<p>Biological invasions are an important issue of global change and an increased understanding of invasion processes is of crucial importance for both conservation managers and international trade. In this thesis, I have studied the invasion of the brown seaweed <i>Fucus evanescens</i>, to investigate the fate and effect of a perennial, habitat-forming seaweed introduced to a coastal ecosystem. A long-term study of the spread of <i>F. evanescens</i> in Öresund (southern Sweden) showed that the species was able to expand its range quickly during the first 20 years after the introduction, but that the expansion has been slow during the subsequent 30 years. Both in Öresund and in Skagerrak, the species is largely restricted to sites where native fucoids are scarce. Laboratory experiments showed that the restricted spread of <i>F. evanescens</i> cannot be explained by the investigated abiotic factors (wave exposure and salinity), although salinity restricts the species from spreading into the Baltic Sea. Neither did I find evidence for that herbivores or epibiota provide biotic resistance to the invader. On the contrary, <i>F. evanescens</i> was less consumed by native herbivores, both compared to the native fucoids and to <i>F. evanescens</i> populations in its native range, and little overgrown by epiphytes. Instead, the restricted spread may be due to competition from native seaweeds, probably by pre-occupation of space, and the establishment has probably been facilitated by disturbance. </p><p>The studies provided little support for a general enemy release in introduced seaweeds. The low herbivore consumption of <i>F. evanescens</i> in Sweden could not be explained by release from specialist herbivores. Instead, high levels of chemical anti-herbivore defence metabolites (phlorotannins) could explain the pattern of herbivore preference for different fucoids. Likewise, the low epibiotic colonisation of <i>F. evanescens </i>plants could be explained by high resistance to epibiotic survival. This shows that colonisation of invading seaweeds by native herbivores and epibionts depends on properties of the invading species. The large differences between fucoid species in their quality as food and habitat for epibionts and herbivores imply that invasions of such habitat-forming species may have a considerable effect on a number of other species in shallow coastal areas. However, since <i>F. evanescens</i> did not exclude other fucoids in its new range, its effect on the recipient biota is probably small.</p>

Les invasions biologiques sont un fardeau économique important si elles affectent des ressources critiques pour l’alimentation, la sante humaine ou les productions agricoles. Les ravageurs de la pomme de terre sont un challenge économique important tant ce tubercule est un aliment clé dans les pays andins. Il est possible de suivre la dispersion récente de la teigne du Guatemala, T. solanivora en Amérique du Sud depuis son introduction au Vénézuela à sa propagation progressive vers le sud. Par ailleurs, les invasions récentes fournissent un modèle unique pour analyser les processus d’adaptation de tout l’écosystème receveur au nouveau venu. Cette introduction de T. solanivora et sa coexistence avec la teigne endémique Phthorimaea operculella, nous offre la possibilité d’étudier l’adaptation de populations virales inféodées à P. operculella au nouvel hôte T. solanivora. Une étude de terrain a été engagée dans les régions productrices de pomme de terre en Colombie. A partir des larves de T. solanivora collectées sur 5 sites distincts, des infections à granulovirus ont été détectées. Tous les isolats viraux sont apparentés au Phthorimaea operculella granulovirus (PhopGV) précédemment décrit. Des différences de pathogénicité envers les deux hôtes ont été observées. Une variabilité a été détectée pour certains isolats au niveau de deux marqueurs génétiques. Les populations présentant une diversité génétique s’avèrent plus pathogènes sur les deux hôtes que des populations génétiquement homogènes. Elles offrent une opportunité