Journal of Petrology Advance Access published online on October 7, 2008
Journal of Petrology, doi:10.1093/petrology/egn046
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Age and Geochemistry of the Central American Forearc Basement (DSDP Leg 67 and 84): Insights into Mesozoic Arc Volcanism and Seamount Accretion on the Fringe of the Caribbean LIP
1Dynamics of the Ocean Floor, Ifm-Geomar Leibniz-Institut Für Meereswissenschaften, Wischhofstr. 1–3, d-24148 Kiel, Germany
2Fachbereich Geowissenschaften, University of Bremen, Postfach 33 04 40, D-28334 Bremen, Germany
Received February 5, 2008; Revised typescript accepted September 2, 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONREGIONAL GEOLOGY AND PREVIOUS...SAMPLING AND PETROGRAPHYANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
The igneous forearc basement along the Pacific coast of northern Central America (between southern Mexico and Costa Rica) comprises a highly tectonized accretionary assemblage of igneous and ultramafic rocks. Volcanic and gabbroic rocks with primitive arc geochemical signatures formed between
100 and 
180 Ma and are interpreted to have originated by arc magmatism resulting from subduction of the Pacific–Farallon plate. Geochemically enriched ocean island basalt (OIB)-like units are interpreted as accreted seamounts and islands of a hotspot track, which was active between 220 and 100 Ma and originated from a hotspot located in the central Pacific. Based on their combined Pb, Nd and Hf isotopic compositions an affiliation of these rocks with the Caribbean Large Igneous Province or the present-day Galápagos hotspot appears unlikely. Rocks of similar age and geochemistry are exposed on the Santa Elena Peninsula in Costa Rica, suggesting that a similar forearc basement is accreted to the continental Chortis Block from southern Mexico to Costa Rica. KEY WORDS: Central America; Chortis Block; DSDP; forearc; Pacific margin
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND PREVIOUS...SAMPLING AND PETROGRAPHYANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
In this study we investigate the 40Ar/39Ar age and comprehensive geochemistry (major, trace element and Sr–Nd–Pb–Hf isotopes) of igneous rocks drilled during Deep Sea Drilling Project (DSDP) Legs 67 and 84 into the forearc basement of Guatemala's Chortis Block. The results provide important constraints on the nature, composition and long-term evolution of the forearc. In the light of accumulating evidence that the Central American margin is an erosional rather than an accretional margin (e.g. Meschede et al., 1999
; Ranero & von Huene, 2000; Vannucchi et al., 2004), studies of the eroding forearc can provide crucial information on the composition of the subduction input, which can directly affect the composition of the mantle wedge and arc magmatism. In addition, the age and composition of the forearc rocks help to constrain the geodynamic evolution of the active margin and also reveal information about previously existing (now subducted) ocean crust and in particular intraplate volcanism on that crust and/or extinct arc volcanism that has accreted to the margin.
Another important question that this study addresses is if the composition of recovered geochemically enriched ocean island basalt (OIB)-like forearc rocks can be linked to the nearby Galápagos hotspot or the Caribbean Large Igneous Province (CLIP). This province consists of the Caribbean oceanic plateau and associated igneous (‘ophiolitic’) complexes at the edges of the Caribbean plate, mainly along the Pacific coast of Central America and western Colombia (Fig. 1a). Based on 40Ar/39Ar age determinations, similar Sr, Nd, Pb, and Hf isotope ratios, and plate tectonic reconstructions, the CLIP was interpreted to have formed from the Galápagos starting plume head at c. 83–95 Ma (e.g. Duncan & Hargraves, 1984; Kerr et al., 1996; Hauff et al., 1997, 2000a, 2000b; Sinton et al., 1997; 1998; Thompson et al., 2003; Hoernle & Hauff, 2007) followed by accretion of portions of the Galápagos paleo-hotspot track to its margins (Hoernle et al., 2002; Geldmacher et al., 2003). Based on seismic reflection studies, Mauffret & Leroy (1997) argued that the CLIP may be formed from up to three distinct volcanic plateaux (with ages ranging from 76 to 113 Ma). Recently, samples up to 140 Myr old with CLIP-like isotope compositions have been reported from the Nicoya Peninsula, placing the classical short-term plume head model further into question and suggesting instead that the CLIP represents rather an amalgamation of several smaller oceanic plateaux and other intraplate structures formed over long periods of time, probably through multiple pulses of the Galápagos hotspot (Hoernle et al., 2004; Hoernle & Hauff, 2007). Identification of similar or even older accreted material in the Guatemalan forearc basement that can be associated with the Galápagos hotspot would therefore have great importance for reconstructing the early history of the Galápagos hotspot (e.g. if it moved southward, similar to the model suggested for the Hawaiian hotspot in its earlier history). Alternative models have been proposed suggesting that the CLIP formed between the Americas in an intra-Caribbean setting without any connection to the Galápagos hotspot (e.g. Frisch et al., 1992; Meschede, 1998; Meschede & Frisch, 1998; Pindell et al., 2006). The discovery of CLIP fragments accreted to the Pacific side of the Chortis Block would provide a powerful argument against these models.
|
|
REGIONAL GEOLOGY AND PREVIOUS DATA |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND PREVIOUS...SAMPLING AND PETROGRAPHYANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Geological overview
The Caribbean plate is composed of several genetically different terranes assembled as a result of various collisions of volcanic arcs, continental blocks and the Caribbean oceanic plateau. The western edge of the Caribbean plate is formed by the Chorotega and Chortis blocks (Fig. 1a). The Chorotega Block represents a subaerially exposed part of the thickened oceanic Caribbean plateau and is largely covered by Cenozoic arc lavas. Its northern boundary to the mainly continental Chortis Block is not well constrained, but is believed to run south of the Nicoya Peninsula in Costa Rica to the Hess Escarpment (e.g. Meschede & Frisch, 1998
; Hoernle et al., 2004; Hoernle & Hauff, 2007). The northern border of the Chortis Block, the Motagua–Polóchic transform fault, separates the Caribbean from the North American plate (Maya Block). This fault marks an 110 Myr old convergence zone at which a proto-Caribbean ocean was southwestwardly subducted under the Chortis Block until its collision with the Maya Block (at present the Yucatan Peninsula) at 75 Ma. This suture zone, which hosts several small accreted Cretaceous ophiolites composed of oceanic crust, sub-arc mantle and arc volcanic rocks, was reactivated in the Late Cretaceous as a sinistral strike-slip fault (e.g. Donnelly et al., 1990; Beccaluva et al., 1995). Thereafter, the Chortis Block has traveled >1000 km eastward along this fault from its original position in SW Mexico to its present location (e.g. Pindell & Barrett, 1990; Elming & Rasmussen, 1997), resulting in the opening of the Cayman Trough (Fig. 1a). Recent studies indicate that the southeastern part of the Chortis Block is composed of deformed rocks of oceanic crust and island arc affinity (Siuna Terrane; Rogers et al., 2007) that were accreted to the southern margin of the Chortis Block in the late Cretaceous (Venable, 1994).
Many subaerially exposed CLIP complexes are located along the Pacific coast of South and Central America, extending from Colombia to the Nicoya Peninsula in northern Costa Rica (Fig. 1a). Northward, outboard the Chortis Block, the continental shelf noticeably broadens and forms a wide, submarine forearc basin until the shelf becomes narrow again off the coast of southern Mexico. Early seismic studies implied a dense basement of high acoustic velocity below the forearc basin sediments, suggesting igneous oceanic crust or an ophiolitic complex similar to the igneous CLIP terranes that are subaerially exposed further south (Ibrahim et al., 1979).
During DSDP Leg 67 and Leg 84 the trench and the broad continental shelf were drilled in front of Guatemala to evaluate the style of convergence (accretionary or erosive margin) and the nature of the dense forearc basement. Basement was reached at one site during Leg 67 (Site 494) and at four sites during Leg 84 (Sites 567, 566, 569, 570) with Site 567 at almost the same location as Site 494 (110 m apart) but with increased basement penetration (e.g. Aubouin & von Huene, 1985). Recovered basement rocks consist of ultramafic and mafic igneous rocks including basalts, dolerites, gabbros (Sites 494, 567), amphibolites (569), and serpentinized peridotites and harzburgites (566, 567, 570). All drill sites lie along a narrow transect perpendicular to the slope (Fig. 1b). Recent geophysical data from the lower Nicaraguan margin wedge suggest the presence of a similar dense basement reaching from the front of the margin, some 80 km to the middle of the shelf (Walther et al., 2000). The measured seismic velocities (3·5–6·0 km/s) and calculated densities (2·6–2·9 g/cm3) are too high to reflect accreted sediments or upper continental crust, and imply that the entire shelf area between Costa Rica and southern Mexico is underlain by igneous basement that might be similar to the subaerially exposed igneous complex at Santa Elena and Nicoya (Ye et al., 1996; Walther et al., 2000). Attributing this entire 1000 km long basement to the CLIP would not only greatly expand the province but also raise questions about the age of the igneous basement rocks and their possible affinity to the Galápagos mantle plume (both having important implications for the origin of the CLIP and large igneous provinces in general).
Major element and a few trace element analyses of the moderately to heavily altered drilled basement samples and four 40K/39Ar age determinations, conducted by the Leg 67 and 84 scientific parties, however, are not sufficient to answer these questions. Among the few samples that have been analyzed for major and trace element concentrations, two different lithologies have been described: (1) quartz-normative rocks with low TiO2 and depleted light rare earth element (LREE) signatures characteristic of tholeiites or calc-alkaline arc volcanic rocks; (2) nepheline-normative rocks with high TiO2 and LREE, Ba, and Sr enrichment characteristic of alkali basalts (Maury et al., 1982; Bellon et al., 1985; Bourgois et al., 1985) (Fig. 2). Rocks of both lithologies have been recovered from a single hole (Hole 567A) from which four samples were dated with the 40K/39Ar method. Whereas one basalt with depleted tholeiitic affinity gave a K/Ar age of 78·7 ± 3·9 Ma, three dolerites with alkaline affinities, from deeper sections of the hole, yielded a wide range of ages (91 ± 4·5 Ma, 132 ± 6·6 Ma and 169 ± 8·4 Ma) that was interpreted to reflect the effects of alteration (Bellon et al., 1985). As the sample with the oldest age shows the least secondary alteration and lowest loss on ignition (LOI), it was believed that the 169 Ma age is nearest to the true crystallization age. Because of the surprising discovery of Jurassic alkaline rocks, comparisons of the drilled forearc basement have been made with the nearby, subaerially exposed Santa Elena igneous terrane (e.g. Aubouin & von Huene, 1985; Bellon et al., 1985) which is characterized by the occurrence of harzburgites, depleted arc-related tholeiites and enriched Jurassic alkaline rocks (e.g. Hauff et al., 2000a).
|
The Santa Elena complex and the western Chorotega–Chortis Block boundary
Three igneous complexes can be found at the assumed Chorotega–Chortis Block boundary in NW Costa Rica: Nicoya, Tortugal and Santa Elena, with Nicoya being the largest and best studied of them (e.g. Dengo, 1962
; Kuijpers, 1980; Hauff et al., 1997; Sinton et al., 1997). The complex is mainly composed of olivine tholeiitic lavas, minor intrusive rocks (dikes, gabbros) and subordinate hyaloclastic rocks and pillow breccias; it ranges from 70 to 140 Ma in age (e.g. Hauff et al., 1997; Hoernle et al., 2004).
