Education in Cape Verde has come a long way since its humble colonial beginnings with the opening of the first school in 1525. However, despite notable progress and the existence of some schools which are perceived to be superior, the quality of the education provided in the public school system is poor overall. Unfortunately the dearth of research on education in Cape Verde - and indeed the rest of the developing world - available to guide qualitative improvement forces school officials and change agents to use models from developed countries to guide change efforts. The purpose of this study is to examine the validity of that practice based on experiences accumulated during part of a seven year technical aid intervention.

Les changements anthropiques peuvent affecter de manière inattendue les écosystèmes, y compris les agroécosystèmes. Une technologie qui pourrait provoquer de tels effets est l'éclairage artificiel utilisé pour maintenir la productivité des cultures en serre. L'allongement de la photopériode et la qualité spectrale artificielle des luminaires pourraient ainsi affecter les insectes ravageurs et leurs ennemis naturels. L'hypothèse de cette étude est que l'utilisation de lampes DELs peut améliorer l'efficacité de la lutte biologique par le parasitoïde Aphidius matricariae (Hymenoptera: Braconidae) contre le puceron Myzus persicae (Hemiptera: Aphididae), tout en promouvant la croissance du poivron Capsicum annuum (Solanales: Solanaceae). Afin de vérifier cette hypothèse, l'impact de quatre photopériodes (14, 16, 18 et 20 heures) et trois qualités spectrales lors de l'allongement de la photopériode sur l'activité locomotrice du parasitoïde et la fécondité du puceron a été étudié en chambres de croissance. Afin de déterminer les effets de la lumière sur ces organismes et leurs interactions dans un contexte tritrophique, l'impact de l'allongement du jour par deux régimes lumineux de qualité spectrale différente sur la croissance du poivron, la densité des pucerons et la production de momies par le parasitoïde a été étudié en serre. Mes résultats ont démontré que les conditions lumineuses ont influencé la distribution temporelle de l'activité du parasitoïde, mais n'ont pas affecté son activité totale, ni la fécondité du puceron. En serre, l'éclairage supplémentaire a accru de 127 % la matière sèche moyenne des plantes, sans affecter la densité des pucerons. Les parasitoïdes ont réduit de 93 % la densité moyenne des pucerons, peu importe le traitement lumineux étudié. Mes résultats suggèrent donc que les insectes peuvent être résilients aux changements dramatiques de l'environnement lumineux en serre, ce qui démontre qu'il est possible d'ajuster les conditions lumineuses pour optimiser la croissance des plantes sans compromettre la lutte biologique. / Anthropogenic changes to ecosystems can have unexpected side effects, including in agroecosystems. One technology that may produce such side effects is supplemental light, which is used to maintain productivity in greenhouse crops. The lengthened photoperiods and artificial spectral qualities of the lamps used may also affect greenhouse pests and their natural enemies. The hypothesis of this research is that LED greenhouse illumination can improve efficacy of biological control by the parasitoid Aphidius matricariae (Hymenoptera: Braconidae) against the aphid Myzus persicae (Hemiptera: Aphididae) while promoting the growth of the bell pepper Capsicum annuum (Solanales: Solanaceae). To test this hypothesis, the impact of four photoperiods (14, 16, 18, and 20 hours) and three spectral qualities for day length extension on the locomotor activity of the parasitoid and the fecundity of the aphid were tested in growth chambers. To examine the effects of light on the organisms and their interactions in a tri-trophic context, the effects of day extensions of two spectral qualities on bell pepper dry biomass, aphid population density, and parasitoid mummy formation were evaluated in a greenhouse. My results showed that the light conditions influenced the timing of the parasitoid's daily locomotor activity, but not its total activity. Light conditions also had no effect on aphid fecundity. In the greenhouse, supplemental LED lighting promoted plant growth, increasing dry biomass by 127% on average, without affecting aphid densities. Parasitoids reduced aphid populations by 93% on average, regardless of the LED treatment. Therefore, my findings suggest that greenhouse insects can be resilient to dramatic changes in greenhouse lighting, which shows that it is possible to adjust light conditions to optimize plant growth without compromising biocontrol.

