Lithos 114 (2010) 168–185
Contents lists available at ScienceDirect
Lithos
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / l i t h o s
Geochemical characteristics of the Río Verde Complex, Central Hispaniola:
Implications for the paleotectonic reconstruction of the Lower Cretaceous
Caribbean island-arc
Javier Escuder-Viruete a,⁎, Andrés Pérez-Estaún b, Dominique Weis c, Richard Friedman c
a
b
c
Instituto Geológico y Minero de España, C. La Calera 1, Tres Cantos, 28760 Madrid, Spain
Instituto Ciencias Tierra Jaume Almera-CSIC, Lluís Solé Sabarís s/n, 08028 Barcelona, Spain
Pacific Centre for Isotopic and Geochemical Research, University of British Columbia, 6339 Stores Road Vancouver, Canada BC V6T 1Z4
a r t i c l e
i n f o
Article history:
Received 20 November 2008
Accepted 12 August 2009
Available online 23 August 2009
Keywords:
Island-arc
Back-arc basin
Mantle melting
Hispaniola
Caribbean plate
a b s t r a c t
New geochronological, trace element and Sr–Nd isotope data for metabasalts, dolerites and amphibolites
from the Río Verde Complex, Central Hispaniola, are integrated with existing geochemical data for mafic
volcanic rocks and metamorphic derivatives from the Los Ranchos, Amina and Maimón Formations, giving
new insights into magma petrogenesis and paleotectonic reconstruction of the Lower Cretaceous Caribbean
island-arc–back arc system. U–Pb and 40Ar/39Ar age data show that the Río Verde Complex protoliths were in
part coeval with volcanic rocks of the Los Ranchos Formation (Upper Aptian to Lower Albian). The
geochemical data establish the existence of gradients in trace element parameters (Nb/Yb, Th/Yb, Zr/Yb, Zr/
Ba, and normalized Ti, Sm, Y and Yb abundances) and Nd isotope compositions from throughout Hispaniola,
which reflect differences in the degree of mantle wedge depletion and contributions from the subducting
slab. The Río Verde Complex mafic rocks and some mafic sills and dykes intruding in the Loma Caribe
Peridotite, have a transitional IAT to N-MORB geochemistry and a weak subduction-related signature, and
are interpreted to form in a rifted arc or evolving back-arc basin setting. The Los Ranchos, Amina and
Maimón Formations volcanic rocks have arc-like characteristics and represent magmatism in the volcanic
front. Trace element and Nd isotope modeling reproduce observed data trends from arc to back-arc and
suggest that the variations in several geochemical parameters observed in a SW direction across the
Caribbean subduction system can be explained from the progressively lower subduction flux into a
progressively less depleted mantle source. The low Nb contents and high (εNd)i values in both arc and backarc mafic rocks imply, however, the absence of a significant Lower Cretaceous plume enriched component. In
order to explain these observations, a model of proto-Caribbean oceanic lithosphere subducting to the SW at
least in the 120–110 Ma interval, is proposed to cause the observed magmatic variations in the Lower
Cretaceous Caribbean island-arc–back-arc system. In this context, arc rifting and initial sea-floor spreading to
form the Río Verde Complex protoliths occurred in the back-arc setting of this primitive island-arc, built on
the NE edge of the Caribbean plate.
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
Island-arcs develop because of subduction of oceanic lithosphere
beneath another oceanic plate. The aqueous fluids and/or hydrous
melts released from the subducting slab and their reaction with the
overlying mantle wedge provide the prime control on arc magma
genesis (Hawkesworth et al., 1993; Pearce and Peate, 1995; Woodhead et al., 1998; Stern, 2002). Magma genesis processes along
convergent plate boundaries mainly include: (1) adiabatic upwelling
of asthenospheric mantle induced by slab penetration (Peacock and
⁎ Corresponding author. Instituto Geológico y Minero de España. C. La Calera 1, 28760
Tres Cantos, Madrid. Spain. Tel.: +34 917287242.
E-mail address: [email protected] (J. Escuder-Viruete).
0024-4937/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.lithos.2009.08.007
Wang, 1999; Gerya et al., 2004); (2) partial melting of the mantle
wedge as a result of the addition of slab-derived fluids (Arculus and
Powell, 1986; Pearce and Parkinson, 1993; Schmidt and Poli, 1998;
Hochstaedter et al., 2001; Martinez and Taylor, 2002); and (3) melting
of the subducted slab and addition of the resultant melts to the mantle
wedge (Defant and Drummond, 1990; Yogodzinski et al., 2001;
Tatsumi and Hanyu, 2003). The compositions of arc lavas can vary
across and along individual arcs. This probably results from: (1)
compositional differences in subducted slab rocks (Plank and Langmuir,
1993); (2) differences in the dehydration or melting conditions of slab
materials (Defant and Drummond, 1990); (3) differences in degree of
partial melting in the mantle wedge (Pearce and Parkinson, 1993); (4)
differences in the volume of slab-derived components added to the
overlying mantle wedge (Kelemen et al., 2003; Singer et al., 2007); and
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
(5) pre-existing mantle heterogeneity (Leat et al., 2004). As such, the
composition of island arc lavas can be formed by a variety of processes;
however, identifying which processes are involved in the petrogenesis
of any particular lava is often difficult.
Outcrops of Late Aptian to Lower Albian volcanic rocks in the
Greater Antilles are the oldest known arc-related strata of the
primitive Caribbean island-arc. Geochemical studies of these rocks
indicate a broad compositional spectrum across and along the arc that
make their interpretation difficult in terms of subduction-related
petrogenetic models (Lebrón and Perfit, 1994; Kerr et al., 1999; Jolly
et al., 2001, 2006; Marchesi et al., 2006, 2007; Jolly et al., 2007, for
example in Hispaniola (Lewis et al., 2002; Escuder-Viruete et al., 2006,
2007a,b,c). Further, these volcanic rocks are of considerable significance in the debated tectonic reconstructions of initial subduction
polarity and magma genesis processes, along the long-lived (Early
Cretaceous to Mid-Eocene) destructive plate margin separating the
North American and Caribbean plates.
In this paper, we present new regional petrologic, U–Pb/Ar–Ar
geochronological, trace element and Sr–Nd radiogenic isotope data for
the Río Verde Complex mafic igneous rocks in Central Hispaniola, that
allow us to relate it to the Lower Cretaceous Caribbean island-arc–
back arc system. These data in conjunction with published data of
coeval subduction-related units in Hispaniola allow us to address
three main questions. These are: (1) nature and age of the Río Verde
Complex protoliths; (2) tectonic setting of origin of the complex and
relations with coeval magmatic units; and (3) polarity in the intraoceanic Caribbean subduction system.
2. Geodynamic setting
2.1. The Caribbean island-arc
The Caribbean island-arc is subdivided into three domains: (1) the
extinct Early Cretaceous to Paleogene Greater Antilles in the north,
including Cuba, Jamaica, Hispaniola, Puerto Rico, and the Virgin
Islands; (2) northern South America, including Tobago, Margarita, and
Colombian/Venezuelan allochthons in the south; and (3) the
volcanically active Lesser Antilles in the east, which rest on buried
remnants of the south-eastern extension of the Cretaceous arc. In the
Greater Antilles, Early Cretaceous (Aptian) to mid-Eocene island-arc
volcanic rocks are traditionally subdivided (Donnelly et al., 1990) into
a lower primitive island-arc suite (PIA), consisting predominantly of
spilitized tholeiitic basalt and dacitic–rhyolitic lavas, and an overlying
basaltic to intermediate calc-alkaline suite (CA). PIA lavas typically
have low large-ion lithophile elements (LILE), rare earth elements
(REE), and high field strength elements (HFSE) abundances, low
radiogenic Pb, and near-horizontal primitive-mantle normalized REE
patterns; younger CA lavas are distinguished from PIA by elevated
incompatible element abundances and variably enriched REE patterns. Recent studies, however, have demonstrated that Caribbean
island-arc volcanism produced basalt compositions with a broad
range of LREE/HREE values and Sr–Nd–Pb isotope compositions,
reflecting a wide variation in mantle sources and proportions of
subducted sediments during its 80 Ma long eruptive history, from
Lower Cretaceous to Late Eocene (c.a. 125 to 45 Ma; Kerr et al., 1999;
Lewis et al., 2000; Jolly et al., 2001; Lewis et al., 2002; Jolly et al., 2006;
Marchesi et al., 2006; Escuder-Viruete et al., 2006; Jolly et al., 2007;
Marchesi et al., 2007; Escuder-Viruete et al., 2007a, 2008).
The PIA suite is represented by the Water Island Formation in the
Virgin Islands (Rankin, 2002; Jolly and Lidiak, 2006), volcanic phases I
and II in Central and Northeastern Puerto Rico (pre-Robles and preSanta Olaya Lava units; Jolly et al., 2001, 2006), clasts of PIA rocks in
the pre-Camujiro sedimentary rocks near the province de Camagüey
and Los Pasos Formation in Central Cuba (Kerr et al., 1999; Proenza
et al., 2006), and the Los Ranchos, Amina and Maimón Formations in
the Central and Eastern Cordilleras of Hispaniola (Kesler et al., 1990;
169
Draper and Lewis, 1991; Lebrón and Perfit, 1994; Kesler et al., 2005;
Escuder-Viruete et al., 2006; Fig. 1). Recent geochemical investigations reveal many PIA basalts in the Greater Antilles, including the
Téneme Formation in Eastern Cuba (Proenza et al., 2006; Marchesi
et al., 2007), and the Los Ranchos (Escuder-Viruete et al., 2006),
Maimón (Lewis et al., 2000, 2002) and Amina (Escuder-Viruete et al.,
2007b) Formations in Hispaniola, as well as some Water Island
basalts, are regionally comparable low-Ti island-arc tholeiites (IAT)
and boninites. Taken together, the timing and geochemical characteristics in the PIA suite suggest a supra-subduction zone setting during
the earliest stages of the Aptian to Lower Albian Caribbean island arc
development (Escuder-Viruete et al., 2006). In Hispaniola, the Hatillo
Formation, a massive reef limestone of upper Lower Albian age
(Myczynski and Iturralde-Vinent, 2005), unconformably overlies the
Los Ranchos Formation.
2.2. The geology of Central Hispaniola
Located on the northern margin of the Caribbean plate, the
tectonic collage of Hispaniola results from the WSW to SW-directed
oblique-convergence of the continental margin of the North American
plate with the Greater Antilles island-arc system, which began in
Cretaceous and continues today. The arc-related rocks are regionally
overlain by Upper Eocene to Holocene siliciclastic and carbonate
sedimentary rocks that post-date island-arc activity, and record the
oblique arc-continent collision in the north, as well as the active
subduction along the southern Hispaniola margin (Mann, 1999).
Central Hispaniola is a composite of oceanic derived units bound by
the left-lateral strike-slip Hispaniola and San José–Restauración fault
zones (Fig. 1). Accreted units mainly include serpentinized Loma
Caribe peridotites, MORB-type gabbros and basalts, Late Jurassic deepmarine sediments, volcanic units related to Caribbean–Colombian
oceanic plateau (CCOP; e.g. the Duarte Complex; Lapierre et al., 1997;
Escuder-Viruete et al., 2007c), and Late Cretaceous arc-related
igneous and sedimentary rocks (Lewis et al., 2002). These units are
variably deformed and metamorphosed to prenhite–pumpellyte,
greenschist and amphibolite facies, but the textures of the protoliths
are often preserved.
In the study area (Fig. 2), the macrostructure is characterized by
several main NNW–SSE to WNW–ESE trending fault zones that bound
different crustal domains or tectonic blocks, e.g. Hispaniola (HFZ),
Hato Mayor (HMFZ) and Bonao–La Guácara (BGFZ) fault zones. To the
north of the HFZ, the Maimón Formation forms a NW-trending belt of
schists separating the Los Ranchos Formation, the Hatillo limestone,
the Late Cretaceous Las Lagunas Formation and, locally, the Paleocene–
Eocene sedimentary rocks of the Don Juan Formation, from the Loma
Caribe Peridotite. The belt consists mainly of sub-greenschist and
greenschist-facies metabasalt and metadacite/rhyolite, and minor
intercalated carbonaceous schist, iron formation and marble. Development of penetrative foliation increases toward the contact with the
HFZ, where the rocks are converted to mylonitic–phyllonitic schists
(Draper et al., 1996). To the NW, and in a similar structural position,
the mafic and felsic Amina schists occur under the neogene sediments
of the Cibao Basin (Fig. 1). On the basis of geochemical and Sr–Nd
isotopic data, Escuder-Viruete et al. (2007b) argue that the mafic and
felsic schists of the Amina and Maimón Formations are foliated and
metamorphosed equivalents of the Los Ranchos Formation volcanics.
