 Iridescence and reproductive success
We found that the number of WPY and EPY produced explained a similar amount of variance
in male reproductive success although tree swallows have one of the highest known rate of
extra-pair paternity (reviewed in Griffith et al. 2002). Contrary to our expectation that
iridescence attributes would affect both WPY and EPY in the same direction, we showed that
they were related to either the number of WPY or EPY produced. Indeed, a male displaying
back plumage with a wider angular breadth increased his reproductive success via his number
of WPY, whereas a male with brighter back plumage increased his reproductive success by
obtaining more EPY. These positive relationships observed of either angular breadth or
brightness with their respective reproductive success component (number of WPY or EPY)
were not accompanied by a negative relationship with their complementary component
(number of EPY or WPY, respectively), thus their effect was also detected on annual
reproductive success. These results contrast with those of previous studies (Delhey et al. 2003;
McFarlane et al. 2010) that reported antagonistic effects of ornamentation on the WPY versus
EPY component of male reproductive success. Instead, our results support the contention that
coloration is a multicomponent trait (Møller and Pomiankowski 1993; Grether et al. 2004),
where the components can convey different information and thus be reflected in different
aspects of reproductive success.
Male characteristics had more influence on the number EPY than on WPY. Nonetheless, back
angular breadth increased the number of WPY, %WPY, as well as annual reproductive
success. It supports the idea that males can advertise their paternal care capacity, which should
65
be an important criterion for females seeking direct benefits from their social mate choice
(Kokko 1998). The iridescence of a male with a wider angular breadth value could be
perceived over a larger surface by his conspecifics, whereas males with narrower angular
breadth would have a larger area of its back surface perceived as black. Interestingly, in the
butterfly Colias eurytheme, the wider angular breadth of UV-emission was associated with
better condition (Kemp and Rutowski 2007). However, the first reports on that phenomenon in
birds suggested that high directionality (narrower angular breadth) would represent a male in
better condition as its feathers’ reflecting structures would be more uniformly oriented
(Madsen et al. 2007; Meadows et al. 2012). Clearly, more research should be directed towards
describing the structural mechanism that causes directionality to better understand its link with
condition and ultimately, paternal care.
The number of EPY was influenced by overall body brightness, but this relationship was
variable depending on breeding density. Contrary to our expectations, we found a positive
effect of back brightness on the number of EPY at low density fading with increasing density.
This result contrasts with a study on common yellowthroats, where the selection on plumage
attributes via the number of EPY was present only at high density. The authors suggested that
a high density may increase the capacity of females to assess and compare male plumage and
select the preferred one (Taff et al. 2013). In tree swallows, females have a high control over
extra-pair fertilizations (Leffelaar and Robertson 1984; Lifjeld et al. 1993; Venier et al. 1993;
Whittingham et al. 1993) and a study conducted in a different system than ours also found that
bright males tree swallow were more successful at obtaining EPY (Bitton et al. 2007). We can
therefore speculate that brightness could be preferred by females only at low breeding
densities because of the high cost of preference when intra-specific competition is strong.
Indeed, the cost of refusing an extra-pair copulation may be cheap (Venier et al. 1993), but
with an increasing density, the cumulative cost may be high. For instance, females could
engage in extra-pair copulations independently of male coloration at high density in order to
limit sexual harassment or repeated aggressive attempts from males, as often observed in
66
mammals (Wolff and MacDonald 2004). In birds, male coercion is thought to influence
female extra-pair behavior (Westneat and Stewart 2003; Arnqvist and Kirkpatrick 2005),
although such a constraint was reported in only few cases (Low 2005; Wells et al. 2014).
Moreover, higher densities typically favor aggressiveness (Knell 2009), which in male and
female tree swallow is an important behavior shaped by limited nesting sites (Robertson et al.
1992). If females at higher densities allocate more time to aggressive behavior related to
nesting site defense (Leffelaar and Robertson 1984), less time and energy are left for the
choice of extra-pair partners (Dunn and Whittingham 2007).
3.6.2 Iridescence and %WPY in the nest
Modeling the effect of male iridescence on %WPY rather than on the absolute number of
young in the nest provides a new angle regarding the importance of female choice on
iridescence. All iridescence variables affected %WPY in our system: back brightness, back
angular breadth and crown hue all increased %WPY, while crown brightness decreased
%WPY at high density. The fact that iridescence metrics that increased fitness by acting on the
number of WPY or EPY also increased %WPY supports female preference for these signals.
Ornamentation has indeed been found to increase %WPY in other species (Balenger et al.
2008; Reudink et al. 2009; Jacobs et al. 2014). Male iridescence could influence %WPY
through female choice, but male sperm competition could also play a role given that
copulation frequency of male tree swallows with their social mate has recently been shown to
increase with the proportion of paternities they own within their social nest (Crowe et al.
2009). More ornamented males may thus have more energy to allocate to fertilize their mate.
Our study also revealed that crown iridescence was an important determinant of %WPY and
the number of EPY independently of back iridescence, and that these effects were influenced
by breeding density. These results support our prediction that crown iridescence can be used
as a sexual signal, in accordance with our field observations of males erecting their crown
67
towards females (see Supplementary Figure 3.7). Our study thus stresses the importance of
sampling different body regions even if differences are not necessarily perceived by the
human eye. The fact that the relationship between crown brightness and %WPY as well as
with the number of EPY and annual reproductive success became negative at high breeding
density is unexpected and remains unclear. We suggest two alternative mechanisms that could
explain this result. First, it is possible that at high density, males with a duller crown allocate
less time in aggressive behaviors in favor of pursuing more extra-pair copulations.
Alternatively, female preferences could be plastic and depend on the social environment
(Rodríguez et al. 2013; Atwell and Wagner Jr 2014), favoring bright crowns at low density
and dull crowns at high density. More studies are required to test these two possibilities.
3.6.3 Methodological considerations
Coloration can be quantified using several metrics (Montgomerie 2006) and the amount of
redundancy between those metrics varies widely among species. Researchers typically use
Principal Component Analysis (PCA) to reduce the number of variables in subsequent
analyses, as we did. The drawback of using a PCA in our case was that back hue contributed
equally to PC1 and PC2, and its effect could therefore not be clearly disentangled from that of
the other iridescence variables by our analysis. Nevertheless, we can say that males with a
greener back have a higher overall reproductive success (because both back PC1 and PC2
were selected in our final model), and that males with greener crown had a higher %WPY.
Hue was highly correlated with UV-chroma (S1 in Montgomerie (2006), r=0.8, data not
shown), because the greener a feather is, the more shifted away the spectrum peak is from the
UV-region. The positive effect of greener hue on reproductive success contrasts substantially
with many studies of structural coloration, where males with a more UV-shifted hue, or higher
UV-chroma, were in better condition (Keyser and Hill 2000; Siitari and Huhta 2002; Doucet
and Montgomerie 2003) or had fitness benefits (Balenger et al. 2008; Lehtonen et al. 2009;
68
O’Brien and Dawson 2010). Importantly, however, the species studied generally have a
plumage with a considerable contribution of reflectance in the UV region. Tree swallow’s
spectrum on the other hand reaches a minimum in the UV region. It may therefore be
important to consider longer wavelengths when studying species showing a low contribution
of UV reflectance, as such coloration attributes could contribute to signaling.
3.6.4 Conclusion
Altogether, our results showed that in a socially monogamous bird, both components of
reproductive success (WPY and EPY) can be affected by sexual selection through coloration
traits. We also reported that sexual selection might be variable depending on breeding density.
Taking into account ecological heterogeneity will help to better understand why selection
patterns are so variable even among studies of the same species. Importantly, this study
pointed towards new questions that will benefit to the understanding of iridescence evolution.
We suggested that directionality may play a role in the signaling of parental care, but we have
yet to assess the mechanism behind it. Studying how feather microstructure causes variation in
directionality independently from brightness will help validating the role of iridescence as a
multicomponent trait.
3.7 Aknowledgements
We are grateful to Myriam Cadotte for helping with feathers analyses. We thank all the
graduate students and field assistants who helped collect feathers, as well as the farm owners
who gave us access to their land. This work was supported by the Natural Sciences and
Engineering Research Council of Canada (NSERC) discovery grants to D.G., M.B., and F.P.,
69
the Fonds de Recherche du Québec – Nature et Technologies (FRQNT) “Projet de recherche
en équipe” grant awarded to D.G., M.B., and F.P., and by the Canada Research Chair held by
F.P. and M.B. S.V.W was supported by NSERC and FRQNT scholarships.