Le but de l'étude consistait à prédire le comportement viscoélastique longitudinal du bois vert dans la période correspondant à la vie d'un arbre et d'explorer la diversité de ces propriétés. Le travail s'est déroulé en deux temps. Une analyse exploratoire des propriétés vibratoires a été effectuée sur un large échantillon incluant divers types de bois, y compris des bois de réaction, issus de dix espèces tropicales. Dans un deuxième temps, une sélection restreinte des échantillons a été utilisée pour une étude approfondie du comportement en fluage à long terme. Par ailleurs, la relation avec les paramètres structuraux tels que densité, angle des microfibrilles et pourcentage des éléments anatomiques a été étudiée. Une procédure d'évaluation du fluage à long terme a été mise au point à partir d'essais à différentes températures. L'occurrence du vieillissement physique suite au refroidissement consécutif au chauffage au dessus de la température de transition vitreuse du bois a été mise en évidence. L'applicabilité du principe d'équivalence temps-température a été remise en question par l'analyse des résultats dans le plan complexe approché (PCA). L'hypothèse supplémentaire de la dépendance de la complaisance initiale à la température, similaire à l'élasticité entropique des polymères amorphes, a été proposée et appliquée, permettant de prédire avec succès le comportement à long terme à partir d'essais courts. Le fluage thermoactivé, ainsi que le phénomène de vieillissement physique, ont été décrits par un modèle de Maxwell parabolique identifié à partir de la représentation des résultats dans le PCA. Le comportement en fluage est apparu non corrélé au coefficient d'amortissement mesuré lors des essais vibratoires, suggérant une dissociation entre les mécanismes rhéologiques qui contrôlent le comportement viscoélastiques aux échelles de temps acoustiques (quelques centaines de Hertz) et biologiques (plusieurs années). Enfin, l'hypothèse d'un effet prépondérant de la lamelle mitoyenne sur le processus de fluage a été suggérée pour expliquer les faibles corrélations observées entre le fluage relatif et la structure des parois cellulaires.

Au travers de ce travail de thèse, nous nous sommes fixé comme objectif, de rechercher des solutions pour augmenter la durée de service des outils de coupe du bois, plus spécialement ceux utilisés en première transformation (usure prématurée et ébréchures des arêtes de coupe). Les améliorations apportées à ces outils en laboratoire proviennent des modifications de surfaces par applications de films durs en couches minces (CrN en l'occurrence) ou bien par traitements thermochimiques (nitruration ionique basse température). Par ailleurs, afin de vérifier l'effet des modifications d'outils sous des conditions réelles d'utilisation, nous avons réalisé du déroulage et fraisage industriel de bois vert, avec des couteaux modifiés en surface par des traitements classiquement utilisés (TiN, CrN, DLC et duplex (nitruration + dépôt dur)). Ces tests d'usinage ont été réalisés dans le cadre d'un projet européen CRAFT. Au regard des résultats obtenus, nous avons ensuite débuté le développement de nouveaux films ternaires (CrAlN et ZrBN) pour un éventuel emploi en usinage du bois. La modification de la géométrie par pré-rodage de l'arête de coupe des couteaux réalisée grâce à un traitement mécanique de sablage a également été étudiée. Pour réaliser cette étude un grand nombre d'équipements de caractérisations physicochimiques et tribologiques des films a été utilisé ainsi que des machines de laboratoire et industrielles d'usinage du bois.

Cette thèse, financée par la Fondation des Sciences du Patrimoine, Patrima, a pour but la compréhension du mécanisme chimique responsable du brunissement de certains pigments verts à base de cuivre, sous forme de complexes organométalliques, utilisés en peinture de l'Antiquité à la Renaissance. Il s'agit du « vert de gris » et du « résinate de cuivre ». La variation chromatique de ces pigments engendre une perte de lisibilité de l'œuvre et pose un important problème de conservation. La détermination des réactions impliquées dans l'altération est fondamentale pour définir une méthodologie d'intervention. Afin de clarifier le mécanisme responsable du changement de couleur il est nécessaire d'étudier d'une part la géométrie du centre métallique et d'autre part la nature des liaisons cuivre-ligand. Dans ce but une stratégie multianalytique a été développée. Elle est basée sur la reconstitution des composés utilisés par les artistes et de ceux qui peuvent se former dans les couches picturales, et sur leur caractérisation, avant et après un vieillissement accéléré (thermique oupar exposition à la lumière). La synthèse des ces systèmes modèles a permis d'évaluer l'influence de plusieurs paramètres (solvant, pH, température et temps de réaction) sur la structure des complexes. Les produits de départ et les composés de dégradation ont été analysés par différentes techniques. L'étude des morphologies et la répartition chimique du cuivre au sein des systèmes est réalisée par MEB -EDS. La structure moléculaire a été caractérisée par IRTF diffraction des rayons X, spectroscopie d'absorption des rayons X et par résonance paramagnétique électronique. La caractérisation des produits de dégradation présents dans la phase organique est effectuée par chromatographie en phase gazeuse couplée à la spectrométrie de masse. En parallèle l'étude de leurs propriétés optiques a été réalisée par spectroscopie UV-visible. Les données collectées sont comparées à celles obtenues, par les mêmes méthodes analytiques, sur des échantillons de couches picturales prélevés sur des œuvres peintes plus ou moins altérées. Ce