To the south of the HFZ, Central Hispaniola domain is also bounded
by the Hato Mayor fault zone, and comprises the Loma Caribe Peridotite,
several related gabbro and dolerite bodies, the Río Verde Complex, and
the Peralvillo Sur Formation. Due to the fact that the block is composed
of a peridotite basement intruded and/or covered by volcanic mafic
rocks, it has been considered to be an ophiolite (Lewis et al., 2002, 2006).
The Loma Caribe Peridotite is mainly composed of spinel harzburgite,
but clinopyroxene-rich harzburgite, dunite, lherzolite and small bodies
of podiform chromitites also occur (Lewis et al., 2006). The peridotites
170
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
Fig. 1. (a) Map of the northeastern Caribbean plate margin. Box shows location of the study area. (b) Schematic geological map of Central, Septentrional and Eastern Cordilleras in
Hispaniola. SFZ, Septentrional fault zone; HFZ, Hispaniola fault zone; BGFZ, Bonao–La Guácara fault zone; SJRFZ, San José–Restauración fault zone; EPGFZ, Enriquillo–Plantain Garden
fault zone. Box shows location of the Fig. 2.
are typically extensively serpentinized and variably sheared, in particular toward the upper structural levels. The overlying rocks consist of
hundred-meter-sized bodies of modal layered and foliated gabbro that
pass structurally upward into massive, isotropic gabbro. Individual
dolerite dykes and sills intrude serpentinized peridotites and gabbroic
rocks, showing chilled margins. They become more abundant upwards
in the sequence and to the NE. The Peralvillo Sur Formation forms a
narrow belt immediately northeast of the Loma Caribe Peridotite
(Fig. 2). It is composed of a 1500–2300 m-thick basaltic sequence of
massive flows and pillow lavas that host massive sulfide deposits, and is
overlain by ∼1000 m of volcaniclastic sediments, tuffaceous mudstone
and Campanian radiolarian cherts (Lewis et al., 2002). South of the
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
171
Fig. 2. Schematic geological map of the Bonao 1:100,000 quadrangle and A–A' geological cross section, showing the stratigraphic and structural relationships of the Río Verde Complex. HFZ, Hispaniola fault zone; HMFZ, Hato Mayor fault zone;
and HT, Hatillo thrust. Stars show locations of samples for U–Pb and Ar–Ar geochronology and obtained ages, as well as some other relevant regional data (Lewis et al., 2002; Escuder-Viruete et al., 2007a and unpublished).
172
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
HMFZ, the Lower Cretaceous Duarte Complex is intruded by the Arroyo
Caña batholith, the Late Cretaceous foliated Hbl-tonalites and the
Middle Campanian to Maastrichtian Siete Cabezas Formation.
3. Geology of the Río Verde Complex
The Río Verde Complex forms a NW-trending belt, of 4–5 km
width and 25 km length, with a strongly deformed contact against the
Loma Caribe Peridotite. Internally, it consists of several imbricated
thin, but laterally extensive, NW-trending slices of mafic igneous
rocks, subordinate sedimentary rocks and metamorphic derivatives.
Felsic volcanic rocks are absent. These units are variably deformed and
metamorphosed to prenhite–pumpellyte, greenschist and low-P
amphibolite facies conditions. In the lower structural levels and to
the SW, it is composed of a 1500–2500 m-thick basaltic sequence of
massive flows that host minor massive sulfide deposits, and is
overlain by ∼ 1000 m of mafic tuffs, volcaniclastic sediments,
tuffaceous mudstone and green cherts. Upwards in the structural
sequence and to the NE an increase in deformation occurs and rocks
are intensely sheared and transformed into fine-grained actinolite,
epidote, chlorite, and white mica-bearing mafic schists to mediumand coarse-grained foliated amphibolites (Fig. 3b). However, gabbroic
and doleritic textures such as coarse-grained, intersectal to subophitic
intergrowths of clinopyroxene and plagioclase have been preserved in
low strain pods (Fig. 3a). In the uppermost structural levels, the main
structural elements in the amphibolites are a synmetamorphic
foliation (S) and mineral or stretching lineation (L). S is defined by
the planar alignment of hornblende and plagioclase prismatic grains
and, locally, by alternating Hbl-rich and Pl-rich segregations; L is
commonly defined by elongate Hbl nematoblasts. Regionally, S planes
have a consistently NW- to WNW-trend and dip a low to high-angle to
the NE. Syn- to late-kinematic dolerite dykes and Ep ± Qtz ± Cal veins
occur throughout the Río Verde Complex, particularly in the upper
structural levels. Dolerite dykes can be traced into high-shear strain
domains, where they are rotated subparallel to the regional foliation
in the amphibolites and transformed into L–S mylonitic tectonites
(Fig. 3b). Field relations therefore indicate that deformation and
metamorphism of the Río Verde Complex protoliths was in part coeval
with the syn-kinematic emplacement of the dolerite magmas and
hydrothermal activity.
4. Geochronology
Map in Fig. 2 shows the sample locations. Analytical procedures
are in Appendix 1 and results are reported in Appendices 2 and 3. All
ages are quoted at the 2σ level of uncertainty.
4.1. U–Pb samples
The selected U–Pb sample was a clinopyroxene + plagioclase
foliated gabbro (sample 6JE93A) collected in the core of a texturally
zoned sill ∼ 10 m-thick, intruded in the serpentinized Loma Caribe
Peridotite at Loma Peguera, Falcondo plant, Bonao. Typically, gabbros
have a magmatic foliation in the coarse-grained core of the sill that
grades to a strong magmatic to solid state foliation in the fine-grained
rim, which is subparallel to the intrusive contact and the foliation in
the enclosing serpentinites. The sample has MORB geochemical
characteristics with a weak subduction signature (Section 5.3.,
geochemistry). Only a few zircon grains were recovered from the
sample; they were clear, pale pink to colourless, stubby to equant
prisms, with aspect ratios of ∼ 1.5–2.0. Of these, two fractions (one
single grain and one two grain fraction) yielded concordant, overlapping and precise results (Fig. 4) that give a Concordia age (Ludwig,
Fig. 3. Río Verde Complex rocks at Balneario Ledesma outcrop. (a) Gabbros and
metagabbros (108 ± 20 Ma; Sm–Nd whole rock isochron) in an undeformed meterscale boudin surrounded by foliated amphibolites. (b) Syn- to late-kinematic intrusion
of a dolerite dyke in the amphibolites with a 110 Ma old S–L fabric (Ar–Ar in Hbl).
Amphibolites and dolerite are similar BABB-like magmas and suggest coeval intrusion
and deformation of Río Verde Complex magmas.
Fig. 4. Concordia diagrams for microgabbro sill intrusive at Loma Peguera in the Loma
Caribe serpentinized peridotite (6JE93A). U–Pb procedures and analytical data are in
the Appendices 1 and 2. See text for discussion.
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
2003) of 125.4 ± 0.4 Ma, which is interpreted as the crystallization age
of the gabbro. One other analysed grain gave overlapping, but very
imprecise results that do not affect the calculated age and two other
processed grains were not successfully analysed. The attempts of
dating the Río Verde Complex mafic protoliths by U–Pb method were
unsuccessful due to the absence of zircons.
4.2.
40
Ar/39Ar samples
Amphibolites of the Río Verde Complex with a penetrative S–L
fabric were selected for dating by 40Ar/39Ar method (Fig. 5). Samples
show a strong mineral lineation defined by 0.5–4 mm long hornblende
nematoblasts consistent with the simultaneity of ductile deformation
and low-P amphibolite facies metamorphism. Sample 6JE34B is a
173
blastomylonitic S–L amphibolite collected in the upper structural
levels of the complex at Balneario Ledesma outcrop. The protolith is
locally preserved in low-strain domains surrounded by an S–C fabric
developed in the amphibolites (Fig. 3a), and consists of isotropic
gabbro with MORB geochemical characteristics and a weak subduction
signature. The obtained hornblende plateau age is 110.3 ± 1.4 Ma
(MSWD = 0.36) for five steps (7–11) and 70.2% of the 39Ar released.
The inverse isochron age (MSWD = 0.29) is 107 ± 27 Ma. Sample
6JE34D is an amphibolite with a strong S–L fabric also from the
Balneario Ledesma. Mafic protoliths are also coarse-grained MORBlike tholeiitic gabbros. The obtained hornblende plateau age is 110.7 ±
1.6 Ma (MSWD = 0.33) for five steps (6–10) and 73.2% of the 39Ar
released. The inverse isochron age is 111 ± 15 Ma (MSWD = 0.34).
Sample 2JE38 is a medium to coarse-grained amphibolite with a
foliation defined by alternating Hbl-rich and Pl-rich bands. It was
collected near the tectonic contact with the Loma Caribe Peridotite in
the unpaved road to Río Verde town. The obtained hornblende plateau
age is 118.6 ± 1.3 Ma (MSWD = 0.44) for four steps (4–7) and 76.8% of
the 39Ar released. The inverse isochron age for the same plateau steps
is 119.2 ± 5.9 Ma (MSWD = 0.35). For the six high temperature steps
(4–9), the obtained inverse isochron age of 113.2 ± 8.4 Ma
(MSWD = 1.8) is younger but within error of the original plateau.
4.3. Interpretation
Fig. 5. The 40Ar/39Ar spectrum of hornblende from amphibolites of the Río Verde
Complex. The plateau ages were calculated following techniques described in Appendix
1. A summary of 40Ar–39Ar incremental heating experiments is in Appendix 3. See text
for discussion.
Fig. 6 includes the U–Pb and Ar–Ar ages obtained for this study and
other relevant regional data, which permits us to constrain the origin
and cooling history of the Río Verde Complex and to establish
correlations with coeval units in Hispaniola. Two points are related in
Fig. 6. (1) The U–Pb 125.4 ± 0.4 Ma age of the gabbro sill is the oldest
to date obtained in the Central Cordillera and suggests that tholeiitic
magmas with a subduction-related signature are as old as the
lowermost Aptian. (2) The Ar–Ar plateau ages obtained in the
amphibolites indicates metamorphic thermal peak, ductile deformation and cooling of the complex between 120 and 110 Ma, due to the
peak temperatures of 635–562 °C obtained from Hbl-Pl thermobarometry and the slightly lower closure temperature of hornblende
(525–450 °C). Thus, field and geochronological data supports the
synchronicity of the MORB-like magmatism with a subduction
signature and the syn-metamorphic deformation of the complex, at
least during the 120–110 Ma interval. This interpretation is consistent
with the imprecise but within error whole-rock Sm/Nd isochron age
of 108 ± 20 Ma (MSWD = 0.54; [143Nd/144Nd]i = 0.512834), obtained
from seven samples of amphibolites and metabasalts of the Río Verde
Complex (Escuder–Viruete et al., unpublished). The dated gabbro sill
is likely to belong to the same tholeiitic suite or to a different slightly
older suite with similar geochemical characteristics.
The range of Ar–Ar plateau ages of the Río Verde Complex
amphibolites are coeval, within error, to the porphyritic dacite/
rhyolite flows of the Los Ranchos Formation, and the genetically
related gabbros, diorites and the hornblende-bearing tonalites of the
Eastern Cordillera and Cotuí area batholiths (Fig. 6; data from Kesler
et al., 2005; Escuder-Viruete et al., 2006). These results allow us to
establish the arc-related felsic volcanism of the Los Ranchos
Formation at 118–110 Ma (Late Aptian–Early Albian), and that this
volcanism was coeval with the ductile deformation and the last
batches of tholeiitic magmas of the Río Verde Complex, recorded by
the syn- to late-kinematic dolerite dikes. Also, these ages are
consistent with the late Lower Albian age of the unconformably
overlying Hatillo Limestone (∼ 107–105 Ma; Myczynski and IturraldeVinent, 2005). In summary, the data presented indicates that the Río
Verde Complex mafic magmas were in part coeval with volcanic rocks
of the Los Ranchos Formation, i.e. they are part of the Lower
Cretaceous Caribbean island-arc–back-arc system. Note that, south
of the HFZ, the Duarte Complex records Lower Cretaceous mantle
plume magmatism without subduction influences.