At Santa Elena, a southwestward verging, predominantly ultramafic nappe overthrusts a basalt–radiolarite assemblage (e.g. Azéma et al., 1985a; Hauff et al., 2000a; Denyer et al., 2006) (Fig. 2a). The nappe, consisting of peridotitic rocks and gabbroic intrusions, both cut by doleritic dikes, was emplaced after the Cenomanian (the age of the youngest underlying sediments, DeWever et al., 1985) but before the late Campanian (age of rudist reefs growing on the top of the exhumed nappe; Schmidt-Effing, 1980), corresponding to an age range between 75 and 93 Ma. This age range is supported by a radiometric K/Ar age of 88 ± 4·5 Ma from a secondary amphibole sampled from a recrystallized doleritic dike (Bellon & Touron, 1978) and most probably dates amphibolite metamorphism along shear zones related to the nappe emplacement.
The subdivision of the underlying basalt–radiolarite assemblage and its genetic relationship to the ultramafic and mafic nappe is controversial. Based on trace element and isotope data, Hauff et al. (2000a) have defined four lithological units at Santa Elena. The first unit (also named the ‘Santa Rosa Accretionary Complex’; Baumgartner & Denyer, 2006; Denyer et al., 2006) comprises alkaline pillow lavas and dikes discordantly associated with radiolarian cherts, which range in age biostratigraphically from Cenomanian to late Pliensbachian (93–190 Ma). The alkaline rocks have enriched incompatible trace element and isotope compositions similar to intraplate OIB but differ from other CLIP rocks by their distinctly lower Nd isotope ratios (Hauff et al., 2000a). The second structural unit is made up of columnar and pillow basalts, for which a 40Ar/39Ar total fusion age of 109 ± 1 Ma was obtained. An intrusive complex of layered gabbros forms the third unit, from which plagioclase separates yielded a total fusion 40Ar/39Ar age of 124 ± 2 Ma (Hauff et al., 2000a). The fourth unit, according to this subdivision, comprises the ultramafic nappe. In contrast to the alkaline Unit I, basalts, gabbros and dikes from Units II, III, and IV range from tholeiitic basalts to basaltic andesites and trachy-andesites, and show depleted trace element and isotopic signatures consistent with their generation from arc volcanoes or from depleted mantle sources in a subduction zone environment (e.g. Hauff et al. 2000a; Hoernle & Hauff, 2007).
The much smaller igneous complex of Tortugal, located about 100 km SE of Santa Elena, consists of a picritic basal complex, overlain by alkali basaltic and tholeiitic flows. As no clear contacts between these different units are exposed, the stratigraphy and structural relationships are unclear (Alvarado et al., 1997). Plagiclase crystals from the basal picrites, which appear to be subvolcanic cumulates of the alkali basaltic dike magmas (Hauff et al., 2000a), yielded an 40Ar/39Ar plateau age of 89·7 ± 1·7 Ma (Alvarado et al., 1997). The moderately depleted geochemical characteristics of the tholeiitic lava sequence indicate derivation from the same source as the other CLIP tholeiites, such as those from Nicoya and Herradura. Incompatible trace element patterns and isotope data for the alkaline Tortugal rocks, however, are similar to those of the alkaline rocks from Santa Elena, except for their less enriched (slightly more radiogenic) Nd isotope ratios (Hauff et al., 2000a). The enriched isotopic compositions of the Tortugal and Santa Elena alkaline rocks are unique in Central America and extremely rare within the CLIP. Similar enriched compositions have been reported from only one Caribbean DSDP drill site (Site 151, Sinton et al., 1998), from a single location in the Central Cordillera of Colombia (El Encenillo, Kerr et al., 2002) and at Gorgona Island (Kerr, 2005). This would argue for the existence of a rare, geochemically enriched component in the early CLIP. On the other hand, Gorgona and some terranes in Colombia and Ecuador might have been formed by a different hotspot unrelated to the CLIP (Kerr et al., 2002; Kerr & Tarney, 2005).
The subduction-related upper units of Santa Elena were interpreted to be associated with an old subduction system or island arc in front of the Chortis Block (Hauff et al., 2000a; Baumgartner & Denyer, 2006), consistent with nearby Albian pyroclastic deposits (Calvo, 2003, and references therein; Baumgartner & Denyer, 2006). This implies that the boundary between the Chortis Block (southernmost part of the Mesozoic Chortis subduction zone) and the Chorotega Block (CLIP) runs south of Santa Elena; this is also supported by a distinct change in the magnetic field across the shelf between the Santa Elena and Nicoya peninsulas (Berhorst, 2006). The geochemically enriched alkaline basalts of Tortugal (similar to Santa Elena Unit I) and possibly the older parts of Nicoya (>100 Myr old rocks?) can also be interpreted to have accreted to the southern end of the Chortis subduction system (Hoernle et al., 2004; Hoernle & Hauff, 2007).
In summary, the common occurrence of two geochemically distinct rock types (depleted, subduction-related and enriched OIB-type volcanic rocks) reported from the Guatemalan forearc basement drill cores and also subaerially exposed along the southern boundary of the Chortis Block (Santa Elena, Tortugal) suggests that the entire forearc of the Central American subduction zone records a similar history from southern Mexico to northern Costa Rica.
We have re-sampled the DSDP Leg 67 and 84 drill cores from the Guatemalan forearc basement and present a comprehensive state-of-the-art geochemical dataset (new major and trace element, Sr, Nd, Pb, Hf isotope data), as well as 40Ar/39Ar laser step heating age determinations of the recovered rocks, to determine the structural origin of the igneous forearc basement, its geodynamic evolution and possible relationship to the CLIP and Central American geology.
|
SAMPLING AND PETROGRAPHY |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND PREVIOUS...SAMPLING AND PETROGRAPHYANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Cores from all DSDP sites that reached the igneous forearc basement have been re-sampled for this study. Rocks selected for geochemical investigation and age determinations are from Hole 494A (Leg 67) and Holes 566C, 569A, 567A (Leg 84). Two mid-ocean ridge basalt (MORB) samples of the incoming Cocos plate from Holes 499B and 500B (Leg 67) have also been analyzed for comparison (Fig. 1b). Additionally, Hf isotope ratios from subaerially exposed Santa Elena rocks have been determined on the same rock powders for which major and trace element and Sr, Nd, Pb isotope data have been reported previously (Hauff et al., 2000a
). We also report new trace element and Sr, Nd, Pb, Hf data for two additional samples from the alkaline Santa Elena sequence (Unit I or Santa Rosa Accretionary Complex) for which only major element data for the same rock powder had been reported previously (Hauff et al., 2000a).
All the different lithological units recovered during Legs 67 and 84 have been described petrologically in great detail in the DSDP Initial Reports (e.g. Maury et al., 1982; Bellon et al., 1985; Bourgois et al., 1985; Shipboard Scientific Party, 1985); these are available online (http://www.deepseadrilling.org/i_reports.htm) and therefore the results will be summarized only briefly below. Detailed petrographic descriptions of the harzburgite and amphibolite samples used in this study have also been given by Gärtner (2005).
Ultramafic rocks at Site 566
Whereas serpentinized peridotites have been recovered at Site 494 and 567 in sequences intercalated with mafic rocks, the entire recovered basement at Sites 566 and 570 consists of ultramafic rocks that commonly display internal shearing. All the peridotites are highly serpentinized but have been petrologically and geochemically identified as harzburgites (Bourgois et al., 1985; Gärtner, 2005) comprising olivine, orthopyroxene with clinopyroxene exsolution and spinel. Two groups can be distinguished petrographically: harzburgites having panxenomorphic textures (Sites 567 and 566) and cumulative peridotites (Site 570). The latter (cumulative harzburgites) are similar to the harzburgitic cumulates at Santa Elena.
Amphibolites at Site 569
Hole 569A encountered high-temperature metamorphic amphibolites that have been overprinted by a retrograde low-temperature greenschist-facies assemblage. Tschermakitic hornblende (50 vol. %) and an albite-rich plagioclase (40 vol. %) make up the major constituents (Gärtner, 2005). Both minerals form small hypidiomorphic to xenomorphic crystals in a granoblastic texture. The whole-rock major element compositions of the amphibolites indicate a basaltic origin (Bourgois et al., 1985).
Mafic rocks at Sites 567 and 494
The basement recovered at Site 567 (Leg 85) and nearby Site 494 (Leg 67) (Fig. 2a) is composed of an unordered sequence of gabbro, dolerite, basalt and serpentinite in variable states of alteration and metamorphism. Compared with the other sites, the basement is more tectonically deformed and fragmented at theses two sites, resulting in a poor recovery of less than 20%. No consecutive cores were recovered and the mafic rocks are mostly pebble-size, angular drill cuttings, with no recovered contacts between lithological types (Shipboard Scientific Party, 1985). No apparent sequences (e.g. variations of grain size from basalt to gabbro) have been identified. Most of the samples have undergone various grades of zeolite- to greenschist-facies metamophism and/or are often fractured, brecciated or strongly weathered. The dominant minerals in the sampled mafic rocks are plagioclase, clinopyroxene and olivine (Bellon et al., 1985; Bourgois et al., 1985). Olivine and clinopyroxenes are often strongly altered to serpentine or iddingsite and actinolite or chlorite, respectively.
Although the dissected state of the recovered rocks precludes the identification of any larger volcanic structures (such as individual lava flows), quenched textures and progressive transitions from microlitic to fine-grained doleritic textures observed in thin sections from basaltic samples from the lower part of Hole 567A [below 430 m below sea floor (mbsf)] have been interpreted as pillow fragments (Bellon et al., 1985). A submarine formation for the lower sequence of this hole is further supported by the abundant palagonitic groundmass (altered glass) seen in many basaltic samples. In contrast, cores 17, 18, and 24, all from the upper part of the hole above 430 mbsf, host rare clay beds containing volcanic glass, xenoliths and apatite, and are interpreted as strongly altered tuffaceous material (Shipboard Scientific Party, 1985).
|
ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND PREVIOUS...SAMPLING AND PETROGRAPHYANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Samples were wrapped in plastic foil and carefully manually crushed, washed with deionized water in an ultrasonic bath and hand-picked under a binocular microscope to avoid the most altered parts, veins and secondary fillings. Selected rock-chips were then split and one fraction was manually powdered in an agate mortar for major and trace element analyses; the other split of chips was dissolved for isotope chemistry.
Major and some trace element determinations were performed on fused glass beads by X-ray fluorescence spectroscopy (XRF) at IFM-GEOMAR using a Philips X’Unique PW 1480 with Rh-tube. H2O and CO2 concentrations were analyzed in a Rosemount CSA 5003 infrared photometer. Average accuracy (n = 7) of international reference standards JB-2, JB-3 and JA-2 measured with the samples is better than 1% (SiO2), 3% (TiO2, Al2O3, MgO, Na2O, K2O), 5% (Fe2O3, CaO, P2O5, Sr, V), 8% (MnO), 20% (Co, Zr) and 32% (Cr, Ni).
