International Geology Review, Vol. **, ****, p. **8 790.

Copyright 2006 by V. H. Winston & Son, Inc. All rights reserved.

P-T Path for Ultrahigh-Pressure Garnet Ultramafic Rocks of the

Cuaba Gneiss, Rio San Juan Complex, Dominican Republic

RICHARD N. ABBOTT, JR.,1

Department of Geology, Appalachian State University, Boone, North Carolina 28608

GRENVILLE DRAPER,

Department of Earth Sciences, Florida International University, Miami, Florida 33199

BONNIE N. BROMAN

AND

Department of Geology, Appalachian State University, Boone, North Carolina 28608

Abstract

Ultrahigh-pressure (UHP) rocks in the Cuaba Gneiss include Grt ultramafic rocks, mafic eclog-

ite, and partially retrograded equivalents. The Grt ultramafic rocks (Spl-bearing Grt peridotite, Spl-

bearing Grt clinopyroxenite, Crn-Spl bearing Grt clinopyroxenite) are of igneous origin, with mag-

matic conditions of P > 3.5 GPa, T > 1550 C. The magmatic history took place in the asthenosphere.

New chemical analyses of minerals give the following subsolidus conditions: Grt peridotite, 3.0 4.2

GPa, 838-867 C; Grt clinopyroxenite (Grt + Cpx + Spl + Crn), 2.75 GPa, 807 C. The ultramafic

rocks are associated with partially retrograded eclogite, interpreted as deep-subducted oceanic

crust. New chemical analyses of minerals in the eclogite give conditions that relate to the retrograde

decomposition of Grt + Omp + Qtz to Grt + Pl + Di + Qtz, 1.8 GPa, 730 C.

The P-T path for the Grt ultramafic rocks is modeled in three parts: (1) slow, isobaric (> 4 GPa)

cooling in the mantle, >1550 C down to ~850 C; (2) relatively rapid, nearly adiabatic decom-

pression, ~4 GPa (~850 C) down to ~1 GPa (~700 C); and (3) relatively rapid, non-adiabatic

decompression and cooling. The first part of the path (1) took place in the mantle above the subduc-

tion zone and relates to delivery of the Grt ultramafic rocks to the subduction zone. Incorporation of

Grt ultramafic rocks in the deep-subducted oceanic crust (eclogite) marked the end of this part of

the path. The second part of the path (2) was associated with transport up the subduction zone.

Retrograde P-T conditions for the eclogite fall on this part of the path, supporting the idea that the

Grt ultramafic rocks were transported as blocks in the eclogite. The third part of the path (3) relates

to final uplift to the surface.

Introduction et al., 1975). The terrane itself was delivered to the

surface at an ocean-ocean convergent margin, con-

THE PURPOSE of this contribution is to develop a

founding any simple explanation for uplift.

P-T path for ultrahigh-pressure (UHP) rocks of the

New chemical analyses of minerals were used to

Cuaba Gneiss in the Dominican Republic (Abbott et

estimate P-T conditions for three principal rock

al., 2001a, 2001b, 2003a, 2003b, 2005a, 2005b;

types (Grt peridotite, Crn-bearing Grt clinopyroxen-

Draper et al., 2002; Broman et al., 2005). Among

ite, and eclogite). The estimated conditions con-

globally scarce UHP rocks, the Dominican example

is unusual in many respects. Grt ultramafic rocks strain retrograde P-T paths for the Grt ultramafic

show clear and exciting evidence for magma cham- rocks and associated eclogite. The results are

ber processes under asthenospheric conditions (P > important in developing a tectonic model for (1)

3.4 GPa, T > 1550 C; Abbott et al., 2003a, 2005a, assembling the parts of the terrane and (2) deliver-

2005b; Broman et al., 2005). Garnet clinopyroxenite

ing the terrane to the surface of the earth.2

shows the only known, natural occurrence of coex-

isting garnet, spinel and corundum (see Ackermand

2Mineral and component abbreviations are consistent with

1Corresponding author; email: abp907@r.postjobfree.com Kretz (1983).

