Le processus de sélection intersexuelle peut impliquer plusieurs mécanismes différents dans le choix du partenaire. Chez la pyrale Achroia grisella, les femelles choisissent les mâles en se basant sur des caractères de leur chant d'avertissement ultrasonique, et de précédentes études ont montré que ce choix peut procurer, à une femelle discriminante, certains bénéfices génétiques indirects. Dans cette thèse, j'ai continué à explorer le mécanisme de bénéfices indirects chez cette espèce en 1) déterminant avec plus de précision un paramètre acoustique, l‟amplitude du chant, qui influence le choix des femelles et en 2) examinant la covariance génétique entre les traits du chant mâle et ceux de la préférence femelle. J'ai constaté qu'en plus de préférer des chants mâles émis avec une plus grande amplitude et une plus grande puissance acoustique, les femelles font également une discrimination en faveur des chants comprenant des forts « pics d'amplitude ». Cependant, mes analyses génétiques moléculaires de la relation entre les traits du signal mâle et ceux de la préférence femelle, n'ont révélé aucune covariance significative, comme prédit par la théorie de la sélection sexuelle. Ces derniers résultats fournissent un aperçu important concernant l‟existence d‟un tempo réduit de la sélection sexuelle, dans les populations naturelles. / The process of inter-sexual selection may involve several different mechanisms of mate choice. In the pyralid moth Achroia grisella, females choose males based on characters of their ultrasonic advertisement song, and previous work has indicated that such choice may afford a discriminating female certain indirect genetic benefits. In this thesis I continued to probe the indirect benefits mechanism in this species by 1) determining with more precision one acoustic character, song amplitude, that influences female choice and by 2) examining the genetic covariance between the male song trait and the female preference trait. I found that in addition to preferring male songs delivered at greater amplitude and with greater acoustic power, females also discriminate in favor of songs including higher 'amplitude peaks'. However, my molecular genetic analyses of the relationships between the male signal and female preference traits did not reveal any significant covariance, as predicted by sexual selection theory. These latter findings provide important insight on the reduced tempo of sexual selection that may exist in natural populations.

Pendant de nombreuses années, les femelles ont été décrites comme passives dans les interactions mâles-femelles et ont parfois été négligées dans les études sur les comportements reproducteurs et la communication animale. Pourtant, il est de plus en plus évident que les comportements des femelles influencent ceux des mâles et que les mâles peuvent ajuster leurs comportements de parades aux comportements des femelles. L’objectif principal de cette thèse est de mieux comprendre les signaux que les femelles émettent lors d’interactions avec un mâle dans un contexte de reproduction chez le canari domestique. L’ensemble de mes résultats montre tout d’abord que les femelles utilisent des signaux de parades visuels et acoustiques, c’est-à-dire des postures de sollicitation à l’accouplement et des trilles-spécifiques de femelles, comme une invitation à s’accoupler, mais qu’elles peuvent aussi les utiliser pour inciter le mâle à parader et l’aider à échantillonner la qualité de partenaires potentiels. D’autre part, ces deux signaux n’auraient pas la même efficacité selon le contexte d’émission. Ensuite, il semblerait que les signaux véhiculés par la modalité visuelle pourraient jouer un rôle plus important qu’on ne le pensait jusque-là dans les interactions intersexuelles chez cette espèce. Enfin, cette thèse a aussi pu apporter de nouveaux éléments concernant les préférences des femelles pour les chants de mâles et a montré que les méthodes utilisées en laboratoire pour tester les préférences des femelles étaient fiables et congruentes. Cette thèse permet de contribuer aux recherches grandissantes mettant en évidence le rôle important de la femelle dans les interactions intersexuelles. / Much studies on reproduction and animal communication have considered the female as the passive sex; the role of the female during male-female interactions have often been overlooked. However, there is growing evidence that female behaviours can affect those of the males and that males can adjust their courtships to female behaviours. The aim of this thesis is to understand the signals produced by females during interactions with a male in a reproductive context in the domestic canary. Overall, my results show that females not only use their visual and acoustic signals, the copulation solicitation display and the female-specific trills, as an invitation to copulate but also to incite male to sing as an aid to sample potential mates. Then, these two signals could not have the same efficacy in different contexts of transmission. Moreover, the visual components of the communication seem to be more important than previously thought during intersexual interactions in this species. Finally, this thesis provides new elements about the female preferences for male songs and shows that methods used to test female preferences in laboratory are reliable and congruent. This thesis contributes to the growing number of researches showing that females play an active role in intersexual interactions.