174
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
Fig. 6. (a) Summary of geochronological ages for the Los Ranchos Formation and intrusive Hbl-bearing tonalite batholiths in the Eastern Cordillera. Sources; a, Kesler et al. (2005); b,
Escuder-Viruete et al. (2006); c, this work; and d, Myczynski and Iturralde-Vinent (2005). Rectangles are the error bars (in 2σ). Time scale from Gradstein et al. (2004). See text for
discussion.
5. Geochemistry
5.2. Chemical changes due to alteration and metamorphism
5.1. Analytical methods
The analyzed mafic igneous rocks have been variably altered,
deformed and metamorphosed. Consequently, changes of the bulkrock chemistry are expected as a consequence of selected mobility of
relevant elements during these processes. Many major (e.g., Si, Na, K,
Ca) and trace (e.g., Cs, Rb, Ba, Sr) elements are easily mobilised by late
and/or post-magmatic fluids and under metamorphism; however, the
HFSE (Y, Zr, Hf, Ti, Nb and Ta), REE, transition elements (V, Cr, Ni and
Sc) and Th, are generally unchanged under a wide range of
metamorphic conditions, including seafloor alteration at low to
moderate water/rock ratios (Bédard, 1999). Therefore, the geochemical characterization and the petrogenetic discussion will be based
mostly on the HFSE and REE, as well as Nd-isotopes, as it can be
assumed that they were not significantly affected by alteration or
metamorphism. We focus on the Río Verde Complex mafic igneous
rocks and their relation with the Los Ranchos, Amina and Maimón
Formations (PIA suite), whose geochemical characteristics have been
described (Escuder-Viruete et al., 2006, 2007b).
Samples were powdered in an agate mill, and analysed for major
oxides and trace elements by inductively-coupled plasma-mass
spectrometry (ICP-MS). This analytical work was done at ACME
Analytical Laboratories Ltd in Vancouver and results reported in
Table 1, as well as details of analytical accuracy and reproducibility in
Appendix 1. For major elements the detection limits are in general
<0.01% (Appendix 1). The detection limits for trace elements are
typically <0.1 ppm, except for Ba, Ce, Th (0.2), Ga and Zr (0.5 ppm);
for some trace elements, they are as low as 0.05 ppm. A representative
subset of samples (Table 2) was also analysed for Sr and Nd isotopic
compositions at the Pacific Centre for Isotopic and Geochemical
Research at the University of British Columbia. Rb, Sr, Sm and Nd
concentrations were measured by a Thermo Finnigan Element2, a
double focussing (i.e., high resolution) Inductively Coupled PlasmaMass Spectrometer. For isotopic analysis, samples were repeatedly
leached with HCl6N to remove secondary alteration. Sr and Nd were
separated using the method described in Weis et al. (2006). Isotope
ratios were measured on a Thermo Finnigan Triton-TI TIMS in static
mode with relay matrix rotation on single Ta filament and double Re–
Ta filament for Sr and Nd isotopic analyses respectively. Sr and Nd
isotopic compositions were corrected for fractionation using 86Sr/
88
Sr = 0.1194 and 146Nd/144Nd = 0.7219. During the course of the
analyses, the NBS987 Sr standard gave an average of 0.710259 ±
0.000021 (n = 10) and the La Jolla Nd standard gave an average
value of 0.511857 ± 0.000008 (n = 13). 147Sm/144Nd ratio errors are
approximately ∼ 1.5%, or ∼0.006.
5.3. Geochemical characteristics of the Río Verde Complex rocks and
comparisons
In the Nb/Y versus Zr/TiO2 plot (Fig. 7), the Lower Cretaceous igneous
rocks and metamorphic equivalents from throughout Hispaniola can be
geochemically grouped in two main clusters of mafic and felsic
compositions. Mafic rocks from Los Ranchos, Amina and Maimón
Formations, as well as Río Verde Complex, cluster mainly in the
subalkaline andesite/basalt field, with minor subalkaline basalt compositions, generally at Nb/Y < 0.1; felsic rocks plot along the boundary
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
175
Table 1
Major and trace element data of representative rocks from the Río Verde Complex, Amina Formation and mafic intrusives in the Loma Caribe Peridotite.
Unit
RVC
RVC
RVC
RVC
RVC
RVC
RVC
RVC
RVC
RVC
RVC
X (UTM)
388982
377250
377800
379148
379148
378169
377250
378174
393139
393139
393139
Y (UTM)
2070035
2080100
2077050
2078379
2078379
2077480
2080100
2077479
2064916
2064916
2064916
Rocka
AMPH
AMPH
MBAS
AMPH
AMPH
MBAS
AMPH
AMPH
DOL
AMPH
AMPH
Sample
6JE34D
2JE31
2JE33
2JE38
7JE38
2JE34
2JE31B
2JE35
2JE112
6JE113
2JE114
wt.%
SiO2
TiO2
Al2O3
Fe2O3
MgO
CaO
Na2O
K2O
P2O5
MnO
Cr2O3
LOI
Total
Mg#b
Cr
Co
Ni
V
Rb
Ba
Th
Nb
Ta
La
Ce
Pb
Pr
Sr
Nd
Sm
Zr
Hf
Eu
Gd
Tb
Dy
Y
Ho
Er
Tm
Yb
Lu
48.49
1.08
12.22
10.66
11.74
10.87
1.85
0.18
0.09
0.18
0.07
1.55
99.0
69
493
55
227
292
0.3
10
0.1
1.1
0.25
2.44
8.27
0.6
1.40
40.0
7.8
2.6
18.3
0.7
0.98
3.50
0.67
4.36
23.8
0.92
2.75
0.45
2.53
0.39
50.64
0.73
16.46
8.2
7.29
11.21
2.72
0.53
0.06
0.13
0.04
1.95
100.0
64
266
36
90
242
1.2
44
0.1
1.0
0.25
1.9
4.9
0.4
0.94
86.5
4.3
1.5
42.0
1.0
0.67
2.59
0.43
2.87
17.0
0.67
1.77
0.29
1.90
0.28
50.88
0.73
15.59
8.8
7.46
11.3
2.39
0.28
0.06
0.15
0.04
2.4
100.1
63
259
37
92
278
0.5
58
0.06
0.65
0.25
1.39
4.59
0.8
0.81
75.6
4.6
1.6
37.1
1.2
0.66
2.32
0.45
3.07
16.5
0.65
1.94
0.29
1.87
0.28
51.58
1.18
15.3
10.02
6.44
10.56
3.02
0.36
0.1
0.17
0.03
1.3
100.1
56
201
35
76
310
1.3
78
0.2
1.0
0.25
2.8
8.3
1.0
1.57
89.4
6.5
2.2
65.0
2.0
1.08
4.03
0.76
4.99
25.0
1.11
3.16
0.47
3.00
0.45
53.13
1.07
14.97
9.19
6.15
9.59
3.34
0.38
0.097
0.15
0.024
1.7
99.8
57
–
62.2
31
259
6.2
42
0.05
1.0
0.3
2.5
7.5
0.5
1.33
140.8
7.4
2.4
61.4
2.3
0.93
3.76
0.67
4.49
24.6
0.94
2.68
0.41
2.45
0.38
51.35
1.15
15.39
10.5
6.54
9.93
3.52
0.31
0.1
0.17
0.005
1.3
100.3
55
79
38
51
325
1.1
37
0.2
1.1
0.25
2.6
7.5
0.9
1.44
68.6
6.5
2.1
61.0
2.0
1.05
3.74
0.68
4.51
22.0
0.98
2.91
0.42
3.00
0.44
50.41
1.57
14.29
11.79
6.48
9.95
3.28
0.33
0.14
0.2
0.02
1.75
100.2
52
147
41
50
367
1.0
46
0.1
0.9
0.25
3.4
10.2
0.2
1.94
78.7
8.9
3.0
87.0
4.0
1.36
5.05
0.99
6.33
32.0
1.41
3.91
0.62
4.00
0.58
49.79
1.38
14.34
11.71
6.43
11.08
3.03
0.15
0.12
0.18
0.02
1.45
99.7
52
134
38
56
423
0.2
39
0.1
1.0
0.25
2.7
8.2
0.3
1.57
69.9
6.1
2.0
64.0
2.0
1.09
4.39
0.83
5.20
26.0
1.17
3.41
0.47
3.30
0.53
50.66
0.99
15.78
9.5
7.55
11.08
2.79
0.4
0.07
0.16
0.04
1.3
100.3
61
284
38
95
298
1.3
25
0.05
1.1
0.25
2.1
5.6
0.8
1.12
82.2
5.1
1.8
50.0
2.0
0.9
3.17
0.58
3.85
20.0
0.82
2.64
0.37
2.40
0.37
52.91
1.49
14.8
11.22
5.35
8.55
3.47
0.7
0.12
0.17
0.01
1.25
100.0
49
93
35
32
389
3.6
59
0.15
1.0
0.25
3.3
9.9
0.6
1.84
91.1
8.4
2.8
76.0
3.0
1.18
4.74
0.84
5.44
28.0
1.19
3.45
0.5
3.40
0.52
53.46
1.5
14.89
11.21
5.08
8.55
3.77
0.39
0.12
0.17
0.02
1.1
100.3
47
157
33
38
402
1.5
44
0.2
2.0
0.25
3.6
10.1
1.0
1.89
52.7
8.4
2.6
77.0
3.0
1.25
4.46
0.83
5.33
29.0
1.20
3.49
0.51
3.50
0.49
Unit
LCP
LCP
LCP
LCP
LCP
AmF
AmF
AmF
AmF
AmF
AmF
X (UTM)
360773
358500
358500
358500
335187
244368
242263
280317
243458
244368
276851
Y (UTM)
2090364
2090500
2090500
2090500
2120924
2161685
2163095
2149186
2164257
2161685
2146790
Rocka
GAB
GAB
GAB
GAB
DIQ
MSCH
MSCH
MSCH
MSCH
MSCH
MSCH
Sample
6JE92B
6JE93A
6JE93B
6JE94
6JE99
FC9068B
MJ9122
FC9110
MJ9049
MJ9068
FC9106
wt.%
SiO2
TiO2
Al2O3
Fe2O3
MgO
CaO
Na2O
K2O
P2O5
MnO
Cr2O3
LOI
Total
Mg#b
Cr
Co
44.36
1.48
15.45
11.25
7.87
11.26
3.04
0.11
0.14
0.18
0.024
4.7
99.9
58
164
58
51.01
1.37
15.45
11.11
5.74
8.73
4.58
0.28
0.11
0.18
0.006
1.4
100.0
51
41
52
50.07
1.94
14.93
12.82
5.25
8.12
4.57
0.31
0.17
0.2
0.007
1.6
100.0
45
48
58
49.48
1.0
15.97
9.61
8.15
9.75
2.97
0.28
0.04
0.16
0.046
2.3
99.8
63
315
42
51.3
0.85
14.63
9.62
8.49
8.56
3.72
0.41
0.08
0.17
0.047
2.0
99.9
64
322
48
47.89
1.05
16.89
9.04
8.24
11.21
2.82
0.17
0.14
0.15
0.031
2.1
99.6
64
212
45
50.58
0.95
16.17
10.47
5.08
8.84
2.61
1.73
0.31
0.17
0.005
2.7
99.6
49
34
34
48.1
0.4
18.45
9.15
4.44
13.1
0.74
0.05
0.04
0.15
0.015
5.2
99.8
49
103
35
49.85
0.92
17.38
11.85
4.72
5.98
5.06
0.09
0.1
0.21
0.001
3.4
99.6
44
7
36
55.88
0.74
16.75
8.92
3.11
5.33
5.44
0.38
0.08
0.15
0.005
2.7
99.5
41
34
20
52.9
0.17
12.44
6.47
7.26
9.08
5.17
0.36
0.02
0.16
0.051
5.6
99.7
69
349
31
(continued on next page)
176
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
Table 1 (continued)
Unit
RVC
RVC
RVC
RVC
RVC
RVC
RVC
RVC
RVC
RVC
RVC
X (UTM)
388982
377250
377800
379148
379148
378169
377250
378174
393139
393139
393139
Y (UTM)
2070035
2080100
2077050
2078379
2078379
2077480
2080100
2077479
2064916
2064916
2064916
Rocka
AMPH
AMPH
MBAS
AMPH
AMPH
MBAS
AMPH
AMPH
DOL
AMPH
AMPH
Sample
6JE34D
2JE31
2JE33
2JE38
7JE38
2JE34
2JE31B
2JE35
2JE112
6JE113
2JE114
wt.%
Ni
V
Rb
Ba
Th
Nb
Ta
La
Ce
Pb
Pr
Sr
Nd
Sm
Zr
Hf
Eu
Gd
Tb
Dy
Y
Ho
Er
Tm
Yb
Lu
24
326
1.3
9
0.1
1.9
0.2
3.3
11.0
0.2
1.93
143.3
10.5
3.5
95.1
2.9
1.24
4.62
0.98
5.84
33.0
1.16
3.76
0.56
3.18
0.50
271
359
2.5
70
0.1
1.6
0.3
3.4
8.8
0.5
1.57
154.1
8.6
3.0
74.6
2.4
1.12
4.24
0.86
5.16
31.4
1.06
3.41
0.52
3.04
0.48
103
456
2.6
35
0.1
2.6
0.3
4.7
13.3
0.4
2.31
140.3
12.9
4.2
114.8
3.6
1.42
5.81
1.19
7.00
40.6
1.41
4.39
0.68
4.06
0.65
817
258
3.9
28
0.05
1.6
0.3
2.9
7.8
0.2
1.5
155.0
8.4
2.7
63.6
2.0
0.97
4.03
0.78
4.66
31.5
0.92
2.94
0.43
2.55
0.38
29
300
5.1
22
0.2
1.0
0.1
2.9
5.3
0.2
1.04
131.8
5.2
1.9
43.4
1.6
0.68
2.63
0.55
3.40
19.8
0.67
2.23
0.34
2.07
0.33
56
193
4.6
75
0.3
2.0
0.1
4.0
11.1
0.4
1.73
270.5
9.9
3.0
84.4
2.4
1.24
3.91
0.74
4.24
26.2
0.88
2.73
0.4
2.48
0.41
15
294
39.3
677
3.2
3.5
0.2
16.4
33.8
0.9
4.09
508.7
19.7
4.4
82.7
2.6
1.3
4.39
0.82
4.16
23.9
0.81
2.33
0.35
2.23
0.35
12
261
<0.5
14
<0.1
<0.5
<0.1
0.6
2.1
0.3
0.36
26.1
2.4
0.9
11.9
0.5
0.42
1.69
0.3
1.78
11.8
0.41
1.28
0.20
1.32
0.19
10
362
0.5
27
0.4
0.7
<0.1
2.4
6.5
0.2
0.99
100.9
5.9
1.8
37.0
1.3
0.74
2.83
0.51
2.97
17.9
0.69
1.9
0.27
1.68
0.28
7
313
3.9
125
0.3
0.8
< 0.1
2.9
8.4
0.2
1.28
143.7
7.3
2.1
54.7
2.0
0.75
3.13
0.71
3.64
23.2
0.79
2.46
0.35
2.29
0.37
36
153
5.1
46
< 0.1
< 0.5
< 0.1
0.6
1.1
0.2
0.2
114.7
1.1
0.5
13.1
< 0.5
0.19
0.56
0.11
0.64
3.6
0.13
0.36
0.06
0.65
0.09
Major oxides recalculated to an anhydrous basis. Total Fe as Fe2O3.