travail comparatif a permis à la fois de mieux comprendre les mécanismes de formation et de transformation au cours du vieillissement de ces systèmes hybrides organique-inorganique, d'évaluer quels paramètres ont plus d'influence sur l'altération et de valider ou d'invalider certaines hypothèses proposées concernant les changements des propriétés optiques observées. / This thesis, funded by Patrima, Foundation for Cultural Heritage Sciences, aims to understand the chemical mechanisms responsible for the darkening of some green copper organometallic pigments (Verdigris and copper resinate) widely used in paintings from Antiquity to the Renaissance. The color variation of these pigments involves a loss of legibility of the artwork and is a major conservation issue. The determination of the reactions involved in the alteration is fundamental to define an intervention methodology. In order to clarify the discoloration process it is necessary to get information on the geometry of the copper cluster and the nature of the copper-ligand bonds. With this aim, a multi-analytical methodology has been developed, based on the reconstitution of ancient pigment and of the compounds that can be formed in the paint layer. These model systems have been characterized before and after an accelerated ageing (by heat or light). Their synthesis has permitted to evaluate the influence of several parameters (solvent, pH, temperature and reaction time) on the resulting structure. The original materials and their degradation products have been analyzed by different techniques. The study of the morphology and chemical distribution of copper in the system was performed by SEM-EDS. The molecular structure was characterized by FTIR, XRD spectroscopy, X-ray absorption and EPR (Electronic Paramagnetic Resonance). The characterization of organic degradation products has been carried out by gas chromatography coupled to mass spectrometry. At the same time a study of optical properties has been performed by UV-visible spectroscopy. The data collected are compared to those obtained, by the same analytical methods, on samples of paint layers taken from ancient paintings. This comparative work has permitted to better understand the mechanisms of formation and transformation of these organic-inorganic systems, to assess which parameters influence the alteration, and to validate or invalidate some preliminary hypothesis about the chromatic changes.

Dans le domaine pharmaceutique, la voie orale demeure la voie d’administration de prédilection, car plus simple et mieux acceptée par les patients. Cependant, ce mode d’administration pose problème pour de nombreux principes actifs (PA) présentant une faible solubilité, une faible perméabilité et/ou une instabilité dans l’environnement gastro-intestinal. Leur micro-encapsulation dans des matrices polymériques peut permettre d’y répondre, notamment si les microparticules générées résistent aux environnements rencontrés lors du tractus gastro-intestinal et jouent alors un rôle protecteur, tant pour le principe actif que pour les muqueuses rencontrées. La recherche de nouveaux excipients, issus des agro-ressources tels que les polymères naturels, est en plein essor. Les protéines végétales, grâce à leurs propriétés fonctionnelles telles qu’une bonne solubilité, une viscosité relativement basse, et des propriétés émulsifiantes et filmogènes, représentent des candidats privilégiés. De plus, la grande diversité de leurs groupements fonctionnels permet d’envisager des modifications chimiques ou enzymatiques variées. L’objectif de ce travail était d’étudier l’intérêt de la protéine de soja en tant que matériau enrobant de principes actifs pharmaceutiques destinés à la voie orale, et plus particulièrement en tant que candidat pour l’élaboration de formes gastro-résistantes. Un isolat protéique de soja (SPI) été utilisé comme matière enrobante et l’atomisation comme procédé. L’ibuprofène, anti-inflammatoire non stéroïdien, a été choisi comme molécule modèle du fait de sa faible solubilité nécessitant une amélioration de sa biodisponibilité, et de ses effets indésirables gastriques nécessitant une mise en forme entérique. Deux modifications chimiques des protéines (l’acylation et la succinylation) ont été étudiées dans le but de modifier la solubilité de la protéine de soja. Ces modifications ont été effectuées dans le respect des principes de la Chimie Verte, notamment en absence de solvant organique. Les microcapsules obtenues par atomisation ont été caractérisées en termes de taux et efficacité d'encapsulation, morphologie et distribution de tailles des particules, état physique du PA encapsulé et capacité de libération en milieu gastrique et intestinal simulé. Les résultats obtenus ont permis de valider l’intérêt des modifications chimiques de la protéine de soja pour moduler les cinétiques de libération d’actif. Les modifications chimiques sont apparues particulièrement adaptées pour l’encapsulation de principes actifs hydrophobes, et ont permis de l’obtention de cinétiques de libération d’ibuprofène ralenties à pH acide (gastrique). La dernière partie de ce travail a permis de valider cette dernière hypothèse par la réalisation de formes gastro-résistantes sur le modèle des comprimés MUPS (multiple unit pellet system). Les résultats de ce travail exploratoire démontrent que les protéines de soja, associées à un procédé de mise en forme multi-particulaire couplé à de la compression directe, peuvent constituer une alternative biosourcée, respectueuse de l’environnement (manipulation en solvant aqueux, temps de séchage et étapes de compression réduits) et sûre à l’enrobage utilisé dans les formes gastro-résistantes traditionnelles.