pour le contrôle biologique de ces ravageurs. Des populations artificielles ont été construites pour mimer des populations naturelles mélangées. Elles se comportent de la même manière d’un point de vue biologique, mais l’évolution de la fréquence des marqueurs n’est pas liée à l’efficacité biologique, ce qui suggère que des différences non détectées dans le génome pourraient être responsables de l’adaptation de l’hôte. La productivité des infections dans les deux hôtes a été étudiée car elle est la clé de voute du développement d’un agent de contrôle biologique. Les productivités sur P. operculella (1,36 à 2,69 × 108 OBs/ mg) et T. solanivora (0,48 à 3,64 × 108 OBs/mg) ne sont pas très différentes. Les populations génétiquement mélangées ne se distinguent pas des populations homogènes par leur production totale dans l’un ou l’autre des deux hôtes, cependant, les rendements (production virale/inoculum) montrent des différences claires, les populations mélangées (naturelles ou artificielles) sont plus performantes sur les deux hôtes. Aucune réduction de la pathogénicité sur l’hôte d’origine n’a été observée après plusieurs cycles de réplication de la population virale sur l’hôte alternatif. Les populations virales originellement adaptées à P. operculella ont évolué pour infecter T. solanivora. Dans les régions où les deux hôtes sont présents, les populations virales développent une stratégie pour être efficaces sur les deux hôtes. / Biological invasions constitute an important economical burden when they affect key resources for human alimentation, health or agronomic productions. Potato pests are important as this tuber is a key food source in Andean countries. The recent dispersion of the Guatemalan potato tuber moth, T. solanivora in South America can be traced back to the introduction in Venezuela, with progressive dispersion towards the South. Recent invasions provide, in addition, a unique model to analyse the process of adaptation of the whole receiving ecosystem to the new comers. This introduction of T. solanivora and its coexistence with the endemic potato tuber moth, Phthorimaea operculella offer us the possibility of studying the adaptation to T. solanivora of virus populations infeodated to the later. A survey has been carried out in the potato-producing regions of Colombia. From the T. solanivora larvae collected, granulovirus infections were detected in five different locations. All virus isolates are related to the previously described Phthorimaea operculella granulovirus (PhopGV). Differences in the pathogenicity against the two hosts were observed. Variability was detected in some isolates at two genetic markers. Genetically diverse populations appear to be more pathogenic for both hosts than genetically homogeneous populations. They provide a possible solution for the biological control of these insect pests. Artificial populations were constructed to mimic the mixed natural populations. They behave similarly from a biological point of view, but the evolution of the markers frequencies is not related to the biological efficacy, suggesting that undetected differences in the genomes could be responsible of this host adaptation. The productivity of the infections in both hosts has been studied as it constitutes a key point for the development of a biocontrol agent. The productivity in P. operculella (1.36 to 2.69 × 108 OBs/ mg) and T. solanivora (0.48 to 3.64 × 108 OBs/mg) are not very different. Genotypically mixed populations cannot be differentiated from homogeneous populations by their total production in one or the other host, however, the yields (virus output/doses to infect) show clear differences, mixed populations (natural or artificial) perform better in both hosts. No reduction in the pathogenicity for one host was observed after few cycles of replication of the virus population in the second host. Virus populations originally adapted to P. operculella had evolved to infect T. solanivora. In regions where both host are present, the populations developed a strategy to be efficient on both hosts.