Trace elements (Sc, Cu, Ga, Zn, Rb, Y, Cs, Nb, Ba, Hf, Ta, Tl, Pb, Th, U and all REE) were analyzed by inductively coupled plasma mass spectrometry (ICP-MS) using a ThermoFinnigan Element2 at the Institute of Geosciences, University of Bremen. Mixed acid (HF–aqua regia) pressure digests were prepared at 210°C using a MLS Ethos microwave. About 50 mg of the sample was processed, and the analyte solution (having a final dilution factor of 5000) was spiked with 2·5 ng/ml indium as internal standard. Precision (n = 8) of multiple digests of international reference standard BCR2 analysed along with the samples is better than 5% for most elements; accuracy is better than 5% except for Hf (6%), Y (7%), Zn (9%), Cs and Pb (10%), and Cu (11%). Results of major and trace element analyses of samples and standards are given in Table 1.
|
All isotope analyses were carried out at IFM-GEOMAR on rock chips by thermal ionization mass spectrometry (TIMS) using Thermo Finnigan TRITON (Sr, Nd isotopes) and Finnigan MAT262 (Pb, Sr isotopes) systems and an AXIOM multiple collector (MC)-ICP-MS system (Hf isotopes), all operating in static mode. Samples for Sr, Nd, and Pb isotope determinations were leached with 6 N HCl for 1 h at 125°C and dissolved in a hot HF–HNO3 mixture. Sr, Nd, and Pb chromatography followed the procedure outlined by Hoernle & Tilton (1991
). Sr and Nd ratios were fractionation corrected within run to 86Sr/88Sr = 0·1194 and 146Nd/144Nd = 0·7219. During three analysis sessions the Sr standard NBS 987 averaged 86Sr/87Sr = 0·710241 ± 6 (external 2
deviation, n = 5, TRITON) and 0·710226 ± 11 (n = 4, MAT 262) and 0·710239 ± 12 (n = 3, MAT 262). All Sr isotope data were normalized to 0·710250. The in-house SPEX Nd monitor gave 143Nd/144Nd = 0·511717 ± 6 (external 2 deviation, n = 7) and 0·511720 ± 6 (n = 3). All Nd isotope data were normalized to SPEX = 0·511715 corresponding to 0·511850 for the La Jolla standard. During the analysis periods, the Pb standard NBS 981 yielded a long-term average of 206Pb/204Pb = 16·899 ± 7 (external 2 deviation, n = 169), 207Pb/204Pb = 15·436 ± 9 and 208Pb/204Pb = 36·524 ± 28 corresponding to an external 2 reproducibility of better than 0·021%/a.m.u. All Pb data were fractionation corrected to the values of Todt et al. (1996). Total Pb chemistry blanks in the analysis period averaged 68 ± 20 pg (n = 26) and are thus insignificant.
Hf isotopes were determined on a subset of samples. Rock chips (200–500 mg) were digested for 60 h at 130°C in a HF–HNO3 mixture; chemical separation followed the procedures outlined by Blichert-Toft et al. (1997). Detailed descriptions of the MC-ICP-MS setup and analyzing modes have been given by Geldmacher et al. (2006). The JMC 475 Hf-reference standard gave 176Hf/177Hf = 0·282121 ± 8 (external 2 deviation, n = 9) and all measured Hf values were corrected to JMC 475 = 0·282163 (Blichert-Toft et al., 1997). To monitor performance throughout the measurements our in-house Hf SPEX standard was additionally run every two or three samples (yielding a JMC 475 corrected average of 0·282171 ± 12, n = 34). All isotope ratios are presented in Table 2.
|
|
40Ar/39Ar incremental heating experiments were conducted on amphibole and feldspar phenocryst separates and on matrix chips at the IFM-GEOMAR Tephrochronology Laboratory. The particles were hand-picked from crushed and sieved splits. All separates and chips were cleaned using an ultrasonic disintegrator. Phenocrysts were then etched in 15% hydrofluoric acid for 10 min (amphibole) and 15 min (feldspar). Samples were neutron irradiated at the 5 MW reactor of the GKSS Reactor Center (Geesthacht, Federal Republic of Germany), with crystals and matrix chips in aluminum trays and irradiation cans wrapped in 0·7 mm cadmium foil. Samples were step-heated by laser. Purified gas samples were analyzed using a MAP 216 noble gas mass spectrometer. Raw mass spectrometer peaks were corrected for mass discrimination, background and blank values determined every fifth analysis. The neutron flux was monitored using TCR sanidine (Taylor Creek Rhyolite = 27·92 Ma; Dalrymple & Duffield, 1988
; Duffield & Dalrymple, 1990) and internal standard SAN6165 (0·470 Ma; van den Bogaard, 1995). Vertical variations in J values were quantified by a cosine function fit. Lateral variations in J were not detected. Corrections for interfering neutron reactions on Ca and K are based on analyses of optical grade CaF2 and high-purity K2SO4 salt crystals that were irradiated together with the samples. Ages derived from step-heating analyses are based on plateau portions of the age spectra (Table 3). Plateau regions generally comprise >50% of the 39Ar released and more than three consecutive heating steps that yield the same ages (within 2 error). |
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND PREVIOUS...SAMPLING AND PETROGRAPHYANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Major and trace elements
As revealed from petrographic examination, most of the recovered rocks have undergone seawater alteration and experienced varying degrees of greenschist-facies metamorphism. Therefore the concentrations of major and trace elements known to be mobile during alteration and (open-system) metamorphism (such as K, Na, Ca, Mg, Si, Rb, Ba and Sr) will not be considered in subsequent data interpretation. Instead, only elements that are relatively immobile, including Hf, Th, Nb, Ta, Y, V, Zr, Ti and all REE are used. U and Pb receive special consideration. Although U can be mobile in aqueous solutions (in an oxidized environment) and Pb is mobilized under hydrothermal conditions, the necessity for age correction of Pb isotope data prevented their exclusion (see below for potential impacts on the Pb isotope system).
In rock classification diagrams, using only minor and trace elements that are generally considered to remain inert during a number of secondary processes, including metamorphism and submarine alteration, the Guatemalan igneous forearc basement samples fall into two groups (Fig. 3). The first group, comprising all the recovered basement rocks from Holes 494A, 569A (amphibolites), and the upper part of the penetrated basement at Hole 567A (above 430 mbsf), has lower concentrations of highly to moderately incompatible elements and LREE (such as Th, La, Ce and Nd), lower Ti, and a marked depletion in high field strength elements (HFSE), in particular Nb and Ta. This group either plots in or largely overlaps with the field for volcanic arc basalts in tectonomagmatic discrimination diagrams. In particular, Th is useful for distinguishing between arc basalts and depleted MORB when dealing with altered samples (Fig. 3c), as it is the least fluid-mobile large ion lithophile element (and is immobile during greenschist-facies metamorphism) but also behaves as a non-conservative element during subduction processes (becomes enriched in arc magmas) (Elliott, 2003; Kessel et al., 2005). The second group, defined by all igneous rocks sampled from the lower part of Hole 567A (below 430 mbsf until its termination at 501 mbsf), shows enriched incompatible elements, no HFSE depletion and consistently plots in the intraplate basalt field in all discrimination diagrams. Multi-element patterns of depleted group samples (Fig. 4a and b) with (La/Yb)n = 0·12–0·76, (La/Sm)n = 0·19–0·81 and (Sm/Lu)n = 0·41–1· 06 are noticeably distinct from the enriched patterns of the lower Hole 567A samples (Fig. 4c and d) showing (La/Yb)n = 3·15–4·46, (La/Sm)n = 1·63–2·31 and (Sm/Lu)n = 1·46–2·48. The most distinct feature of the depleted group is their marked depletion in Nb and Ta [(Nb/La)n = 0·07–0·83] characteristic of arc volcanic rocks generated in a subduction zone environment. In contrast, the enriched samples from the lower part of Hole 567A display enriched Nb and Ta concentrations [(Nb/La)n = 1·15–1·84] and a uniform depletion of heavy REE (HREE), both consistent with intraplate or OIB compositions. Interestingly, the subaerially exposed igneous units from the Santa Elena Peninsula can be divided into the same two geochemical groups with nearly identical trace element compositions to the drill samples (Figs 3 and 4e, f). Depleted Santa Elena samples SE 6 (Unit II) and SE 27 (Unit IV), however, display less steep incompatible trace element patterns and possess relatively high (Nb/Th)n ratios (1· 0–1· 4), placing them somewhere intermediate between arc and MORB compositions (Hauff et al., 2000a). A few depleted forearc drill samples display similar trace element patterns (e.g. sample 84-567A-20-1, 115–117 cm), including (Nb/Th)n ratios >1 (Fig. 4a).
|
|
Age determinations
Depleted group samples
As a result of the generally higher degree of alteration and the lower potassium content of the depleted forearc basement rocks, selecting suitable samples for age determination and generating reliable data is difficult. Only four 40Ar/39Ar analyses from depleted group samples meet the quality criteria outlined above (Table 3). The data, however, include at least one sample from each drill site from which depleted rocks have been recovered (Fig. 5). Feldspar phenocrysts from Hole 494A (67-494A-29CC) yield a broad age plateau (comprising 81% of 39Ar) at 112·1 ± 9·0 Ma. Matrix analyses from several other depleted arc rock samples from this site produced only disturbed age spectra (below 50% of total 39Ar in the plateau, and therefore not included in Table 3) but seem to reflect a similar age of about
100 Ma. Likewise, all matrix samples from the depleted (upper portion) of Hole 567 failed to produce reliable plateaux. Instead, their age spectra seem to indicate a significant disturbance of their 40Ar/39Ar system. This is demonstrated by samples from both Hole 567A and Hole 494A (Fig. 6). The age spectrum of sample 84-567A-19CC, which produced a small high-temperature plateau (32%) at 100 Ma and a small low-temperature plateau (42%) at 70 Ma could be interpreted to reflect a mid-Cretaceous crystallization age that was overprinted by a late Cretaceous reheating event. Likewise, sample 67-494A-31CC, 3–4 cm produced a limited high-temperature plateau (41%) at 111 Ma and an almost acceptable low-temperature plateau (47%) at 88 Ma. Based on the reliable 112 Ma plateau age of feldspar crystals from depleted sample 67-494A-29CC and the apparent evidence for similar mid-Cretaceous crystallization ages indicated by the disturbed spectra of all other depleted group samples, an average age of 110 Ma is used for the radiometric age correction of radiogenic isotopes for all depleted samples that did not produce a reliable plateau age (Table 2). To evaluate the effects of possible imprecise age assumptions, an age vector illustrating the impact for over- or under-correction of radioactive decay ages (with steps of 10 Myr) is added to each isotope plot in Fig. 7 (see figure caption for details).
|
|
|
If a mid-Cretaceous crystallization age is assumed for all the depleted group samples, the inability of the analyzed matrix samples to preserve this signal confirms known observations that fine-grained crystalline groundmass is more susceptible to disturbance of pristine magmatic 40Ar/39Ar signatures than coarse-grained phenocrysts, which better preserve the crystallization age of the rock (e.g. Koppers et al., 2000
). Following this interpretation, the relatively young age of 83 ± 15 Ma produced by matrix sample 84-567A-20-1, 115–117 cm could result from a complete degassing of radiogenic Ar and thereby record the age of the thermal event (Fig. 5). In accordance with this view, the reported similar K/Ar age of 78·7 ± 3·9 Ma (Bellon et al. 1985) from a sample taken just above (567A-19CC) could be interpreted likewise. A thermal event at around 80 Ma is also supported by the age of the metamorphic amphibolites from Hole 569A. Hornblende crystals from sample 84-569A-10-1, 13–22 cm, yield an age of 79·6 ± 1·1 Ma and most probably reflect the closure age of the crystals after their formation during metamorphism. Interestingly, feldspar phenocrysts from the lowermost depleted sample analyzed from Hole 567 (sample 84-567A-25-1, 1–10 cm, #1) preserved a much older age of 182·0 ± 14·0 Ma, which could serve as an upper age limit for the origin of some of the depleted rocks.