778

0020-6814/06/891/778-13 $25.00

779

UHP GARNET ULTRAMAFIC ROCKS

FIG. 1. Geology of the southern part of the Rio San Juan Complex. The Island of Hispaniola is shown in the inset.

The vertical dashed boundary separates Haiti (west) from the Dominican Republic (east). The Cordillera Septentrional

occupies the land north of the Septentrional fault (SF) in the Dominican Republic. Movement between the North Amer-

ican plate and Caribbean plate is distributed over the Septentrional fault (SF), Camu fault (CF), and an offshore strongly

oblique covergent zone (Mann et al., 2002; Jansma et al., 2000). The study area is marked by the filled rectangle. The

geologic map shows the southern part of the Cretaceous Rio San Juan Complex (re-mapped 2002 2005, cf. Draper and

Nagle, 1991). The Cuaba Gneiss consists of hornblende schist (Kc1), garnet hornblende gneiss (Kc2), garnet metadiorite

(Kc3). Small, filled circles in Kc2 represent sites where garnet ultramafic rocks were observed or sampled from boulders.

The Cuaba Gneiss is intruded by the Rio Bobo gabbro complex (g). Younger sedimentary rocks and sediment are Upper

Eocene Miocene clastic sedimentary rocks (Tcs), Neogene limestones (Tl), and Quaternary alluvium (Q). Reverse faults:

bold, toothed lines (teeth, on hang wall). Left-lateral strike-slip faults: bold un-ornamented lines, dashed where uncer-

tain. S.F. de M. is the city of San Francisco de Macoris.

Geologic Setting arc widened by accretion, in response to SW-

directed subduction of oceanic parts of the North

The Dominican Republic occupies the eastern and South American plates (Pindell and Barrett,

two-thirds of the Island of Hispaniola (Fig. 1). The 1990; Pindell, 1995; Draper et al., 1996; Pindell et

island is located at the northern edge of the Carib- al., 2005). The original island-arc system has since

bean plate. Like the other islands of the Greater been modified by E-W, left-lateral, transcurrent

Antilles (Jamaica, Puerto Rico, southeasternmost tectonics (Fig. 1) that began in the mid-Eocene and

Cuba), the basement began as an Early Cretaceous, continues today (Mann et al., 1990; Draper et al.,

intra-oceanic island arc complex above a NE-dip- 1996). GPS studies indicate that most of the current

ping, Pacific-derived plate (Pindell and Barrett, movement between the North American and Carib-

1990; Draper et al., 1994, 1996; Pindell, 1994; bean plates, in the vicinity of the study area, takes

Pindell et al., 2005). In the mid-Cretaceous (~120 place along the Septentrional fault (Jansma et al.,

100 Ma; Pindell and Barrett, 1990), the plate 2000; Mann et al., 2002).

boundary was relocated from the Pacific side to the The Cordillera Septentrional occupies the land

Atlantic side of the island-arc complex and the north of the Septentrional fault (Fig. 1) in the

polarity of subduction reversed to the present-day Dominican Republic. Upper Eocene to Lower

configuration (i.e., SW- to W-dipping subduction, Miocene siliciclastic and carbonate sedimentary

plate boundary on the Atlantic side). From the Late rocks cover most of the metamorphic and igneous

Cretaceous to mid-Eocene, the intra-oceanic island basement complex (Eberle et al., 1982; Lewis and

780 ABBOTT ET AL.

FIG. 2. Digital images of thin sections. A. Garnet peridotite (DR03-10). B. Crn-bearing garnet clinopyroxenite

(DR03-12). C. Partially retrograded eclogite (DR03-7).

Draper, 1990). Only locally, where the cover has mm- to dm-scale layers in otherwise symplectite-

been eroded away, is the Cretaceous basement free gneiss. Garnet peridotite and garnet clinopyrox-

exposed in a number of stratigraphic windows, or enite (Abbott et al., 2001a, 2003a, 2003b, 2005a,

inliers. The Cuaba Gneiss (Fig. 1) is the southern- 2005b) occur as stream boulders (up to 4 meters in

most unit in the largest of these inliers, known as the diameter) eroded out of the hornblende gneiss (Kc2).