3.8 References
Able DJ (1996) The contagion indicator hypothesis for parasite-mediated sexual selection.
Proc Natl Acad Sci U S A 93:2229–33.
Akçay E, Roughgarden J (2007) Extra-pair paternity in birds : review of the genetic benefits.
Evol Ecol Res 9:855–868.
Andersson M (1994) Sexual Selection. Princeton University Press, Princeton, New Jersey
Arnqvist G, Kirkpatrick M (2005) The evolution of infidelity in socially monogamous
passerines: the strength of direct and indirect selection on extrapair copulation behavior
in females. Am Nat 165 Suppl :S26–S37. doi: 10.1086/429350
Atwell A, Wagner Jr WE (2014) Female mate choice plasticity is affected by the interaction
between male density and female age in a field cricket. Anim Behav 98:177–183. doi:
10.1016/j.anbehav.2014.10.007
Balenger SL, Johnson LS, Masters BS (2008) Sexual selection in a socially monogamous bird:
male color predicts paternity success in the mountain bluebird, Sialia currucoides. Behav
Ecol Sociobiol 63:403–411. doi: 10.1007/s00265-008-0674-5
Bateman AJ (1948) Intra-sexual selection in Drosophila. Heredity 2:349–368.
Bates D, Maechler M, Bolker BM (2014) lme4: Linear mixed-effects models using Eigen and
S4.
Bennett ATD, Cuthill IC, Partridge JC, Lunau K (1997) Ultraviolet plumage colors predict
mate preferences in starlings. Proc Natl Acad Sci U S A 94:8618–8621.
Bergeron P, Baeta R, Pelletier F, et al (2011) Individual quality: tautology or biological
reality? J Anim Ecol 80:361–4. doi: 10.1111/j.1365-2656.2010.01770.x
70
Bitton P-P, O’Brien EL, Dawson RD (2007) Plumage brightness and age predict extrapair
fertilization success of male tree swallows, Tachycineta bicolor. Anim Behav 74:1777–
1784. doi: 10.1016/j.anbehav.2007.03.018
Cleasby IR, Nakagawa S (2012) The influence of male age on within-pair and extra-pair
paternity in passerines. Ibis 154:318–324. doi: 10.1111/j.1474-919X.2011.01209.x
Cohen R (1984) Criteria for distinguishing breeding male tree swallows from brightly colored
females prior to capture. North Am Bird Bander 9:2–3.
Crawley M (2007) The R book. Wiley, West Sussex
Crowe SA., Kleven O, Delmore KE, et al (2009) Paternity assurance through frequent
copulations in a wild passerine with intense sperm competition. Anim Behav 77:183–
187. doi: 10.1016/j.anbehav.2008.09.024
Dakin R, Montgomerie R (2013) Eye for an eyespot: how iridescent plumage ocelli influence
peacock mating success. Behav Ecol 24:1048–1057. doi: 10.1093/beheco/art045
Darwin C (1871) The descent of man, and selection in relation to sex. Appleton, New York
Delhey K, Johnsen A, Peters A, et al (2003) Paternity analysis reveals opposing selection
pressures on crown coloration in the blue tit (Parus caeruleus). Proc Biol Sci 270:2057–
63. doi: 10.1098/rspb.2003.2460
Doucet M, Montgomerie R (2003) Multiple sexual ornaments in satin bowerbirds : ultraviolet
plumage and bowers signal different aspects of male quality. Behav Ecol 14:503–509.
Dunn PO, Whittingham LA (2007) Search costs influence the spatial distribution, but not the
level, of extra-pair mating in tree swallows. Behav Ecol Sociobiol 61:449–454. doi:
10.1007/s00265-006-0272-3
Fitzpatrick S (1998) Colour schemes for birds : structural coloration and signals of quality in
feathers. Ann Zool Fenn 35:67–77.
Ghilain A, Bélisle M (2008) Breeding success of tree swallows along a gradient of agricultural
intensification. Ecol Appl 18:1140–1154.
Grether GF, Kolluru GR, Nersissian K (2004) Individual colour patches as multicomponent
signals. Biol Rev Camb Philos Soc 79:583–610. doi: 10.1017/S1464793103006390
Griffith SC, Owens IPF, Thuman K (2002) Extra pair paternity in birds : a review of
interspecific variation and adaptative function. Mol Ecol 11:2195–2212.
71
Hailman J (1977) Optical signals: animal communication and light. Indiana University Press,
Bloomington
Hamilton WD, Zuk M (1982) Heritable true fitness and bright birds: a role for parasites?
Science 218:384–7.
Hart NS (2001) The visual ecology of avian photoreceptors. Prog Retin Eye Res 20:675–703.
Hill GE (2006) Female mate choice for ornamental coloration. In Bird Coloration Vol.2.
Function and Evolution, Hill GE, McGraw KJ, eds. (Cambridge: Harvard University
Press), pp 137–200
Hill GE, Doucet SM, Buchholz R (2005) The effect of coccidial infection on iridescent
plumage coloration in wild turkeys. Anim Behav 69:387–394. doi:
10.1016/j.anbehav.2004.03.013
Hoelzer GA. (1989) The good parent process of sexual selection. Anim Behav 38:1067–1078.
doi: 10.1016/S0003-3472(89)80146-0
Houde AE, Endler JA (1990) Correlated evolution of female mating preferences and male
color patterns in the guppy Poecilia reticulata. Science 248:1405–8. doi:
10.1126/science.248.4961.1405
Jacobs AC, Fair JM, Zuk M (2014) Coloration, paternity, and assortative mating in western
bluebirds. Ethology 120:1–11. doi: 10.1111/eth.12327
Jennions MD, Petrie M (2000) Why do females mate multiply? A review of the genetic
benefits. Biol Rev Camb Philos Soc 75:21–64.
Kalinowski ST, Taper ML, Marshall TC (2007) Revising how the computer program CERVUS
accommodates genotyping error increases success in paternity assignment. Mol Ecol
16:1099–106. doi: 10.1111/j.1365-294X.2007.03089.x
Kemp DJ, Rutowski RL (2007) Condition dependence, quantitative genetics, and the potential
signal content of iridescent ultraviolet butterfly coloration. Evolution 61:168–83. doi:
10.1111/j.1558-5646.2007.00014.x
Kemp DJ, Vukusic P, Rutowski RL (2006) Stress-mediated covariance between nanostructural architecture and ultraviolet butterfly coloration. Funct Ecol 20:282–289. doi:
10.1111/j.1365-2435.2006.01100.x
Kempenaers B, Everding S, Bishop C, et al (2001) Extra-pair paternity and the reproductive
role of male floaters in the tree swallow (Tachycineta bicolor). Behav Ecol Sociobiol
49:251–259. doi: 10.1007/s002650000305
72
Keyser AJ, Hill GE (2000) Structurally based plumage coloration is an honest signal of quality
in male blue grosbeaks. Behav Ecol 11:202–209.
Knell RJ (2009) Population density and the evolution of male aggression. J Zool 278:83–90.
doi: 10.1111/j.1469-7998.2009.00566.x
Kokko H (1998) Should advertising parental care be honest? Proc R Soc B Biol Sci 265:1871–
1878. doi: 10.1098/rspb.1998.0515
Kokko H, Rankin DJ (2006) Lonely hearts or sex in the city? Density-dependent effects in
mating systems. Philos Trans R Soc Lond B Biol Sci 361:319–334. doi:
10.1098/rstb.2005.1784
Leclaire S, Pierret P, Chatelain M, Gasparini J (2014) Feather bacterial load affects plumage
condition, iridescent color, and investment in preening in pigeons. Behav Ecol 25:1192–
1198. doi: 10.1093/beheco/aru109
Leffelaar D, Robertson RJ (1984) Do male tree swallows guard their mates ? Behav Ecol
Sociobiol 16:73–79.
Legagneux P, Théry M, Guillemain M, et al (2010) Condition dependence of iridescent wing
flash-marks in two species of dabbling ducks. Behav Processes 83:324–30. doi:
10.1016/j.beproc.2010.01.017
Lehtonen PK, Primmer CR, Laaksonen T (2009) Different traits affect gain of extrapair
paternity and loss of paternity in the pied flycatcher, Ficedula hypoleuca. Anim Behav
77:1103–1110. doi: 10.1016/j.anbehav.2009.01.014
Lessard A, Bourret A, Bélisle M, et al (2014) Individual and environmental determinants of
reproductive success in male tree swallow (Tachycineta bicolor). Behav Ecol Sociobiol
68:733–742. doi: 10.1007/s00265-014-1686-y
Lifjeld J, Dunn P, Robertson RJ, Boag PT (1993) Extra-pair paternity in monogamous tree
swallows. Anim Behav 45:213–229.