A partir des années 90, de nombreux professionnels du paysagisme ont participé en France et en Europe à la naturalisation de l'espace vert public dans la ville; ce mouvement a été qualifié parfois de "retour au naturel". Animés par le souci d'appliquer des méthodes plus respectueuses de l'environnement et soucieux d'offrir aux citadins une variété d'espaces de nature allant du jardin sophistiqué à l'espace d'aspect sauvage, les ingénieurs des services d'espaces verts des agglomérations ont expérimenté la "gestion différenciée" de leur patrimoine vert. Parallèlement à cette nouvelle approche essentiellement technique, l'évolution de l'idée de nature urbaine revendiquée par les gestionnaires des services espaces verts a entraîné la création de nouveaux projets de parcs publics alternatifs, dont les images et les usages sont à fortes références champêtre et sauvage. En l'absence de recherche antérieure sur la signification culturelle de la gestion différenciée, cette étude porte autant sur l'origine que sur la construction matérielle et symbolique de ce nouveau type de nature. L'hypothèse que nous voulons vérifier est que la mise en oeuvre de la gestion différenciée, avatar philosophique et esthétique du wild garden anglosaxon du XIXe siècle imprégné des valeurs contemporaines durables, a contribué à la construction de nouveaux territoires urbains publics aux références champêtres et sauvages. Ceux-ci ont été dessinés et gérés par les ingénieurs des services d'espaces verts de nombreuses villes dans la dernière décennie selon de nouveaux modèles de parcs intégrant des intentions écologiques et esthétiques.

Un nouveau procédé d'extraction des huiles essentielles a été développé, mis au point et optimisé nommé "Vapo-diffusion Assistée par Micro-ondes (MSDf)". Le procédé MSDf à été appliqué à l'extraction des huiles essentielles des peaux d'oranges et de fleurs de lavande. Ce procédé est basé sur la combinaison d'une technique d'extraction classique : l'hydro-diffusion et le chauffage micro-ondes comme technologie d'intensification. Une deuxième nouvelle approche pour l'extraction des huiles essentielles à partir de matrices aromatiques sèches et sans ajout de vapeur ni d'eau nommé "Diffusion à sec générée par micro-ondes (MDG)". Le procédé MDG à été appliqué à l'extraction des huiles essentielles des graines de carvi. Le MSDf et le MDG apparaissent comme des procédures "douces" permettant un gain de temps et d'énergie considérable et la diminution des rejets de CO2 dans l'atmosphère

<p>Dette prosjektet har satt søkelyset på to problemstillinger knyttet til samevolusjonen mellom parasitt og vert; 1) utvikling av vertstilpasninger som mottrekk mot tilpasninger hos parasitten, med spesiell fokus på eggtilpasninger, og 2) mekanismer som kan forklare den store variasjonen i forsvarsatferd mot kullparasittisme blant ulike verter.</p><p>1) Flere gjøk- (Cuculus canorus) stammer eller gentes har utviklet egg som er veldig like vertens egne egg, såkalt eggmimikry, for å vanskeliggjøre vertenes eggavvisning. For å svare på dette har mange verter på sin side gjort det vanskeligere for parasitten ved å utvikle en lavere variasjon i utseende mellom sine egg innen kullet (innenkull-variasjon), og en høyere variasjon mellom egg fra kull til kull (mellomkull-variasjon). Mange nordamerikanske spurvefugler blir benyttet som verter av brunhodetrupialen (Molothrus ater), men denne parasitten har ikke utviklet eggmimikry i forhold til vertseggene. Vi sammenlignet kullvariasjonen mellom spurvefugler i Europa og Nord-Amerika og fant en høyere innenkull-variasjon og en lavere mellomkull-variasjon i eggutseende hos nordamerikanske spurvefugler, selv om forskjellen i innenkull-variasjon mellom kontinentene var mindre enn forventet.Hos europeiske spurvefugler er det i tidligere eksperimenter funnet at det er en sammenheng som forventet mellom avvisningsraten overfor parasittiske ikkemimikry egg og kullvariasjonen i eggutseende. Vi fant at det ikke var noen slik sammenheng hos spurvefugler i Nord-Amerika. Resultatene gir støtte til hypotesen om at parasitter med eggmimikry utøver et betydelig seleksjonstrykk for utvikling av bestemte eggkarakterer hos sine verter.