Enriched group samples
From the enriched OIB-like rocks (lower part of Hole 567A, below 430 mbsf) feldspar phenocrysts from seven samples produced acceptable plateau ages ranging from 104 to 219 Ma (Fig. 5). One additional analyzed matrix sample (84-567A-25-1, 91–96 cm) produced a limited plateau of just below 50% but with an age that fits well into this range (143·7 ± 1·6 Ma). These new 40Ar/39Ar step heating ages confirm the ambiguous Cretaceous and Jurassic ages previously determined from the enriched group samples by K/Ar age determinations (Bellon et al., 1985).
Radiogenic isotopes
Consistent with their trace element differences, the two groups form distinct fields on isotope diagrams (Fig. 7) with the incompatible element depleted group also having depleted isotope ratios (low 206Pb/204Pbin = 18·088–19·337 and 207Pb/204Pbin = 15·479–15·564 but elevated 143Nd/144Ndin = 0·512904–0·513050 and 176Hf/177Hfin = 0·283056–0·283161). Except for two samples from Hole 567A, the depleted drill samples from all holes and the similar depleted Santa Elena units II–IV form a dense cluster that overlaps the compositional field for Pacific normal (N)-MORB (age corrected to 125 Ma, to reflect the average age of all the analyzed samples) in all isotope diagrams. Two samples (567A-21CC, 23–27 cm and 567A-23-1, 1–4 cm), however, possess more radiogenic Pb isotope ratios but have Nd and Hf isotope ratios similar to other depleted group samples. Because analytical error can be ruled out (duplication of sample 567A-21CC, 23–27 cm reproduced the same enriched Pb isotope values), a post-eruptive addition of Pb-enriched material (e.g. by Pb-rich hydrothermal fluids) is assumed. In fact, these samples have been recovered near the transition to the underlying geochemically enriched units (having higher Pb isotope ratios and concentrations), and an exchange of hot, Pb-carrying fluids along this transition zone during tectonic juxtaposition (see discussion below) with the enriched group rocks appears possible.
In contrast, the enriched group samples show a more widespread compositional range with distinctly more enriched signatures (high 206Pb/204Pbin = 19·361–21· 696 and 207Pb/204Pbin = 15·647–15·858 but low 143Nd/144Ndin = 0·512715–0·512836 and 176Hf/177Hfin = 0·282892–0·282917). In general, the samples describe linear arrays in all isotope ratio plots (Fig. 7), suggesting mixtures of two source components with the enriched end-member having a HIMU-like isotope composition. The most enriched sample (84-567A-29-2, 144–146 cm) has similar Pb isotope ratios to present-day HIMU type-locality lavas from Tubai and St. Helena. Although no radiometric age is available for this sample, an age of 150 Ma is assumed (see above). A 30 or even 60 Myr younger (or older) age, however, would not result in any significant revision of the HIMU affiliation for this sample as demonstrated by the age/enrichment vector in Fig. 7a and b). Compared with the isotopic compositions of other CLIP rocks and present-day Galápagos archipelago lavas, the enriched forearc samples have more radiogenic 206Pb/204Pb and 207Pb/204Pb, and less radiogenic 176Hf/177Hf ratios [all CLIP isotope data plot entirely within the age-corrected Galápagos fields (e.g. Hauff et al., 2000b; Thompson et al. 2003) and therefore are not separately shown in Fig. 7].
The immobility of the refractory Lu–Hf system during submarine alteration and strong metamorphism even in the presence of fluids (e.g. Weaver & Tarney, 1981; Thompson et al., 2008) makes Hf isotope ratios particularly suitable for the interpretation of the variably altered samples of this study. Similarly, the Sm–Nd system is generally also considered to be highly resistant to alteration, although in some cases REEs can be redistributed into different secondary phases during low-temperature alteration (Thompson et al., 2008). In contrast, the Pb isotope compositions of altered or greenschist-facies metamorphed rocks can be susceptible to alteration processes and therefore must be treated with caution. Any increase of U (from seawater) or decrease of Pb (during hydrothermal alteration), or a combination of both, inevitably causes a higher 238U/204Pb (µ) ratio than the initial magmatic value. If this happened shortly after solidification of the lava and the system remained closed for U, Th and Pb to the present, correction for radioactive ingrowth over time will give a reasonably good approximation of the initial Pb isotope ratio. If significant U and/or Pb mobilization occurs more than several million years after rock formation, any age correction using the altered parent/daughter element ratios will result in an over- or under-correction of the data, resulting in less or more radiogenic initial Pb values, respectively. A way to assess the extent of this effect is to check internal trace element correlations for the samples and to compare the magnitude of the applied age correction with the alteration state [expressed as loss on ignition (LOI) value]. Most forearac basement rocks show relatively little deviation from their average U/Th ratio (= 3·3, R2 = 0·96). Samples from the depleted group, however, generally tend to have gained post-magmatic U (Fig. 8a), whereas samples from the enriched group seem to be more vulnerable to U loss (assuming that Th behaves incompatibly during submarine alteration or metamorphism). As shown in Fig. 7a–c, the exceptional radiogenic initial Pb isotope values for at least two samples from the depleted group (23-1, 1–4 cm and 21CC, 23–27 cm) and 2–4 samples from the enriched group (in particular 29-2, 144–146 cm and 25-2, 31–35 cm) of Hole 567A could reflect disturbance of their U–Pb system by post-magmatic alteration. As the two depleted samples show no significant U gain (Fig. 8b), post-magmatic addition of Pb has to be assumed. On the other hand, these two samples do not have particularly low Nd/Pb ratios (Fig. 8c), compared with the average of the other depleted group samples (being considerably lower than N-MORB as expected for subduction-related rocks). The relatively similar initial Pb isotope composition of almost all other depleted group samples, despite varying degrees of age correction (and possible U and Pb gain or loss) suggests that the initial Pb isotopic composition of the depleted arc rocks was relatively uniform and that any U–Pb system disturbance occurred shortly after their formation. Therefore, the deviation of depleted samples 23-1, 1–4 cm and 21CC, 23–27 cm towards more radiogenic Pb values as seen in the isotope plots of Fig. 7a–c probably results from late-stage disturbance of the Pb system leading to age under-correction. This is also consistent with them showing some of the highest LOI values but the smallest applied age correction (Fig. 8d).
|
H2O + CO2 contents of samples; see Inter-element systematics provide a less clear picture for the two exceptionally radiogenic samples from the enriched group. For the most extreme sample (29-2, 144–146 cm) moderate loss of U is coupled with no significant increase in Pb (compared with other enriched OIB sources); this caused only a minor under-correction (also indicated by the smallest difference in measured and initial 206Pb/204Pb values among all the enriched samples, Fig. 8d). Contamination towards extremely radiogenic Pb isotope ratios by the added Pb also seems unlikely, as the melts or fluids from depleted mantle, subducted sediments or seawater would all have significantly lower 206Pb/204Pb ratios. The other sample (25-2, 31–35 cm) shows only a very small gain in U but a significant loss of Pb (Fig. 8c), resulting in a strong over-correction and implying that the true initial value of this sample would be even more radiogenic as shown in Fig. 7.
It can be concluded that some post-magmatic disturbance of the U–Pb system of several of the forearc samples has almost certainly occurred, but that the general picture of a radiogenically depleted and an enriched group is not affected. The Pb isotope systematics of both groups are also consistent with other, generally immobile, isotope systems, such as Hf and Nd (Fig. 7c–f) and their trace element composition (Figs 3 and 4
). The trend towards extreme HIMU-like initial isotope compositions in the enriched group samples is probably not caused by post-magmatic U–Pb disturbance, but seems to reflect the source composition of these magmas. This interpretation is also consistent with almost all the enriched group samples having Nd and Hf isotope ratios identical or very similar to present-day St. Helena or Tubai compositions (Fig. 7c–f).
In comparison with the enriched forearc samples, the three samples from the enriched Santa Elena unit I display similar trends in Pb isotope space towards the HIMU mantle end-member (although with slightly more enriched 208Pb/204Pb relative to 206Pb/204Pb) but their Hf and Nd isotope ratios are distinctly lower [similar to the Enriched Mantle (EM) II end-member], therefore suggesting a different (or heterogeneous) magma source.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND PREVIOUS...SAMPLING AND PETROGRAPHYANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Stratigraphic correlations
The new age and geochemical data confirm that the igneous forearc basement off Guatemala cannot be considered as a monolithic structure, but is heterogeneous even at a single drill site (e.g. Hole 567A). The mafic (basaltic to gabbroic) rocks can be subdivided into two groups: (1) a geochemically strongly depleted group with subduction-related geochemical signatures that comprises all samples from Site 494 and from the upper part of the basement at Hole 567A (above 430 mbsf); (2) a geochemically enriched group having HIMU-like OIB compositions and found only in the lower basement (below 430 mbsf) at Hole 567A (Fig. 2a). Because of the geochemical similarity of the depleted rocks from Holes 494A and 567A (Figs 3, 4 and 7) and the close proximity of the two sites, it is safe to assume that both holes penetrate the same (shallow) depleted rock units. These depleted rock units are covered at both sites by late Cretaceous sediments of late Campanian–early Maastrichtian age (
71 Ma) (Shipboard Scientific Party, 1982, 1985), consistent with the older mid-Cretaceous 40Ar/39Ar ages here reported for the underlying basement.
The temporal ambiguity seen in the disturbed age spectra obtained from matrix chip analyses of depleted group samples, and their failure to produce acceptable plateau ages, contrasts with the few available, but >100 Ma, phenocryst ages. Considering the evidence for a post-magmatic thermal overprint at 80 Ma and the 80 Ma age of the secondary amphibolite analyzed from Hole 569A (sample 84-569-10-1, 13–22 cm, #2), we propose a multi-stage scenario for the history of the depleted parts of the Guatemalan forearc basement. The depleted rocks were formed by arc volcanism between the mid-Jurassic and mid-Cretaceous (180 to 100 Ma). At around 80 Ma, they experienced a pervasive tectonic event, resulting in various stages of metamorphism from zeolite facies observed in most basalts and dolerites to greenschist facies in the gabbros (e.g. Holes 494A, 567A) and locally amphibolite facies (Hole 569A). The age of the amphibolites records the timing of the metamorphic overprint (resetting the thermally susceptible Ar/Ar system of their protolith). The Ar/Ar system in the matrix of less metamorphosed greenschist-facies rocks became severely disturbed (resulting in different high- and low-temperature plateaux); however, their coarse-grained phenocrysts were able to retain their radiogenic Ar and have preserved their initial crystallization ages.
The age range of the enriched group samples generally overlaps with the reported (phenocryst) and assumed (disturbed matrix age spectra) crystallization ages of the depleted rocks, although one sample (e.g. 84-567A-25-2, 88–104 cm, #11) yields a surprisingly old late Triassic age. As the distinct geochemical differences between the depleted arc rock samples and the enriched OIB samples require different mantle processes, the two groups must have been formed in different tectonic settings (geographical areas?) and were subsequently brought together in their present location in the Guatemalan forearc basement.