Low M (2005) Female resistance and male force: Context and patterns of copulation in the
New Zealand stitchbird Notiomystis cincta. J Avian Biol 36:436–448. doi:
10.1111/j.0908-8857.2005.03460.x
Loyau A, Gomez D, Moureau B, et al (2007) Iridescent structurally based coloration of
eyespots correlates with mating success in the peacock. Behav Ecol 18:1123–1131. doi:
10.1093/beheco/arm088
73
Madsen V, Dabelsteen T, Osorio D, Osorno JL (2007) Morphology and ornamentation in male
magnificent frigatebirds: variation with age class and mating status. Am Nat 169
Suppl:S93–S111. doi: 10.1086/510096
Maia R, Eliason CM, Bitton P-P, et al (2013) PAVO : an R package for the analysis,
visualization and organization of spectral data. Methods Ecol Evol 4:906–913. doi:
10.1111/2041-210X.12069
Marshall T, Slate J, Kruuk L, Pemberton J (1998) Statistical confidence for likelihood-based
paternity inferrence in natural populations. Mol Ecol 7:639–655.
McFarlane ML, Evans MR, Feldheim KA, et al (2010) Long tails matter in sugarbirds positively for extrapair but negatively for within-pair fertilization success. Behav Ecol
21:26–32. doi: 10.1093/beheco/arp147
McGraw KJ, Mackillop EA, Dale J, Hauber ME (2002) Different colors reveal different
information: how nutritional stress affects the expression of melanin- and structurally
based ornamental plumage. J Exp Biol 205:3747–55.
Meadows MG, Morehouse NI, Rutowski RL, et al (2011) Quantifying iridescent coloration in
animals: a method for improving repeatability. Behav Ecol Sociobiol 65:1317–1327. doi:
10.1007/s00265-010-1135-5
Meadows MG, Roudybush TE, McGraw KJ (2012) Dietary protein level affects iridescent
coloration in Anna’s hummingbirds, Calypte anna. J Exp Biol 215:2742–2750. doi:
10.1242/jeb.069351
Milinski M, Bakker TCM (1990) Female sticklebacks use male coloration in mate choice and
hence avoid parasitized males. Nature 344:330–333. doi: 10.1038/344330a0
Møller A, Birkhead TR (1994) The Evolution of Plumage Brightness in Birds is Related to
Extrapair Paternity. Evolution 48:1089–1100.
Møller A, Pomiankowski A (1993) Why have birds got multiple sexual ornaments ? Behav
Ecol Sociobiol 32:167–176.
Montgomerie R (2006) Analyzing Colors. In Bird Coloration Vol. 1. Mechanisms and
Measurements, Hill GE, McGraw KJ, eds. (Cambridge: Harvard University Press), pp
90–147
Nakagawa S, Schielzeth H (2013) A general and simple method for obtaining R 2 from
generalized linear mixed-effects models. Methods Ecol Evol 4:133–142. doi:
10.1111/j.2041-210x.2012.00261.x
74
Neff BD, Pitcher TE (2005) Genetic quality and sexual selection: an integrated framework for
good genes and compatible genes. Mol Ecol 14:19–38. doi: 10.1111/j.1365294X.2004.02395.x
O’Brien EL, Dawson RD (2010) Plumage color and food availability affect male reproductive
success in a socially monogamous bird. Behav Ecol 22:66–72. doi:
10.1093/beheco/arq167
O’Brien EL, Dawson RD (2007) Context-dependent genetic benefits of extra-pair mate choice
in a socially monogamous passerine. Behav Ecol Sociobiol 61:775–782. doi:
10.1007/s00265-006-0308-8
Osorio D, Ham AD (2002) Spectral reflectance and directional properties of structural
coloration in bird plumage. J Exp Biol 205:2017–27.
R Core Team (2014) R: A language and environment for statistical computing.
Rahman MM, Kelley JL, Evans JP (2013) Condition-dependent expression of pre- and
postcopulatory sexual traits in guppies. Ecol Evol 3:2197–213. doi: 10.1002/ece3.632
Reudink MW, Marra PP, Boag PT, Ratcliffe LM (2009) Plumage coloration predicts paternity
and polygyny in the American redstart. Anim Behav 77:495–501. doi:
10.1016/j.anbehav.2008.11.005
Rioux Paquette S, Pelletier F, Garant D, Bélisle M (2014) Severe recent decrease of adult
body mass in a declining insectivorous bird population. Proc R Soc B Biol Sci 281.
Robertson R, Stutchbury B, Cohen R (1992) Tree Swallow. In: Poole A, Stettenheim P, Gill F
(eds) Birds North Am. Philadelphia, p No. 11
Robinson MR, van Doorn GS, Gustafsson L, Qvarnström A (2012) Environment-dependent
selection on mate choice in a natural population of birds. Ecol Lett 15:611–8. doi:
10.1111/j.1461-0248.2012.01780.x
Rodríguez RL, Rebar D, Fowler-Finn KD (2013) The evolution and evolutionary
consequences of social plasticity in mate preferences. Anim Behav 85:1041–1047. doi:
10.1016/j.anbehav.2013.01.006
Savard J-F, Keagy J, Borgia G (2011) Blue, not UV, plumage color is important in satin
bowerbird Ptilonorhynchus violaceus display. J Avian Biol 42:80–84. doi:
10.1111/j.1600-048X.2010.05128.x
Schielzeth H (2010) Simple means to improve the interpretability of regression coefficients.
Methods Ecol Evol 1:103–113. doi: 10.1111/j.2041-210X.2010.00012.x
75
Siitari H, Huhta E (2002) Individual color variation and male quality in pied flycatchers
(Ficedula hypoleuca): a role of ultraviolet reflectance. Behav Ecol 13:737–741.
Stamps JA. (2011) Density bias in behavioral ecology. Behav Ecol 22:231–232. doi:
10.1093/beheco/arq174
Taff CC, Freeman-Gallant CR, Dunn PO, Whittingham LA (2013) Spatial distribution of nests
constrains the strength of sexual selection in a warbler. J Evol Biol 26:1392–405. doi:
10.1111/jeb.12141
Van Wijk S, Bélisle M, Garant D, Pelletier F (in review) A reliable technique to quantify
iridescence in birds. J. Avian Biol.
Venier L, Dunn PO, Lifjeld JT, Robertson RJ (1993) Behavioural patterns of extra-pair
copulation in tree swallows. Anim Behav 45:412–415.
Weatherhead PJ, Robertson RJ (1979) Offspring quality and the polygyny treshold: “the sexy
son hypothesis.” Am Nat 113:201–208.
Webster MS, Pruett-Jones S, Westneat D, Arnold SJ (1995) Measuring the effects of pairing
success, extra-pair copulations and mate quality on the opportunity for sexual selection.
Evolution 49:1147–1157.
Wells DA., Jones DN, Bulger D, Brown C (2014) Male brush-turkeys attempt sexual coercion
in
unusual
circumstances.
Behav
Processes
106:180–186.
doi:
10.1016/j.beproc.2014.06.002
Westneat DF, Sherman PW, Morton M (1990) The ecology and evolution of extra-pair
copulations in birds. Curr Ornithol 7:331–369.
Westneat DF, Stewart IRK (2003) Extra-pair paternity in birds : causes, correlates, and
conflict. Annu Rev Ecol Evol Syst 34:365–396. doi: 10.1146/132439
Whittingham L, Dunn PO, Robertson RJ (1993) Confidence of paternity and male parental
care: an experiment study in tree swallow. Anim Behav 46:139–147.
Whittingham LA, Dunn PO (2005) Effects of extra-pair and within-pair reproductive success
on the opportunity for selection in birds. Behav Ecol 16:138–144. doi:
10.1093/beheco/arh140
Whittingham LA, Dunn PO (2014) Extra-pair mating and sexual selection on male traits
across populations. Wilson J Ornithol 126:9–18.
76
Wolff JO, MacDonald DW (2004) Promiscuous females protect their offspring. Trends Ecol
Evol 19:127–134. doi: 10.1016/j.tree.2003.12.009
Zahavi A (1975) Mate Selection - A Selection for a Handicap. J Theor Biol 53:205–214.
Zirpoli JA, Black JM, Gabriel PO (2013) Parasites and plumage in Steller’s jays : an
experimental field test of the parasite-mediated handicap hypothesis. Ethol Ecol Evol
25:103–116.
