A Complex History for the Caribbean Oceanic Plateau: Petrology, Geochemistry and Geochronology of the Beata Ridge

Abstract

The Caribbean province is an area where the thickness of oceanic crust is between 8 and more than 20 kin, well in excess that of 'normal' oceanic crust. This has led to the conclusion, supported by geochemical studies and tectonic reconstructions, that much of the Caribbean plate formed as an oceanic plateau in the eastern Pacific. Most authors accept that the Caribbean plateau formed through melting of the Galapagos plume head. Through eastward movement of the Farallon plate, part of this plateau was entrained between the two American plates while the southern part collided with northern South America. At least two major phases have been recognized in the plateau building. The first, volumetrically most important event was around 90-88 Ma, and the second around 76 Ma. New petrological, geochemical and geochronological data from the Beata topographic ridge, south of Hispaniola, highlight the complexi ty of the Caribbean plateau construction and may require a revision of this model. The Beata ridge is an area of particularly thick crust (up to 20 km) between the Dominican and Haitian sub-basins. The ridge is 2000 to 5000 m deep and is bounded to the west by a normal fault. In this study we focus on the magmatic samples recovered by submersible during the Nautica-Beata cruise. The ridge is mainly composed by gabbros and dolerites, with minor diorite and rare pillow basalt. The gabbros and dolerites contain euhedral plagioclase (An54_62) and clinopyroxene (Wo37_4o En44_47 Fs13 19) grains with a variable but minor amount of olivine and/or orthopyroxene relicts, ilmenite and acicular apatite. The presence of orthopyroxene and the early crystallisation of plagioclase indicate a tholeiitic affinity. Textures that vary significantly, from coarse-grained gabbros to very fine-grained dolerites, reflect differences in cooling rates and suggest a subsurface, hypabyssal environment. The rare basaltic flows contain phenocrysts of weathered olivine, plagioclase (Anso-61) and clinopyroxene (Wo44-47 En40-44 Fs12-16) in a matrix of plagioclase microlites (An5o-55), clinopyroxene microcrysts (Wo45-46 En37-4o Fs15-16) and a glassy, vesicular mesostasis. The early crystallisation of olivine, an absence of orthopyroxene and the late crystallisation of plagioclase suggest an alkaline affinity. The major element compositions of gabbros and dolerites plot on simple trends that correspond to fractional crystallisation of olivine, clinopyroxene and plagioclase (MgO 5-11%; A1203 12-17%; CaO 8-15%). Trace-element ratios are close to chondritic [(Nb/Zr)N 0.85-1.1 ] and rare-earth-element patterns are almost flat [(La/Yb)N 0.63-1.02]. The source was isotopically depleted [~Na +7.4 to +9.5]. The basalts have higher trace-element ratios and enriched rare earth element patterns [(Nb/Zr)N 3.45; (La/Yb)N 28-30]. Their source was less depleted than that of the gabbros and dolerites [eNa +5]. Several samples were dated by the 39Ar/4~ incremental heating method, either on whole rocks or separated plagioclases. Six samples have ages between 84 to 70 Ma and these correspond with previous dates within the province. But others are surprisingly young, with ages at 62, 56 and 50 Ma. There is no apparent order to the ages: all are found in different rock types and from one end of the ridge to the other. Within the younger age group, consistent results were obtained for both whole rocks and plagioclases. In contrast, within the older age group, whole-rocks and plagioclase gave conflicting ages. For this reason, we consider that the young ages are not due to thermal resetting but probably represent a true magmatic event. In contrast, the ages for older