Les parasites exercent une pression de sélection quasiment universelle. Cette thèse aborde les relations hôte-parasite dans le système Lymnaeidae (Gastropoda) Fasciolidae (Trematoda, douves) sous différents aspects, afin de brosser une image de cette interaction et de son évolution. J'ai tout d'abord établi les relations phylogénétiques entre les espèces de Lymnaeidae, puis retracé l'évolution de différents caractères, tels que la susceptibilité à l'infestation par Fasciola hepatica et F. gigantica. Alors que F. hepatica est un parasite généraliste, capable d'infester des mollusques de presque tous les clades de la famille Lymnaeidae, l'infestation par F. gigantica est plus restreinte à un clade. J'ai ensuite étudié plus finement la coévolution entre le parasite F. hepatica et deux de ses hôtes Lymnaeidae (Galba truncatula et Omphiscola glabra) au sein d'une métapopulation, ce qui a confirmé la stratégie généraliste de F. hepatica. En plus, il semblerait que les mollusques parasités et non parasités de G. truncatula aient des différences génétiques, au moins dans cinq des huit populations étudiées. J'ai caractérisé la diversité génétique de deux espèces de mollusques envahissantes, préférentiellement autogames et impliquées dans la transmission de F. hepatica : Pseudosuccinea columella et Lymnaea sp. On trouve une diversité génétique très réduite, chez ces deux espèces, ce qui pourrait faciliter leur expansion géographique et leur infestation par F. hepatica. Ce travail m'a ensuite conduit à mesurer le temps d'attente avant l'autofécondation et la dépression de consanguinité chez ces deux espèces. J'ai trouvé que ces deux espèces sont caractérisées par une dépression de consanguinité très faible et un temps d'attente nul, ce qui confirme les résultats obtenus lors d'une collaboration dans une étude à plus large échelle. Cette thèse souligne l'importance des études en évolution pour comprendre l'épidémiologie des maladies parasitaires. / Parasites constitute a selective pressure to almost all living beings. This thesis addresses the host-parasite interaction in the Lymnaeidae (Gastropoda) Fasciolidae (Trematoda; liver flukes) system through different approaches, with the aim to give a comprehensive image of this interaction and its evolution. First, I established the phylogenetic relationships among Lymnaeidae species, and then mapped the evolution of different characters such as the susceptibility to the infection by Fasciola hepatica and F. gigantica. While F. hepatica is a generalist parasite, capable to infect snails from almost all clades of the Lymnaeidae, infection by F. gigantica is restricted to one clade. Next, I studied the co-evolution between the parasite F. hepatica and two of its intermediate host species (Galba truncatula and Omphiscola glabra) at a finer scale: within a metapopulation. This study confirmed the generalist strategy of F. hepatica. In addition, it seems that parasitized and non-parasitized G. truncatula snails exhibit genetic differences, at least in five out of eight studied populations.I also characterized the genetic diversity of two species of invasive snails involved in the transmission of F. hepatica: Pseudosuccinea columella and Lymnaea sp. We discuss the possible reasons of invasion success in these snails, despite their low genetic diversity, which could facilitate their infection by F. hepatica. Their capacity to respond to parasitism is certainly reduced, all the more that these species are preferential selfers. This work has then led me to measure the waiting time before self-fertilization and inbreeding depression in these two snails. I found that these two species are characterized by low inbreeding depression and present no waiting time, which confirms the results obtained in a collaborative project at larger phylogenetic scale. This thesis strengthens the importance of evolutionary studies to understand the epidemiology of parasitic diseases.

A invasão por gramíneas africanas é uma das principais ameaças à diversidade biológica e funcional das formações savânicas do cerrado. Uma questão central em invasões por plantas é entender como espécies oriundas de sistemas com diferentes pressões seletivas são capazes de adquirir e utilizar recursos em um novo ambiente. O objetivo deste estudo foi de comparar os padrões de uso luz e água de gramíneas nativas e invasoras em duas condições de cobertura arbórea ao longo de um gradiente de invasão em uma área de cerrado localizada no Parque Estadual do Juquery - SP. O efeito da presença da cobertura sobre os indivíduos foi investigado através de amostragens em fisionomia de campo (dossel ausente) e de campo cerrado (dossel presente). A hipótese de que o desempenho no uso de recursos de invasoras supera o de nativas em presença de cobertura arbórea foi testada através da medição de aspectos ligados à capacidade de