Unit: RVC, Río Verde Complex; LCP, Loma Caribe Peridotite; AmF, Amina Formation.
a
Rock type abbreviations: DOL, dolerite; GAB, gabbro; MBAS, metabasalt; AMPH. Amphibolite; DIQ, mafic dyke; MSCH, mafic schist.
b
Mg# = 100 ⁎ mol MgO/ mol (FeO + MgO); for Fe2O3/FeO = 0.2.
between andesite and rhyodacite fields, also at low Nb/Y ratios. This
bimodal association is characteristic in all PIA units in the Greater
Antilles as well as in other Cenozoic arcs, including Izu-Bonin, the
Kermadecs, and South Sandwich (see discussion in Jolly and Lidiak,
2006). As a suite, the metabasalts, dolerites and amphibolites of the Río
Verde Complex have a restricted range in SiO2 content, ranging from
48.5 to 52.8 wt.% (Table 1), for TiO2 contents between 0.7 and 1.5 wt.%.
These rocks show an increase of SiO2, Fe2O3T, alkalis, TiO2, Zr and Nb, and
a decrease in Cr and Ni for decreasing MgO (not shown). Al2O3 and CaO
increase slightly to reach a maximum at about 6 wt.% MgO, then
decrease in the evolved basalts. These trends are tholeiitic and can be
attributed to the fractionation of olivine plus Cr-spinel, plagioclase and
clinopyroxene, which is compatible with the observed mineralogy and
relics in the amphibolites. In the Th–Co discrimination diagram of Hastie
et al. (2007), the mafic rocks of the Río Verde Complex plot as tholeiitic
basalts and basaltic andesites.
Comparison of Río Verde Complex mafic rocks with contemporaneous rocks of the Los Ranchos, Amina and Maimón Formations, as
well as mafic dykes and sills intrusive in the Loma Caribe Peridotite, is
made through TiO2 content (Fig. 8a). Ti is a conservative element in
Table 2
Sr–Nd isotope ratios for representative samples of Río Verde Complex and Amina Formation.
Group
Rock type
Sr
87
Sr/86Sr
Muestra
Rb
Río Verde Complex
BABB
Amphibolite
BABB
Amphibolite
BABB
Metabasalt
BABB
Metabasalt
BABB
Amphibolite
BABB
Amphibolite
BABB
Dolerite
BABB
Amphibolite
BABB
Amphibolite
BABB
Amphibolite
2JE31
2JE31B
2JE33
2JE34
2JE35
2JE38
2JE112
2JE114
6JE34B
6JE34D
1.2
1.0
0.5
1.07
0.23
1.3
1.3
1.5
3.6
0.26
86
79
76
69
70
89
82
53
91
40
0.703468
0.703247
0.703298
0.703457
0.703001
0.703583
0.703187
0.703624
0.703764
0.703238
Amina Formation
I
Act + Chl schist
I
Act + Chl schist
II
Act + Chl schist
III
Act + Chl + Ep schist
MJ9068
MJ9049
FC9110
FC9106
3.9
0.5
0.3
5.1
143.7
100.9
26.1
114.7
0.705098
0.705057
0.704401
0.704973
(87Sr/86Sr)i
Sm
Nd
143
Nd/144Nd
(9)
(9)
(8)
(12)
(8)
(10)
(9)
(9)
(9)
(11)
0.703403
0.703185
0.703268
0.703383
0.702985
0.703514
0.703115
0.703493
0.703575
0.703207
1.45
3.01
1.2
2.13
2.03
2.17
1.77
2.6
2.75
2.25
4.3
8.9
3.4
6.5
6.06
6.5
5.1
8.4
8.4
6.5
0.513147
0.513161
0.513145
0.513165
0.513174
0.513153
0.513165
0.513141
0.513152
0.513163
(8)
(8)
(9)
(6)
0.704969
0.705033
0.704355
0.704763
2.1
1.8
0.9
0.5
7.3
5.9
2.4
1.10
0.513107
0.513101
0.513139
0.513072
(143Nd/144Nd)i
(εNd)i
(6)
(6)
(6)
(8)
(8)
(6)
(6)
(6)
(6)
(6)
0.512993
0.513007
0.512987
0.513016
0.513022
0.513001
0.513006
0.512999
0.513004
0.513005
9.81
10.09
9.70
10.27
10.37
9.97
10.07
9.94
10.02
10.05
(5)
(5)
(11)
(6)
0.512976
0.512962
0.512968
0.512865
9.49
9.21
9.33
7.32
Calculated initial ratios (i) and εNd-values calculated at t = 115 Ma. Number in brackets is the absolute 2σ error in the last decimal places.
εNd values are relative to 143Nd/144Nd = 0.512638 and 147Sm/144Nd = 0.1966 for present day CHUR (e.g. Weis et al., 2005, 2006) and lambda 147Sm = 6.54 × 10− 12/year.
Amina Formation groups: I, IAT; II, low-Ti IAT; and III, boninites.
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
177
Fig. 7. Nb/Y versus Zr/TiO2 diagram (Winchester and Floyd, 1977) for the diverse Lower Cretaceous igneous rocks in Hispaniola.
Fig. 8. Plot TiO2 versus MgO for the diverse geochemical groups of Lower Cretaceous
igneous rocks in Hispaniola. NVTZ, CG and SR fields are Northern Volcano–Tectonic
Zone, Central Graben and Spreading ridge fields of the Mariana Arc–Trough system
from Gribble et al. (1998), which are shown for comparisons with a modern analog. Los
Ranchos and Amina Formation data are from Escuder-Viruete et al. (2006) and this
work. The Maimón Formation field includes data from Lewis et al. (2000, 2002). Also
indicated are 5% fractional crystallization vectors for olivine (Ol), clinopyroxene (Cpx),
and plagioclase (Pl), determined from the average Río Verde Complex composition.
the subduction environment (Pearce and Peate, 1995), and the TiO2
concentrations should therefore provide an indicator of the extent of
source depletion, taking into account the Fe–Ti oxides fractionation.
Several points are shown in this figure. (1) The trend of increasing
TiO2 with decreasing MgO in the Río Verde Complex samples is not
seen in the PIA suite, which shows a low-Ti trend. (2) The mafic sills
and dykes of the Loma Caribe Peridotite have similar TiO2 contents to
the Río Verde Complex samples and both are significantly TiO2
enriched relative to the PIA suite, which show a progressive TiO2
increase from the boninites, to low-Ti IAT to normal IAT. (3) The Río
Verde Complex basalts, dolerites and amphibolites have similar TiO2
contents to the basalts and basaltic andesites from the Central Graven
and Spreading Ridge of the Northern Mariana Trough (Gribble et al.,
1998), but are less fractionated than the Northern Volcano-Tectonic
Zone of the rifted Mariana Arc. Therefore, the diverse TiO2 content in
the PIA suite and Río Verde Complex magmas suggests different
Caribbean mantle sources.
Comparisons are also made through patterns in normal mid-ocean
ridge basalt (N-MORB) normalized trace elements diagrams (Fig. 9),
which are all characterized by significant enrichment in LILE (Rb, Ba,
Th, U, Pb and K) and LREE relative to the HFSE (Nb, Ta, Zr, Hf, Ti and Y)
and HREE. Río Verde Complex samples display a slight LREE depletion
or enrichment and a flat HREE pattern. The obtained values in the
primitive mantle normalized ratios (La/Nd)N = 0.6–0.9 and (Sm/
Yb)N = 0.98–1.1 are characteristic of N-MORB (e.g. Su and Langmuir,
2003). Relative to N-MORB, however, these rocks have Nb–Ta
negative anomalies and higher abundances of LILE such as Rb, Ba, K
and Pb. Such anomalies in intra-oceanic settings are widely interpreted to reflect supra-subduction zone magmatism, involving mantle
wedge sources that have been contaminated by mass transfer (melts
or fluids) from the subducting slab (Pearce and Peate, 1995). In the
arc-related Los Ranchos Formation, these aforementioned geochemical signatures increase from the boninites through the low-Ti IATs to
the IAT (Escuder-Viruete et al., 2006). The IAT have almost flat (NMORB-like) HFSE profiles, whereas the boninites show the greatest
degrees of depletion of these elements. In the Río Verde Complex
metabasalts and amphibolites, the trace element patterns are subparallel to those of the IAT of the Amina and Los Ranchos Formations,
although these volcanics are more enriched in the subduction mobile
elements Th, LILE and LREE (Fig. 9b–c). A weak subduction signature
is also established by Nb/Th ratio values of 5–11.5 in Fig. 8b, whereas
178
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
Fig. 9. MORB-normalized multi-element plots for: (a) amphibolites of the Río Verde Complex; (b) basalts and dolerites of the Río Verde Complex and metavolcanic rocks of Amina
Formation; (c) volcanic rocks of the Los Ranchos Formation (Escuder-Viruete et al., 2006); and (d) mafic rocks intruded in the Loma Caribe Peridotite (Table 1). MORB-normalizing
values are from Sun and McDonough (1989).
PIA suite samples generally present Nb/Th < 5 values more typical
of arc-related rocks (Swinden et al., 1997). By their transitional IAT to
N-MORB geochemistry and weak subduction-related signature, we
interpret the mafic rocks of the Río Verde Complex to form in a rifted
arc or evolving back-arc basin setting.
The mafic sills and dykes intruded in the Loma Caribe Peridotite
have a range of 41.9–51.0 wt.% SiO2 for 8.1–5.5 wt.% MgO (Table 1)
and cluster with the Río Verde Complex rocks in the subalkaline
andesite/basalt field of Fig. 7. These rocks also show a tholeiitic trend
of increasing of SiO2, Fe2O3T, TiO2, Nb and Zr for decreasing MgO (not
shown). They define a mid-Ti trend with Río Verde Complex samples
in Fig. 8, but the evolved rocks are slightly more Ti-rich (TiO2 = 0.7–
1.9 wt.%). These plutonic rocks also display sub-horizontal multielement patterns similar to N-MORB (Fig. 9d); with a slight LREEdepletion ([La/Nd]N = 0.6–1.1) and flat HREE ([Sm/Yb]N = 1.0–1.2).