The ultimate goal of my thesis is to develop a better understanding of the contribution of indirect benefits (i.e. good genes) to the evolution of female mate preferences. It is genetic variance in, and genetic correlations (covariances) among, male sexual displays, female preferences for them, and fitness that in part determine the degree to which females preferring certain male displays over others will gain an indirect benefit by having higher fitness offspring. Recent advances in quantitative genetic theory provide the mathematical means for quantifying the strength of indirect selection for female mate preferences (Kirkpatrick and Hall 2004), at least under certain conditions, but there are few empirical systems for which such data exist (Brooks and Endler 2001; Qvarnström et al. 2006). I have undertaken a classic half-sibling breeding design with the ultimate goal of estimating the specific parameters of this model in a population of the Australian fruit fly Drosophila serrata. The breeding design was performed across two environments - one to which the population was well adapted and a novel environment to which it was not - thereby also providing insight into genotype-by-environment interactions for this suite of traits and their effects on good genes indirect benefits in a novel environment. General insight is also gained into the genetic covariance of male and female fitness and the prevalence of intralocus sexual conflict, the quantitative genetic basis of female mate preferences for multiple male traits, the condition-dependence of these traits, and the genetic association between sexual displays and fitness when mutation-selection balance is inferred. My results advocate caution in the application of existing theory to quantify the strength of indirect selection, suggesting that a good genes process may be fundamentally different when the exaggeration of sexual displays is eventually halted and an equilibrium is reached between opposing selection.

The ultimate goal of my thesis is to develop a better understanding of the contribution of indirect benefits (i.e. good genes) to the evolution of female mate preferences. It is genetic variance in, and genetic correlations (covariances) among, male sexual displays, female preferences for them, and fitness that in part determine the degree to which females preferring certain male displays over others will gain an indirect benefit by having higher fitness offspring. Recent advances in quantitative genetic theory provide the mathematical means for quantifying the strength of indirect selection for female mate preferences (Kirkpatrick and Hall 2004), at least under certain conditions, but there are few empirical systems for which such data exist (Brooks and Endler 2001; Qvarnström et al. 2006). I have undertaken a classic half-sibling breeding design with the ultimate goal of estimating the specific parameters of this model in a population of the Australian fruit fly Drosophila serrata. The breeding design was performed across two environments - one to which the population was well adapted and a novel environment to which it was not - thereby also providing insight into genotype-by-environment interactions for this suite of traits and their effects on good genes indirect benefits in a novel environment. General insight is also gained into the genetic covariance of male and female fitness and the prevalence of intralocus sexual conflict, the quantitative genetic basis of female mate preferences for multiple male traits, the condition-dependence of these traits, and the genetic association between sexual displays and fitness when mutation-selection balance is inferred. My results advocate caution in the application of existing theory to quantify the strength of indirect selection, suggesting that a good genes process may be fundamentally different when the exaggeration of sexual displays is eventually halted and an equilibrium is reached between opposing selection.

The ultimate goal of my thesis is to develop a better understanding of the contribution of indirect benefits (i.e. good genes) to the evolution of female mate preferences. It is genetic variance in, and genetic correlations (covariances) among, male sexual displays, female preferences for them, and fitness that in part determine the degree to which females preferring certain male displays over others will gain an indirect benefit by having higher fitness offspring. Recent advances in quantitative genetic theory provide the mathematical means for quantifying the strength of indirect selection for female mate preferences (Kirkpatrick and Hall 2004), at least under certain conditions, but there are few empirical systems for which such data exist (Brooks and Endler 2001; Qvarnström et al. 2006). I have undertaken a classic half-sibling breeding design with the ultimate goal of estimating the specific parameters of this model in a population of the Australian fruit fly Drosophila serrata. The breeding design was performed across two environments - one to which the population was well adapted and a novel environment to which it was not - thereby also providing insight into genotype-by-environment interactions for this suite of traits and their effects on good genes indirect benefits in a novel environment. General insight is also gained into the genetic covariance of male and female fitness and the prevalence of intralocus sexual conflict, the quantitative genetic basis of female mate preferences for multiple male traits, the condition-dependence of these traits, and the genetic association between sexual displays and fitness when mutation-selection balance is inferred. My results advocate caution in the application of existing theory to quantify the strength of indirect selection, suggesting that a good genes process may be fundamentally different when the exaggeration of sexual displays is eventually halted and an equilibrium is reached between opposing selection.