77
ANNEXE
Electronic Supplementary Material
Breeding density influences the sexual selection of iridescence in male tree swallows,
Tachycineta bicolor
Sonia Van Wijk, Audrey Bourret, Marc Bélisle, Dany Garant and Fanie Pelletier
Table 3.3 PC loadings of iridescence variables (hue, brightness, saturation and directionality)
from a principal component analysis performed separately on 2 body regions: crown and back.
Percent variance explained by PC1 and PC2 are shown last.
Crown
PC1
Back
PC2
PC1
PC2
Hue
-0.01 0.95
0.50
0.61
Brightness
0.86
0.12
0.86
-0.17
Saturation
0.84
0.25
0.84
0.07
Directionality
0.74
0.41
-0.26
0.85
% of variance
50
29
44
28
78
Table 3.4 Mean and standard deviation of iridescence and other individual and environmental
variables in the 202 male tree swallows used to model the influence of iridescence on
reproductive success. Iridescence variables are the four variables entered in PCA (hue,
brightness, saturation, directionality).
Variable
Mean
Standard Deviation
Crown Hue (nm)
497.82
14.56
Crown Brightness (%)
16.35
2.27
Crown Saturation
0.85
0.14
Crown Directionality (°)
36.22
3.21
Back Hue (nm)
473.42
13.84
Back Brightness (%)
18.33
2.32
Back Saturation
0.92
0.13
Back Directionality (°)
30.26
1.82
Minimal Age (years)
1.57
1.07
Mass (g)
19.96
1.10
Wing Length (mm)
118.45
2.95
Tarsus Length (mm)
12.15
0.43
Density (Nb. of occupied nests in a 15-km radius)
30.64
15.79
Laying Date (Julian Day)
141.40
8.72
Iridescence
Other variables
79
Figure 3.5 Variation in plumage iridescence variables measured on 202 male tree swallows
breeding in southern Québec, Canada, 2013-2014: a) crown brightness, b) crown hue, c) back
brightness, and d) back directionality between years. Boxplots indicate 1st and 3rd quartiles
(edges), median (line) and 1.5 times the interquartile range (whiskers).
80
Figure 3.6 Correlation matrix of explanatory variables included in models explaining
reproductive success of 202 male tree swallows breeding in southern Québec, Canada, 20132014. More intense color and narrower ellipses on the upper panel represent stronger
correlation. Positive and negative correlations are represented by blue and red, respectively.
Correlation coefficients are reported on the lower panel. The package corrplot (Wei, 2013)
was used with R to draw the figure.
81
Figure 3.7 A social pair of tree swallows. Male erecting his crown feathers (left) towards a
female SY (right).
References
Wei T (2013) CORRPLOT:Visualization of a correlation matrix.
82
CHAPITRE 4
DISCUSSION GÉNÉRALE ET CONCLUSION
4.1 Retour sur les résultats
4.1.1 Mesure et répétabilité de l’iridescence
Dans la première étape du projet, j’ai dirigé la fabrication de l’appareil de mesure décrit dans
Meadows et al. (2011) afin de quantifier l’iridescence des mâles Hirondelle bicolore de
manière répétable. Dans un sous-échantillon de 30 individus, les mesures répétées sur les
mêmes plumes ont montré que l’appareil de mesure permettait d’obtenir des mesures
répétables des variables d’iridescence (R=0,48-0,89). Meadows et al. (2011) avait testé
l’appareil seulement chez une espèce de colibri, dont la structure des barbules amène une
directionnalité rarement observée chez les oiseaux (Greenewalt et al., 1960). Notre
contribution réside donc dans la vérification de la répétabilité de l’appareil chez une espèce,
l’Hirondelle bicolore, dont l’iridescence se compare mieux avec un plus grand nombre
d’espèces. Dans l’échantillon complet de mâles des deux années d’étude, les mesures de trois
plumes différentes d’une même région ont permis de vérifier que l’appareil permettait de
capter des différences individuelles dans toutes les composantes d’iridescence mesurées, dont
la directionnalité. J’ai ainsi effectué une étape essentielle dans l’étude de cette composante
unique de l’iridescence, promouvant l’incorporation de cette variable dans de futures études
sur sélection sexuelle de l’iridescence.
83
J’ai aussi découvert qu’une étape dans le procédé de quantification permettait de limiter un
biais de mesure. En fait, les barbules des plumes iridescentes des mâles Hirondelle bicolore
sont généralement parallèles au plan de la plume, comme le montre la Figure 1.3. J’ai vérifié
si l’inclinaison des barbules était bel et bien nulle pour tous les individus. Bien que la
moyenne de l’inclinaison était en effet 0 (comme c’est le cas pour l’individu 1 dans la Figure
4.1a), j’ai observé de légères différences individuelles dans cette mesure. J’ai montré que ces
différences peuvent amener un biais, car si cette inclinaison n’est pas contrôlée, les mâles
ayant une inclinaison extrême peuvent être classés comme étant parmi les moins brillants alors
qu’en réalité ils étaient parmi les plus brillants. En effet, si on n’effectue pas la rotation de la
plume afin de contrôler cette inclinaison, le spectre obtenu a une réflexion nettement inférieure
à la réalité puisque l’appareil ne capte pas le spectre à l’angle où le plus de barbules sont
alignées vers le récepteur (Figure 4.1b et c). Avec ce résultat, j’ai mis l’emphase sur
l’importance de considérer l’inclinaison des barbules afin d’obtenir une mesure non-biaisée de
l’iridescence. J’ai constaté que la variabilité en directionnalité était plus grande que la
variabilité en inclinaison des barbules. Comme la directionnalité est ainsi plus susceptible
d’être détectée par les congénères, j’ai ajouté seulement cette variable dans l’étude de la
sélection sexuelle de l’iridescence chez l’Hirondelle bicolore. De cette manière, j’ai suivi le
principe de parcimonie en limitant le nombre de variables de coloration à quatre. Bien
entendu, la variabilité de l’inclinaison des barbules que j’ai détectée pourrait aussi être
impliquée dans la sélection sexuelle, et mérite qu’on s’y attarde dans de futures études.
84
Figure 4.1 Contrôle de l’angle d’inclinaison des barbules. a) Individu 1 dont les barbules ont
une inclinaison nulle. b) Individu 2 avec des barbules inclinées et chez qui aucune rotation n’a
été effectuée. c) Individu 2 chez qui la rotation a été effectuée afin de contrôler l’inclinaison
des barbules. θ correspond à l’angle d’inclinaison des barbules et à la rotation devant être
effectuée. Ligne noire pleine épaisse superposée de traits : plan de la plume et barbules. Ligne
noire pleine mince : Normale par rapport aux barbules. En jaune : faisceau d’illumination. En
bleu : Réflexion de la plume. La sonde grise et rouge est la fibre de réception du
spectrophotomètre. Pour des fins d’illustration, le faisceau n’illumine qu’une barbule, alors
qu’en réalité il en éclaire plusieurs.
Finalement, en ayant collecté des plumes de trois régions différentes, soit la couronne, le
manteau et le croupion, j’ai pu vérifier les différences de coloration et de répétabilité entre
celles-ci. J’ai observé qu’aucune des régions n’avait une répétabilité supérieure aux autres
pour l’ensemble des variables d’iridescence. Ces trois régions ont donc pu être incorporées
dans mon étude de sélection sexuelle. J’ai aussi observé que l’iridescence de la couronne
différait plus du croupion que le manteau du croupion, en étant plus verte, moins brillante,
moins saturée et en ayant une plus faible directionnalité, malgré des différences peu
perceptibles à l’œil humain. Étant donné que les oiseaux ont une vision plus élaborée que la
nôtre en ayant entre autres un cône de plus que les humains permettant la vision des rayons
UV (Cuthill, 2006), il est important de ne pas baser la quantification de ces régions seulement
sur l’appréciation humaine.