Rio San Juan Complex (Eberle et al., 1982; Draper

and Nagle, 1991). The Cuaba Gneiss unit is approx-

imately 30 km long and up to 6 km wide. On its Petrography

north side, the Cuaba Gneiss was intruded by

Details of the petrography of the garnet ultrama-

gabbroic to quartz dioritic rocks of the Rio Boba

fic rocks are reported elsewhere (Abbott et al.,

Intrusive Complex (Draper and Nagle, 1991). On its

2001a, 2001b, 2005a). A summary is given here

southern side the Cuaba Gneiss is faulted against

for examples used in this study. The eclogite is

Tertiary siliciclastics.

discussed in somewhat more detail, as it has not

The Cuaba Gneiss is divided into three mem-

been described in the context of the present exam-

bers, according to the dominant rock type. From

ple. The relevant rock types and mineral assem-

west to east (Fig. 2), the members are hornblende

blages are as follows:

schist (Kc1), garnet hornblende gneiss (Kc2), and

garnet metadiorite (Kc3). The common mineral

Garnet peridotite (DR03-10),

assemblage in all three units is Hbl + Pl (andesine) +

Cpx + Ol + Grt + Spl + hornblende + serpentine

Qtz + Rt Grt Bt Ep. The units differ mainly in

texture and modal abundance of minerals. Draper Garnet peridotite (DR00-3), locally on a cm-scale,

and Nagle (1991) suggested a mafic protolith Cpx + Grt + Spl + Crn + hornblende

(basalt/diabase/gabbro) of oceanic crustal origin.

Corundum-bearing garnet clinopyroxenite (DR03-12),

This study focuses on minor constituents, garnet

Cpx + Grt + Spl + Crn + hornblende

ultramafic rocks, and eclogite in the garnet horn-

blende gneiss (Kc2). Evidence for eclogite is in the

Eclogite (DR03-7),

form of Pl-Di(symplectite) + Grt, with greater or

Pl-Cpx symplectite + Grt + Qtz + zoisite +

lesser amounts of hornblende depending on the

hornblende

extent of retrograde hydration (Abbott and Draper,

1998, 2002). The retrograded eclogite occurs as Accessory minerals include magnetite and pyrite.

781

UHP GARNET ULTRAMAFIC ROCKS

Garnet peridotite Partially retrograde eclogite

Coarse, granoblastic garnet peridotite (DR03-10, Partially retrograded eclogite is fairly common in

the garnet hornblende gneiss (Kc2). Foliation is

Fig. 2A) is the most common of the garnet ultramafic

developed weakly to strongly, depending on the

lithologies. Crystals of pink garnet, up to 2 cm in

amount of hornblende. The sample described here

diameter, are the most conspicuous feature. The

(DR03-7, Fig. 2C) is typical. Euhedral to subhedral

crystals are anhedral, but crudely equidimensional.

porphyroblasts (1 3 mm) of pink garnet are evenly

The garnet contains fine (

distributed in a matrix of symplectic intergrowth of

transparent, emerald-green spinel. Locally, on a

micron-scale plagioclase and diopside. Garnet has

sub cm scale, where olivine is absent (DR00-3), the

sub mm-scale inclusions of quartz and zoisite.

garnet may also contain fine ( 3.4 GPa, T>1550 C; Abbott et al., 2003a, independently on some of the same minerals at the

2005a, 2005b). Florida Center for Analytical Microscopy, Florida

TABLE 1. Garnet Peridotite1

782

Garnet peridotite, DR03-10b Garnet peridotite, DR00-3

Mineral: Grt rim Grt (FIU)2 Grt core Spl core Spl rim Oli Oli (FIU) Cpx (FIU)3 Hbl (FIU)4 Grt Spl Grt Spl