Moreover, they have small enrichments in some subduction-mobile
elements (Rb, Ba and K), slight depletions in Nb–Ta and no significant
negative Zr–Hf anomalies ([Zr/Sm]N = 0.9–1.1). All these characteristics are typical of back-arc basin basalts (BABB; e.g. Hawkins, 1995).
These features also suggest that the mantle source for these rocks was
similar to both those of the Río Verde Complex protoliths and to NMORB source (i.e. depleted mantle).
New and published Sr and Nd isotope ratios for Río Verde
Complex, Los Ranchos and Amina Formations are listed in Table 2
and plotted in Fig. 10. Initial (87Sr/86Sr)i versus (143Nd/144Nd)i
variation in the metabasalts, dolerites and amphibolites of the Río
Verde Complex are restricted to high (εNd)i values between + 9.7 and
+10.3 (i = 115 Ma), similar to N-MORB (Su and Langmuir, 2003).
(87Sr/86Sr)i ratios vary between 0.70299 and 0.70358, and values are
located near the MORB array, which probably reflects the source
composition little modified by subsolidus, hydrothermal alteration.
The high (εNd)i values of Río Verde Complex samples are compatible
with a homogeneous source dominated by depleted mantle, similar to
DMM composition (Su and Langmuir, 2003), with minimal incorpo-
ration of a sedimentary component. Initial Sr–Nd isotope variations in
Los Ranchos and Amina Formations (Table 2) also display a horizontal
trend and are also restricted to high (εNd)i values between +8.0 and
+10.6 (except + 7.32 for FC9106 boninite). (87Sr/86Sr)i ratios are
highly variable (0.702966–0.705684), similar to altered rocks in
modern intraoceanic arcs, and consistent with seawater alteration
that shift the samples from the MORB array to the right (e.g. Pearce
et al., 1995). In the PIA suite, (87Sr/86Sr)i values tend to decrease from
boninites through to low-Ti IAT and IAT.
5.4. Discussion
5.4.1. Trace element ratio variations and implications
Differences between island-arc related Los Ranchos, Amina and
Maimón Formations and BABB-like Río Verde Complex rocks are also
seen in incompatible element plots, in which the relative contribution
of source composition and subduction component can be evaluated.
The inference is that Zr/Yb and Nb/Yb ratios are little or unaffected by
additions of components during subduction, whereas increase in LILE/
HREE ratios, for example Th/Yb, reflects addition of slab-derived
components (Pearce et al., 1995). Following the approach of Pearce
and Parkinson (1993), a MORB-OIB array of increasing Zr/Yb and Th/
Yb with increasing Nb/Yb, is defined by the subduction-unmodified
lavas from the East Pacific Ridge (Fig. 11; data from Su and Langmuir,
2003; PETDB, 2007; and references herein), considered on the basis of
trace element/isotopic fingerprinting, onset of SW-dipping subduction (Krebs et al., 2007) and plate reconstruction models (Pindell
et al., 2005), to belong to the same Pacific mantle domain. The regional
MORB-OIB array is completed by samples from the CCOP (Lapierre
et al., 1997; Hauff et al., 2000; Lapierre et al., 2000; Révillon et al.,
2000; Kerr et al., 2002), which generally have higher Nb/Yb ratios
than average N-MORB, and suggest the influence of a Late Cretaceous
Caribbean plume in the source enrichment. The Caribbean islandarc trend is represented by the volcanic rocks of Puerto Rico, which
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
179
(vector A; Pearce et al., 1995). PIA suite samples follow this vertical
trend. Fig. 11c shows that subduction vector A extends vertically from
the Caribbean MORB-OIB array, with the subduction contribution
estimated by contour lines drawn parallel to the array. Fig. 10d reveals
that the subduction contributions for Th range up to 90% for Los
Ranchos, Amina and Maimón Formations, being generally lower
(<50%) for Río Verde Complex and indicating a lower subduction
input. Addition of a subduction component followed by variable
degrees of melting, such as dynamic melting beneath the arc, gives a
trend parallel to but displaced from the MORB-OIB array (vector C;
Pearce et al., 1995). The trends formed by PIA-related samples run subparallel to the regional MORB-OIB array for Th (Fig. 11d). Therefore,
the rocks of the Los Ranchos, Amina and Maimón Formations result
from a combination of variable subduction components added to a
variably depleted mantle wedge composition, which was particularly
depleted in terms of Zr/Yb values for low-Ti IAT and boninites. The Río
Verde Complex samples plot close to the Zr/Yb ratio of mean N-MORB,
within the depleted part of the array. Some samples extend to slightly
higher contents of Th indicating a small subduction input. Therefore,
the Río Verde Complex samples are interpreted as being derived from
depleted mantle with minor subduction component addition. The
relative contributions of both components is quantitatively modeled
below.
Fig. 10. (a) Initial Sr–Nd isotopes ratios (i = 115 Ma) for the different geochemical
groups of Lower Cretaceous igneous rocks in Hispaniola. The fields for the Duarte
Complex, the CCOP (except Gorgona) and Caribbean island-arc lavas from Northeastern
and Central Puerto Rico, are taken from Escuder-Viruete et al. (2007a), Hastie et al.
(2008; and references herein), Hauff et al. (2000), Jolly et al. (2001, 2006, 2007) and
Thompson et al. (2004). The MORB-OIB array is defined by the subduction-unmodified
lavas from the East Pacific Ridge (data from PETDB, 2007; and references herein).
Depleted MORB mantle (DMM) Sr–Nd isotopic compositions are taken from Su and
Langmuir (2003): DMM average for MORBs far from plumes; D-DMM is 2σ depleted
and E-DMM is 2σ enriched over the average. CAM is Cretaceous Atlantic MORB (Janney
and Castillo, 2001). Fields and mantle components are not age corrected to 115 Ma. (b)
Caribbean island-arc field is subparallel to a calculated mixing line between pelagic
sediments and representative arc basalt taken from Jolly et al. (2001). The Sr–Nd
isotopic data for Río Verde Complex, Amina and Los Ranchos Formations suggest a
minor subducted sedimentary component (< 0.5%; data from Escuder-Viruete et al.,
2006; and this work).
constitute a complete record of subduction-related volcanism in the
area, spanning >70 Ma from Aptian to the Eocene (data from Jolly
et al., 2001, 2006, 2007). The Caribbean island-arc lavas are displaced
from the MORB-OIB array to higher concentrations of the subductionmobile element Th.
In the Zr/Yb versus Nb/Yb plot (Fig. 11), samples of Los Ranchos,
Amina and Maimón Formations are collectively located near average
N-MORB and extend along the depleted part of the Caribbean islandarc trend. Therefore, for these rocks both Zr and Nb are not present in
significant concentrations in the subduction component (Pearce et al.,
1995). The plot shows that samples from the Río Verde Complex are
generally similar to the PIA suite in that they also have low Zr/Yb and
Nb/Yb ratios, but they have a more restricted composition. The mantle
source for all units is variably depleted relative to average N-MORB
and interpreted to have experienced previous partial melt extraction,
and hence depletion in incompatible elements (Pearce et al., 1995).
Their low Nb/Yb ratios (and higher [εNd]i values) discard the influence
of a Caribbean plume component in the Lower Cretaceous Caribbean
island-arc–back-arc system, which displace CCOP samples from the
MORB-OIB array to higher Nb/Yb values in an opposite sense to vector
B in Fig. 11d.
Addition of a variable subduction component to a mantle source
of constant composition results in a vertical trend on the diagrams of
the Fig. 11, as Th is non-conservative and Nb and Yb are conservative
5.4.2. Sr–Nd isotope variations and implications
Sr–Nd isotope variations permit similar petrogenetic interpretations. In the (87Sr/86Sr)i versus (εNd)i diagram of Fig. 10b, the MORBOIB array defined by lavas from the East Pacific Ridge gives a welldefined mantle trend, which extends from a depleted DMM composition towards that of enriched DMM (data from PETDB, 2007; and
references herein). Caribbean island-arc volcanism is represented by a
diagonal field that includes lavas from Northeastern and Central
Puerto Rico (Jolly et al., 2001, 2006). The field is subparallel to a
calculated mixing line between sediments and a representative arc
basalt, and reflect a wide variation in mantle sources and proportions
of pelagic sediment subducted beneath the arc from the Albian to
Maastrichtian (Jolly et al., 2007). In this figure, however, isotopic data
for Los Ranchos and Amina Formations, as well as the Río Verde
Complex, cluster at high (εNd)i values and are together indicative of
minimal incorporation of subducted sediments. Additionally, these
high (εNd)i values of all samples are also inconsistent with the role of an
enriched Caribbean plume component in their genesis, at least in the
Lower Cretaceous and in Hispaniola. In the Sr–Nd isotope variation
diagram, additions of this component to the mantle source would
displace the samples from the MORB array toward the CCOP (data from
Thompson et al., 2004; Hastie et al., 2008; and references herein) or the
Duarte Complex fields, which has been interpreted as a CCOP unit in
the Lower Cretaceous (Escuder-Viruete et al., 2007c).
6. Petrogenesis and comparisons
6.1. Modeling of mantle melting and magma sources
6.1.1. Introduction
To evaluate the nature of the mantle source region that gave rise to
the arc magmas, as well as the general subduction flux on arc melt
geochemistry, selected compositions for each geochemical group of the
Caribbean island-arc–back-arc system were compared with the pooled
fractional melting calculations for various possible mantle sources. These
sources were modeled by Pearce and Parkinson (1993) and Ewart et al.
(1998), and start up from spinel lherzolite corresponding to fertile MORB
mantle (FMM). In the calculations, we used here non-modal fractional
melting and rates of phase melting following Pearce and Parkinson
(1993), i.e. clinopyroxene disappear at 25% and orthopyroxene at 40%
melting. Details of the source mineralogy and partition coefficients
(Brenan et al., 1995) are given in Appendix 4. Progressive source
180
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
Fig. 11. Plots of (a, b) Zr/Yb and (c, d) Th/Yb versus Nb/Yb for the Lower Cretaceous igneous rocks in Hispaniola. The Caribbean island-arc trend is represented by the Aptian to Eocene
volcanic rocks of Puerto Rico (data from Jolly et al., 2001, 2006, 2007). Caribbean MORB-OIB array is defined by the subduction-unmodified lavas from the East Pacific Ridge (data from
Su and Langmuir, 2003, PETDB, 2007; and references herein) and completed by samples from the Late Cretaceous Caribbean–Colombian oceanic plateau (Hauff et al., 2000; Kerr et al.,
1997, 2002; Sinton et al., 1998; Lapierre et al., 1999, 2000). The Caribbean plume enriched component probably increases in magnitude with proximity to the plume. N-MORB, E-MORB
and OIB values are from Sun and McDonough (1989). In the plot, there are three principal types of trend (vector), described in detail by Pearce et al. (1995): A = variable subduction
component; B = variably enriched mantle wedge; C = variable melt extraction. See text for explanation.
depletion is calculated after 5, 10, 15, 20, and 25% prior fractional melting
“events” (referred to as RMM5 to RMM25, respectively, where RMM is
residual MORB mantle), with appropriate adjustment of the phase
mineralogy during the calculated melting steps; 1% porosity is assumed.
Trace element compositions of the melts produced, by further fractional
melting and pooling of melts of the RMM5 to RMM25 source
compositions, are then calculated for 5, 10, 15 and 25% remelting (i.e.
second stage melts), believed to simulate the range of Caribbean
subduction system melt zones.
To model the possible effects of subduction input, the composition of
a subduction-derived “H2O-rich component” following Stolper and
Newman (1994) is used, due to the uncertainty in the composition of
most fluid-mobile elements in the original Caribbean magmas. This
implies modifying the RMM5 to RMM25 source compositions by
addition of 0.001, 0.005, and 0.01 weight fractions of this H2O-rich
component. The second stage melting calculations are then repeated for
5 and 15% melting, as before. The resulting curves (Fig. 12) for
subduction modified source compositions are believed to approximate
the effects of remelting of mantle wedge which has been modified by
prior, variable melt extraction (i.e. source depletion) and then selective
subduction-derived enrichment. To minimize the effects of fractional
crystallization and crystal accumulation, a MgO value of 9 wt.% was
chosen for parental magmas and the theoretical trace element content
of each sample at this MgO value was calculated following the
methodology of Pearce and Parkinson (1993). The modeled melt
compositions for different percentage of partial melting, different source
compositions, and different subduction enrichment are plotted in
Fig. 12. These plots allow us to distinguish more conservative elements
(Y and Ti) from non-conservative elements (Sm and Ba), in which Yb is
treated as fully conservative (Pearce and Parkinson, 1993).