85
4.1.2 L’iridescence et le succès reproducteur des mâles
En 2013 et 2014, nous avons récolté les plumes iridescentes des mâles et obtenu le nombre
d’oisillons dans le couple et hors couple à l’aide d’analyses génétiques. Avec ces données, j’ai
pu vérifier comment l’iridescence influençait les deux composantes du succès reproducteur
des mâles. J’ai aussi ajouté deux autres composantes : le succès reproducteur annuel, puisque
c’est ce paramètre qui donne la meilleure information quant aux phénotypes sélectionnés, et le
pourcentage d’oisillons dans le couple du nid, qui représente mieux la préférence de la
femelle. Les variables que j’ai entrées dans les modèles sont des composantes principales
d’une analyse en composantes principales (ACP) effectuée sur deux zones du corps, soit la
couronne et le dos (regroupant manteau et croupion). J’ai jugé que cette méthode était la plus
adéquate, puisqu’elle avait l’avantage de diminuer le nombre de variables d’iridescence et
ainsi permettre de faire un seul modèle comprenant à la fois les variables de la couronne et
celles du dos. Contrairement à mon hypothèse de départ voulant que les mâles plus bleus,
brillants et saturés auraient un meilleur succès à la fois au niveau du nombre de jeunes dans le
couple et hors couple, ce ne sont pas les mêmes variables de l’iridescence qui affectaient les
deux composantes du succès reproducteur. En fait, l’étendue angulaire d’iridescence, soit la
nouvelle variable quantifiant le changement de brillance avec un mouvement de l’animal,
augmentait le nombre d’oisillons dans le couple produits par le mâle, et cela se reflétait aussi
dans le succès reproducteur annuel. D’un autre côté, les mâles brillants avaient un meilleur
succès hors couple que les mâles peu brillants à faible densité seulement et cela se reflétait
aussi dans le succès reproducteur annuel. Ces variables d’iridescence augmentant le succès
reproducteur augmentaient aussi le pourcentage de jeunes dans le couple, ce qui supporte la
préférence de la femelle pour ces signaux. Dans l’ensemble, cela suggère que les critères des
femelles pour choisir leur partenaire social ne sont pas les mêmes que pour leur(s)
partenaire(s) hors couple. L’étendue angulaire pourrait ainsi représenter la capacité du mâle à
donner des soins parentaux, une contribution importante du mâle pour la survie et l’envol des
oisillons. Le meilleur succès hors couple des mâles brillants à faible densité de reproducteurs
porte à penser que la préférence des femelles pour la brillance est compromise à haute densité.
86
Une haute densité pourrait limiter le choix des femelles en occasionnant un coût cumulatif de
refuser des copulations hors couple, par exemple à cause des comportements agressifs des
femelles lors de ces refus. Le mécanisme derrière la préférence des femelles pour les mâles
brillants peut simplement être un biais sensoriel préexistant pour la brillance ayant évolué
autrement que sous sélection sexuelle (Ryan et Keddy-Hector, 1992) ou la théorie des bons
gènes, puisque de manière générale la brillance est corrélée avec la condition (Hill et al., 2005;
Keyser et Hill, 2000; McGraw et al., 2002; Zirpoli et al., 2013). Une littérature abondante ne
supporte pas la théorie des bons gènes chez l’Hirondelle bicolore, puisqu’aucune différence en
masse, taille, immunité ou survie entre les jeunes dans le couple et les jeunes hors couple n’a
été détectée (Dunn et al., 2009; Kempenaers et al., 1999; Whittingham et Dunn, 2001). Ainsi,
les bénéfices de choisir un mâle brillant pourraient ne pas être détectés au stade oisillon, mais
au stade adulte. Les jeunes dans le couple pourraient bénéficier d’une survie à un an plus
élevée ou pourraient obtenir les gènes de brillance combinés aux bénéfices de la préférence
des femelles, comme prédit par l’hypothèse du fils sexy (Weatherhead et Robertson, 1979).
4.2 Perspectives futures
Du volet technique de mon projet a émergé une réflexion approfondie à propos de la méthode
optimale pour quantifier la variation de couleur perçue par l’oiseau. J’ai utilisé des variables
de couleur, brillance et saturation, trois variables caractérisant ce que l’œil humain perçoit. La
similarité du fonctionnement de l’œil aviaire et de l’œil humain rend ces variables aussi
pertinentes à utiliser pour l’ensemble des oiseaux (Montgomerie, 2006). Comme les oiseaux
perçoivent les UV, il s’agit simplement de calculer ces variables en prenant en compte
l’émission dans les UV (Montgomerie, 2006). Mais l’œil aviaire comporte d’autres
différences, comme la présence de gouttes d’huile spécifiques à chaque cône qui filtreraient
différemment les longueurs d’onde ainsi que la présence d’un double-cône (Cuthill, 2006).
Compte tenu de ces différences, et que la meilleure estimation de la couleur vue par l’oiseau
87
est de prendre en compte la lumière ambiante, une technique de quantification sophistiquée
nommée modèle tétrahédral a été mise de l’avant. Cette technique permet de quantifier le
nombre de photons excitant chaque cône de l’œil (Endler et Mielke, 2005). Appliquer cette
technique sur l’Hirondelle bicolore requiert d’approximer le fonctionnement de son œil
comme étant celui de l’espèce étudiée la plus proche génétiquement, soit la Mésange bleue
(Cyanistes caeruleus) et de déterminer quel spectre de lumière ambiante est présent lors de la
perception de la couleur. J’ai exploré cette technique avec le logiciel
PAVO
(Maia et al., 2013)
et remarqué que les variables retirées du modèle tétrahédral étaient fortement corrélées aux
variables traditionnelles (coefficient de corrélation; couleur : 0,93, brillance : 0,94, saturation :
0,83). Étant donné qu’un modèle tétrahédral avec l’Hirondelle bicolore fait beaucoup de
suppositions à propos du fonctionnement de son œil, j’ai appliqué le principe de précaution et
conservé la méthode traditionnelle. Le modèle tétrahédral est la voie du futur, mais je suggère
de l’utiliser que lorsque les connaissances sur l’œil des oiseaux ainsi que sur les différences
entre espèces seront plus approfondies. Ce modèle permettra aussi de déterminer si la
coloration de deux régions similaires à l’œil humain sont distinguables ou non pour un oiseau
(Endler et Mielke, 2005).
L’utilisation de l’appareil de mesure proposé par Meadows et al., (2011) m’a fait réaliser les
limites de celui-ci. En particulier, j’ai remarqué que les différences en étendue angulaire entre
les régions et entre les années étaient peu fiables. Ceci est dû à l’alignement des faisceaux de
la lumière et du récepteur. Pour passer de la mesure des plumes de croupion aux plumes de
couronne, il est nécessaire de diminuer la taille des faisceaux car la zone de mesure est plus
petite. Mais cette étape requiert de réaligner les faisceaux puisque l’axe de rotation n’est pas
parfait. Le réalignement se fait à l’œil humain, donc entre chaque réalignement, les faisceaux
ne sont pas alignés exactement de la même manière. Cela fait en sorte que lorsqu’on effectue
une rotation de la plume pour mesurer l’étendue angulaire, il peut y avoir un léger décalage
entre les deux faisceaux plus ou moins grand dépendant de l’alignement. Ainsi, lorsqu’un
réalignement s’effectue entre deux séries de mesures, la comparaison de la moyenne d’étendue
88
angulaire des deux séries est peu fiable. Par contre, les différences entre individus d’une même
série restent fiables, et c’est ce qui était important pour mon projet. Cette constatation est
importante à considérer pour l’utilisation future de cette variable. En effet, je n’ai pas observé
de désalignement graduel durant une série de mesures, mais cela pourrait être le cas avec le
vieillissement de l’appareil.
Du volet biologique de mon projet a émergé plusieurs questions auxquelles il m’apparait
important de répondre dans le futur. Il s’agit premièrement de la condition-dépendance de
l’iridescence. Le lien entre les différentes variables et la microstructure causant l’iridescence
commence à être étudiée (Doucet et al., 2006; Maia et al., 2012), mais beaucoup d’éléments
sont manquants. Mon étude a permis de voir que la brillance et l’étendue angulaire sont
impliquées différemment dans la sélection sexuelle, l’étape suivante est de savoir si elles
signalent différents aspects de la condition des individus reflétés par différents aspects de la
microstructure de la plume. Une des avenues possibles absente de la littérature est de
quantifier la proportion de zones dégradées par les bactéries sur les plumes avec la
microscopie en réflexion (Figure 4.2). Sur les photos, on remarque beaucoup de zones floues
causée par la faible profondeur de champ du microscope, pouvant être éliminées par la
technique du « Focus Stacking ».
89
Figure 4.2 Photographie des barbules de deux plumes iridescentes de croupion de mâle par
microscopie en réflexion. a) Barbules en grossissement 500X, un filament segmenté
correspond à une barbule, b) Barbules en grossissement 1000X, dont un des segments contient
une tache brune probablement causée par la dégradation de la kératine par des bactéries. Le
microscope utilisé provient de l’Infrastructure des Matériaux et Dispositifs Quantiques
(IMDQ) du département de physique de l’Université de Sherbrooke. Crédit photo : Sonia Van
Wijk.