The transition between the geochemically depleted and geochemically enriched units in Hole 567A is marked by a 30 m thick interval (Cores 22–24) of serpentinitic mud containing pieces of serpentinite breccia and altered mafic rocks from sand size up to several centimeters (Shipboard Scientific Party, 1985). The first enriched rocks were recovered in Core 25. The uppermost sample in this core (84-567A-25-1, 1–10 cm, #1) still possesses a depleted geochemical signature (Tables 1 and 2). Its position in the core (above 15 cm of gabbroic drill breccia at the very top of the core), however, indicates that this small piece of depleted rock could have been derived from the superjacent depleted section of this hole and could have fallen into the open hole (during the exchange of the core barrel) before coring continued with Core 25. The thick serpentinitic mud zone above Core 25 is interpreted as a major shear zone (Shipboard Scientific Party, 1985), but, because the entire hole shows an unordered sequence of gabbro, dolerite, basalt and serpentinite (all variably deformed, brecciated and metamorphosed), the entire basement has to be considered as highly tectonized. This is supported by the description of multiple shear zones from micrometer to decimeter scale in most recovered rocks (Ogava et al., 1985), and the lack of any noticeable relationship between drilling depths and radiometric Ar/Ar age of the samples (compare Table 3). In fact, the youngest age of all the enriched group samples comes from the deepest sample on which age determination was carried out (84-567A, 27-1, 92–99 cm, #1), whereas the oldest age comes from a sample recovered in the first (uppermost) core penetrating enriched rock units (84-567A, 25-2, 88–104 cm, #11). Because of the relatively undisturbed Cenozoic sedimentary cover (e.g. Aubouin & von Huene, 1985), this deformation cannot be attributed to the present-day subduction setting, but must have occurred along the landward slope of the trench before the early Eocene (Shipboard Scientific Party, 1985).
At Santa Elena, the mafic and ultramafic rocks show a similar unordered sequence (e.g. Gazel et al., 2006) comprising single thrust sheets with no primary contacts between lithological or stratigraphically different units (e.g. Hauff et al., 2000a; Baumgartner & Denyer, 2006). Similar to the deformation in the Guatemalan forearc basement, thrusting and tectonization of the Santa Elena units occurred during (or prior to) their accretion between 94 and 71 Ma. This is consistent with the 88 ± 4·5 Ma age of the secondary hornblende from an amphibolitic shear zone on Santa Elena indicating a major thermotectonic event.
Geochemical constraints
The similar complex structural relationships between the tectonic slivers of the different lithologies in the Guatemalan forearc basement and at Santa Elena, the similar isotopic and trace element composition of the depleted arc rocks and their overlapping age lead to the conclusion that the depleted Guatemalan forearc basement and Units II–IV of the Santa Elena complex represent a similar, if not the same, dismembered accretionary arc complex (Fig. 2a and b), as previously suggested (e.g. Azéma et al., 1985a). The extreme depletion of incompatible trace elements is typical for relatively young, immature island arc lavas and resembles the signature of West Pacific Izu–Bonin island arc tholeiites from the volcanic front (Fig. 4).
As the primary source of (young) island arc volcanism is the asthenospheric mantle wedge above the subducting slab, its isotopic signature generally overlaps with N-MORB compositions as seen in Fig. 7. Similar ophiolitic bodies containing mixed volcanic (MORB and arc chemistry), gabbroic and serpentinized ultramafic rocks also occur in the Izu–Bonin–Mariana, Tonga and other forearc trench slopes (e.g. Bloomer & Hawkins, 1983; Bloomer & Fisher, 1987). Similar deformation patterns (sheared serpentinite breccias, cataclastic gabbro breccias, repeated successions of volcanic, gabbroic and highly serpentinized ultramafic lithologies) as seen in the Guatemalan and Costa Rican (Santa Elena) forearcs have been documented; for example, at Hahajima seamount, cropping out in the Izu–Bonin forearc (Ishiwatari et al., 2006). In the Izu–Bonin system, early stages of forearc volcanism are characterized by Mg- and Si-rich boninitic andesites resulting from primary melts of refractory (harzburgitic) mantle caused by volatile addition above a subducting slab; these are interpreted as marking the initiation of subduction at a given site (e.g. Dobson et al., 2006). In contrast to the Izu–Bonin forearc, boninitic rocks cannot be identified among the depleted arc samples of this study. Although the lack of high silica contents (boninites have 57–69 wt % SiO2) in the recovered depleted group samples (SiO2 = 46–57 wt %, if normalized on a volatile-free base) could be a result of alteration, their Ti/Zr ratios (both extremely immobile) are higher (97–656) than those characteristic of boninites (23–63; Le Maitre et al., 2002).
The geochemically enriched group possesses an OIB-like trace element and isotopic signature, and can be interpreted as remnants of intraplate volcanism (e.g. seamounts and ocean islands). In contrast to CLIP samples, which have isotope ratios that plot within the extensive age-corrected Galápagos field, drilled Guatemalan forearc rocks do not consistently fall within the Galápagos isotopic fields but trend toward the HIMU mantle end-member. Therefore, it is unlikely that the enriched OIB-type Guatemalan forearc (and enriched Tortugal and Santa Elena Unit I) rocks are related to the CLIP or were derived from the Galápagos plume, unless the plume had more radiogenic Pb (a more HIMU-type composition) in the early Mesozoic (see below). Alternatively, derivation from more concentrated enriched Galápagos plume material (higher proportion of today's enriched end-member composition in the magma source) also appears to be unlikely if the present geochemical systematics of the Galápagos plume heterogeneity are compared with the drilled forearc basement samples. The Galápagos plume is isotopically heterogeneous (e.g. White et al., 1993; Blichert Toft & White, 2001; Harpp & White, 2001) and four end-members are called upon to account for the Sr, Nd, Pb, Hf and He isotopic variations observed across the island group: Northern, Southern, Central, and Eastern domain components (Hoernle et al., 2000). This heterogeneity can be traced back at least 20 Myr (Hoernle et al., 2000) and possibly even further to the early stages of CLIP volcanism (Geldmacher et al., 2003; Thompson et al., 2003). The average compositions of these domains in the various isotopic spaces are shown in Fig. 7 (encircled letters). As shown in the figure, the enriched Guatemalan forearc group samples form trends that extend from different enriched domains or end-members in each plot. In Fig. 7a the trend appears to originate from the Central or Eastern end-member, in Fig. 7b and c the enriched group lies on a continuation of the Southern domain end-member, whereas in Fig. 7c and d, f the enriched group extends the Northern domain end-member composition. This contrasting style of isotopic enrichment precludes any affiliation of the enriched forearc basement samples with the Cenozoic Galápagos plume.
Despite their similar age range and Pb isotope composition, the distinctly lower Hf and Nd isotope ratios of enriched OIB rocks of the Costa Rican forearc (Santa Elenas Unit I) compared with the Guatemalan forearc samples would point to distinct magma sources. On the other hand, if the enriched OIB rocks at both locations originate from accretion of a former Pacific hotspot track, the geochemistry of this track could have been spatially heterogeneous (e.g. with a southern portion having lower Hf and Nd isotope ratios) similar to the spatial heterogeneity observed in the Galápagos and Hawaiian hotspot chains.
Although speculative, one could consider that the geochemically enriched Guatemalan forearc rocks, Santa Elena Unit I and alkaline Tortugal lavas were all produced by an earlier stage of the Galápagos hotspot, which was isotopically more enriched in 206Pb/204Pb and 176Hf/177Hf but heterogeneous in Nd compared with its post-100 Ma composition. To accrete seamounts or terranes that originated at the present site of the Galápagos hotspot to the Chortis Block (which was even further northward before its translation) a strongly northward-directed Farallon plate motion has to be assumed for the Mesozoic, which seems unlikely. On the other hand, the hotspot itself could have been located further to the north in the Mesozoic, consistent with palaeomagnetic evidence, radiometric age dating and numerical models showing that other Pacific hotspots (e.g. Hawaii, Louisville) moved southward during the Cenozoic (e.g. Tarduno et al., 2003; Koppers et al., 2004; Steinberger et al., 2004).
Alternatively, the enriched OIB material accreted in the Guatemalan forearc and Santa Elena did not originate from the Galápagos plume at all but stems from a different hotspot that was located on the Farallon plate prior to 100 Ma. The only long-lived hotspot of similar latitude to the palaeo-position of the Chortis Block is Hawaii. Although reconstructions of Pacific plates and plate boundaries are difficult and ambiguous for times prior to the end of the Cretaceous superchron (118–83 Ma), the spreading center between the West Pacific and the Farallon plates could have been near (compare Steinberger & Gaina, 2007) or even to the west of the Hawaiian hotspot before 100 Ma, resulting in an eastward transportation of any erupted material from the Hawaiian hotspot with the eastward-moving Farallon plate. This idea is consistent with none of the analyzed age data for the enriched group forearc samples being younger than 100 Ma, which would require an odd coincidence for an unknown Farallon hotspot to become extinct right at the time when the Hawaiian hotspot crossed the spreading center. The counter argument would be that the isotopic compositions of the Hawaiian islands and their Emperor Seamount hotspot track generally have far less radiogenic Pb than the enriched Guatemalan and Costa Rican (Santa Elena) forearc samples (Fig. 7). As speculated for the Galápagos hotspot, however (see above), changes in isotopic composition over time spans of >100 Myr cannot be ruled out for long-lived mantle plumes such as the Hawaii plume (which might have had HIMU-type isotopic composition in the past).
Geodynamic reconstruction
Figure 9 shows a simplified geotectonic model for the generation of the Guatemalan forearc. In general, the convergent margins bordering the Farallon plate evolved from overall extensional to compressional strain regimes through the Mesozoic and early Cenozoic as typical for arc systems facing large ocean basins. This transition is characterized by progressive inboard migration of the arc axis, collision of island arcs with continental terranes and increased coupling between the subducting and overriding plates (Busby, 2004). Since at least Mid-Jurassic times, arc volcanism was active in front of the North and Central American continental blocks above the Pacific–Farallon plate subduction zone (Fig. 9a). This is recorded in the subduction-related igneous rocks and associated sedimentary units of the Guerrero terrane (Tardy et al., 1994), the Guatemalan forearc basement (Bellon et al., 1985; this study), Santa Elena (e.g. Hauff et al., 2000a; Gazel et al., 2006) and the Siuna terrane (Flores et al., 2007; Rogers et al., 2007). It is unclear if this volcanism formed a more offshore arc with associated back-arc basins (e.g. Servais et al., 1986). However, the recovery of some samples with similar trace-element patterns to oceanic crust (e.g. forearc sample 84-567A-20-1, 115–117 cm or Santa Elena samples SE 27 and SE 6; Hauff et al., 2000a), the large quantities of depleted serpentinized harzburgite, the pelagic Upper Cretaceous limestone cover above the forearc basement at Site 567, the lack of any seismic or geochemical evidence for continental crust below the Guatemalan forearc basement and the absence of a geochemical trend towards upper crustal ‘EM II’-like isotope compositions in the depleted group samples support this model and point towards formation of arc volcanism in an oceanic setting. Occasionally islands or seamounts of a hotspot track (existing to the east of the Pacific–Farallon ridge between 100 and 220 Ma and formed by intraplate volcanism with a strong HIMU signature) collided with the arc and fragments from them were accreted. No geochemically enriched OIB-like rocks have been reported from the Siuna terrane in the south (e.g. Flores et al., 2007) or from the Guerrero terrane in the north (Tardy et al., 1994) so far, supporting the idea of a relatively narrow hotspot track responsible for the accretion of the enriched group samples in the forearc basement of Guatemala and at Santa Elena.
|
Field evidence also requires the existence of a collision zone to the east of the Guerrero and Chortis terranes (Fig. 9a) recorded by the relicts of arc-magmatism, high-pressure–low-temperature metamorphism of MORB ophiolites, and supra-subduction mantle units all located along the Motagua–Polóchic suture zone (Beccaluva et al., 1995
; Giunta et al., 2002; Harlow et al., 2004). This subduction zone can be interpreted either as the continuation of the ancestral Greater Antilles arc (e.g. Burke et al., 1968; Tardy et al., 1994; Meschede & Frisch, 1998) or as short-lived regional subduction after cessation of rifting along a southward extension of the Arperos Basin (Pindel & Kennan, 2001) or along a failed northwestward rift arm of the proto-Caribbean spreading center (Mann et al., 2007) (Fig. 9a). Subduction along this margin, however, ceased after collision of the Chortis Block with the North American continent in the mid-Cretaceous (Harlow et al., 2004; J. Geldmacher et al., unpublished data). This collision preludes the transition to a more compressive stress regime in the whole region during the Cretaceous mainly caused by the progressively decreasing age of the incoming Farallon lithosphere (decreasing distance to the Pacific–Farallon spreading center) and the cessation of the proto-Caribbean spreading (leading to the beginning of convergence between the two Americas).