6.1.2. Source characterization
Comparison of the Caribbean island-arc–back-arc system compositions, with the calculated melting curves suggests that concentrations of Ti, Y and Yb depend on the degree of source depletion and are
little modified by subduction input (Fig. 12a–b). This allows
evaluating the source of the three main Caribbean magma groups:
BABB of the Río Verde Complex and mafic sills intruded in the Loma
Caribe Peridotite; IAT; and low-Ti IAT and boninites of the Los
Ranchos, Amina, and Maimón Formations. BABB group of magmas
(Yb9 = 1.5–3.2; Ti9 = 3700–7200; Y9 = 15–30; Sm9 = 1.2–2.8; Zr/
Ba = 0.6–2.6) requires a depleted shallow spinel lherzolite mantle
(FMM) affected by <7% of previous melt extraction and by relative
low degrees of remelting (1–10%). IAT group of magmas (Yb9 = 1.0–
2.8; Ti9 = 2900–4400; Y9 = 10–25; Sm9 = 1.0–1.5; Zr/Ba = 0.12–0.56)
are consistent with a spinel lherzolite source affected by ∼5–15% of
prior melt extraction and low to moderate degrees of remelting (5–
10%). This modeled source is slightly depleted than those of the BABB
group. Group of low-Ti IAT and boninites (Yb9 = 0.8–1.4; Ti9 = 1015–
3200; Y9 = 2.8–12; Sm9 = 0.45–0.9; Zr/Ba = 0.1–1.6) can be modeled
by moderate to high degrees of remelting (10–22%) of a source
previously depleted by 10–25% of extracted melt. Therefore, this
mantle source is more depleted and underwent higher degrees of
remelting than the sources of groups BABB and IAT.
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
181
Fig. 12. Calculated melt composition for 1, 5, 10, 15 and 25% remelting of previously depleted model mantle wedge compositions (corresponding to FMM to RMM25 compositions, as
defined in the text; numbers along curves are percentage of prior melt extraction) for: Yb versus (a) Y, (b) Ti (both unaffected by subduction), (c) Sm (minor subduction effect), and
(d) Zr/Ba (high subduction effect). These curves are compared, in each plot, with the normalized (MgO = 9%) data, from selected compositions for Los Ranchos, Amina and Maimón
Formations, Río Verde Complex, and mafic dykes and sills. In (c) and (d), calculated effects of addition of 0.001 wt fractions of “H2O-rich component” to the variously depleted model
mantle wedge sources, followed by 5 and 15% remelting, are shown (discontinuous curves). FMM and RMM are fertile and residual MORB mantle, respectively. See text for
explanation.
An increase in subduction component is suggested when variably
non-conservative elements are plotted against each other. Fig. 12c–d
show varying degrees of departure from the curves of the calculated
melts. Sm is an example of an element that shows moderately
different values to the calculated melts whereas the trend for Ba
(expressed as Zr/Ba) departs markedly from the calculated depletion
trends (Fig. 12c–d). In all cases, however, these diverging trends can
be broadly matched with the calculated melt trends from the
“subduction modified” sources containing ∼ 0.001 weight fraction of
the modeled H2O-rich component. Comparison of the Caribbean
island-arc–back-arc system compositions, with the calculated melting
curves suggests a higher subduction component in IAT and low-Ti IAT
and boninites groups than in BABB group, particularly for Sm. As Ba is
mobile in aqueous fluids (Elliott et al., 1997), the generally lower
values of the Zr/Ba ratio obtained for the PIA group also suggest a
higher enrichment in fluid-mobile elements than in BABB group.
However, the results obtained for the Ba should be considered as
qualitative, since this element could be also mobilized during the late
alteration and metamorphism.
6.1.3. Isotope-trace element results
To investigate possible balances between differing subduction inputs
(slab derived components) and differing degrees of mantle source
depletion, Nd isotopic data have also been modeled. Two endmembers
were chosen (Appendix 3): the mantle wedge, based on the average
isotopic compositions of AEPR (average Eastern Pacific Ridge MORB, or
mantle wedge unmodified by subduction components); and the
modeled subduction flux composition (SF). Abundance of Nd follow
the Stolper and Newman, (1994) estimates for the “H2O-rich component” (as used above), and weight fractions of 0.001, 0.005, and 0.01 of
this component are again added to the modeled mantle wedge sources
(FMM, RMM5 to RMM15), and second stage melt compositions are
calculated. Results are shown in Fig. 13, in which calculated Nd isotope
ratios are plotted against Nd and Ti. These trends are compared with the
normalized element abundances (MgO = 9%) and Nd isotopic composition of different geochemical groups. Each plot shows the effects of
progressive subduction input on the second stage melts derived from
each variably depleted mantle source composition, and the dashed lines
mark the combined trends of variable source depletion with superimposed constant subduction input. The compositionally diverse groups
derived from different mantle sources can be illustrated by Ti, least
affected by subduction input (Fig. 13b). As previously deduced from
trace element modeling, the source of BABB group magmas was relative
enriched compared to sources of the PIA groups, particularly for low-Ti
IAT and boninites. The subduction inputs can be evaluated in Fig. 13a,
where observed melt trends follow a similar H2O-rich component
between 0.001 and 0.005 weight fractions, but where abundances still
reflect variations in source depletion composition. Melt modeling
indicates that subduction input in BABB group were slightly lower
than in PIA groups, particularly for IAT. As the LREE (and Th) are often
enriched in clastic sediments but are not highly mobile in fluids (Stalder
et al., 1998), a minor sediment addition in the low-Ti IAT and boninite
groups than in the IAT group is suggested by their lower Nd contents and
Nd isotopic ratio values. A lower sediment addition is also shown by the
decrease in the Th/Yb ratio values from IAT and low-Ti IAT and boninites
to BABB groups (Fig. 11d).
182
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
Fig. 13. Calculated isotopic and trace element melt compositions for 15% remelting of
previously depleted model mantle wedge compositions (FMM to RMM15), assuming
zero, 0.001, 0.005, and 0.01 weight fraction of a “H2O-rich component” added to these
sources. The continuous lines show increasing weight fractions of this component to
the source, up to 0.01 from each given starting model mantle composition (FMM to
RMM15). The discontinuous curves mark combined source depletion plus constant
subduction input to source trends. The calculated curves are compared with normalized
(MgO = 9%) data for Los Ranchos, Amina and Maimón Formations, and Río Verde
Complex. In all cases, the altered oceanic crust isotopic endmember is used. See text for
explanation.
In summary, the combined trace element and trace elementisotope modeling are consistent, and suggest that the coupled
processes of mantle wedge depletion with input of a subduction
component (modeled between 0.001 and 0.005 weight fraction
abundances of a H2O-rich component) can explain the geochemistry
of the Caribbean island-arc–back-arc system magmas. The correlations between geochemical source depletion indicators (e.g. Yb, Y, Ti),
together with the observed Nd isotopic composition variations along
each geochemical groups, are inferred to result from a progressively
lower subduction flux into a mantle wedge which is progressively less
depleted from the arc to the back-arc units. Also, Nd-isotope and
incompatible trace element patterns are diverse in the mafic rocks
groups, but all are consistent with sources unrelated to a Lower
Cretaceous Caribbean mantle plume. Both can be considered as two
significant conclusions that have important tectonic implications.
6.2. Tectonic implications
The Los Ranchos Formation has been proposed as a record of the
volcanic activity in the suprasubduction zone setting of the primitive
Caribbean island-arc (Escuder-Viruete et al., 2006). In this tectonic
context, the data presented in this work imply: (1) Los Ranchos, Amina
and Maimón Formations are petrological and geochemical equivalents, form part of the same volcanic front, and include similar
boninites, IAT and felsic volcanic rocks; (2) the protoliths of the Río
Verde Complex, exclusively of BABB-like mafic composition, were
extruded/intruded during arc rifting and the early stages of a back-arc
basin development; (3) the Río Verde Complex was deformed by a
heterogeneous syn-metamorphic shearing at ∼ 110 Ma; (4) the latest
batches of the BABB-like magmas are syn- to late-kinematic in relation
to this low-P/low to middle-T deformation; (5) the spatial distribution
of arc and back-arc related geological units in Hispaniola, in actual NE
and SW positions, respectively, indicates a SW-directed subduction
polarity in the Lower Cretaceous; (6) a same subduction polarity can
be deduced from the progressively lower subduction flux into a
progressively less depleted mantle source of the PIA suite and the Río
Verde Complex; and (7) the shallow limestones of the Hatillo
Formation were deposited in the upper Lower Albian on top of the
eroded arc.
In order to explain these observations, a model of proto-Caribbean
oceanic lithosphere subducted at least in the 120–110 Ma interval, is
proposed as the cause of tectonic and magmatic variations in the
Lower Cretaceous Caribbean island-arc–back-arc system (Fig. 14),
which is supported by the onset of SW-dipping subduction in
northern Hispaniola and in eastern Cuba at ∼120 Ma (Krebs et al.,
2007; Lázaro et al., 2008). In this context, arc rifting and sea-floor
spreading to form the Río Verde Complex protoliths occurred in the
back-arc setting of this NE-facing primitive island-arc, built onto the
NE edge of the Caribbean plate. This spatial configuration explains, on
the one hand, the existence of a back-arc area not affected by slabderived geochemical components and, on the other hand, it precludes
the presence of a Caribbean plume component in the petrogenesis of
the PIA magmas. However, this plume component is present in the
picrites and high-Mg basalts of the Lower Cretaceous Duarte Complex
(Escuder-Viruete et al., 2007c), probably advected by lateral mantle
flow of the CCOP source. Therefore, the Caribbean island-arc–back-arc
system includes three different magma sources related to three
different mantle domains. From the volcanic front toward the backarc (Fig. 14), these melt source regions are a suprasubduction mantle
wedge, a back-arc spreading centre, and a deep mantle containing
garnet influenced by an enriched plume. These mantle domains were
originally separated by an undetermined distance and their structural
juxtaposition took place later, during the closure of the back-arc basin,
probably in the Middle Eocene arc–continent collision.
Finally, it remains to explain the relations between the Río Verde
Complex and the BABB-like magmas intruded/extruded in Central
Hispaniola during the Late Cretaceous. Based on the Turonian–
Campanian volcanic history and the geochemical composition of their
constituent igneous rocks, the tectonic blocks that made up Central
Hispaniola have been recently interpreted as remnants of extended
island-arc and oceanic plateau, transitional and oceanic crust, which
formed part of a Loma Caribe back-arc basin (Escuder-Viruete et al.,
2008). The data presented in this work suggest that the back-arc
spreading system that formed the Río Verde Complex protoliths at
∼120–110 Ma was not sufficiently separated from the volcanic front,
and it was still affected by slab-derived geochemical components. After
an interval of arc inactivity and erosion in the upper Lower Albian, the
ridge system propagated toward the NW into the actual Central
Cordillera, rifting at ∼90 Ma the Albian to Turonian arc and opening
during the Santonian to Lower Campanian the Loma Caribe back-arc
basin. Later, the Central Hispaniola terrain was tectonically juxtaposed
with the Lower Cretaceous arc by arc-parallel, large-scale sinistral
strike-slip shearing along the Hato Viejo and Hispaniola fault zones.