J’ai aussi exploré la relation entre l’iridescence et différentes caractéristiques de l’individu et
de l’environnement pour connaître quelle information l’iridescence pourrait véhiculer. Ces
analyses préliminaires suggèrent qu’en effet la brillance semble avoir plus de potentiel de
condition-dépendance (Prum, 2006), puisque j’ai observé que les mâles avec un dos brillant
étaient moins parasités par les mallophages et ceux avec une couronne brillante avaient une
plus grande masse. Je m’attendais à voir que les « bons » individus s’agrègent dans le paysage
compte tenu des résultats de Porlier et al. (2009). Ses résultats ont montré que les individus
avec une plus grande diversité génétique se trouvaient dans les habitats de moins bonne
qualité, où l’agriculture intensive prédominait. Je n’ai pas trouvé de différences d’iridescence
selon la proportion de terres agricoles intensives ou la densité en reproducteurs la ferme, sauf
une légère tendance de l’augmentation de la brillance avec une diminution de la proportion de
terres agricoles intensives dans un rayon de 15 km. J’ai par contre observé un lien relativement
90
fort de toutes les variables d’iridescence avec le jour julien de capture, par exemple la couleur
(Figure 4.3a). L’iridescence était plus bleue, moins brillante, moins saturée et avec une plus
petite étendue angulaire au fil du temps. Mon hypothèse est qu’il y a un changement
populationnel avec le temps causé par le vieillissement de la plume, mais que les individus
diffèrent par leur entretien des plumes et leur investissement dans la reproduction, amenant
des différences individuelles en pente. Les mesures répétées d’iridescence prises chez
quelques individus le suggèrent (Figure 4.3b). Dans notre système d’étude, répéter la prise de
plumes pour les femelles recapturées lorsqu’on tente de capturer le mâle (plus de 100 femelles
par année) pourrait être réalisé afin de mieux comprendre le phénomène. D’autant plus que le
peu d'études s’y attardant ont vu une augmentation de la couleur (longueurs d’onde plus
élevées) avec le jour julien chez la Mésange bleue (Delhey et al., 2006; Ornborg et al., 2002),
alors qu’ici c’est l’inverse qui a été observé. Il pourrait aussi s’agir d’un biais de capture des
individus plus bleus vers la fin de la saison de reproduction.
Figure 4.3 Relation entre la couleur et le jour julien de capture. a) Régression de la couleur
centrée réduite (moyenne de 0 et variance de 1) du dos selon le jour julien. Ligne pleine :
prédiction. Lignes pointillées : intervalle de confiance à 95 %. Les données sont représentées
par des points, 1 point par mâle (202 mâles). b) Couleur du croupion selon le jour julien pour
14 individus (9 femelles, 5 mâles) dont les plumes iridescentes ont été récoltées deux fois dans
la même saison de reproduction. Une ligne de couleur correspond à un individu. La ligne relie
les moyennes des 3 plumes. Chaque plume est représentée par un point.
91
La récolte des plumes iridescentes chez les femelles durant les deux années de cette présente
étude va permettre d’étudier les mêmes questions que chez les mâles. Un modèle d’étude
comprenant les deux sexes ornementés et des soins biparentaux est idéal pour tester le choix
mutuel (Johnstone et al., 1996). Valérie Lemieux, étudiante aux crédits de recherche à l’Hiver
2015, a trouvé un lien entre la brillance du mâle et celle de la femelle, comme observé chez
Bitton et al. (2008). Il faudrait élargir l’analyse à l’étendue angulaire compte tenu que la
qualité parentale de la femelle pourrait être un critère important au choix du mâle. Les
femelles brunes d’un an sont aussi un très bon sujet d’étude de la coloration agissant comme
signal d’habileté compétitive (Bentz et Siefferman, 2013). En effet, l’Hirondelle bicolore est
une des deux seules espèces en Amérique du Nord chez qui la maturation retardée du plumage
est observée chez les femelles et non chez les mâles (Lozano et Handford, 1995). De plus, le
plumage blanc de la poitrine a un fort potentiel de sélection sexuelle puisqu’il a une
contribution importante d’émission dans les UV (Bennett et al., 1997), peut être coûteux à
maintenir (Kose et Møller, 1999), et a déjà été relié au succès reproducteur (Doucet, 2005). Il
sera donc échantillonné à partir de la saison 2015.
92
4.3 Conclusion
Mon étude a permis une amélioration de la technique de quantification de l’iridescence ainsi
qu’une meilleure compréhension de la sélection sexuelle de ce type de coloration répandu
chez les oiseaux. En effet, j’ai pu mieux comprendre la fonction des diverses composantes de
l’iridescence en ayant décortiqué le succès reproducteur dans le couple et hors couple. Grâce
au système d’étude à grande échelle et à la grande taille d’échantillon, j’ai pu observer que la
sélection sexuelle semblait s’opérer différemment dépendamment de la densité de
reproducteurs. La tendance à n’étudier que des populations très denses à des fins économiques
peut biaiser les résultats et ainsi mal représenter les vrais patrons de sélection (Stamps, 2011).
Globalement, mon projet a contribué à l’avancement des connaissances fondamentales sur les
forces de sélection sexuelle agissant en milieu naturel.
93
BIBLIOGRAPHIE
Able, D.J. (1996). The contagion indicator hypothesis for parasite-mediated sexual selection.
Proc. Natl. Acad. Sci. U. S. A. 93, 2229–2233.
Andersson, M. (1994). Sexual Selection (Princeton, New Jersey: Princeton University Press).
Andersson, S., and Prager, M. (2006). Quantifying Colors. In Bird Coloration: Vol. 1.
Mechanisms and Measurements, G. Hill, and K.J. McGraw, eds. (Cambridge: Harvard
University Press), pp. 41–89.
Balenger, S.L., Johnson, L.S., and Masters, B.S. (2008). Sexual selection in a socially
monogamous bird: male color predicts paternity success in the mountain bluebird, Sialia
currucoides. Behav. Ecol. Sociobiol. 63, 403–411.
Barber, C.A., Robertson, R.J., and Boag, P. (1996). The high frequency of extra-pair paternity
in Tree Swallows is not an artifact of nestboxes. Behav. Ecol. Sociobiol. 38, 425–430.
Bateman, A.J. (1948). Intra-sexual selection in Drosophila. Heredity (Edinb). 2, 349–368.
Bennett, A.T.D., Cuthill, I.C., Partridge, J.C., and Lunau, K. (1997). Ultraviolet plumage
colors predict mate preferences in starlings. Proc. Natl. Acad. Sci. U. S. A. 94, 8618–8621.
Bentz, A.B., and Siefferman, L. (2013). Age-dependent relationships between coloration and
reproduction in a species exhibiting delayed plumage maturation in females. J. Avian Biol. 44,
80–88.
Bitton, P.-P., O’Brien, E.L., and Dawson, R.D. (2007). Plumage brightness and age predict
extrapair fertilization success of male tree swallows, Tachycineta bicolor. Anim. Behav. 74,
1777–1784.
Bitton, P.-P., Dawson, R.D., and Ochs, C.L. (2008). Plumage characteristics, reproductive
investment and assortative mating in tree swallows Tachycineta bicolor. Behav. Ecol.
Sociobiol. 62, 1543–1550.
Bortolotti, G.R. (2006). Natural selection and coloration. In Bird Coloration: Vol.2. Function
and Evolution, G.E. Hill, and K.J. McGraw, eds. (Cambridge: Harvard University Press), pp.
3–35.
Butler, R.W. (1988). Population Dynamics and Migration Routes of Tree Swallows,
Tachycineta bicolor, in North America. J. F. Ornithol. 59, 395–402.
94
Cuthill, I.C. (2006). Color perception. In Bird Coloration: Vol. 1. Mechanisms and
Measurements, G. Hill, and K. McGraw, eds. (Cambridge: Harvard University Press), pp. 3–
40.
Dakin, R., and Montgomerie, R. (2013). Eye for an eyespot: how iridescent plumage ocelli
influence peacock mating success. Behav. Ecol. 24, 1048–1057.
Darwin, C. (1859). On the origin of species by means of natural selection (London, UK:
Murray.).
Darwin, C. (1871). The descent of man, and selection in relation to sex (New York: Appleton).
Delhey, K., Peters, A., Johnsen, A., and Kempenaers, B. (2006). Seasonal changes in blue tit
crown color: do they signal individual quality? Behav. Ecol. 17, 790–798.
Doucet, S.M. (2002). Structural plumage coloration, male body size, and condition in the blueblack grassquit. Condor 104, 30–38.
Doucet, S.M. (2005). Achromatic plumage reflectance predicts reproductive success in male
black-capped chickadees. Behav. Ecol. 16, 218–222.