interceptação da luz e ao desempenho de uso da luz e água associado ao potencial de ganho de carbono. As gramíneas africanas apresentaram padrões de arranjo espacial das folhas e estrutura foliar que diferem dos de nativas. As copas das invasoras apresentaram maior grau de sobreposição entre as folhas. As folhas das invasoras exibiram menores inclinações, menor espessura e menor razão de massa seca por área. A presença do dossel arbóreo favoreceu o transporte fotoquímico de elétrons de invasoras e desfavoreceu o de nativas. As principais diferenças nas dinâmicas de trocas gasosas entre nativas e invasoras foram observadas quando as razões foram expressas por massa seca foliar. Diferenças nas relações entre as variáveis indicam que a relação do padrão de investimento em tecidos foliares com o potencial fotossintético pode implicar em respostas contrastantes de nativas e invasoras aos regimes de irradiância nas fisionomias de campo sujo e campo cerrado / Invasions by African grasses constitutes one of the greatest threats to the biological and functional diversity of the cerrado savannas. A central question to understand plant invasions is how species originating from systems with distinct selective pressures are able to acquire and utilize resources in their new environments. The objective of this study was to compare native and invasive grasses with respect to their light and water use patterns in two sites with differing canopy tree cover along an invasion gradient in a cerrado area in the Juquery State Park, SP. The effect of the canopy tree cover on the grass individuals was investigated through sampling in the cerrado physiognomies of \'campo sujo\' (tree cover absent) and \'campo cerrado\' (tree cover present). The hypothesis that invaders outperform natives under the presence of canopy tree cover in terms of resource-use was tested through field measurements of traits related to light interception capacity and to the performance in light and water use associated with carbon gain potentials. Invasive African grasses showed distinct patterns of crown architecture and leaf structure when compared to natives. Invaders showed a greater degree of leaf overlapping in the crown. Invaders leaves were thinner, relatively more planophile, and showed smaller leaf dry mass per area ratios. The presence of canopy tree cover favored the invaders photochemical electron transport while disfavored natives. The main differences concerning native and invaders gas-exchange dynamics were observed on a leaf dry mass basis. The relation between patterns of leaf tissue investment and photosynthetic potentials could imply in contrasting responses of native and invasive grasses to the irradiance regime in differing cerrado vegetation types

As invasões por espécies exóticas vêm acarretando a homogeneização das comunidades naturais tornando-se; atualmente, a segunda principal causa da perda da biodiversidade. Entre os domínios brasileiros mais afetados está o Cerrado, que possui mais de 11 fitofisionomias e 33% de toda biodiversidade vegetal do país. As alterações pela atividade humana foram facilitadas em decorrência das áreas de Cerrado estarem dispostas em quase todo o território nacional e, na fisionomia de campo sujo, em particular, pela facilidade de mecanização e consequentemente pelo emprego da agricultura em larga escala. Dessa forma, foram inseridas gramíneas exóticas africanas para a melhoria da qualidade e produtividade do forrageio. Essas gramíneas se ajustaram tão bem em áreas de campo sujo que se expandiram desordenadamente e conduziram a um desequilíbrio estrutural e funcional destas áreas. O objetivo deste estudo foi investigar os padrões de utilização de recursos em gramíneas nativas e invasoras em um campo sujo levando em conta os períodos de maior e menor disponibilidade hídrica e os possíveis efeitos das variáveis microclimáticas sobre as respostas fisiológicas de gramíneas nativas e invasoras. A área de estudo localiza-se no Parque Estadual do Juquery, município de Franco da Rocha, Estado de São Paulo. As espécies utilizadas para o estudo foram as nativas: Tristachya leiostachya, Imperata brasiliensis, Axonopus pressus e Axonopus siccus. E as invasoras: Urochloa brizantha e Melinis minutiflora. As respostas fisiológicas frente às flutuações entre os períodos chuvoso e seco foram obtidas através de análises de trocas gasosas, análise de fluorescência e de medições das variáveis microclimáticas ao longo de 23 meses entre 8:00 e 12:00. Os resultados indicaram que gramíneas nativas investem 16% menos em fotoproteção do que gramíneas invasoras, sendo estas mais sensíveis aos fotodanos. No entanto, gramíneas nativas investem 13% a mais na síntese de clorofila total e se equivalem na capacidade de assimilação de carbono (A) e quanto as taxas de transporte de elétrons (ETR) indicando uma aprovável equivalência