</p><p>Vi undersøkte om det var noen forskjell i innenkull-variasjon hos avvisere og akseptorer av parasittegg innen bestemte populasjoner av tre europeiske spurvefugler; rørsanger (Acrocephalus scirpaceus), bokfink (Fringilla coelebs) og munk (Sylvia atricapilla). Det ble funnet at det var en signifikant forskjell i innenkull-variasjon i eggutseende mellom avvisere og akseptorer av kunstige ikke-mimikry gjøkegg i en rørsanger-populasjon i Tsjekkia; avviserne hadde en lavere innenkull variasjon enn akseptorer av slike egg. Denne vertspopulasjonen har en intermediær avvisningsrate overfor ikke-mimikry egg. Et tilsvarende forsøk ble utført hos en bokfink-populasjon i Norge og en munk-populasjon i Tsjekkia. Begge artene er meget gode avvisere av ikke-mimikry egg, noe som indikerer at de aller fleste individer er i stand til å avvise slike egg. Vi valgte derfor å benytte egg fra artsfrender i forsøkene med disse artene. I motsetning til hos rørsangeren fant vi at det ikke var noen forskjell i innenkullvariasjon mellom akseptorer og avvisere av fremmede egg hos bokfink og munk. Hos begge artene ble det funnet at avvisningen av fremmede egg i stor grad avhenger av kontrasten (grad av mimikry) mellom egne egg og parasittegget. Dette viser at selv om individene er i stand til å avvise parasittegg, så finnes det kognitive begrensninger som medfører at egg som utseendemessig ligger under en viss terskelverdi med hensyn til likhet med egne egg vil bli akseptert. Det ble ikke funnet noen indikasjoner på at avvisningsatferden var avhengig av vertenes alder eller av kondisjonelle stimuli for noen av de tre artene. Dette kan tyde på at det er en genetisk basert kobling mellom det å kunne gjenkjenne fremmede egg og innenkull-variasjon. </p><p>2) Mange vertsarter viser ingen eller kun intermediære avvisningsrater overfor fremmede ikke-mimikry egg. En slik tilsynelatende suboptimal atferd kan skyldes at det er kostnader forbundet med avvisningen som forhindrer evolusjon av perfekt avvisningsatferd. Slike kostnader kan være feilaktig avvisning av egne egg i uparasitterte reir (gjenkjenningsfeil), eller avvisning av egne egg i tillegg til parasittegget i parasitterte reir (avvisningskostnader). Hos gjøkverter, som ved suksessfull gjøkparasittisme har en reproduktiv suksess tilnærmet lik null, vil kun gjenkjenningsfeil være kilde til et potensielt seleksjonstrykk mot utvikling av høy avvisning av fremmede egg. Vi undersøkte om slike kostnader forekommer hos bokfink og munk; to arter som antas å ha blitt benyttet av gjøken tidligere, men som i dag ikke blir regelmessig parasittert. På grunn av at avvisningsatferden opprettholdes i fravær av parasittisme, forventet vi at disse artene begår få gjenkjenningsfeil.</p><p>Undersøkelsen gav støtte til denne prediksjonen; avvisningskostnader i parasitterte reir var relativt høye, men gjenkjenningsfeil i uparasitterte reir var meget sjeldent forekommende.</p><p>En hypotese ("spatiell habitat-struktur hypotesen") basert på metapopulasjonsdynamikk og med vekt på karakteristikker vedrørende vertsartenes hekkebiotop ble framsatt for å forklare de store variasjonene i avvisning hos europeiske spurvefugler. Hypotesen bygger på at gjøken benytter de verter som hekker nær utkikkspunkter for parasitten, dvs. nær trær. Arter som hekker både nær og langt fra trær er de beste gjøkvertene, i og med at genflyt fra uparasitterte populasjoner vil forhindre utvikling av perfekt avvisning i parasitterte populasjoner. Arter som alltid hekker nær trær har høye avvisningsrater fordi få eller ingen populasjoner har unnsluppet parasittering, og det har derfor vært sterk seleksjon for utvikling av vertsforsvar. Data for gjøkverter i Europa gav god støtte til hypotesen. Grad av parasitt eggmimikry og parasitteringsrater er høyest hos de vertsarter som kan hekke både langt fra trær og nær trær, noe som tyder på at gjøken har størst suksess hos slike arter.</p> / <p>This thesis deals with two topics in the coevolution between brood parasites and their hosts: 1) evolution of host adaptations against parasite egg mimicry, and 2) sources that could explain the considerable variation in rejection behaviour found among various passerines.</p><p>1) Several common cuckoo (<i>Cuculus canorus</i>) tribes or gentes in Europe have evolved eggs that are remarkably similar to the host eggs in both size and appearance (i.e. egg mimicry). To counter this adaptation in the parasite, hosts can produce eggs with similar appearance within clutches (low intraclutch variation) as well as eggs with diverging appearance between different clutches (high interclutch variation). Many North American passerines are utilised as hosts by the brown-headed cowbird (<i>Molothrus ater</i>). However, this parasite generally lays non-mimetic eggs. As predicted, we found that European passerines had a lower intraclutch variation and a higher interclutch variation in egg appearance than North American passerines.