Any comprehensive geodynamic reconstruction of the Caribbean region for the late Cretaceous is hampered by the competing models about the origin of the Caribbean oceanic plateau. According to many paleotectonic reconstruction models, this vast plateau was formed on the Farallon plate in the mid-Cretaceous (e.g. Duncan & Hargraves, 1984; Pindell & Barrett, 1990; Hauff et al., 2000a; Pindell et al., 2006) with its peak magmatic activity occurring at around 90 Ma (Hoernle & Hauff, 2007). The plateau (moving eastwards with Farallon plate drift) then collided with the Proto Greater Antilles island arc during the Campanian (84–71 Ma) and choked off the subduction zone. We propose that this collision led to the accretion (and deformation) of island arc complexes onto the western and southern edges of the continental Chortis Block (Fig. 9b), forming the present-day forearc basement in front of Guatemala–El Salvador–Nicaragua and the Siuna terrane in southern Nicaragua. This is consistent with the timing of the secondary thermal overprint recorded in many forearc samples, the age of amphibolite metamorphism and the onset of Late Campanian–Maastrichtian sedimentation at the drill sites.
In the Greater Antilles, the collision caused the reversal of arc polarity, resulting in SW-dipping subduction that allowed the insertion of the Caribbean Plateau (Chorotega Block) between the merging Americas by trench rollback (e.g. Duncan & Hargraves, 1984; Pindell & Barrett, 1990; Hauff et al., 2000a; Hoernle et al., 2002). The collision and insertion of the CLIP transferred the eastward-directed Farallon plate motion to the Chortis Block (Fig. 9c), leading to reactivation of the Motagua–Polóchic suture zone (the rheologically weakest part of the Chortis Block) as a sinistral strike-slip fault (e.g. Donnelly et al., 1990; Beccaluva et al., 1995). The resulting eastward translation of the Chortis Block (including the accreted forearc basement) subsequently triggered the opening of the Cayman Trough in the Eocene (Leroy et al., 2000) (Fig 9d) after parts of the Antilles arc collided with the Bahaman platform and continuing northeastward migration of the arc was prevented (Mann et al., 2007). The timing of the Chortis Block movement is also documented by the systematic decrease of cooling ages of subduction-related plutons in southern Mexico (from 80 to 11 Ma) interpreted to reflect the progressive uplift of the newly formed active continental margin in response to the development of the SE-progressive trench along Mexico's Pacific coast (Guerrero-Garcia & Herrero-Bervera, in preparation) (Fig. 9c and d).
At the southwestern edge of the Chortis Block the accretionary assemblage was partly thrust faulted over the northern rim of the Caribbean plateau (Chorotega Block), resulting in further tectonic dismembering as demonstrated by the large-scale south-directed nappe emplacement at Santa Elena. Any determination of the exact boundary between the Chortis and Chorotega blocks therefore remains ambiguous.
The increasingly compressional regime might explain subsequent uplift of the shelf edge (thereby forming the forearc basin controlling the deposition of the Cenozoic sediments) and further tectonic disruption of accreted arc rocks, OIB and sediment units that form the present-day forearc basement. Alternatively, uplift of the shelf edge and further disruption of the forearc basement was caused by crustal thickening beneath the margin slope by underplating of imbricated oceanic crustal segments detached from the subducting plate until the Eocene (von Huene, 1989). It is possible that these segments represent the latest arriving seamounts and islands from the postulated mid-Pacific hotspot track (having drifted on the incoming plate for several tens of million years).
Starting from at least the Neogene, the subduction style below the forearc wedge changed from accretionary to erosional (possibly as a result of faster convergence rates and arrival of younger oceanic crust; Ranero et al., 2000) leading to long-term margin subsidence, thinning of the upper plate and landward migration of the trench axis and the volcanic chain (Vannuchi et al., 2004; Berhorst, 2006). It is therefore assumed that large parts of the lower forearc basement have been removed and recycled into the mantle.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND PREVIOUS...SAMPLING AND PETROGRAPHYANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
The forearc basement rocks recovered during DSDP Legs 67 and 84 represent a highly tectonized mélange that underlies the broad shelf zone stretching from southern Mexico to the Chortis–Chorotega boundary, where it is subaerially exposed and forms the Santa Elena Peninsula. The depleted volcanic and gabbroic rocks have primitive arc signatures and are interpreted to have formed between
180 and 100 Ma during the development of arc magmatism in front of the western margin of the Chortis Block. The geochemically enriched OIB-like rock units are interpreted as seamounts or islands that have been obducted onto the arc, and originated from a hotspot located in the central Pacific, which was active between 219 and 100 Ma. We can only speculate on whether these units reflect the earlier history of the Hawaiian hotspot when the spreading center was located to the west of it. Based on their isotopic composition, however, there is no evidence for any affiliation of these rock units with the CLIP or the Galápagos hotspot. Therefore, the occurrence of these geochemically enriched volcanic rocks in the Central American forearc basement neither supports nor rejects any Pacific or intra-Caribbean model for the origin of the CLIP. Tectonization–shearing, amphibolite metamorphism and the timing of the secondary thermal overprint of the forearc basement rocks are proposed to originate from the collision of the Chortis Block with the North American Maya Block in the east and the incoming Caribbean plateau in the south at 80 Ma, leading to compression, imbrication and accretion of the arc onto the seaward margins of the Chortis Block. |
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND PREVIOUS...SAMPLING AND PETROGRAPHYANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Supplementary data for this paper are available at Journal of Petrology online.
|
ACKNOWLEDGEMENTS |
|---|
This research used samples provided by the Integrated Ocean Drilling Program (IODP). We thank Jerry Bode (IODP west coast repository) for his friendly help during sampling of the drill cores. D. Rau, M. Thöner, J. Fietzke, J. Sticklus, and S. Hauff (IFM-GEOMAR) and H. Anders (University of Bremen) are acknowledged for outstanding analytical support. This study was funded by the Deutsche Forschungsgemeinschaft (DFG, German Resarch Council), grants HO1833/13-1 and SFB 574 (Contribution 147). Formal Journal reviews by A. Kerr, R. von Huene and A. Saunders are gratefully acknowleged.
*Corresponding author. Present address: IODP, Texas A&M University, 1000 Discovery Drive, College Station, TX 77840, USA. Telephone: +1-979-845-0506. Fax: +1-979-845-0876. E-mail: geldmacher{at}iodp.tamu.edu
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND PREVIOUS...SAMPLING AND PETROGRAPHYANALYTICAL METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Alvarado G, Denyer P, Sinton CW. The 89 Ma Tortugal komatiitic suite, Costa Rica: Implications for a common origin of the Caribbean and Eastern Pacific region from a mantle plume. Geology (1997) 25:439–442.
Aubouin JA, von Huene R. Summary: Leg 84, Middle America trench transect off Guatemala and Costa Rica. In: Deep Sea Drilling Project, Initial Reports, 84—von Huene R, Aubouin J, et al, eds. (1985) Washington DC: US Government Printing Office. 939–957.
Aubouin J, von Huene R, Azéma J, Coulbourn WT, Cowan DS, et al. Premiers résultants des forages profonds dans le Pacifique au niveau de la fosse du Guatemala (fosse d’Amérique Central) Leg 67 du Deep Sea Drilling Project: mai–juin, 1979. Comptes Rendus de l’Académie des Sciences (1979) 289:1215–1220.
Azéma J, Bourgois J, Tournon J, Baumgartner PO. L’orogène pré-sénonien supérieur de la marge pacifique de Costa Rica (Amérique Centrale). Bulletin de la Société Géologique de France (1985a) 1(2):173–179.[Web of Science]
Azéma J, Bourgois J, Baumgartner PO, Tournon J, Desmet A, Aubouin J. A tectonic cross-section of the Costa Rican Pacific littoral as a key to the structure of the landward slope of the Middle America trench off Guatemala. In: Deep Sea Drilling Project, Initial Reports, 84—von Huene R, Aubouin J, et al, eds. (1985b) Washington, DC: US Government Printing Office. 831–849.
Baumgartner PO, Denyer P. Evidence for middle Cretaceous accretation at Santa Elena peninsular (Santa Roas Accretionary Complex), Costa Rica. Geologica Acta (2006) 4(1–2):179–191.
Beccaluva L, Bellia S, Coltori M, Dengo G, Giunta G, Mendez J, Romero J, Rotolo S, Siena F. The northwestern border of the Caribbean plate in Guatemala: new geological and petrological data on the Motagua ophiolitic belt. Opfioliti (1995) 20(1):1–15.
Bellon H, Tournon J. Contribution de la géochronométrie K–Ar à l’étude du magmatism de Costa Rica, Amérique Central. Bulletin de la Société Géologique de France (1978) 6(7):955–956.
Bellon H, Maury RC, Joron JL, Bourgois J, Aubouin J. Geochemistry, mineralogy, and 40K-40Ar radiometric dating of Leg 84 basalts—Guatemala trench. In: Deep Sea Drilling Project, Initial Reports, 84—von Huene R, Aubouin J, et al, eds. (1985) Washington, DC: US Government Printing Office. 655–663.
Berhorst A. Die Struktur des aktiven Kontinentalhangs vor Nicaragua und Costa Rica: marin-seismische Steil- und Weitwinkelmessungen. In: PhD thesis (2006) University of Kiel.
Blichert-Toft J, White WM. Hf isotope geochemistry of the Galápagos islands. Geochemistry, Geophysics, Geosystems (2001) 2. paper number 2000GC000138.