ID nos:. 1 5 *-**-*-**-**-** 1 5 1 17 1 10 18- 27 11 16

No. of analyses: 5 4 *-**-*-*-*-**-*-** 10 10 6

wt% wt%

MgO 10.24 9.47 8.06 16.63 16.27 39.15 40.12 15.49 16.48 8.85 16.04 8.11 16.29

22.34 22.57 22.49 62.98 63.22 0.68 0.03 2.47 11.40 23.12 63.31 23.12 63.01

Al2O3

39.4 38.86 39.08 0.48 0.49 38.28 37.62 50.90 46.95 39.35 0.67 39.67 1.12

SiO2

CaO 9.01 8.57 12.72 0.31 0.24 0.35 0.01 23.05 12.48 12.61 0.46 13.24 0.63

0.12 0.01 0.16 0.13 0.13 0.15 0.00 0.19 0.48 0.14 0.15 0.14 0.12

TiO2

MnO 0.27 0.81 0.26 0.10 0.1 0.16 0.23 0.15 0.05 0.25 0.16 0.29 0.15

FeO 18.63 21.10 17.23 19.37 19.54 21.23 23.21 5.29 7.09 15.68 19.21 15.43 18.69

Total 100 101.40 100-***-***-*** 101.27 97.76 96.99 100-***-***-***

Cations p.f.u Cations p.f.u

Mg 1.15 1.05 0.91 0.64 0.63 1.51 1.53 0.87 3.51 0.99 0.62 0.91 0.63

Al 1.98 1.99 2.00 1.92 1.93 0.02 0.00 0.13 1.92 2.04 1.93 2.05 1.92

Si 2.96 2.90 2.95 0.01 0.01 0.99 0.96 1.90 6.72 2.95 0.02 2.98 0.03

Ca 0.73 0.69 1.03 0.01 0.01 0.01 0.00 0.93 1.91 1.01 0.01 1.07 0.02

Ti 0.01 0.00 0.01 0.00 0.00 0.00 0.00 0.01 0.05 0.01 0.00 0.01 0.00

Mn 0.02 0.05 0.02 0.00 0.00 0.00 0.00 0.00 0.01 0.02 0.00 0.02 0.00

Fe 1.17 1.32 1.09 0.42 0.42 0.46 0.49 0.16 0.85 0.98 0.42 0.97 0.40

Total 8.00 8.00 8.00 3.00 3.00 3.00 3.00 4.0 15.45 8.00 3.00 8.00 3.00

0.00

ABBOTT ET AL.

Q5 23.91 23.79 23.92 7.95 7.96 8.00 7.88 11.94 23.96 7.97 24.03 7.98

Fe 0.09 0.21 0.08 0.05 0.04 0.00 0.12 0.06 0.04 0.03 0.00 0.02

Fe 1.08 1.11 1.01 0.36 0.38 0.46 0.37 0.10 0.94 0.38 0.97 0.38

Mg# 0.49 0.49 0.45 0.60 0.60 0.77 0.78 0.84 0.81 0.50 0.60 0.48 0.61

prp 0.37 0.36 0.30 0.33 0.31

grs 0.24 0.24 0.34 0.34 0.36

alm 0.38 0.38 0.36 0.33 0.33

spl 0.64 0.63 0.62 0.63

hc 0.34 0.36 0.35 0.36

mag 0.03 0.02 0.02 0.01

wo 0.47

en 0.44

fs 0.08

jd 0.01

1Garnet standard #5 (NBS 110752)

2FIU = analyses performed at Florida Center for Analytical Electron Microscopy, Florida International University; Abbott et al., 2001a, 2005a, 2005b.

3Cpx (FIU) includes Na O = 0.12 wt.% (Na = 0.01 pfu).

2

4Hbl (FIU) = includes Na O = 1.49 wt% (Na = 0.41 pfu).

2

5Q = sum of cation charges.