</p><p>However, the difference in intraclutch variation between the continents was less than expected. A relationship has previously been found among European passerines between the rejection rate of non-mimetic eggs and clutch variation in egg appearance, and this is thought to reflect the stage in the coevolution between parasite and host. We found no evidence of such patterns among North American species. These results provide support for the hypothesis that specific host clutch variation is a counteradaptation against parasite egg mimicry.</p><p>We investigated whether there was any difference in clutch variation between acceptors and rejecters of parasitic eggs within populations of three European passerines; reed warblers (<i>Acrocephalus scirpaceus</i>), chaffinches (<i>Fringilla coelebs</i>), and blackcaps (<i>Sylvia atricapilla</i>). In a Czech reed warbler population with an intermediate rejection rate of non-mimetic cuckoo eggs, it was found that rejecters had a statistically significant lower intraclutch variation than acceptors of such eggs. Age or conditional stimuli did not seem to have any influence on the rejectionbehaviour. A similar experiment was carried out in a Norwegian chaffinch population and a Czech blackcap population, which, however, were experimentally parasitised with foreign conspecific eggs because they are both very good rejecters of nonmimetic parasitic eggs. We found no difference in intraclutch variation among acceptors and rejecters of foreign eggs in chaffinches and blackcaps. However, it was found that the rejection of conspecific eggs greatly depends upon the contrast (i.e. mimicry) between the parasitic and their own eggs. It therefore seems that even though individuals have the ability to reject foreign eggs, limitations in their cognitive system entails that parasitic eggs that are too similar to the host eggs will be accepted. We also looked for potential effects of age on rejection behaviour and intraclutch variation, but no relationship between these variables was found. The results indicate that in these three species both rejection behaviour and clutch variation are more or less innate features, and also that there is a genetically based linkage between recognition of odd eggs and intraclutch variation in egg appearance.</p><p>2) Many hosts of brood parasites show no or only intermediate rejection rates of foreign non-mimetic eggs. Evolution of proper rejection behaviour could be prevented by costs related to egg rejection. Important in this respect are erroneous rejection of their own eggs in non-parasitised nests (recognition errors) and rejection of their own eggs in addition to the parasitic egg in parasitised clutches (rejection costs). Because successful cuckoo parasitism usually is detrimental to the breeding success of the host, only recognition errors are believed to be important as an opposing selective pressure against proper host defence in cuckoo hosts. We examined whether such costs exist in chaffinches and blackcaps. These species maintain a high rejection rate of foreign eggs, even though they are not currently used as hosts by the cuckoo. We therefore predicted that recognition errors should be absent or at least rare in these species. We found support for this prediction; rejection costs were relatively high but recognition errors were at best rare events.</p><p>In another investigation, we proposed a hypothesis (the "spatial habitat structure hypothesis") based upon metapopulation dynamics and characteristics concerning host breeding habitats to explain the variation in rejection behaviour found among European passerines. This hypothesis is based upon the fact that the cuckoo, as well as other avian brood parasites, needs access to vantage points in trees to monitor host nests, and thus only species breeding near trees are available as hosts. Our results were very much in accordance with this hypothesis. Species that breed both near and far away from trees are the best cuckoo hosts, because gene flow from non-parasitised populations breeding far from trees will prevent the evolution of proper rejection behaviour in parasitised populations breeding near trees. However, species that always breed near trees have high rejection rates because the majority of the populations have been utilised as hosts, and thus there has been a strong selection for the evolution of host defences. Furthermore, the level of parasite egg mimicry and the level of parasitism was found to be highest among hosts breeding both near and far away from trees, indicating that the cuckoo is most successful when utilising such species as hosts.</p> / Paper VI is not included as a paper in this thesis, but is included as the introduction.