Blichert-Toft J, Chauvel C, Albarède F. Separation of Hf and Lu for high-precision isotope analyses of rock samples by magnetic sector-multiple collector ICP-MS. Contributions to Mineralogy and Petrology (1997) 127:248–260.[CrossRef][Web of Science]
Bloomer SH, Fisher RL. Petrology and geochemistry of igneous rocks from the Tonga trench—a non-accreting plate boundary. Journal of Geology (1987) 95:469–495.[Web of Science]
Bloomer SH, Hawkins JW. Gabbroic and ultramafic rocks from the Mariana trench: An island arc ophiolite. In: The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands (Part II). Geophysical Monograph, American Geophysical Union—Hayes DE, ed. (1983) 27:294–317.
Bourgois J, Desmet A, Tournon J, Aubouin J. Mafic and ultramafic rocks of Leg 84: petrology and mineralogy. In: Deep Sea Drilling Project, Initial Reports, 84—von Huene R, Aubouin J, et al, eds. (1985) Washington, DC: US Government Printing Office. 633–642.
Burke K, Coates AG, Robinson E. Geology of the Benbow Inlier and surrounding areas. In: Transactions of the 4th Caribbean Geological Conference, Port-of-Spain, Trinidad—Saunders JB, ed. (1968) Port of Spain: Ministry of Petroleum and Mines. 299–309.
Busby C. Continental growth at convergent margins facing large ocean basins: a case study from Mesozoic convergent-margin basins of Baja California, Mexico. Tectonophysics (2004) 392:241–277.[CrossRef][Web of Science]
Calvo C. Provenance of plutonic detritus in cover sandstones of Nicoya Complex, Costa Rica: Cretaceous unroofing history of a Mesozoic ophiolite sequence. Geological Society of America Bulletin (2003) 115:832–844.
Chaffey DJ, Cliff RA, Wilson BM. Characterization of the St. Helena source. In: Magmatism in the Ocean Basins. Geological Society, London, Special Publications—Saunders AD, Norry MJ, eds. (1989) 42:257–276.
Dalrymple GB, Duffield WA. High precision 40Ar/39Ar dating of Oligocene tephra from the Mogollon–Datil volcanic field using a continuous laser system. Geophysical Research Letters (1988) 15:463–466.
Dengo G. Tectonic–igneous sequence in Costa Rica. In: A Volume to Honor A. F. Budington. Geological Society of America, Special Papers—Engel A. EJ, James HJ, Leonhard BF, eds. (1962) 133–161.
Denyer P, Baumgartner PO, Gazel E. Characterization and tectonic implications of Mesozoic–Cenozoic oceanic assemblages of Costa Rica and Western Panama. Geologica Acta (2006) 4:219–235.
DePaolo DJ. Neodymium Isotope Geochemistry. Minerals and Rocks Series 20 (1988) New York: Springer.
DeWever P, Azéma J, Tournon J, Desmet A. Découverte de matériel océanique du Lias–Dogger inférieur dans la péninsule de Santa Elena (Costa Rica, Amérique Centrale). Comptes Rendus de l’Académie des Sciences (1985) 300(15):759–764.
Dobson PF, Blank JG, Maruyama S, Liou JG. Petrology and geochemistry of boninite series volcanic rocks, Chichi-jima, Bonin Islands, Japan. International Geology Review (2006) 48:669–701.[CrossRef][Web of Science]
Donnelly TW, Horne GS, Finch RC, Lopez-Ramos E. Northern Central America: The Maja and Chortis blocks. In: The Caribbean Region. Volume H: The Geology of North America—Dengo G, Case JE, eds. (1990) Boulder, CO: Geological Society of America. 37–76.
Duffield WA, Dalrymple GB. Taylor Creek rhyolite of New Mexico, a rapidly emplaced field of domes and flows. Bulletin of Volcanology (1990) 52:475–487.[CrossRef][Web of Science]
Duncan RA, Hargraves RB. Caribbean region in the mantle reference frame. In: The Caribbean–South American Plate Boundary and Regional Tectonics. Geological Society of America, Memoirs—Bonini W, Hargraves RB, Shagam R, eds. (1984).
Elliott T. Tracers of the slab. In: Inside the Subduction Factory. Geophysical Monograph, American Geophysical Union—Eiler J, ed. (2003) 138:23–45.
Elming SÅ, Rasmussen T. Results of magnetotelluric and gravimetric measurements in western Nicaragua, Central America. Geophysical Journal International (1997) 128:647–658.[CrossRef][Web of Science]
Flores K, Baumgartner PO, Skora S, Baumgartner L, Muntener O, Cosca M, Cruz D. The Siuna serpentinite mélange: an early Cretaceous subduction/accretion of a Jurassic arc. EOS Transactions, American Geophysical Union (2007) 88((52)). Fall Meeting Supplement, Abstract T11D-03.
Frisch W, Meschede M, Sick M. Origin of the Central American ophiolites: Evidence from paleomagnetic results. Geological Society of America Bulletin (1992) 104:1301–1314.
Galer SJG, O’Nions RK. Residence times of thorium, uranium and lead in the mantle with implications for mantle convection. Nature (1985) 316:778–782.[CrossRef][Web of Science]
Gärtner C. Petrologisch-petrografische Untersuchung von magmatischen Gesteinsproben des Deep Sea Drilling Programs (DSDP) Leg 84 (Guatemala). Bachelor Thesis (2005) Institut für Mineralogie, Westfälische Wilhelms Universität Münster.
Gazel E, Denyer P, Baumgartner PO. Magmatic and geotectonic significance of Santa Elena Peninsula, Costa Rica. Geologica Acta (2006) 4(1–2):193–202.
Geldmacher J, Hanan BB, Blichert-Toft J, Harpp K, Hoernle K, Hauff F, Werner R, Kerr AC. Hafnium isotopic variations in volcanic rocks from the Caribbean Large Igneous Province and Galápagos hotspot tracks. Geochemistry, Geophysics, Geosystems (2003) 4. number 1062, doi:10.1029/2002GC000477.
Geldmacher J, Hoernle K, Klügel A, van den Bogaard P, Wombacher F, Berning B. Origin and geochemical evolution of the Madeira–Tore Rise (eastern North Atlantic). Journal of Geophysical Research (2006) 111. B09206, doi:10.1029/2005JB003931.
Giunta G, Beccaluva L, Coltorti M, Mortellaro D, Siena F, Cutrupla D. The peri-Caribbean ophiolites; structure, tectono-magmatic significance and geodynamic implications. Caribbean Journal of Earth Sciences (2002) 36:1–20.
Harlow GE, Hemming SR, Avé Lallemant HG, Sisson VB, Sorensen SS. Two high-pressure–low-temperature serpentinite-matrix mélange belts, Motagua fault zone, Guatemala: A record of Aptian and Maastrichtian collisions. Geology (2004) 32:17–20.
Harpp KS, White WM. Tracing a mantle plume: Isotopic and trace element variations of Galápagos seamounts. Geochemistry, Geophysics, Geosystems (2001) 2(6). doi:10.1029/2000GC000137.
Hart SR. The Dupal anomaly: a large-scale isotope anomaly in the Southern Hemisphere mantle. Nature (1984) 309:753–757.[CrossRef][Web of Science]
Hart SR, Gaetani GA. Mantle Pb paradoxes: the sulfide solution. Contributions to Mineralogy and Petrology (2006) 153:295–308.
Hauff F, Hoernle K, Schmincke H-U, Werner R. A mid Cretaceous origin for the Galápagos hotspot: volcanological, petrological, and geochemical evidence from Costa Rican oceanic crustal segments. Geologische Rundschau (1997) 86:141–155.
Hauff F, Hoernle KA, van den Bogaard P, Alvarado GE, Garbe-Schönberg D. Age and geochemistry of basaltic complexes in western Costa Rica: contributions to the geotectonic evolution of Central America. Geochemistry, Geophysics, Geosystems (2000a) 1(5). doi:10.1029/1999GC000020.
Hauff F, Hoernle KA, Tilton G, Graham D, Kerr AC. Large volume recycling of oceanic lithosphere: geochemical evidence from the Caribbean Large Igneous Province. Earth and Planetary Science Letters (2000b) 174:247–263.[CrossRef][Web of Science]
Hoernle K, Hauff F. Oceanic igneous complexes. In: Central America: Geology, Resources and Hazards—Bundschuh J, Alvarado GA, eds. (2007) London: Taylor & Francis–Balkema. 523–547.
Hoernle K, Tilton GR. Sr–Nd–Pb isotope data for Fuerteventura (Canary Islands) basal complex and subaerial volcanics: application to magma genesis and evolution. Schweizerische Mineralogische und Petrographische Mitteilungen (1991) 71:5–21.
Hoernle K, Werner R, Morgan JP, Bryce J, Mrazek J. Existence of complex spatial zonation in the Galápagos Plume for at least 14 m.y. Geology (2000) 28:435–438.
Hoernle KA, van den Bogaard P, Werner R, Lissinna B, Hauff F, Alvarado G, Garbe-Schönberg D. The missing history (16–71 Ma) of the Galápagos hotspot: implications for the tectonic and biological evolution of the Americas. Geology (2002) 30:795–798.
Hoernle K, Hauff F, van den Bogaard P. 70 Myr history (139–69 Ma) for the Caribbean Large Igneous Province. Geology (2004) 32:697–700.
Hofmann AW. Chemical differentiation of the Earth: the relationship between mantle, continental crust, and oceanic crust. Earth and Planetary Science Letters (1988) 90:297–314.[CrossRef][Web of Science]
Ibrahim AK, Latham GV, Ladd JW. Seismic refraction and reflection measurements in the Middle America trench offshore Guatemala. Journal of Geophysical Research (1979) 84(B10):5643–5649.[CrossRef]
Ischwatari A, Yanagida Y, Li Y-B, Ishii T, Haraguchi S, Koizumi K, Ichiyama Y, Umeka M. Dredge petrology of the boninite- and adakite-bearing Hahajima seamount of the Ogasawara (Bonin) forearc: An ophiolite or serpentinite seamount? Island Arc (2006) 15:102–118.[CrossRef]
Janney PE, Castillo PR. Geochemistry of Mesozoic Pacific mid-ocean ridge basalt: Constraints on melt generation and the evolution of the Pacific upper mantle. Journal of Geophysical Research (1997) 102:5207–5229.[CrossRef]
Kerr AC. La Isla de Gorgona, Colombia: a petrological enigma? Lithos (2005) 84:77–101.[Web of Science]
Kerr AC, Tarney J. Tectonic evolution of the Caribbean and northwestern South America; The case for accretion of two Late Cretaceous oceanic plateaus. Geology (2005) 33:269–272.
Kerr AC, Tarney J, Marriner GF, Klaver GT, Saunders AD, Thirlwall MF. The geochemistry and petrogenesis of the late-Cretaceous picrites and basalts of Curaçao, Netherlands Antilles: a remnant of an oceanic plateau. Contributions to Mineralogy and Petrology (1996) 124:29–43.[CrossRef][Web of Science]
Kerr AC, Tarney J, Kempton PD, Spadea P, Nivia A, Mariner GF, Duncan RA. Pervasive mantle plume head heterogeneity: Evidence from the late Cretaceous Caribbean–Colombian oceanic plateau. Journal of Geophysical Research (2002) 107(B7). doi:10.1029/2001JB000790.
