En raison de sa remarquable diversité végétale, de son taux d’endémisme, parmi les plus élevés au niveau mondial (79%) et de son extrême vulnérabilité, la Nouvelle-Calédonie est un des premiers hotspots de la biodiversité mondiale. L’une des remarquables originalités de la flore Calédonienne repose sur la présence de nombreux taxa reconnus, en raison de leurs positions phylogénétiques, comme appartenant aux lignées les plus anciennes des plantes à fleurs. Ces lignées d’Angiospermes « reliques » ont une valeur scientifique et patrimoniale importante, puisqu’elles sont de véritables fenêtres sur le passé. A travers l’étude des traits foliaires, de l’anatomie du bois (e.g. type d’éléments conducteurs, perforations, diamètre et longueur des vaisseaux), de l’architecture (e.g. sympoldialité vs monopodialité, phénomène de réitération, rythmicité de croissance) et de la biomécanique des axes, nous chercherons à caractériser les formes et fonctions de ces taxons. Ce projet de thèse vise à examiner les déterminants structurels et fonctionnels de la répartition spatiale actuelle des Angiospermes basales (grade ANITA + Magnoliidae) en Nouvelle-Calédonie. Dans cette perspective nous chercherons à comprendre quels sont les facteurs qui ont conduit au confinement de certains taxons à des habitats restreints ou au contraire à leur dispersion dans des milieux contrastés. Un premier volet du projet en cours, a pour objectif de caractériser la dynamique de croissance d’Amborella trichopoda et sa plasticité architecturale sous différents régimes lumineux. Certains traits écologiques de ces espèces d’Angiospermes de divergence précoce, s’ils n’ont pas contribué à la radiation des plantes à fleurs actuelles, ont pu permettre son enracinement écologique au sein de la flore du Mésozoïque et fournir un répertoire développemental pour l’explosion subséquente de leur diversité. L’identification et l’étude de ces caractères sont donc déterminantes pour la compréhension de l’évolution structurelle et fonctionnelle des plantes à fleurs. / New Caledonia (NC) is one of the main biodiversity hotspots (Myers 1988), this is because of its remarkable plant diversity, its endemism rates, among the highest in the world (79%), and because of the vulnerability of its flora. One of the main originalities of New Caledonia flora is based on the presence of a large number of taxa recognized, due to their phylogenetical positions, as the most ancient extant representatives of angiosperms. For a long time, New Caledonia has been considered as an early upset fragment of the Gondwana (Pelletier 2006) that suffered an interrupted history of isolation which conferred the evolutionary particularities that we observe today. However, recent evidences show that NC has derived from the Australian land during the late Mesozoic (~80 Mya), the island was then submerged during the first half of the Cenozoic (Pelletier 2006) and a reemergence of the island seems to have occurred ~37 Mya (Cluzel et al. 1998). After the reappearance of the island above the sea level, several events of recolonisation occurred and they wrought the biodiversity that we observe nowadays (Pillon 2012). NC presents humid forests which are unique relics; under the influence of climate changes, these forests have virtually disappeared from other regions of the globe (Morat et al. 1986). The lineages of “relictual” angiosperms, mainly subservient to these humid forests, have a great scientific and patrimony value, as they can be considered as genuine windows on the past. These taxa are susceptible to contain primitive characters which have either disappeared in most of the existing flowering plants, or that are still shared by a narrow number of them. The identification and the study of these characters are therefore determinants for the comprehension of angiosperms evolution. Some ecological features of these panchronic species, may have either contributed to the huge radiation of extant angiosperms, or they may have contributed to the ecological settling of angiosperms within the Mesozoic flora, providing them with a developmental repertoire for the subsequent explosion of their diversity. This PhD project aims to study the ecological, anatomical and functional diversity of basal angiosperms and it seeks to analyze the evolutionary patterns of these structural and functional features. We will consider here as “basal” angiosperms a great group of flowering plants that has diverged before the monocot and eudicot node. This group is conformed by the ANITA grade, formed by Amborella (a single species endemic to NC), Nymphaeales (waterlilies and other herbaceous aquatic plants) and Austrobaileyales (aromatic woody plants). The Magnoliid subclass, a clade of flowering of early divergence, which contains plants considered as paleodicots by Cronquist (1988), will be also included in the analysis of the « basal » taxa. More recently, the Magnoliids have been redefined as a clade comprising Chloranthales, Canellales, Laurales, Magnoliales, et Piperales (APG III, 2009). In a second part of the project, a fieldtrip to Mexico will be held in order to include speces belonging to the Chloranthaceae and Schisandraceae, as well as Cabombaceae et Nymphaeaceae, by this means, we will incorporate species belonging to all the orders of the “basal” angiosperms, reinforcing the comparative analysis. This research work will lean on the recent publications of the phylogenetic relations within basal angiosperms...