Doucet, M., and Montgomerie, R. (2003). Multiple sexual ornaments in satin bowerbirds :
ultraviolet plumage and bowers signal different aspects of male quality. Behav. Ecol. 14, 503–
509.
Doucet, S.M., and Meadows, M.G. (2009). Iridescence: a functional perspective. J. R. Soc.
Interface 6 Suppl 2, S115–S132.
Doucet, S.M., Shawkey, M.D., Hill, G.E., and Montgomerie, R. (2006). Iridescent plumage in
satin bowerbirds: structure, mechanisms and nanostructural predictors of individual variation
in colour. J. Exp. Biol. 209, 380–390.
Dunn, P.O., and Whittingham, L.A. (2007). Search costs influence the spatial distribution, but
not the level, of extra-pair mating in tree swallows. Behav. Ecol. Sociobiol. 61, 449–454.
Dunn, P.O., Robertson, R.J., Michaud-freeman, D., Boag, P.T., Dunn, P., and Boag, D.M.P.T.
(1994). Extra-pair paternity in tree swallows : why do females mate with more than one male ?
Behav. Ecol. Sociobiol. 35, 273–281.
Dunn, P.O., Lifjeld, J.T., and Whittingham, L.A. (2009). Multiple paternity and offspring
quality in tree swallows. Behav. Ecol. Sociobiol. 63, 911–922.
Eliason, C.M., and Shawkey, M.D. (2010). Rapid, reversible response of iridescent feather
color to ambient humidity. Opt. Express 18, 21284–21292.
95
Eliason, C.M., and Shawkey, M.D. (2011). Decreased hydrophobicity of iridescent feathers: a
potential cost of shiny plumage. J. Exp. Biol. 214, 2157–2163.
Endler, J.A., and Mielke, P.W. (2005). Comparing entire colour patterns as birds see them.
Biol. J. Linn. Soc. 86, 405–431.
Fisher, R. (1958). The genetical theory of natural selection (New York: Oxford University
Press).
Fitzpatrick, S. (1994). Colorful migratory birds : evidence for a mechanism other than parasite
resistance for the maintenance of “good genes” sexual selection. Proc. Biol. Sci. 257, 155–
160.
Fitzpatrick, S. (1998). Colour schemes for birds : structural coloration and signals of quality in
feathers. Ann. Zool. Fenn. 35, 67–77.
Forstmeier, W., Martin, K., Bolund, E., Schielzeth, H., and Kempenaers, B. (2011). Female
extrapair mating behavior can evolve via indirect selection on males. Proc. Natl. Acad. Sci. U.
S. A. 108, 10608–10613.
Greenewalt, C., Brandt, W., and Friel, D. (1960). The iridescent colors of hummingbird
feathers. Proc. Am. Philos. Soc. 104, 249–253.
Griffith, S.C., Owens, I.P.F., and Thuman, K. (2002). Extra pair paternity in birds : a review of
interspecific variation and adaptative function. Mol. Ecol. 11, 2195–2212.
Griffith, S.C., Parker, T.H., and Olson, V.A. (2006). Melanin- versus carotenoid-based sexual
signals: is the difference really so black and red? Anim. Behav. 71, 749–763.
Gunderson, A.R., Forsyth, M.H., and Swaddle, J.P. (2009). Evidence that plumage bacteria
influence feather coloration and body condition of eastern bluebirds Sialia sialis. J. Avian
Biol. 40, 440–447.
Hailman, J. (1977). Optical signals: animal communication and light (Bloomington: Indiana
University Press).
Hamilton, D. (1990). Mate Choice near or Far. Am. Zool. 30, 341–352.
Hamilton, W.D., and Zuk, M. (1982). Heritable true fitness and bright birds: a role for
parasites? Science (80-. ). 218, 384–387.
Hedrick, A., and Dill, L. (1993). Mate choice by female crickets is influenced by predation
risk. Anim. Behav. 46, 193–196.
96
Hill, G.E. (2006). Female mate choice for ornamental coloration. In Bird Coloration: Vol.2.
Function and Evolution, G.E. Hill, and K.J. McGraw, eds. (Cambridge: Harvard University
Press), pp. 137–200.
Hill, G.E., Doucet, S.M., and Buchholz, R. (2005). The effect of coccidial infection on
iridescent plumage coloration in wild turkeys. Anim. Behav. 69, 387–394.
Hoelzer, G. a. (1989). The good parent process of sexual selection. Anim. Behav. 38, 1067–
1078.
Hussel, D.J.T. (1983). Age and Plumage Color in Female Tree Swallows. J. F. Ornithol. 54,
312–318.
Isawa, Y., Pomiankowski, A., and Nee, S. (1991). The Evolution of Costly Mate Preferences
II. The “Handicap” Principle. Evolution (N. Y). 45, 1431–1442.
Jennions, M.D., and Petrie, M. (2000). Why do females mate multiply? A review of the
genetic benefits. Biol. Rev. Camb. Philos. Soc. 75, 21–64.
Johnstone, R.A., Reynolds, J.D., and Deutsch, J.C. (1996). Mutual mate choice and sex
differences in choosiness. Evolution 50, 1382–1391.
Kempenaers, B., Lanctot, R.B., and Robertson, R.J. (1998). Certainty of paternity and paternal
investment in eastern bluebirds and tree swallows. Anim. Behav. 55, 845–860.
Kempenaers, B., Congdon, B., Boag, P., and Robertson, R.J. (1999). Extrapair paternity and
egg hatchability in tree swallows : evidence for the genetic compatibility hypothesis ? Behav.
Ecol. 10, 304–311.
Kempenaers, B., Everding, S., Bishop, C., Boag, P., and Robertson, R.J. (2001). Extra-pair
paternity and the reproductive role of male floaters in the tree swallow (Tachycineta bicolor).
Behav. Ecol. Sociobiol. 49, 251–259.
Keyser, A.J., and Hill, G.E. (2000). Structurally based plumage coloration is an honest signal
of quality in male blue grosbeaks. Behav. Ecol. 11, 202–209.
Kodric-Brown, A., and Brown, J.H. (1984). Truth in advertising: the kinds of traits favored by
sexual selection. Am. Nat. 124, 309–323.
Kokko, H., and Rankin, D.J. (2006). Lonely hearts or sex in the city? Density-dependent
effects in mating systems. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 361, 319–334.
Komdeur, J., Oorebeek, M., van Overveld, T., and Cuthill, I.C. (2005). Mutual ornamentation,
age, and reproductive performance in the European starling. Behav. Ecol. 16, 805–817.
97
Kose, M., and Møller, A.P. (1999). Sexual selection, feather breakage and parasites: The
importance of white spots in the tail of the barn swallow (Hirundo rustica). Behav. Ecol.
Sociobiol. 45, 430–436.
Kotiaho, J.S., LeBas, N.R., Puurtinen, M., and Tomkins, J.L. (2008). On the resolution of the
lek paradox. Trends Ecol. Evol. 23, 1–3.
Lack, D. (1968). Ecological Adaptations for Breeding in Birds (London: Methuen).
Leclaire, S., Pierret, P., Chatelain, M., and Gasparini, J. (2014). Feather bacterial load affects
plumage condition, iridescent color, and investment in preening in pigeons. Behav. Ecol. 25,
1192–1198.
Leffelaar, D., and Robertson, R.J. (1984). Do male tree swallows guard their mates? Behav.
Ecol. Sociobiol. 16, 73–79.
Lessard, A., Bourret, A., Pelletier, F., Bélisle, M., and Garant, D. (2014). Individual and
environmental determinants of reproductive success in male tree swallow (Tachycineta
bicolor). Behav. Ecol. Sociobiol. 68, 733-742.
Lifjeld, J., Dunn, P., Robertson, R.J., and Boag, P.T. (1993). Extra-pair paternity in
monogamous tree swallows. Anim. Behav. 45, 213–229.
Llambıas, P.E., Wrege, P., and Winkler, D.W. (2008). Effects of site fidelity and breeding
performance on mate retention in a short-lived passerine, the tree swallow Tachycineta
bicolor. J. Avian Biol. 39, 493–499.
Loyau, A., Gomez, D., Moureau, B., Théry, M., Hart, N.S., Jalme, M.S., Bennett, A.T.D., and
Sorci, G. (2007). Iridescent structurally based coloration of eyespots correlates with mating
success in the peacock. Behav. Ecol. 18, 1123–1131.
Lozano, G., and Handford, P. (1995). A Test of an Assumption of Delayed Plumage
Maturation Hypotheses Using Female Tree Swallows. Wilson Bull. 107, 153–164.