Kessel R, Schmidt MW, Ulmer P, Pettke T. Trace element signature of subduction fluids, melts and supercritical liquids at 120–180 km depths. Nature (2005) 437:724–727.[CrossRef][Web of Science][Medline]
Koppers AP, Staudigel H, Wijbrans JR. Dating crystalline groundmass separates of altered Cretaceous seamount basalts by the 40Ar/39Ar incremental heating technique. Chemical Geology (2000) 166:139–158.[CrossRef][Web of Science]
Koppers APP, Duncan RA, Steinberger B. Implications of a nonlinear 40Ar/39Ar age progression along the Lousville Seamount trail for models of fixed and moving hot spots. Geochemistry, Geophysics, Geosystems (2004) 5. Q06L02, doi:10.1029/2003GC000671.
Kuijpers E. The geologic history of the Nicoya ophiolite complex, Costa Rica, and its geotectonic significance. Tectonophysics (1980) 68:233–255.[CrossRef][Web of Science]
Kurz MD, Geist DJ. Dynamics of the Galápagos hotspot from helium isotope geochemistry. Geochimica et Cosmochimica Acta (1999) 63:4139–4156.[CrossRef][Web of Science]
Le Maitre RW, Streckeisen A, Zanettin B, et al. Igneous Rocks: A Classification and Glossary of Terms: Recommendations of the International Union of Geological Sciences Subcommission on the Systematics of Igneous Rocks (2002) Cambridge: Cambridge University Press.
Leroy S, Mauffret A, Patriat P, Mercier de Lépinary B. An alternative interpretation of the Cayman trough evolution from a reidentification of magnetic anomalies. Geophysical Journal International (2000) 141:539–557.[CrossRef][Web of Science]
Mann P, Rogers RD, Gahagan L. Overview of plate tectonic history and its unresolved tectonic problems. In: Central America: Geology, Resources and Hazards—Bundschuh J, Alverado GE, eds. (2007) London: Taylor & Francis. 200–237.
Mauffret, A. & Leroy S. Seismic stratigraphy and structure of the Caribbean igneous province. Tectonophysics| (1997) 283:61–104.[CrossRef][Web of Science]
Maury RC, Bougault H, Joron JL, Girad D, Treuil M, Azema J, Aubouin J. Volcanic rocks from Leg 67 sites: mineralogy and geochemistry. In: Deep Sea Drilling Project, Initial Reports, 67—Aubouin J, von Huene R, et al, eds. (1982) Washington, DC: US Government Printing Office. 557–576.
Meschede M. A method of discriminating between different types of mid-ocean ridge basalts and continental tholeiites with the Nb–Zr–Y diagram. Chemical Geology (1986) 56:207–218.[CrossRef][Web of Science]
Meschede M. The impossible Galapagos connection: geometric constraints for a near-American origin of the Caribbean plate. Geological Rundschau (1998) 87:200–205.[CrossRef]
Meschede M, Frisch W. A plate-tectonic model for the Mesozoic and Early Cenozoic history of the Caribbean plate. Tectonophysics (1998) 296:269–291.[CrossRef][Web of Science]
Meschede M, Zweigel P, Frisch W, Völker D. Mélange formation by subduction erosion: the case of the Osa mélange in southern Costa Rica. Terra Nova (1999) 11:141–148.
Ogava Y, Fujioka K, Nishiyama T, Uehara S, Nakagawa M. Ophiolitic rocks of the Middle America trench landward slope off Guatemala: deformational characteristics and tectonic significance. In: Deep Sea Drilling Project, Initial Reports, 84—von Huene R, Aubouin J, et al, eds. (1985) Washington, DC: US Government Printing Office. 791–809.
Pearce JA. Trace element characteristics of lavas from destructive plate boundaries. In: Andesites—Thorpe RS, ed. (1982) Chichester: John Wiley. 525–547.
Pindell JL, Barrett SF. Geological evolution of the Caribbean region: A plate tectonic perspective. In: The Caribbean Region. Volume H: The Geology of North America—Dengo G, Case JE, eds. (1990) Boulder, CO: Geological Society of America. 405–432.
Pindell J, Kennan L. Kinematic evolution of the Gulf of Mexico and the Caribbean. In. In: Transactions, Petroleum Systems of Deep-water Basins: Global and Gulf of Mexico Experience (2001) Houston, TX: Gulf Coast Section, Society of Economic Paleontologists and Mineralogists Foundation. 193–220.
Pindell J, Kennan L, Stanek KP, Maresch WV, Draper G. Foundations of Gulf of Mexico and Caribbean evolution: eight controversies resolved. Geologica Acta (2006) 4:303–341.
Ranero CR, von Huene R. Subduction erosion along the Middle America convergent margin. Nature (2000) 404:748–752.[CrossRef][Web of Science][Medline]
Ranero CR, von Huene R, Flueh E, Duarte M, Baca D. A cross section of the forearc Sandino Basin, Pacific Margin of Nicaragua. Tectonics (2000) 19:335–357.[CrossRef][Web of Science]
Rogers RD, Mann P, Emmet PA. Tectonic terranes of the Chortis block based on integration of regional aeromagnetic and geological data. In: Geological and tectonic development of the Carribbean Plate Boundary in Northern Central America. Geological Society of America, Special Papers—Mann P, ed. (2007) 428:129–149.[CrossRef]
Saal AE, Kurz MD, Hart SR, Blusztajn JS, Blichert-Toft J, Liang Y, Geist DJ. The role of lithospheric gabbro on the composition of Galapagos lavas. Earth and Planetary Science Letters (2007) 257:391–406.[CrossRef][Web of Science]
Servais M, Cuevas-Perez E, Monod O. Une section de Sinaloa à San Luis Potosí: nouvelle approche de l’évolution du Mexique nord-occidental. Bulletin de la Société Géologique de France (1986) 8(II:1033–1047.
Shipboard Scientific Party. Site 494. In: Deep Sea Drilling Project, Initial Reports, 67—Aubouin J, von Huene R, et al, eds. (1982) Washington, DC: US Government Printing Office. 27–78.
Shipboard Scientific Party. Site 567. In: Deep Sea Drilling Project, Initial Reports, 84—von Huene R, Aubouin J, et al, eds. (1985) Washington, DC: US Government Printing Office. 111–166.
Shervais JW. Ti–V plots and the petrogenesis of modern ophiolitic lavas. Earth and Planetary Science Letters (1982) 59:101–118.[CrossRef][Web of Science]
Schmidt-Effing R. Radiolarien der Mittelkreide aus dem Santa-Helena Massiv von Costa Rica. Neues Jahrbuch für Geologie und Palaeontologie, Abhandlungen (1980) 160:241–257.
Sinton CW, Duncan RA, Denyer P. Nicoya Peninsula, Costa Rica: a single suite of Caribbean oceanic plateau magmas. Journal of Geophysical Research (1997) 102:15507–15520.[CrossRef]
Sinton CW, Duncan RA, Storey M, Lewis J, Estrada JJ. An oceanic flood basalt province within the Caribbean plate. Earth and Planetary Science Letters (1998) 155:221–235.[CrossRef][Web of Science]
Steinberger B, Gaina C. Plate-tectonic reconstructions predict part of the Hawaiin hotspot track to be preserved in the Bering Sea. Geology (2007) 35:407–410.
Steinberger B, Sutherland R, O’Connel JO. Prediction of Emperor–Hawaii seamount locations from a revised model of global plate motion and mantle flow. Nature (2004) 430:167–173.[CrossRef][Web of Science][Medline]
Sun S-s, McDonough WF. Chemical and isotopic composition of oceanic basalts: implications for mantle composition and processes. In: Magmatism in the Ocean Basins. Geological Society, London, Special Publications—Saunders AD, Norry MJ, eds. (1989) 42:313–345.
Tarduno JA, Duncan RA, Scholl DW, Cottrell RD, Steinberger B, Thordarson T, Kerr BC, Neal CR, Frey FA, Torii M, Carvallo C. The Emperor Seamounts: southward motion of the Wawaiian hotspot plume in the Earth's mantle. Science (2003) 301:1064–1069.
Tardy M, Lapierre H, Freydier C, Coulon C, Gill J.-B, Mercier de Lepinary B, Beck C, Martinez -R. J, Talavera -M. O, Ortiz -H. E, Stein G, Bourdier J.-L, Yta M. The Guerrero suspect terrane (western Mexico) and coeval arc terranes (the Greater Antilles and the Western Cordillera of Colombia): a late Mesozoic intra-oceanic arc accreted to cratonal America during the Cretaceous. Tectonophysics (1994) 230:49–73.[CrossRef][Web of Science]
Taylor RN, Nesbitt RW. Isotopic characteristics of subduction fluids in an intra-oceanic setting, Izu–Bonin Arc, Japan. Earth and Planetary Science Letters (1998) 164:79–98.[CrossRef][Web of Science]
Thompson PME, Kempton PD, White RV, Kerr AC, Tarney J, Saunders AD, Fitton JG, McBirney A. Hf-Nd isotope constraints on the origin of the Cretaceous Caribbean plateau and it's relationship to the Galápagos plume. Earth and Planetary Science Letters (2003) 217:59–75.[Web of Science]
Thompson PME, Kempton PD, Kerr AC. Evaluation of the effects of alteration and leaching on Sm–Nd and Lu–Hf systematics in submarine mafic rocks. Lithos (2008) 104:164–176.
Todt W, Cliff RW, Hanser A, Hofmann AW. Evaluation of a 202Pb–205Pb double spike for high precision lead isotope analyses, in earth processes. Earth Processes: Reading the Isotopic Code. Geophysical Monograph, American Geophysical Union—Basu A, Hart S, eds. (1996) 95:429–437.
van den Bogaard P. 40Ar/39Ar ages of sanidine phenocrysts from Laacher See tephra (12,900 yr BP): Chronostratigraphic and petrological significance. Earth and Planetary Science Letters (1995) 133:163–174.[CrossRef][Web of Science]
Vannucchi P, Galeotti S, Clift PD, Ranero CR, von Huene R. Long-term subduction–erosion along the Guatemalan margin of the Middle America Trench. Geology (2004) 32:617–620.
Venable M. A geological, tectonic, and metallogenetic evaluation of the Siuna terrane (Nicaragua). In: PhD dissertation (1994) Tucson: University of Arizona.
von Huene R. The middle America convergent plate boundary, Guatemala. In: The Eastern Pacific Ocean and Hawaii. Volume N: The Geology of North America—Winterer EL, Hussong DM, Decker RW, eds. (1989) Boulder, CO: Geological Society of America. 535–550.
Walther CHE, Flueh ER, Ranero CR, von Huene R, Strauch W. Crustal structure across the Pacific margin of Nicaragua: evidence for ophiolitic basement and shallow mantle sliver. Geophysical Journal International (2000) 141:759–777.[CrossRef][Web of Science]
Weaver BL, Tarney J. Chemical changes during dyke metamorphism in high-grade basement terranes. Nature (1981) 289:47–49.[CrossRef][Web of Science]
White WM. 238U/204Pb in MORB and open system evolution of the depleted mantle. Earth and Planetary Science Letters (1993) 115:211–226.[CrossRef][Web of Science]
White WM, McBirney AR, Duncan RA. Petrology and geochemistry of the Galápagos Islands: Portrait of a pathological mantle plume. Journal of Geophysical Research (1993) 98:19533–19563.[CrossRef]
Ye S, Bialas J, Flueh ER, Stavenhagen A, von Huene R, Leandro G, Hinz K. Crustal structure of the Middle American trench off Costa Rica from wide-angle seismic data. Tectonics (1996) 15:1006–1021.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||