Les insectes phytophages forment un groupe d’organismes très diversifié et la plupart sont considérés comme spécialistes. Les patrons de spécialisation des insectes vis à vis de leurs plantes hôtes dépendent en partie de leur capacité à interagir avec les plantes (niche fondamentale) et aux facteurs environnementaux modulant ces interactions et aboutissant aux observations en milieu naturel (niche réalisée). La spécialisation fondamentale est déterminée par l'évolution conjointe de deux traits : la performance des larves et la préférence des femelles. Pour comprendre cette spécialisation, nous avons étudié une communauté de huit espèces de mouches des fruits (Diptera : Tephritidae) présentes à La Réunion. Dans un premier temps, nous avons déterminé la niche réalisée de chaque espèce et montré que ces niches étaient structurées par la phylogénie des plantes avec D. demmerezi, D. ciliatus et Z. cucurbitae spécialistes des Cucurbitaceae, N. cyanescensspécialiste des Solanaceae et C. catoirii, C. capitata et C. quilicii généralistes attaquant des plantes de différentes familles. Après l'invasion de B. zonata en 2000, C. capitata et C. quilicii ont subi une réduction de leur gamme d'hôtes. Dans un deuxième temps, nous avons déterminé la niche fondamentale de ces espèces (sauf D. ciliatus). Nous avons évalué les préférences des femelles en mesurant la fécondité de chacune d’entre elles sur une gamme de 29 fruits, puis nous avons testé l'existence d'une corrélation entre la préférence des femelles et la performance des larves (mother knows best hypothesis). Nous avons montré une corrélation positive chez spécialistes des Cucurbitaceae qui pondent sur les plantes pour lesquelles les larves survivent le mieux contrairement aux généralistes pondant et survivant sur une large gamme d'hôtes mais sans corrélation entre ces deux traits.Enfin, la sélection de l'hôte par les femelles se faisant principalement grâce aux composés organiques volatils (COVs) émis par les fruits, nous avons montré que les fruits infestés par les généralistes ont pour point commun l'émission de COVs responsables de la maturation des fruits. Au contraire, les fruits de plusieurs Solanaceae émettent des COVs spécifiques suggérant la détection de ceux-ci par les femelles de N. cyanescens. Les Cucurbitaceae émettent des COVs abondants peu présents dans les autres familles suggérant une détection d'un mélange spécifique de ces COVs par les Tephritidae spécialistes des Cucurbitaceae. / Phytophagous insects are a very diverse group of organisms and most of them are considered as specialized. Patterns of specialization regarding their host plants depend on their ability to interact with their hosts (fundamental niche) and on environmental factors which modulate these interactions leading to observed patterns in the field (realized niche). Fundamental specialization is determined by the joint evolution of two traits: larval performance and female preference. To understand this specialization, we studied a community of eight fruit fly species (Diptera: Tephritidae) present in La Réunion.First, we determined the realized niche of each species and showed that they were structured by plant phylogeny with D. demmerezi, D. ciliatus and Z. cucurbitae as Cucurbitaceae specialists, N. cyanescens as Solanaceae specialist and C. catoirii, C. capitata et C. quilicii as generalists feeding on plants belonging to different families. After the invasion of B. zonata in 2000, C. capitata et C. quilicii were subjected to a decrease of their host range.Then, we determined the fundamental niche of these species (except D. ciliatus). We assessed female preferences by measuring their fecundity on 29 fruits, then we tested the presence of a correlation between female preference and larval performance (mother knows best hypothesis). We showed a positive correlation for Cucurbitaceae specialists laying eggs on plants where larvae survive the best, at the opposite of generalist species laying eggs and surviving on many hosts without any correlation between these two traits.Finally, host selection by females being mostly done thanks to volatile organic compounds (VOCs) emitted by fruits, we showed that fruits infested by generalist species have common VOCs responsible for fruit maturation. On the contrary, the fruits of several Solanaceae emit specific VOCs suggesting their detection by N. cyanescens females. Cucurbitaceae species emit abundant VOCs rarely present in other families suggesting a detection of a specific blend of these VOCS by Cucurbitaceae specialists.

The ultimate goal of my thesis is to develop a better understanding of the contribution of indirect benefits (i.e. good genes) to the evolution of female mate preferences. It is genetic variance in, and genetic correlations (covariances) among, male sexual displays, female preferences for them, and fitness that in part determine the degree to which females preferring certain male displays over others will gain an indirect benefit by having higher fitness offspring. Recent advances in quantitative genetic theory provide the mathematical means for quantifying the strength of indirect selection for female mate preferences (Kirkpatrick and Hall 2004), at least under certain conditions, but there are few empirical systems for which such data exist (Brooks and Endler 2001; Qvarnström et al. 2006). I have undertaken a classic half-sibling breeding design with the ultimate goal of estimating the specific parameters of this model in a population of the Australian fruit fly Drosophila serrata. The breeding design was performed across two environments - one to which the population was well adapted and a novel environment to which it was not - thereby also providing insight into genotype-by-environment interactions for this suite of traits and their effects on good genes indirect benefits in a novel environment. General insight is also gained into the genetic covariance of male and female fitness and the prevalence of intralocus sexual conflict, the quantitative genetic basis of female mate preferences for multiple male traits, the condition-dependence of these traits, and the genetic association between sexual displays and fitness when mutation-selection balance is inferred. My results advocate caution in the application of existing theory to quantify the strength of indirect selection, suggesting that a good genes process may be fundamentally different when the exaggeration of sexual displays is eventually halted and an equilibrium is reached between opposing selection.

No description available.

No description available.