no uso da luz que pode estar associada a maiores taxas de eficiência no uso do nitrogênio em nativas tanto no período chuvoso (70%) quanto no seco (70%), além da ausência de diferença estatística no conteúdo de fósforo foliar por massa seca (Pmg/ms) e na eficiência no uso deste recurso (EUP). A condutância estomática (gs) e transpiração (E) foram respectivamente 22 e 60% maiores nas gramíneas nativas, por isso, a eficiência no uso da água foi 60% menor nestas em comparação às invasoras, tanto no período chuvoso quanto no seco. Tal regulação da condutância provavelmente esteve relacionada ao conteúdo de nitrogênio foliar que foi 100% maior nas invasoras no período chuvoso e 85% maior no período seco. Como esperado, gramíneas nativas apresentaram respostas de menor sensibilidade à seca em comparação às invasoras. Gramíneas invasoras, por sua vez, não corroboraram a expectativa de maior oportunismo em relação às nativas no período chuvoso quanto a captação de recursos, assim como, diferentemente do esperado, não se mostraram tão sensíveis à seca devido a um comportamento conservador no uso da água / Invasions by alien species are causing the homogenization of natural communities, becoming; currently the second leading cause of biodiversity loss. And among the most affected areas are the Brazilian Cerrado, which has more than 11 vegetation types and 33 % of all the country\'s biodiversity. The changes were facilitated by human activity due to the Cerrado areas are arranged in almost all the national territory. And in the face of dirty field, in particular, the ease of mechanization and consequently the employment of agriculture. Thus, African exotic grasses to improve the quality and productivity of foraging were inserted. These grasses have adjusted so well in areas of Cerrado dirty field that expanded wildly and led to a structural and functional imbalance of these areas. The aim of this study was to investigate the patterns of resource use of native grasses and weeds in a field Cerrado dirty taking into account periods of high and low water availability and the possible effects of microclimatic variables on the physiological responses of native and invasive grasses . The study area is located in the Juquery State Park, city of Franco da Rocha, State of Sao Paulo. The species used for the study were native: Tristachya leiostachya , Imperata brasiliensis , Axonopus pressus and Axonopus siccus. And the invaders: Urochloa brizantha and Melinis minutiflora. The physiological responses to fluctuations between wet and dry periods were obtained from analyzes of gas exchange, fluorescence analysis and measurements of microclimate variables over 23 months between 8:00 am and 12:00 am. The results indicate that native grasses Invest 16 % less than photoprotection grass weeds, which are more sensitive to photodamage. However, native grasses invest 13 % more in the synthesis of chlorophyll and are equivalent in capacity to carbon assimilation (A) and on the rate of electron transport (ETR) indicating an approvable equivalence in the use of light that can be associated with higher rates of nitrogen use efficiency in native both in the rainy season (70%) and the dry (70 %). Besides the lack of statistical difference in the content of phosphorus of leaf dry mass (Pmg/ms) and efficient use of this resource (EUP). The stomatal conductance (gs ) and transpiration ( E) were respectively 22 and 60 % higher in native grasses , so ; efficiency in water use was 60 % lower compared to these invasive both in the rainy season as in the dry. Such regulation conductance probably was related to leaf N content that was 100 % higher in invasive in the rainy season and 85 % higher in the dry season. As expected, responses of native grasses showed lower sensitivity to drought compared to invasive. Invasive grasses in tur ; does not support the expectation of greater opportunism in relation to native in the rainy season as fundraising , as well as , unlike expected, were not as sensitive to drought due to a more conservative behavior in water use

The red imported fire ant, Solenopsis invicta, is a South American native introduced in Alabama in the early 20th century. This predatory species has rapidly spread throughout the southeastern US and parts of the West Indies, inflicting great ecological and economic damage. For example, Solenopsis invicta is known to attack the eggs and hatchlings of ground nesting birds and reptiles. The ants swarm into the nests attacking hatchlings and diminishing their chance for survival. My thesis research aimed to survey the distribution of ants on sea turtle nesting beaches in South Florida and St. Croix, USVI, and to evaluate the possible threat of Solenopsis invicta and other predatory ants to sea turtle hatchlings. / Includes bibliography. / Thesis (M.S.)--Florida Atlantic University, 2013.