783

UHP GARNET ULTRAMAFIC ROCKS

TABLE 2. Mineral Analyses, Garnet Clinopyroxenite, DR03-121

Mineral: Grt rim Spl Grt core Grt cr-rm Grt rim Spl core Spl rim Cpx

ID nos.: 1 5 1 *-**-**-**-**-*-** 10 15 5 9, 16 17 1 19

No. of analyses: 5 4 5 3 5 6 7 17

wt%

Na2O n.a. n.a. n.a. n.a. n.a. n.a. n.a. 0.65

MgO 6.97 9.48 2.54 6.51 8.31 10.60 10.85 16.99

Al2O3 21.83 57.65 21.83 21.77 21.81 59.91 59.95 1.56

SiO2 38.90 0.63 38.62 38.36 39.42 0.30 0.28 52.11

K2O n.a. n.a. n.a. n.a. n.a. n.a. n.a. 0.05

CaO 11.13 0.34 16.26 12.51 9.67 0.32 0.30 24.38

TiO2 0.13 0.10 0.13 0.09 0.15 0.18 0.16 0.31

MnO 0.71 0.27 1.59 0.77 0.64 0.27 0.29 0.11

FeO 20.33 31.55 19.03 19.98 19.99 28.43 28.18 3.85

Total 100-***-***-*** 100-***-***-***

Cations p.f.u

Na 0.05

Mg 0.79 0.39 0.29 0.74 0.94 0.43 0.43 0.92

Al 1.97 1.86 2.00 1.96 1.95 1.91 1.91 0.07

Si 2.97 0.02 3.00 2.93 3.00 0.01 0.01 1.89

K 0.00

Ca 0.91 0.01 1.35 1.03 0.79 0.01 0.01 0.95

Ti 0.01 0.00 0.01 0.01 0.01 0.00 0.00 0.01

Mn 0.05 0.01 0.10 0.05 0.04 0.01 0.01 0.00

Fe 1.30 0.72 1.24 1.28 1.27 0.64 0.63 0.12

Total 8.00 3.00 8.00 8.00 8.00 3.00 3.00 4.00

Q 23.93 7.90 24.02 23.84 23.96 7.93 7.93 11.82

Fe 0.07 0.10 0.00 0.16 0.04 0.07 0.07

Fe 1.23 0.62 1.24 1.12 1.23 0.57 0.56

Mg# 0.38 0.35 0.19 0.37 0.43 0.40 0.41 0.89

prp 0.26 0.10 0.24 0.31

grs 0.30 0.45 0.33 0.26

alm 0.43 0.41 0.41 0.42

spl 0.39 0.43 0.43

hc 0.56 0.53 0.53

mag 0.05 0.04 0.04

wo 0.47

en 0.45

fs 0.06

jd 0.02

1Garnet standard #5 (NBS 110752) for Grt, Spl; Omphacite standard #9 (NBS 110607) for Cpx; n.a. = not analyzed.

784 ABBOTT ET AL.

International University (FIU analyses reported in TABLE 3. Mineral Analyses, eclogite, DR03-71

Table 1). In the Crn-bearing clinopyroxenite (DR03-

12), USNM garnet standard (#110752) was used for Symplectite

garnet and spinel; and USNM omphacite standard Mineral: Hbl Grt Pla Cpx

(#110607) was used for clinopyroxene. In the eclog- No. of analyses: 3 5 3 5

ite (DR03-7), USNM omphacite standard (#110607)

was used for all minerals (hornblende, garnet, wt%

plagioclase and clinopyroxene). Na2O 2.10 0.26 7.61 1.99

Iron was reported as weight percent FeO. The MgO 11.63 5.74 0.28 11.16

weight percent oxide components were recalculated Al2O3 13.33 22.65 24.98 4.29

as atomic proportions of cations, normalized to 8 cat- SiO2 43.27 36.35 62.78 51.18

