Lors de cette étude, 47 nouveaux âges ont été obtenus par la technique Cassignol-Gillot, complétant à 128 âges le nombre de données disponibles sur l'île de Basse-Terre. La très bonne reproductibilité des âges obtenus dans cette étude, et la cohérence de ces derniers sur l'ensemble des massifs, appuie l'utilisation de la méthode K-Ar pour la datation des laves des Petites Antilles. Les données géochronologiques ont été associés à des analyses géochimiques et géomorphologiques dans le but de contraindre l'évolution spatio-temporelle du volcanisme de Basse-Terre, mais également d'apporter de nouvelles contraintes sur les taux de construction et d'érosion en contexte tropical.Le volcanisme récent de Basse-Terre, i.e. inférieur à 1 Ma, se concentre dans la moitié sud de l'île. Composée de trois massifs volcaniques (Piton de Bouillante, Sud Chaîne Axiale et le Complexe Volcanique de Grande-Découverte), son activité a débuté au nord-ouest par la mise en place du Piton de Bouillante entre 906 ± 13 et 712 ± 12 ka, avec un taux de construction de 0.7 ± 0.2 km3/kyr. Les nouvelles données obtenues lors de cette étude montrent qu'aucun effondrement de flanc majeur n'a affecté cet édifice. L'activité volcanique a ensuite rapidement migrée vers le sud-est pour former entre 681 ± 12 et 509 ± 10 ka les volcans de Moustique, Matéliane, Capesterre et Icaque, qui constitue le massif Sud Chaine Axiale. La contemporanéité des âges obtenus sur l'ensemble des édifices, et le taux de construction calculé à 0.5 km 3/kyr, appuient la mise en place du sud de la Chaine Axiale par un unique massif volcanique, contredisant les hypothèses d'effondrement de flanc précédemment proposées. L'homogénéité géochimique observée sur l'ensemble du massif supporte l'hypothèse d'un seul édifice. Entre 500 et 450 ka, le flanc ouest du massif Sud Chaîne Axial a été affecté par un slump actuellement matérialisé par le volcan d'Icaque. La dépression formée a permis la mise en place du volcan du Sans-Toucher entre 451 ± 13 et 412 ± 8 ka. Entre 400 et 200 ka, très peu d'activité effusive a pu être mise en évidence. Depuis 200 ka, l'activité volcanique se concentre dans le sud de l'île, avec la mise en place du Complexe Volcanique de la Grande-Découverte, par une succession de phases de construction et de destruction. La dernière activité volcanique a permis la construction du dôme actuel de La Soufrière. Les investigations géomorphologiques nous ont également permis de contraindre les taux d'érosion ayant affecté l'île de Basse-Terre. Ainsi, le Piton de Bouillant subit une érosion de 1 250 ± 700 t/km'/an depuis 700 ka. Pour les volcans du Sans-Toucher, et des Monts-Caraïbes, nous avons obtenus un taux d'érosion similaire, respectivement de 940 ± 380 et 610± 550 t/km2/an. Malgré une localisation et une morphologie initiale différentes, la similarité des taux d'érosion obtenus pour les volcans de Basse-Terre met en évidence l'absence d'un effet barrière sur l'érosion à long terme, pourtant majeur à plus courte échelle de temps. / In this study, forty-seven new ages have been obtained by the Cassignol-Gillot technique, increasing to 128, the geochronological database available for the Basse-Terre Island. The very good reproducibility of the ages obtained in this study, added to a strickly consistency observed between the volcanic edifices, support the use of the K-Ar method in the dating of the Lesser Antilles lavas. This new geochronological dataset has been combined with geochemical and geomorphological analyses in order to constrain the volcanic history of Southern Basse-Terre Island as well as to compute construction and erosion rates.Southern part of Basse-Terre hosts the recent volcanic activity since the last 1 Myr. Composed by three volcanic massifs (Piton de Bouillante, Southern Axial Chain and the Grande-Découverte Volcanic Complex), its activity has begun in the northwest part by the construction of the Piton de Bouillante between 906 ± 13 and 712 ± 12 ka, with a construction rate of 0.7 ± 0.2 km3/kyr. Our new data show that no major flank collapse have affected this volcano. Then, volcanic activity has migrated to the southeast, constructing between 681 ± 12 and 509 ± 10 ka the Southern Axial Chain massif, composed by Moustique, Matéliane, Capesterre and Icaque volcanoes. The contemporaneity of the ages for the whole massif together with the construction rate computed at 0.5 km3/kyr suggest the formation of the southern Axial Chain by a unique volcanic edifice, which did not experienced major flank collapses as previously proposed. The geochemical homogeneity observed throughout the massif supports this single