Nous traversons actuellement une crise de perte de la biodiversité sans précédant. La dégradation des sols et la perte d’habitat, la pollution, la surexploitation et les invasions biologiques contribuent à cette perte mondiale de biodiversité. Par ailleurs, le changement climatique et ses interactions avec les autres menaces, sont probablement l’un des défis majeurs des prochaines décennies pour la biodiversité. À l’heure actuelle, en raison de la multiplication des études et des approches employées, il est difficile d’avoir une vision synthétique des conséquences potentielles de ces changements sur la biodiversité. L’objectif principal de ce travail de thèse a été d’améliorer la caractérisation et la quantification des différents impacts des changements climatiques sur la biodiversité, à l’échelle mondiale par des approches de modélisations et de méta-analyses. Une première partie de mes travaux a ainsi porté sur les conséquences potentielles de la hausse du niveau des mers sur les hotspots insulaires, au cours de laquelle j’ai mis en évidence les conséquences majeures d’une telle hausse pour certains de ces hotspots. Je me suis ensuite intéressée à l’étude des effets conjugués des changements climatiques et des changements d’utilisation des sols sur les invasions biologiques à l’échelle mondiale. Cette partie a permis de mettre en évidence que les conséquences des changements climatiques et des changements d’utilisation des sols sur les espèces invasives dépendent de la région, du taxon et de l’espèce considérée. Ainsi, j’ai mis en évidence que certaines régions pourraient être moins favorables à la présence d’espèce invasives dans le futur. En outre, cette partie a également mis en évidence que les hotspots majoritairement composés d’îles étaient particulièrement favorables à la présence de ces espèces invasives. Finalement, dans une dernière partie, j’ai étudié les conséquences des menaces futures pour les hotspots de biodiversité dans une perspective de conservation. Cette partie a notamment permis d’établir des priorités de recherche et de conservation entre les hotspots de biodiversité en tenant compte des futures menaces qui pèsent sur la biodiversité à l’échelle des hotspots, mais également au sein même des hotspots de biodiversité. Cependant, la mise en œuvre de plans de gestion de sauvegarde d’habitats ou d’espèces ne pourra se faire qu’en intensifiant les collaborations avec l’ensemble des acteurs impliqués. Plus généralement, la mise en œuvre de stratégies d’atténuation et d’adaptation efficaces aux changements climatiques ne pourra pas avoir lieu sans un soutien du grand public. / Global biodiversity is changing at an unprecedented rate due to loss of habitat, biological invasions, pollution, overexploitation. Furthermore, climate changes and their synergies with other threats will probably become the main drivers of biodiversity loss in the next century. Nowadays, the multiplicity of approaches and the resulting variability in projections make it difficult to get a clear picture of the future of biodiversity due to climate change. Yet, the majority of models indicate alarming consequences for biodiversity, with the worst- case scenarios leading to an increase of extinction rates. The aim of this thesis was to improve the knowledge about of the different consequences of climate change on biodiversity worldwide. To do that I mainly used modelisation and meta-analyses approaches. The first part of my work was to investigate the consequences of sea level rise for the ten insular biodiversity hotspot and their endemic species, during which I highlighted that between 6 and 19% of the islands would be entirely submerged. Then I studied the effects of climate and land use changes on biological invasions worldwide. The results showed that invasives species response to climate and land use changes depend on region, taxa and species considered. We also emphasized that some regions could lose a significant number of invasive alien species. Besides, we also found that hotspot that are mainly islands or group of islands are highly suitable for invasive species. Finally, in the last part, I quantified the exposure of biodiversity hotspots to the combined effects of climate change, land use change and biological invasions. This work highlighted the pressing need to consider different drivers of global change in conservation planning. In addition, we established some prioritization framework among the hotspot. Finally, conservation strategies to protect habitat and species under global changes, can only be achieved through closed collaboration with park managers. Overall, implementation of effective adaptation strategies to climate change can only succeed with public support.