Madsen, V., Dabelsteen, T., Osorio, D., and Osorno, J.L. (2007). Morphology and
ornamentation in male magnificent frigatebirds: variation with age class and mating status.
Am. Nat. 169 Suppl, S93–S111.
Maia, R., Macedo, R.H.F., and Shawkey, M.D. (2012). Nanostructural self-assembly of
iridescent feather barbules through depletion attraction of melanosomes during keratinization.
J. R. Soc. Interface 9, 734–743.
98
Maia, R., Eliason, C.M., Bitton, P.-P., Doucet, S.M., and Shawkey, M.D. (2013). PAVO : an R
package for the analysis, visualization and organization of spectral data. Methods Ecol. Evol.
4, 906–913.
McGraw, K.J. (2006). Mechanics of Melanin-Based Coloration. In Bird Coloration: Vol. 1.
Mechanisms and Measurements, G. Hill, and K.J. McGraw, eds. (Cambridge: Harvard
University Press), pp. 243–294.
McGraw, K.J., Mackillop, E. a, Dale, J., and Hauber, M.E. (2002). Different colors reveal
different information: how nutritional stress affects the expression of melanin- and structurally
based ornamental plumage. J. Exp. Biol. 205, 3747–3755.
Meadows, M.G., Morehouse, N.I., Rutowski, R.L., Douglas, J.M., and McGraw, K.J. (2011).
Quantifying iridescent coloration in animals: a method for improving repeatability. Behav.
Ecol. Sociobiol. 65, 1317–1327.
Meadows, M.G., Roudybush, T.E., and McGraw, K.J. (2012). Dietary protein level affects
iridescent coloration in Anna’s hummingbirds, Calypte anna. J. Exp. Biol. 215, 2742–2750.
Milinski, M., and Bakker, T.C.M. (1992). Costs Influence Sequential Mate Choice in
Sticklebacks, Gasterosteus aculeatus. Proc. R. Soc. B Biol. Sci. 250, 229–233.
Møller, A. (1988). Female choice selects for male sexual tail ornaments in the monogamous
swallow. Nature 332, 640–642.
Møller, A., and Birkhead, T.R. (1994). The evolution of plumage brightness in birds is related
to extrapair paternity. Evolution (N. Y). 48, 1089–1100.
Montgomerie, R. (2006). Analyzing Colors. In Bird Coloration: Vol. 1. Mechanisms and
Measurements, G.E. Hill, and K.J. McGraw, eds. (Cambridge: Harvard University Press), pp.
90–147.
Neff, B.D., and Pitcher, T.E. (2005). Genetic quality and sexual selection: an integrated
framework for good genes and compatible genes. Mol. Ecol. 14, 19–38.
O’Brien, E.L., and Dawson, R.D. (2007). Context-dependent genetic benefits of extra-pair
mate choice in a socially monogamous passerine. Behav. Ecol. Sociobiol. 61, 775–782.
Ornborg, J., Andersson, S., Griffith, S.C., and Sheldon, B.C. (2002). Seasonal changes in a
ultraviolet structural colour signal in blue tits, Parus caeruleus. Biol. J. Linn. Soc. 76, 237–
245.
Osorio, D., and Ham, A.D. (2002). Spectral reflectance and directional properties of structural
coloration in bird plumage. J. Exp. Biol. 205, 2017–2027.
99
Owens, I.P.F., and Hartley, I.R. (1998). Sexual dimorphism in birds: why are there so many
different forms of dimorphism? Proc. Biol. Sci. 265, 397–407.
Pomiankowski, A., Isawa, Y., and Nee, S. (1991). The Evolution of Costly Mate Preferences I.
Fisher and Biased Mutation. Evolution (N. Y). 45, 1422–1430.
Porlier, M., Bélisle, M., and Garant, D. (2009). Non-random distribution of individual genetic
diversity along an environmental gradient. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 364,
1543–1554.
Prum, R. (2006). Anatomy, Physics, and Evolution of Structural Colors. In Bird Coloration:
Vol. 1. Mechanisms and Measurements, G. Hill, and K. McGraw, eds. (Cambridge: Harvard
University Press), pp. 295–353.
Robertson, R., Stutchbury, B., and Cohen, R. (1992). Tree Swallow. In Birds of North
America, A. Poole, P. Stettenheim, and F. Gill, eds. (Philadelphia), p. No. 11.
Rohwer, S. (1982). The Evolution of Reliable and Unreliable Badges of Fighting Ability. Am.
Zool. 22, 531–546.
Ryan, M.J., and Keddy-Hector, A. (1992). Directional Patterns of Female Mate Choice and the
Role of Sensory Biases. Am. Nat. 139, S4–S35.
Sauer, J.R., Hines, J.E., Fallon, J.E., Pardieck, K.L., Ziolkowski Jr., D.J., and Link, W.A.
(2014). The North American Breeding Bird Survey, Results and Analysis 1966-2013, Version
01.30.2015.
Savard, J.-F., Keagy, J., and Borgia, G. (2011). Blue, not UV, plumage color is important in
satin bowerbird Ptilonorhynchus violaceus display. J. Avian Biol. 42, 80–84.
Shawkey, M.D., Pillai, S.R., Hill, G.E., Siefferman, L.M., and Roberts, S.R. (2007). Bacteria
as an agent for change in structural plumage color: correlational and experimental evidence.
Am. Nat. 169 Suppl , S112–S121.
Stamps, J. a. (2011). Density bias in behavioral ecology. Behav. Ecol. 22, 231–232.
Stapleton, M.K., Kleven, O., Lifjeld, J.T., and Robertson, R.J. (2007). Female tree swallows
(Tachycineta bicolor) increase offspring heterozygosity through extrapair mating. Behav.
Ecol. Sociobiol. 61, 1725–1733.
Taff, C.C., Freeman-Gallant, C.R., Dunn, P.O., and Whittingham, L.A. (2013). Spatial
distribution of nests constrains the strength of sexual selection in a warbler. J. Evol. Biol. 26,
1392–1405.
100
Trivers, R.L. (1972). Parental invesment and sexual selection. In Sexual Selection and the
Descent of Man 1871-1971, B. Campbell, ed. (Chicago: Aldine), pp. 136–179.
Venier, L., Dunn, P.O., Lifjeld, J.T., and Robertson, R.J. (1993). Behavioural patterns of
extra-pair copulation in tree swallows. Anim. Behav. 45, 412–415.
Weatherhead, P.J., and Robertson, R.J. (1979). Offspring quality and the polygyny treshold:
“the sexy son hypothesis.” Am. Nat. 113, 201–208.
Webster, M.S., Pruett-Jones, S., Westneat, D., and Arnold, S.J. (1995). Measuring the effects
of pairing success, extra-pair copulations and mate quality on the opportunity for sexual
selection. Evolution (N. Y). 49, 1147–1157.
Westneat, D.F., Sherman, P.W., and Morton, M. (1990). The ecology and evolution of extrapair copulations in birds. Curr. Ornithol. 7, 331–369.
Wetton, J.H., and Parkin, D.T. (1991). An association between fertility and cuckoldry in the
house sparrow, Passer domesticus. Proc. Biol. Sci. 245, 227–233.
Whittingham, L.A., and Dunn, P.O. (2001). Survival of extrapair and within-pair young in tree
swallows. Behav. Ecol. 12, 496–500.
Whittingham, L., Dunn, P.O., and Robertson, R.J. (1993). Confidence of paternity and male
parental care: an experiment study in tree swallow. Anim. Behav. 46, 139–147.
Winkler, D.W., Wrege, P.H., Allen, P.E., Kast, T.L., Senesac, P., Wasson, M.F., Llambías,
P.E., Ferretti, V., and Sullivan, P.J. (2004). Breeding Dispersal and Philopatry in the Tree
Swallow. Condor 106, 768–776.
Winkler, D.W., Hallinger, K., Ardia, D., Robertson, R., Stutchbury, B., and Cohen, R. (2011).
Tree Swallow (Tachycineta bicolor). In The Birds of North America, A. Poole, ed. (Cornell
Lab of Ornithology, Ithaca, New York, USA),.
Zahavi, A. (1975). Mate Selection - A Selection for a Handicap. J. Theor. Biol. 53, 205–214.
Zirpoli, J.A., Black, J.M., and Gabriel, P.O. (2013). Parasites and plumage in Steller’s jays : an
experimental field test of the parasite-mediated handicap hypothesis. Ethol. Ecol. Evol. 25,
103–116.
101

Similar documents