Dette prosjektet har satt søkelyset på to problemstillinger knyttet til samevolusjonen mellom parasitt og vert; 1) utvikling av vertstilpasninger som mottrekk mot tilpasninger hos parasitten, med spesiell fokus på eggtilpasninger, og 2) mekanismer som kan forklare den store variasjonen i forsvarsatferd mot kullparasittisme blant ulike verter. 1) Flere gjøk- (Cuculus canorus) stammer eller gentes har utviklet egg som er veldig like vertens egne egg, såkalt eggmimikry, for å vanskeliggjøre vertenes eggavvisning. For å svare på dette har mange verter på sin side gjort det vanskeligere for parasitten ved å utvikle en lavere variasjon i utseende mellom sine egg innen kullet (innenkull-variasjon), og en høyere variasjon mellom egg fra kull til kull (mellomkull-variasjon). Mange nordamerikanske spurvefugler blir benyttet som verter av brunhodetrupialen (Molothrus ater), men denne parasitten har ikke utviklet eggmimikry i forhold til vertseggene. Vi sammenlignet kullvariasjonen mellom spurvefugler i Europa og Nord-Amerika og fant en høyere innenkull-variasjon og en lavere mellomkull-variasjon i eggutseende hos nordamerikanske spurvefugler, selv om forskjellen i innenkull-variasjon mellom kontinentene var mindre enn forventet.Hos europeiske spurvefugler er det i tidligere eksperimenter funnet at det er en sammenheng som forventet mellom avvisningsraten overfor parasittiske ikkemimikry egg og kullvariasjonen i eggutseende. Vi fant at det ikke var noen slik sammenheng hos spurvefugler i Nord-Amerika. Resultatene gir støtte til hypotesen om at parasitter med eggmimikry utøver et betydelig seleksjonstrykk for utvikling av bestemte eggkarakterer hos sine verter. Vi undersøkte om det var noen forskjell i innenkull-variasjon hos avvisere og akseptorer av parasittegg innen bestemte populasjoner av tre europeiske spurvefugler; rørsanger (Acrocephalus scirpaceus), bokfink (Fringilla coelebs) og munk (Sylvia atricapilla). Det ble funnet at det var en signifikant forskjell i innenkull-variasjon i eggutseende mellom avvisere og akseptorer av kunstige ikke-mimikry gjøkegg i en rørsanger-populasjon i Tsjekkia; avviserne hadde en lavere innenkull variasjon enn akseptorer av slike egg. Denne vertspopulasjonen har en intermediær avvisningsrate overfor ikke-mimikry egg. Et tilsvarende forsøk ble utført hos en bokfink-populasjon i Norge og en munk-populasjon i Tsjekkia. Begge artene er meget gode avvisere av ikke-mimikry egg, noe som indikerer at de aller fleste individer er i stand til å avvise slike egg. Vi valgte derfor å benytte egg fra artsfrender i forsøkene med disse artene. I motsetning til hos rørsangeren fant vi at det ikke var noen forskjell i innenkullvariasjon mellom akseptorer og avvisere av fremmede egg hos bokfink og munk. Hos begge artene ble det funnet at avvisningen av fremmede egg i stor grad avhenger av kontrasten (grad av mimikry) mellom egne egg og parasittegget. Dette viser at selv om individene er i stand til å avvise parasittegg, så finnes det kognitive begrensninger som medfører at egg som utseendemessig ligger under en viss terskelverdi med hensyn til likhet med egne egg vil bli akseptert. Det ble ikke funnet noen indikasjoner på at avvisningsatferden var avhengig av vertenes alder eller av kondisjonelle stimuli for noen av de tre artene. Dette kan tyde på at det er en genetisk basert kobling mellom det å kunne gjenkjenne fremmede egg og innenkull-variasjon. 2) Mange vertsarter viser ingen eller kun intermediære avvisningsrater overfor fremmede ikke-mimikry egg. En slik tilsynelatende suboptimal atferd kan skyldes at det er kostnader forbundet med avvisningen som forhindrer evolusjon av perfekt avvisningsatferd. Slike kostnader kan være feilaktig avvisning av egne egg i uparasitterte reir (gjenkjenningsfeil), eller avvisning av egne egg i tillegg til parasittegget i parasitterte reir (avvisningskostnader). Hos gjøkverter, som ved suksessfull gjøkparasittisme har en reproduktiv suksess tilnærmet lik null, vil kun gjenkjenningsfeil være kilde til et potensielt seleksjonstrykk mot utvikling av høy avvisning av fremmede egg. Vi undersøkte om slike kostnader forekommer hos bokfink og munk; to arter som antas å ha blitt benyttet av gjøken tidligere, men som i dag ikke blir regelmessig parasittert. På grunn av at avvisningsatferden opprettholdes i fravær av parasittisme, forventet vi at disse artene begår få gjenkjenningsfeil. Undersøkelsen gav støtte til denne prediksjonen; avvisningskostnader i parasitterte reir var relativt høye, men gjenkjenningsfeil i uparasitterte reir var meget sjeldent forekommende. En hypotese ("spatiell habitat-struktur hypotesen") basert på metapopulasjonsdynamikk og med vekt på karakteristikker vedrørende vertsartenes hekkebiotop ble framsatt for å forklare de store variasjonene i avvisning hos europeiske spurvefugler. Hypotesen bygger på at gjøken benytter de verter som hekker nær utkikkspunkter for parasitten, dvs. nær trær. Arter som hekker både nær og langt fra trær er de beste gjøkvertene, i og med at genflyt fra uparasitterte populasjoner vil forhindre utvikling av perfekt avvisning i parasitterte populasjoner. Arter som alltid hekker nær trær har høye avvisningsrater fordi få eller ingen populasjoner har unnsluppet parasittering, og det har derfor vært sterk seleksjon for utvikling av vertsforsvar. Data for gjøkverter i Europa gav god støtte til hypotesen. Grad av parasitt eggmimikry og parasitteringsrater er høyest hos de vertsarter som kan hekke både langt fra trær og nær trær, noe som tyder på at gjøken har størst suksess hos slike arter. / This thesis deals with two topics in the coevolution between brood parasites and their hosts: 1) evolution of host adaptations against parasite egg mimicry, and 2) sources that could explain the considerable variation in rejection behaviour found among various passerines. 