ions per formula unit (pfu) for garnet, 4 cations pfu for

CaO 12.62 8.63 4.14 22.08

clinopyroxene, 3 cations pfu for spinel and olivine, 5

FeO 17.06 26.37 0.21 9.30

cations pfu for plagioclase, and 15 cations pfu for

Total 100-***-***-***

hornblende (excluding Na). Fe3+ was estimated on the

basis of stoichiometry and charge balance. Cations p.f.u

Na 0.59 0.04 0.66 0.14

Data

Mg 2.50 0.66 0.02 0.62

Minerals in two assemblages were analyzed for Al 2.26 2.07 1.31 0.19

the garnet peridotite (Table 1). Minerals in sample Si 6.23 2.81 2.80 1.89

DR03-10 refer to the assemblage Cpx + Ol + Grt + Ca 1.95 0.72 0.20 0.88

Spl + Mg-hornblende + serpentine. No analyses

Fe 2.06 1.71 0.01 0.29

were performed on the serpentine. Minerals in sam-

Total 15.59 8.00 5.00 4.00

ple DR00-3 refer to the assemblage Cpx + Grt + Spl

+ Crn + Mg-hornblende in a cm-scale, olivine-free Q 45.32 23.65 16.26 11.83

part of the garnet peridotite. Garnet is not notably Fe 0.68 0.35 0.00 0.17

zoned (core, rim), and the composition is similar in Fe 1.38 1.36 0.01 0.12

both assemblages. The typical composition has more Mg# 0.64 0.33 0.84

or less equal mole % of prp, alm, and grs compo- Al(iv) 1.77 1.31 0.11

nents, and less than 1 mole % sps. Spinel composi- ts (8-Si-Na) 1.18

tions are remarkably uniform both within individual ed (Na) 0.59

grains (core, rim) and from one grain to another, and ab 0.77

essentially the same in both assemblages. The com-

an 0.23

position is spl63hc36mag01.

prp 0.21

Mineral analyses for the Crn-bearing garnet

grs 0.23

clinopyroxenite (DR03-12) refer to the assemblage

alm 0.55

Cpx + Grt + Spl + Crn + hornblende (Table 2).

wo 0.44

Garnet analyses show some variation, but this

en 0.31

cannot be characterized well without more chemical

fs 0.14

analyses. Compared to the garnet peridotite, the

jd 0.07

garnet in the clinopyroxenite has less Mg (10 31

mole% prp), consistently more Fe (41 43 mole %

1Omphacite standard #9, NBS 110607.

alm), and similar, but more variable Ca (26 45

mole% grs). Spinel is uniform in composition, with

uncertainty, the traverse showed no variation in any

notably less Mg (39 43 mole% spl) and more Fe

element across the phenocryst. The clinopyroxene

(53 56 mole% hc, 4 5 mole% mag) than spinel in

has essentially the same composition as clinopyrox-

the garnet peridotite. The composition of the clino-

ene in garnet peridotite (Table 1).

pyroxene (wo47en 45fs07jd02) is the average of 17

Chemical analyses of minerals in eclogite

single-point analyses, at 0.5 mm intervals along a

(DR03-7) are reported in Table 3. Compositions of

traverse (rim to core to rim) across the phenocryst

the components of the symplectite are plagioclase

near the center of Figure 2B (nearly surrounded by

ab77an23 and clinopyroxene wo44en31fs14jd07. The

garnet). Within reasonable limits of analytical

785

UHP GARNET ULTRAMAFIC ROCKS

TABLE 4. P-T Estimates

Sample Assemblage P, GPa T, C

Thermobarometry, WEBINVEQ

Grt peridotite DR03-10 Grt-Cpx-Oli-Spl 4.15 867

Grt peridotite DR00-3 Grt-Cpx-Spl-Crn 3.00 838

Grt clinopyroxenite DR03-12 Grt-Cpx-Spl-Crn 2.75 807

eclogite DR03-7 Grt-Cpx-Pla-Qtz 1.76 731

Grt-Cpx at P = 1.5 GPa Grt-Cpx at P = 3 GPa Grt-Hbl

Thermometry, T C

Ref.1 A(94) K(98) R(04) A(94) K(88) R(00) GP(84)

Grt peridotite 751-***-***-*** 679 673 664

Grt clinopyroxenite 771-***-***-*** 689 663 732

Eclogite 690-***-***-*** 664 609 727

1A(94) = Ai, 1994; K(88) = Krogh, 1988; R(00) = Ravna, 2000; GP(84) = Graham and Powell, 1984.

