The origin and emplacement of numerous ophiolite complexes (terranes) along the Pacific coast of Central America is still a matter of debate. The volcanology, petrology and geochemistry of three terranes (Nicoya, Herradura and Quepos) were investigated. The Nicoya complex (age uncertain) consists of ophiolite layers 1-3 (sediments, extrusive series and plutonics), separated almost exclusively by tectonic contacts. Herradura (age unknown) consists primarily of aphyric pillow lavas and massive basalts. Field and petrographic observations, as well as studies of volatiles, suggest that both of these complexes were erupted in moderate to deep water, possibly forming part of an oceanic structure similar to the Cocos Ridge. Quepos (66 Ma) consists of olivine (ol) and clinopyroxene (cpx) phyric pillow basalts, overlain by a volcaniclastic series. The volcaniclastic sediments provide evidence for steep topography and shallow water to subaerial eruptions, typical of an emergent ocean island.
Major element and trace element analyses (by XRF, EMS and ICP-MS) of glasses from pillow rinds and whole rock samples, picked under a binocular microscope, show that only the most mobile elements (Na, K, Rb, Cs, Ba and Sr) have been affected by seawater alteration. From SiO2 and immobile incompatible elements, it is clear that rocks from all three areas are tholeiitic. In the Nicoya tholeiites, MgO correlates positively with Al2O3, CaO, Cr and Ni but negatively with FeO* (total iron), TiO2, V, Zn, Y and Zr, consistent with fractionation of ol + plag ± Cr spinel at low pressures (<15 km). In the Quepos tholeiites, CaO, Cr and Ni decrease and Al2O3 increases as MgO decreases indicating fractionation of ol + cpx ± Cr-spinel at higher pressures (>15 km). Lower SiO2 and higher FeO* in Quepos tholeiites than in Nicoya/Herradura tholeiites with MgO = 9 wt. % are consistent with greater depths of melt generation for Quepos parental magmas. The heavy rare earth elements (REE), however, indicate that melts in all three complexes formed within the spinel stability field. Deeper fractionation and melt generation suggest that the Quepos ocean island formed on thicker oceanic lithosphere than the Nicoya/Herradura oceanic ridge, which may have formed near a mid ocean ridge.
Nicoya/Herradura tholeiites have flat chondrite-normalized REE patterns characteristic of enriched mid ocean ridge basalt (MORB) or hotspots lying on or near mid ocean ridges. In contrast, Quepos tholeiites are enriched in the light REE by up to a factor of seven. These large differences in incompatible element concentration in primitive tholeiites cannot alone be explained by differences in degree of partial melting and thus indicate derivation from a more enriched (ocean island-type) mantle source. Surprisingly, Pb and Nd Isotope ratios of tholeiites from all areas are very similar (e.g. 144Nd/143Nd = 0.5129-0.5130 and 206Pb/204Pb = 19.2-19.6), consistent with derivation from a common source. Pb isotopes are more radiogenic than observed in Pacific and Indian MORB, and Nd isotopes are generally less radiogenic than in Pacific and Atlantic MORB. The isotope data, however, fall almost completely within the published fields for the Galápagos Islands. In addition, the range in major and immobile trace elements in the studied tholeiites is very close to that for the Galapagos Islands and seamounts.
Therefore, we propose that at least some of the Central American ophiolites may represent accreted oceanic ridges, seamounts and islands, generated through plume/ridge interactions as illustrated by the Galapagos islands. Presently the Cocos Ridge and seamounts, representing part of the Galapagos hotspot track, are being subducted or accreted along the Pacific margin of Central America. This poses an intriguing question. Could some of the Central American volcanic terranes reflect an earlier history of the Galapagos hotspot or of a possible precursor?