7. Conclusions
Variations of trace elements parameters (Nb/Yb, Th/Yb, Zr/Yb, Zr/
Ba, and normalized Ti, Sm, Y and Yb abundances) and Nd isotopic
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
183
Fig. 14. Schematic tectonomagmatic model for Upper Aptian–Lower Albian Caribbean island-arc–back arc system based on the spatial distribution of igneous rocks in Hispaniola. The
mantle flow convective regimes beneath rifted arcs and evolving back-arc basins are inspired in Gribble et al. (1998) and Taylor and Martinez (2003). The SW-directed motion of the
subducting proto-Caribbean slab drives corner flow advection in the mantle wedge. Water released by the downgoing slab promotes partial melting in the mantle above the solidus
(heavy dashed lines), which is progressively depleted of a melt component toward the volcanic front. Melts rises and gave rise to extrusion of initially boninites and low-Ti IAT and
subsequently normal IAT (Escuder-Viruete et al., 2006). When arc extension commences, the lithosphere rifts near the rheologically weak volcanic front. Hydrated mantle is
advected upward into the stretching and thinning lithosphere, leading to high degrees of melting in the rift phase, and promotes the lower arc crust melting and development of
felsic volcanism and tonalitic plutonism. With increasing extension a seafloor spreading centre is established near the volcanic front advecting highly hydrated mantle. As
consequence, BABB-like Río Verde Complex magmas result in a SW position respect the volcanic front. Melts derived from a deeper Caribbean plume enriched source are
incorporated by lateral flow from the SW and gave rise to the OIB-like off-ridge magmatism of the Duarte Complex, in the back-arc area located SW of the spreading system, which is
not affected by slab-derived geochemical components. The Loma La Monja volcano-plutonic assemblage represents a dismembered fragment of the Late Jurassic Pacific-derived
oceanic crust, in which Duarte Complex melts were intruded. The migration of this propagating back-arc rift system toward the NW produced arc-rifting and back-arc basin
development from ∼90 Ma in the Central Hispaniola domain (Escuder-Viruete et al., 2008). CCOP, Caribbean–Colombian oceanic plateau; SC, Septentrional Cordillera; EC, Eastern
Cordillera; CC, Central Cordillera in Hispaniola. Age of eclogitic metamorphism in the Río San Juan high-P complex is from Krebs et al. (2007). See text for further explanation.
compositions are observed in the Aptian to Lower Albian mafic
igneous rocks throughout Hispaniola. These variations are systematic
and establish gradients that reflect differences in the degree of mantle
source depletion and variations in the subduction flux. Across the
Caribbean island-arc–back-arc system, a progressively lower subduction flux into a progressively less depleted mantle source is recorded
from arc related Los Ranchos, Amina and Maimón Formations, to the
rifted-arc to back-arc related Río Verde Complex and the mafic
intrusions of the Loma Caribe Peridotite. These gradients imply a SWdirected subduction polarity and are consistent with the cartographic
distribution of arc and back-arc geological units. By modeling the
simultaneous effects of source depletion and subduction input, the
calculated trace element and Nd isotope ratio curves for mantle
melting and magma sources reproduce the observed data trends from
arc to back-arc. Modeling suggests that HREE and Ti are least affected
by subduction input, with Sm showing minor modification, whereas
Ba has strong subduction input effects. Using the “H2O-rich component” model of Stolper and Newman (1994), levels of 0.001–0.005
weight fractions are suggested to have been added to the arc sources.
The low Nb contents and high (εNd)i values in both arc and back-arc
related mafic rocks imply the absence in the source of a significant
Lower Cretaceous plume enriched component.
Acknowledgements
The authors would like to thank John Lewis (George Washington
University), Gren Draper (Florida International University) and
Francisco Longo (Falconbridge Dominicana) for discussions on the
igneous rocks in the Dominican Republic. We are also grateful to many
colleagues of the IGME-BRGM team for their help and topic
discussions. Dirección General de Minería of Dominican Government
is also thanked for the support. Elisa Dietrich-Sainsaulieu is thanked
for her help with the Sr–Nd isotopic analyses at PCIGR. This work
forms part of the MCYT projects BTE-2002-00326 and CGL200502162/BTE and also received aid from the cartographic project of the
Dominican Republic funded by the SYSMIN Program of the European
Union. Careful reviews from Dr. Andrew Kerr and two anonymous
reviewers are much appreciated.
184
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.lithos.2009.08.007.
References
Arculus, R.J., Powell, R., 1986. Source component mixing in the regions of arc magma
generation. Journal of Geophysical Research 91B, 5913–5926.
Bédard, J.H., 1999. Petrogenesis of boninites from the Betts Cove Ophiolite, Newfoundland, Canada: identification of subducted source components. Journal of Petrology
40, 1853–1889.
Brenan, J.M., Shaw, H.F., Ryerson, F.J., Phinney, D.L., 1995. Mineral-aqueous fluid
partitioning of trace elements at 900 °C and 2.0 GPa: constraints on the trace
element chemistry of mantle and deep crustal fluids. Geochimica et Cosmochimica
Acta 59, 3331–3350.
Defant, M.J., Drummond, M.S., 1990. Derivation of some modern arc magmas by
melting of young subducted lithosphere. Nature 347, 662–665.
Donnelly, T.W., Beets, D., Carr, M.J., Jackson, T., Klaver, G., Lewis, J., Maury, R.,
Schellenkens, H., Smith, A.L., Wadge, G., Westercamp, D., 1990. History and tectonic
setting of Caribbean magmatism. In: Dengo, G., Case, J. (Eds.), The Caribbean Region.
Vol. H. The Geology of North America. Geological Society of America, pp. 339–374.
Draper, G., Lewis, J.F., 1991. Metamorphic belts in Central Española. In: Mann, P., Draper,
G., Lewis, J.F. (Eds.), Geologic and Tectonic Development of the North AmericaCaribbean Plate Boundary in Española: Geological Society of America Special Paper,
vol. 262, pp. 29–46.
Draper, G., Gutiérrez, G., Lewis, J.F., 1996. Thrust emplacement of the Española
peridotite belt: orogenic expression of the Mid Cretaceous Caribbean arc polarity
reversal. Geology 24 (12), 1143–1146.
Elliott, T., Plank, T., Zindler, A., White, W., Bourdon, B., 1997. Element transport from
slab to volcanic front at the Mariana arc. Journal of Geophysical Research 102,
14,991–15,019.
Escuder-Viruete, J., Díaz de Neira, A., Hernáiz Huerta, P.P., Monthel, J., García Senz, J.,
Joubert, M., Lopera, E., Ullrich, T., Friedman, R., Mortensen, J., Pérez-Estaún, A., 2006.
Magmatic relationships and ages of Caribbean island-arc tholeiites, boninites and
related felsic rocks, Dominican Republic. Lithos 90, 161–186.
Escuder-Viruete, J., Contreras, F., Stein, G., Urien, P., Joubert, M., Pérez-Estaún, A.,
Friedman, R., Ullrich, T.D., 2007a. Magmatic relationships and ages between
adakites, magnesian andesites and Nb-enriched basalt-andesites from Hispaniola:
record of a major change in the Caribbean island arc magma sources. Lithos 99,
151–177.
Escuder-Viruete, J., Contreras, F., Joubert, M., Urien, P., Stein, G., Weis, D., Pérez-Estaún, A.,
2007b. Tectónica y geoquímica de la Formación Amina: registro del arco isla caribeño
primitivo en la Cordillera Central, República Dominicana. Boletín Geológico y Minero
118 (2), 221–242.
Escuder-Viruete, J., Pérez-Estaún, A., Contreras, F., Joubert, M., Weis, D., Ullrich, T.D.,
Spadea, P., 2007c. Plume mantle source heterogeneity through time: insights from the
Duarte Complex, Central Hispaniola. Journal of Geophysical Research 112, B04203.
doi:10.1029/2006JB004323.
Escuder-Viruete, J., Joubert, M., Urien, P., Friedman, R., Weis, D., Ullrich, T., Pérez-Estaún,
A., 2008. Caribbean island-arc rifting and back-arc basin development in the Late
Cretaceous: geochemical, isotopic and geochronological evidence from Central
Hispaniola. Lithos 104, 378–404.
Ewart, A., Collerson, K.D., Regelous, M., Wendt, J.I., Niu, Y., 1998. Geochemical evolution
within the Tonga–Kermadec–Lau Arc–Back-arc system: the role of varying mantle
wedge composition in space and time. Journal of Petrology 39, 331–368.
Gerya, T.V., Yuen, D.A., Maresch, W.V., 2004. Thermomechanical modelling of slab
detachment. Earth and Planetary Science Letters 226, 101–116.
Gradstein, F.M., Ogg, J.G., Smith, A.G., 2004. A geologic time scale 2004. Cambridge
University Press. 610 pp.
Gribble, R.F., Stern, R.J., Newman, S., Bloomer, S.H., O'Hearn, T., 1998. Chemical and
isotopic composition of lavas from the northern Mariana Trough: implications for
magmagenesis in back-arc basins. Journal of Petrology 39, 125–154.
Hastie, A.H., Kerr, A.C., Pearce, J.A., Mitchell, S.F., 2007. Classification of altered volcanic
island arc rocks using immobile trace elements: development of the Th–Co
discrimination diagram. Journal of Petrology 48, 2341–2357. doi:10.1093/petrology/
egm062.
Hastie, A.R., Kerr, A.C., Mitchell, S.F., Millar, I.L., 2008. Geochemistry and petrogenesis of
Cretaceous oceanic plateau lavas in eastern Jamaica. Lithos 101, 323–343.
Hauff, F., Hoernle, K.A., Van den Bogaard, P., Alvarado, G.E., Garbe-Schínberg, C.D., 2000.
Age and geochemistry of basaltic complexes in western Costa Rica: contributions
to the geotectonic of Central America. Geochemistry, Geophysics, Geosystems 1
[1999gc000020].
Hawkesworth, C.J., Gallager, K., Hergt, J.M., McDermott, F., 1993. Mantle and slab contributions
in arc magmas. Annual Reviews of Earth and Planetary Science 21, 175–204.
Hawkins, J.W., 1995. The geology of the Lau Basin. In: Taylor, B. (Ed.), Backarc Basins:
Tectonics and Magmatism. Plenum Press, New York, pp. 63–138.
Hochstaedter, A.G., Gill, J.B., Peters, R., Broughton, P., Holden, P., Taylor, B., 2001. Acrossarc geochemical trends in the Izu-Bonin arc: contributions from the subducting
slab. Geochemistry Geophysics Geosystems 2 Paper number 2000GC000105.
Janney, P.E., Castillo, P.R., 2001. Geochemistry of the oldest Atlantic oceanic crust
suggests mantle plume involvement in the early history of the central Atlantic
Ocean. Earth and Planetary Science Letters 192, 291–302.
Jolly, W.T., Lidiak, E.G., 2006. Role of crustal melting in petrogenesis of the Cretaceous
Water Island Formation (Virgin Islands, northeast Antilles Island Arc). Geologica
Acta 4, 7–33.
Jolly, W.T., Lidiak, E.G., Dickin, A.K., Wu, T.W., 2001. Secular geochemistry of Central
Puerto Rican island arc lavas: constraints on mesozoic tectonism in the eastern
Greater Antilles. Journal of Petrology 42, 2197–2214.
Jolly, W.T., Lidiak, E.G., Dickin, A.P., 2006. Cretaceous to mid-Eocene pelagic sediment
budget in Puerto Rico and the Virgen Islands (northeast Antilles Island Arc). Geologica
Acta 4, 35–62.
Jolly, W.T., Schellekens, J.H., Dickin, A.P., 2007. High-Mg andesites and related lavas
from southwestern Puerto Rico (Greater Antilles Island Arc): petrogenetic links
with emplacement of the Caribbean mantle plume. Lithos 98, 1–26.
Kelemen, P.B., Yogodzinski, G.M., Scholl, D.W., 2003. Along-strike variation in lavas of the
Aleutian island arc: genesis of high Mg# andesite and implications for the continental
crust. In: Eiler J. (Ed.) Inside the subduction factory. American Geophysical Union
Monograph vol. 138, pp. 1–54.
Kerr, A.C., Tarney, J., Marriner, G.F., Nivia, A., Saunders, A.D., 1997. The Caribbean–
Colombian Cretaceous igneous province: The internal anatomy of an oceanic
plateau. In: Mahoney, J.J., Coffin, M.F. (Eds.), Large Igneous Provinces. AGU
Washington DC, pp. 123–144.
Kerr, A.C., Iturralde-Vinent, M.A., Saunders, A.D., Babbs, T.L., Tarney, J., 1999. A new
plate tectonic model of the Caribbean: implications from a geochemical
reconnaissance of Cuban Mesozoic volcanic rocks. Geological Society of America
Bulletin 111, 1581–1599.
Kerr, A.C., Tarney, J., Kempton, P.D., Spadea, P., Nivia, A., Marriner, G.F., Duncan, R.A.,
2002. Pervasive mantle plume head heterogeneity: evidence from the late
Cretaceous Caribbean–Colombian oceanic plateau. Journal of Geophysical Research
107 (B7). doi:10.1029/ 2001JB000790.
Kesler, S.E., Russell, N., Polanco, J., McCurdy, K., Cumming, G.J., 1990. Geology and
geochemistry of the Early Cretaceous Los Ranchos Formation, central Dominican
Republic. In: Mann, P., Draper, G., Lewis, J.F. (Eds.), Geologic and Tectonic Development
of the North America–Caribbean Plate Boundary in Española: Geological Society of
America Special Paper, vol. 262, pp. 187–201.