1) Several common cuckoo (Cuculus canorus) tribes or gentes in Europe have evolved eggs that are remarkably similar to the host eggs in both size and appearance (i.e. egg mimicry). To counter this adaptation in the parasite, hosts can produce eggs with similar appearance within clutches (low intraclutch variation) as well as eggs with diverging appearance between different clutches (high interclutch variation). Many North American passerines are utilised as hosts by the brown-headed cowbird (Molothrus ater). However, this parasite generally lays non-mimetic eggs. As predicted, we found that European passerines had a lower intraclutch variation and a higher interclutch variation in egg appearance than North American passerines. However, the difference in intraclutch variation between the continents was less than expected. A relationship has previously been found among European passerines between the rejection rate of non-mimetic eggs and clutch variation in egg appearance, and this is thought to reflect the stage in the coevolution between parasite and host. We found no evidence of such patterns among North American species. These results provide support for the hypothesis that specific host clutch variation is a counteradaptation against parasite egg mimicry. We investigated whether there was any difference in clutch variation between acceptors and rejecters of parasitic eggs within populations of three European passerines; reed warblers (Acrocephalus scirpaceus), chaffinches (Fringilla coelebs), and blackcaps (Sylvia atricapilla). In a Czech reed warbler population with an intermediate rejection rate of non-mimetic cuckoo eggs, it was found that rejecters had a statistically significant lower intraclutch variation than acceptors of such eggs. Age or conditional stimuli did not seem to have any influence on the rejectionbehaviour. A similar experiment was carried out in a Norwegian chaffinch population and a Czech blackcap population, which, however, were experimentally parasitised with foreign conspecific eggs because they are both very good rejecters of nonmimetic parasitic eggs. We found no difference in intraclutch variation among acceptors and rejecters of foreign eggs in chaffinches and blackcaps. However, it was found that the rejection of conspecific eggs greatly depends upon the contrast (i.e. mimicry) between the parasitic and their own eggs. It therefore seems that even though individuals have the ability to reject foreign eggs, limitations in their cognitive system entails that parasitic eggs that are too similar to the host eggs will be accepted. We also looked for potential effects of age on rejection behaviour and intraclutch variation, but no relationship between these variables was found. The results indicate that in these three species both rejection behaviour and clutch variation are more or less innate features, and also that there is a genetically based linkage between recognition of odd eggs and intraclutch variation in egg appearance. 2) Many hosts of brood parasites show no or only intermediate rejection rates of foreign non-mimetic eggs. Evolution of proper rejection behaviour could be prevented by costs related to egg rejection. Important in this respect are erroneous rejection of their own eggs in non-parasitised nests (recognition errors) and rejection of their own eggs in addition to the parasitic egg in parasitised clutches (rejection costs). Because successful cuckoo parasitism usually is detrimental to the breeding success of the host, only recognition errors are believed to be important as an opposing selective pressure against proper host defence in cuckoo hosts. We examined whether such costs exist in chaffinches and blackcaps. These species maintain a high rejection rate of foreign eggs, even though they are not currently used as hosts by the cuckoo. We therefore predicted that recognition errors should be absent or at least rare in these species. We found support for this prediction; rejection costs were relatively high but recognition errors were at best rare events. In another investigation, we proposed a hypothesis (the "spatial habitat structure hypothesis") based upon metapopulation dynamics and characteristics concerning host breeding habitats to explain the variation in rejection behaviour found among European passerines. This hypothesis is based upon the fact that the cuckoo, as well as other avian brood parasites, needs access to vantage points in trees to monitor host nests, and thus only species breeding near trees are available as hosts. Our results were very much in accordance with this hypothesis. Species that breed both near and far away from trees are the best cuckoo hosts, because gene flow from non-parasitised populations breeding far from trees will prevent the evolution of proper rejection behaviour in parasitised populations breeding near trees. However, species that always breed near trees have high rejection rates because the majority of the populations have been utilised as hosts, and thus there has been a strong selection for the evolution of host defences. Furthermore, the level of parasite egg mimicry and the level of parasitism was found to be highest among hosts breeding both near and far away from trees, indicating that the cuckoo is most successful when utilising such species as hosts. / Paper VI is not included as a paper in this thesis, but is included as the introduction.