volcano hypothesis. Between 500 and 450 ka, a slump has affected the western part of the Southern Axial Chain and forming the Icaque volcano. The resulting depression has allowed the construction of the Sans-Toucher volcano from 451 ± 13 to 412 ± 8 ka. After the construction of the Sans-Toucher volcano, only few evidences for an effusive activity occurring between 400 and 200 ka can be found. Since 200 ka, volcanic activity is present in the southern part with the construction of the Grande-Découverte Volcanic complex (GDVC), alternating constructive and destructive phase. The last volcanic activity formed the 1530 AD La Soufrière dome. Geomorphological investigations have also allowed us to constrain the erosion rates having affected Basse-Terre Island. Thus, Piton de Bouillante volcano have suffered of an erosion rate of 1,250 ± 700 t/km²/yr since 700 ka. We have obtained for the Sans-Toucher and Monts-Caraïbes volcanoes similar rates of 940 ± 380 and 610 ± 550 t/km²/yr, respectively. Despite a different location and different initial morphology, the similarity erosion rates observed for each massif suggest that the barrier effect does not significantly affect the long-term erosion budget while it plays a major role at much shorter time-scale.

The Lyttelton Volcanic Complex, north-western Banks Peninsula, New Zealand, is comprised of five overlapping volcanic cones. Two magma systems are postulated to have fed Banks Peninsula’s basaltic intraplate volcanism, with simultaneous volcanism occurring in both the north-western and south-eastern regions of Banks Peninsula, to form Lyttelton and Akaroa Volcanic Complexes respectively. The elongate form of Banks Peninsula is postulated to relate to the upward constraining of magmatism in a north-west / south-east fault bounded zone. The Lyttelton Volcanic Complex resulted from the development of a pull-apart basin, with a number of releasing bend faults, controlling the location of eruptive sites. Cone structure further influenced the pathway magma propagated, with new eruptive sites developing on the un-buttressed flanks, resulting in the eruption and formation of a new cone, or as further cone growth recorded as an eruptive package. Each cone formed through constructional or eruptive phases, termed an eruptive package. Eruptive packages commonly terminate with a rubbly a’a to blocky lava flow, identified through stratigraphic relationships, lava flow trends and flow types, a related dyking regime, and radial erosional features (i.e. ridges and valleys). Within the overall evolving geochemical trend of the Lyttelton Volcanic Complex, are cyclic eruptive phases, intrinsically linked to eruptive packages. Within an eruptive package, crystal content fluctuates, but there is a common trend of increasing feldspar content, with peak levels corresponding to a blocky lava flow horizon, indicating the role of increased crystalinity and lava flow rheology. Cyclic eruptive phases relate to discreet magma batches within the higher levels of the edifice, with crystal content increasing as each magma batch evolves, limiting the ability of the volcanic system, over time, to erupt. Evolving magmas resulted in explosive eruptions following effusive eruptives, and / or result in the intrusion of hypabyssal features such as dykes and domes, of more evolved compositions (i.e. trachyte). Each eruptive package hosts a radial dyke swarm, reflecting the stress state of a shallow level magma chamber or a newly developed stress field due to gravitational relaxation in the newly constructed edifice, at the time of emplacement. Two distinct erosional structures are modelled; radial valleys and cone-controlled valleys. Radial valleys reflect radial erosion about a cone’s summit, while cone-controlled valleys are regions where eruptive packages and cones from different centres meet, allowing stream development. Interbedded epiclastic deposits within the Lyttelton lava flow sequences indicate volcanic degradation during volcanic activity. As degradation of the volcanic complex progressed, summit regions coalesced, later becoming unidirectional breached, increasing the area of the drainage basin and thus the potential to erode and transport extensive amounts of material away, ultimately forming Lyttelton Harbour, Gebbies Pass, and the infilled Mt Herbert region. Epiclastic deposits on the south-eastern side of Lyttelton Harbour indicate a paleo-valley system (paleo-Lyttelton Harbour) existed prior to 8.1 Ma, while the morphology of the Lyttelton Volcanic Complex directed the eruptive sites, style and resultant morphology of the proceeding volcanic groups.