WEBINVEQ

garnet composition (prp21grs23alm55) pertains to the

edge of a porphyroblast next to symplectic inter- Starting with the mineral analyses, the activities

growth of plagioclase and clinopyroxene. The horn- of end member components of minerals in an assem-

blende analyses are from a 0.5 mm wide rim formed blage were calculated using the program AX, writ-

on a garnet porphyroblast. ten by Holland and Powell (2000). The activities

were converted to single-site mixing models and

then entered, online, into the program WEBINVEQ,

P-T Estimates

developed by Gordon (1999). WEBINVEQ deter-

The garnet peridotite and garnet clinopyroxenite mines the equilibrium conditions for all possible

originated as igneous rocks in the asthenosphere equilibria involving the mineral components. Each

(Abbott et al., 2003b, 2005a). The mineral assem- such equilibrium describes a line in P-T space.

blages constrain the magmatic conditions to P > 3.4 Ideally the lines for all such equilibria would inter-

GPa and T > 1550 C. Previously estimated P-T con- sect at P-T conditions satisfying all of the equilibria.

ditions based on mineral equilibrium show that the Typically the equilibria do not intersect at a well-

minerals reequilibrated under subsolidus condi- defined point, in which case a search is conducted

tions, P = 2.8 3.4 GPa and T = 740-810 C (Abbott for the set of activities that gives the best solution.

et al., 2003b, 2005a). These subsolidus conditions The results provide for error analysis, and assess-

were based on a limited number of chemical analy- ment of uncertainty in the estimated conditions.

ses of minerals, and chemical analyses of the spinel Ellipses in Figure 3 represent 68.3% confidence,

were of questionable accuracy. New chemical analy- assuming an error of 1 kJ for the chemical poten-

ses, presented here, provide for much-improved tial of each mineral component.

estimates of the subsolidus P-T conditions of the Estimated equilibrium P-T conditions for the

garnet ultramafic rocks, and provide for an estimate assemblage Grt + Cpx + Oli + Spl in garnet peridot-

of the P-T conditions for the associated eclogite. ite are 4.15 GPa, 867 C (Fig. 3). Estimated equilib-

P and T were estimated by two means: (1) multi- rium conditions for the assemblage Grt + Cpx + Spl

ple-equilibrium analysis (WEBINVEQ; Gordon, + Crn in garnet peridotite and garnet clinopyroxene

1999) and (2) Fe-Mg exchange thermometry involv- are similar, 3.00 GPa, 838 C, and 2.75 GPa, 807 C,

ing pairs of minerals, Grt-Cpx and Grt-Hbl. Results respectively. Equilibrium conditions for the assem-

are reported in Table 4, and portrayed in Figure 3. blage Grt + Cpx + Pl + Qtz in eclogite are 1.76 GPa,

786 ABBOTT ET AL.

FIG. 3. Estimated P-T conditions. The shaded region (P > 3.5 GPa, T > 1550 C) defines the magmatic conditions for

the garnet ultramafic rocks. Filled circles represent results of WEBINVEQ calculations. The ellipses represent 68.3 %

confidence, assuming 1 kJ error in the chemical potential of each mineral component. Results of Cpx-Grt thermometry

are represented by the steep fine lines. Results of Grt-Hbl thermometry are represented by the fine vertical lines, termi-

nating at about 2.5 GPa, the approximate upper pressure limit for hornblende. Abbreviations: D/G = diamond-graphite;

C/Q = coesite-quartz.

731 C, and relate to the decomposition of omphacite Cpx model agrees most closely with the WEBIN-

to plagioclase + diopsidic clinoproxene. The four VEQ conditions.

WEBINVEQ P-T conditions fall essentially on a Temperatures were also calculated for Fe-Mg

single line of steep P-T slope (Fig. 3). exchange between garnet and hornblende, using the

model of Graham and Powell (1984). Results are

Fe-Mg exchange thermometry shown in Figure 3 as the vertical lines terminating at

Temperatures were calculated at two pressures about 2.5 GPa, the approximate upper pressure limit

(1.5 GPa, 3.0 GPa) for three models of Fe-Mg for hornblende (Gilbert et al., 1982; Poli and Fuma-

exchange between garnet and clinopyroxene (Ai, gali, 2003). At P 1550 C, > 3.4 GPa (Abbott et al., 2003b, other UHP rocks (e.g., Carswell and Zhang, 1999).