Les espèces exotiques envahissantes végétales sont des plantes introduites et naturalisées hors de leur aire de répartition native et capables de maintenir et d’accroitre leur population. Certaines sont considérées comme transformatrices de par leur effet sur les écosystèmes : leur structure, leur fonctionnement ainsi que leur communauté végétale et animale. Ces transformations peuvent rendre certaines de ces espèces nuisibles de par leurs impacts écologiques et économiques important. Les travaux réalisés dans le cadre de cette thèse et présentés ici ont pour objectif d’approfondir les connaissances sur l’impact des invasions biologiques. La faune du sol, la végétation native et leur substrat ainsi que son fonctionnement ont été étudiés à différentes échelles spatiales. Deux espèces exotiques, envahissantes en Europe, ont été considérées comme modèles pour ces travaux : le robinier faux-acacia (Robinia pseudoacacia) et la renouée du japon (Reynoutria japonica). Premièrement, une méta-analyse globale a permis de démontrer l’effet positif des invasions biologiques végétales sur l’abondance de certains groupes de la faune du sol, notamment les consommateurs primaires, en fonction de la structure de l’habitat (ouvert ou fermé). Ensuite, une étude à large échelle sur le robinier faux-acacia a permis d’illustrer les différences qui peuvent exister dans la réponse des écosystèmes forestiers aux invasions le long d’un gradient latitudinal. Ce gradient, composé de quatre régions distinctes en Europe de l’Ouest présente des différences de climat et de végétation dominante, ces différences modifiant l’impact du robinier faux-acacia. Une étude approfondie sur le robinier faux-acacia en Normandie a permis de mieux comprendre l’effet du robinier faux-acacia sur les communautés animales et végétales ainsi que sur le fonctionnement des écosystèmes par comparaison avec deux essences natives dominantes. Finalement, une manipulation expérimentale en laboratoire a démontré l’impact des composés allélopathiques de la renouée du Japon sur une partie de la faune du sol. Cette étude a montré que certaines espèces exotiques envahissantes sont susceptibles d’influencer la faune, et les réseaux trophiques, du sol par leur métabolisme secondaire. Ces travaux illustrent l’intérêt, dans le contexte des invasions biologiques végétales, de l’étude simultanée des compartiments aériens et souterrains à différentes échelles spatiales. / Invasive alien plants are species introduced and naturalized outside of their native distribution range and which have the capacity to maintain and expand their population. Some of these species are considered to be ecosystem transformers by altering their structure, functioning as well as resident animal and plant communities. These induced alterations make some of these species undesirable through their ecological and economical impacts. The work presented in this thesis aimed at a better understanding of the impact of biological invasions by alien plants. The soil fauna, native vegetation and their substrate, as well as ecosystem functioning, were studied at different spatial scales. Two exotic alien species, invasive in Europe, were considered as biological models for this work: the black locust (Robinia pseudoacacia) and the Japanese knotweed (Reynoutria japonica). Firstly, a global meta-analysis demonstrated the positive impact that plant invasions can exert on the abundance of some groups within the soil fauna, notably primary consumers, within different types of habitats (open or closed). Then, a large-scale study on the black locust revealed the differences that can can occur in the response of forest ecosystems to invasions along a latitudinal gradient. Study sites along this gradient, distributed amog four distinct regions in western Europe, exhibit differences in climate and dominant native vegetation which can alter the impact of the black locust. A detailed study on black locust impact in Normandy demonstrated the impact of R. pseudoacacia on native plant and soil fauna communities, as well as some ecosystem functions, in comparison to two native tree species. Finally, a laboraty experiment demonstrated the impact that allelopathic compounds extracted from Japanese knotweed rhizomes can have on some organisms within the soil fauna. This study showed that some invasive alient plants can influence the soil fauna, and soil food webs, through their secondary metabolism. This thesis illustrates that simultaneous study of both aboveground and belowground ecosystem compartments at different spatial scales is of interest in the context of biological invasions.