Kesler, S.E., Campbell, I.H., Allen, C.h.M., 2005. Age of the Los Ranchos Formation,
Dominican Republic: timing and tectonic setting of primitive island arc volcanism
in the Caribbean region. Geological Society of America Bulletin 117, 987–995.
Krebs, M., Maresch, W.V., Schertl, H.-P., Baumann, A., Draper, G., Idleman, B., Münker, C.,
Trapp, E., 2007. The dynamics of intra-oceanic subduction zones: A direct
comparison between fossil petrological evidence (Rio San Juan Complex, Dominican
Republic) and numerical simulation. Lithos 103, 106–137.
Lapierre, H., Dupui, V., Lepinay, B.M., Tardy, M., Ruiz, J., Maury, R.C., Hernandez, J.,
Loubert, M., 1997. Is the Lower Duarte Complex (Hispañiola) a remnant of the
Caribbean plume generated oceanic plateau? Journal of Geology 105, 111–120.
Lapierre, H., Dupuis, V., de Lepinay, B.M., Bosch, D., Monie, P., Tardy, M., Maury, R.C.,
Hernandez, J., Polve, M., Yeghicheyan, D., Cotten, J., 1999. Late Jurassic oceanic crust
and upper cretaceous Caribbean plateau picritic basalts exposed in the Duarte
igneous complex, Hispaniola. Journal of Geology 107, 193–207.
Lapierre, H., Bosch, D., Dupuis, V., Polvé, M., Maury, R., Hernandez, J., Monié, P., Yeghicheyan,
D., Jaillard, E., Tardy, M., de Lepinay, B., Mamberti, M., Desmet, A., Keller, F., Senebier, F.,
2000. Multiple plume events in the genesis of the peri-Caribbean Cretaceous oceanic
plateau province. Journal of Geophysical Research 105, 8403–8421.
Lázaro, C., García-Casco, A., Rojas Agramonte, Y., Kröner, A., Neubauer, F., Iturralde-Vinent,
M., 2008. Fifty-five-million-year history of oceanic subduction and exhumation at the
northern edge of the Caribbean plate (Sierra del Convento mélange, Cuba). Journal of
Metamorphic Geology. doi:10.1111/j.1525-1314.2008.00800.
Leat, P.T., Pearce, J.A., Barker, P.F., Millar, I.L., Barry, T.L., Larter, R.D., 2004. Magma
genesis and mantle flow at a subducting slab edge: the South Sandwich arc-basin
system. Earth and Planetary Science Letters 227, 17–35.
Lebrón, M.C., Perfit, M.R., 1994. Petrochemistry and tectonic significance of Cretaceous
island-arc-rocks, Cordillera Oriental, Dominican Republic. Tectonophysics 229,
69–100.
Lewis, J.F., Astacio, V.A., Espaillat, J., Jiménez, J., 2000. The occurrence of volcanogenic
massive sulfide deposits in the Maimon Formation, Dominican Republic: the Cerro
de Maimón, Loma Pesada and Loma Barbuito deposits. In: Sherlock, R., Barsch, R.,
Logan, A. (Eds.), VMS deposits of Latin America. Geological Society of Canada
Special Publication. 223–249 pp.
Lewis, J.F., Escuder-Viruete, J., Hernaiz Huerta, P.P., Gutiérrez, G., Draper, G., 2002.
Subdivisión Geoquímica del Arco Isla Circum-Caribeño, Cordillera Central Dominicana: Implicaciones para la formación, acreción y crecimiento cortical en un ambiente
intraoceánico. Acta Geológica Hispánica 37, 81–122.
Lewis, J.F., Draper, G., Proenza, J., Espaillat, J., Jiménez, J., 2006. Ophiolite-related ultramafic
rocks (Serpentinites) in the Caribbean: a review of their occurrence, composition,
origin, emplacement and Ni-laterite soil formation. Geológica Acta 4, 7–28.
Ludwig, K.R., 2003. Isoplot 3.00: a geochronological toolkit for Microsoft Excel. Berkeley
Geochronology Center Special Publication No. 4.
Mann, P., 1999. Caribbean sedimentary basins: classification and tectonic setting from
Jurassic to Present. In: Mann, P. (Ed.), Caribbean Basins: Sedimentary Basins of the
Word, vol. 4, pp. 3–31.
Marchesi, C., Garrido, C.J., Godard, M., Proenza, J.A., Gervilla, F., Blanco-Moreno, J., 2006.
Petrogenesis of highly depleted peridotites and gabbroic rocks from the MayaríBaracoa ophiolitic belt (eastern Cuba). Contribution to Mineralogy and Petrology
151, 717–736.
Marchesi, C., Garrido, C.J., Bosch, D., Proenza, J.A., Gervilla, F., Monié, P., Rodríguez-Vega,
A., 2007. Geochemistry of Cretaceous magmatism in eastern Cuba: recycling of
North American continental sediments and implications for subduction polarity in
the Greater Antilles Paleo-arc. Journal of Petrology 48, 1813–1840.
J. Escuder-Viruete et al. / Lithos 114 (2010) 168–185
Martinez, F., Taylor, B., 2002. Mantle wedge control on backarc crustal accretion. Nature
416, 417–420.
Myczynski, R., Iturralde-Vinent, M., 2005. The Late Lower Albian Invertebrate Fauna of
the Río Hatillo Formation of Pueblo Viejo, Dominican Republic. Caribbean Journal of
Science 41 (4), 782–796.
Peacock, S.M., Wang, K., 1999. Seismic consequence of warm versus cool subduction
metamorphism: examples from southwest and northeast Japan. Science 286, 937–939.
Pearce, J.A., Parkinson, I.J., 1993. Trace element models for mantle melting: application
to volcanic arc petrogenesis. In: Pritchard, H.M., Alabaster, T., Harris, N.B.W., Nearly,
C.R. (Eds.), Magmatic Processes and Plate Tectonics: Geological Society of London
Special Publication, vol. 76, pp. 373–403.
Pearce, J.A., Peate, D.W., 1995. Tectonic implications of the composition of volcanic arc
magmas. Earth and Planetary Science Annual Review 23, 251–285.
Pearce, J.A., Baker, P.E., Harvey, P.K., Lui, W., 1995. Geochemical evidence for subduction
fluxes, mantle melting and fractional crystallization beneath the South Sandwich
island arc. Journal of Petrology 36, 1073–1109.
PETDB, 2007. Petrological database of the ocean floor. http://petdb.ldeo.columbia.edu/
petdb.
Pindell, J.L., Kennan, L., Maresch, W.V., Stanek, K.P., Draper, G., Higgs, R., 2005. Platekinematics and crustal dynamics of circum-Caribbean arc–continent interactions:
tectonic controls on basin development in Proto-Caribbean margins. In: Lallemant,
A., Sisson, V.B. (Eds.), Caribbean–South American Plate Interactions: Geological
Society of America Special Paper, vol. 394, pp. 7–52.
Plank, T., Langmuir, C.H., 1993. Tracing trace elements from sediment input to volcanic
output at subduction zones. Nature 362, 739–742.
Proenza, J.A., Díaz-Martínez, R., Iriondo, A., Marchesi, C., Melgarejo, J.C., Gervilla, F., Garrido,
C.J., Rodríguez-Vega, A., Lozano-Santacruz, R., Blanco-Moreno, J.A., 2006. Primitive
Cretaceous island-arc volcanic rocks in eastern Cuba: the Téneme Formation. Geologica
Acta 4, 103–121.
Rankin, D.W., 1631. 2002. Geology of St. John, U. S., Virgin Islands. U.S. Geological.
Survey. Professional Paper, pp. 1–36.
Révillon, S., Hallot, E., Arndt, N., Chauvel, C., Duncan, R.A., 2000. A complex history for
the Caribbean plateau: petrology, geochemistry, and geochronology of the Beata
Ridge, South Hispaniola. Journal of Geology 108, 641–661.
Schmidt, M.W., Poli, S., 1998. Experimentally based water budgets for dehydrating slabs
and consequences for arc magma generation. Earth and Planetary Science Letters
163, 361–379.
Singer, B. S., B. R. Jicha, W. P. Leeman, N. W. Rogers, M. F. Thirlwall, J. Ryan, Nicolaysen,
K.E., 2007. Along-strike trace element and isotopic variation in Aleutian Island arc
basalt: Subduction melts sediments and dehydrates serpentine. Journal of
Geophysical Research 112, B06206. doi:10.1029/2006JB004897.
Sinton, C.W., Duncan, R.A., Storey, M., Lewis, J., Estrada, J.J., 1998. An oceanic flood basalt
province within the Caribbean plate. Earth Planetary Science Letters 155, 221–235.
Stalder, R., Foley, S.F., Brey, G.P., Horn, I., 1998. Mineral-aqueous fluid partitioning of
trace elements at 900–1200 °C and 3.0–5.7 GPa: new experimental data for garnet,
185
clinopyroxene and rutile and implications for mantle metasomatism. Geochimica
Cosmochima Acta 62, 1781–1801.
Stern, R.J., 2002. Subduction zones. Reviews of Geophysics 40, 1012. doi:10.1029/
2001RG000108.
Stolper, E.M., Newman, S., 1994. The role of water in the petrogenesis of Mariana
Trough magmas. Earth and Planetary Science Letters 121, 293–325.
Su, Y., Langmuir, C.H., 2003. Global MORB chemistry compilation at the segment scale,
PhD Thesis, Department of Earth and Environmental Sciences, Columbia University.
http://petdb.ldeo.columbia.edu/documentation/morbcompilation/.
Sun, S.S., McDonough, W.F., 1989. Chemical and isotopic systematics of oceanic basalts:
implications for mantle compositions and processes. In: Saunders, A.D., Norry, M.J.
(Eds.), Magmatism in the Ocean Basins: Geological Society of London Special Publication,
vol. 42, pp. 313–345.
Swinden, H.S., Jenner, G.A., Szybinski, Z.A., 1997. Magmatic and tectonic evolution of the
Cambrian–Ordovician Laurentian margin of Iapetus: geochemical and isotopic
constraints from the Notre Dame subzone, Newfoundland. In: Sinha, A.K., Whalen,
J.B., Hogan, J.P. (Eds.), The Nature of Magmatism in the Appalachian Orogen, 191,
pp. 337–365.
Tatsumi, Y., Hanyu, T., 2003. Geochemical modeling of dehydration and partial melting of
subducting lithosphere: toward a comprehensive understanding of high-Mg andesite
formation in the Setouchi volcanic belt, SW Japan. Geochemistry Geophysics Geosystems
4 [1081GC000080].
Taylor, B., Martinez, F., 2003. Back-arc basin systematics. Earth and Planetary Science
Letters 210, 481–497.
Thompson, P.M.E., Kempton, P.D., White, R.V., Saunders, A.D., Kerr, A.C., Tarney, J.,
Pringle, M.S., 2004. Elemental, Hf–Nd isotopic and geochronological constraints on
an island arc sequence associated with the Cretaceous Caribbean Plateau: Bonaire,
Dutch Antilles. Lithos 74, 91–116.
Weis D., Kieffer B., Maerschalk C., Pretorius W., and Barling J. (2005), Highprecision PbSr-Nd-Hf isotopic characterization of USGS BHVO-1 and BHVO-2 reference
materials. Geochem. Geophys. Geosyst. 6, Q02002. doi:10.1029/2004GC000852.
Weis, D., Kieffer, B., Maerschalk, C., Barling, J., de Jong, J., Willians, G.A., Hanano, D.,
Pretorius, W., Scoates, J.S., Goolaerts, A., Friedman, R.M., Mahoney, J.B., 2006. Highprecision isotopic characterization of USGS reference materials by TIMS and MCICP-MS. Geochemistry Geophysics Geosystems 7 [2006GC001283].
Winchester, J.A., Floyd, P.A., 1977. Geochemical discrimination of different magma series and
their differentiation products using immobile elements. Chemical Geology 20, 325–343.
Woodhead, J.D., Eggins, S.M., Johnson, R.W., 1998. Magma genesis in the New Britain
island arc: further insights into melting and mass transfer processes. Journal of
Petrology 39, 1641–1668.
Yogodzinski, G.M., Lees, J.M., Churikova, T.G., Dorendorf, F., Woerner, G., Volynets, O.N.,
2001. Geochemical evidence for the melting of subducting oceanic lithosphere at
plate edges. Nature 409, 500–504.