La naïveté, définie comme l’absence de reconnaissance d’un nouvel « ennemi » ou de réponses appropriées en raison d’un manque d’expérience, à été longtemps considérée comme une des caractéristiques des espèces insulaires natives, et comme un des principaux facteurs de leur sensibilité face à l’arrivée de prédateurs ou compétiteurs introduits, particulièrement en milieux insulaires. A travers l’analyse des réponses comportementales aux odeurs d’espèces invasives, nous avons étudié la naïveté de différents modèles biologiques d’espèces natives (squamates, oiseaux, mammifères), au niveau insulaire (Nouvelle-Calédonie) et continental (Mali). Ces expérimentations ont concerné différents types d’interactions (prédateurs invasifs vs proies natives, compétiteurs ou agresseurs invasifs vs compétiteurs natifs), avec une gamme d’espèces introduites depuis plusieurs dizaines, centaines, voire milliers d’années, ou face à des espèces non-encore introduites mais avec une forte probabilité d’introduction. La naïveté apparait comme un facteur complexe, présent chez des espèces natives insulaires et continentales, dépendant du couple d’espèces introduite/native considéré, et pouvant être perdu sous l’influence de différents facteurs liés à l’histoire de l’invasion et à l’intensité de l’impact de l’espèce introduite. Etudier les conséquences de la perte de naïveté, permet de mieux comprendre le rôle de la naïveté dans les processus de persistance, de déclin ou d’extinction d’espèces natives suite à l’arrivée de nouveaux « ennemis », et d’éclairer les stratégies de gestion destinées à limiter l’impact des espèces introduites et l’érosion mondiale de la biodiversité. / The introduction of alien predators are causing rapid extinctions or declines in many native prey species, and these impacts are generally attributed to prey naïveté, i.e. their failure to recognize a novel enemy and respond appropriately, due to a lack of experience, particularly on island ecosystems. Here, we tested behavioral responses of different native species towards invasive species odors for several biological models of native species (squamates, mammals, birds), on island (New Caledonia) and mainland (Mali) ecosystems. We tested different types of species interactions (invasive predators vs native prey, invasive competitors/aggressors vs native competitors), for species introduced since decades, centuries, millennium, or not yet introduced but with a high risk of introduction. Naïveté is a complex phenomenon, that occurs both for mainland and island native species, that varies among species probably as a consequence of time since introduction and of the intensity of the threat posed by invasive species. Studying the consequences of the loss of naïveté improve our understanding of the factors underpinning the persistence, declines or extinctions of native species after alien species introduction and provide insights for developing accurate management strategies of invasive species impacts and limiting the current biodiversity loss.

Biological invasions are an important issue of global change and an increased understanding of invasion processes is of crucial importance for both conservation managers and international trade. In this thesis, I have studied the invasion of the brown seaweed Fucus evanescens, to investigate the fate and effect of a perennial, habitat-forming seaweed introduced to a coastal ecosystem. A long-term study of the spread of F. evanescens in Öresund (southern Sweden) showed that the species was able to expand its range quickly during the first 20 years after the introduction, but that the expansion has been slow during the subsequent 30 years. Both in Öresund and in Skagerrak, the species is largely restricted to sites where native fucoids are scarce. Laboratory experiments showed that the restricted spread of F. evanescens cannot be explained by the investigated abiotic factors (wave exposure and salinity), although salinity restricts the species from spreading into the Baltic Sea. Neither did I find evidence for that herbivores or epibiota provide biotic resistance to the invader. On the contrary, F. evanescens was less consumed by native herbivores, both compared to the native fucoids and to F. evanescens populations in its native range, and little overgrown by epiphytes. Instead, the restricted spread may be due to competition from native seaweeds, probably by pre-occupation of space, and the establishment has probably been facilitated by disturbance. The studies provided little support for a general enemy release in introduced seaweeds. The low herbivore consumption of F. evanescens in Sweden could not be explained by release from specialist herbivores. Instead, high levels of chemical anti-herbivore defence metabolites (phlorotannins) could explain the pattern of herbivore preference for different fucoids. Likewise, the low epibiotic colonisation of F. evanescens plants could be explained by high resistance to epibiotic survival. This shows that colonisation of invading seaweeds by native herbivores and epibionts depends on properties of the invading species. The large differences between fucoid species in their quality as food and habitat for epibionts and herbivores imply that invasions of such habitat-forming species may have a considerable effect on a number of other species in shallow coastal areas. However, since F. evanescens did not exclude other fucoids in its new range, its effect on the recipient biota is probably small.