2005a). 3. Relatively rapid, non-adiabatic decompres-

1. Slow, approximately isobaric (~4 GPa) cooling sion and cooling. H2O entering and passing through

in the mantle, down to approximately 850 C. No the system in the formation of hornblende and

signature of magmatic conditions is preserved in the serpentine would enhance heat loss through hydro-

major-element chemistry of the minerals. The lack thermal fluid convection, hence increasing the rate

of chemical zoning in garnet and clinopyroxene of cooling.

crystal requires a very slow cooling rate for complete P-T conditions for the eclogite fall on the second

reequilibration of the mineral chemistry in the large part of the path. Field observations indicate that the

(cm-scale) crystals. The conditions (4.15 GPa, garnet ultramafic rocks are meter-scale pods in the

788 ABBOTT ET AL.

FIG. 5. Simplified plate tectonic interpretation. Three parts of the P-T path for the garnet ultramafic rocks are num-

bered (1, 2, 3) as in Figure 4. Eclogite originated as oceanic crust.

eclogite, and as such must have been delivered to stage was essentially completed prior to mid-crustal

the surface in the eclogite. An oceanic crustal origin intrusion by dioritic to gabbroic rocks (Rio Boba

for the hornblende schist and gneiss of the Cuaba Complex) of uncertain, but presumably Late Creta-

Gneiss indicates that the early history of the greater ceous, age. The third part of the path (3, Fig. 4)

part of the Cuaba Gneiss was independent of the relates to final uplift to the surface, completed by

early history of the garnet ultramafic rocks. The mid-Eocene time, perhaps in response to initiation

hypothetical P-T path shown in Figure 4 for the of transcurrent tectonics (Mann et al., 1990; Pindell

early history of the eclogite is admittedly poorly and Barrett, 1990; Draper et al., 1994, 1996; Pin-

constrained, and meant only to be consistent with dell, 1994; Pindell et al., 2005).

current models for deep-subduction of oceanic crust

(e.g., Gerya et al., 2002; Roselle and Engi, 2002; Acknowledgments

Roselle et al., 2002; Gerya and Yuen, 2003).

The project is supported by National Science

Foundation Grants EAR-8306145, EAR-8509542,

Conclusions

and INT-0139536 to Draper and NSF Grants EAR-

Figure 5 offers a bare-bones tectonic interpreta- 0111471 and INT-0139490 to Abbott. The research

tion. The isobaric part of the P-T path for the garnet was supported by an Appalachian State University

ultramafic rocks (1, Fig. 4) took place in the mantle Research Grant to Abbott. We especially appreciate

and relates to delivery of the rocks to deep-sub- the assistance and patience of Dr. Ruth Dewel,

ducted oceanic crust in mid- to Late Cretaceous Director, College of Arts and Sciences Microscope

time. Garnet ultramafic rocks of the mantle wedge Facility, Appalachian State University.

and deep-subducted oceanic crust were juxtaposed,

and the former cooled, in response to erosion of the

hanging wall of the subduction zone. Alternatively REFERENCES

(not illustrated in Fig. 5), the process may have

Abbott, R. N., Jr., and Draper, G., 1998, Retrograde eclog-

taken place during initiation of subduction, in which

ite in the Cuaba Amphibolite of the Rio San Juan Com-

case the garnet ultramafic rocks may have been

plex, northern Hispaniola, i n 1 5th Caribbean

incorporated, and cooled, at the leading edge of the

Geological Conference, Kingston, Jamaica.

downgoing oceanic crust. In either case, incorpora-

Abbott, R. N., Jr., and Draper, G., 2002, Retrograded

tion of Grt ultramafic rocks in the deep-subducted

eclogite in the Cuaba Amphibolite of the Rio San Juan

oceanic crust (eclogite) marked the end of the first Complex, Northern Hispaniola, in Jackson, T. A., ed.,

part of the path. The second part of the P-T path (2, Caribbean geology: Into the third millennium: Trans-

Fig. 4) was associated with transport up the subduc- actions of the 15th Caribbean Geological Conference:

tion zone, perhaps by reverse flow, as modeled by Kingston, Jamaica, The University of the West Indies

Gerya et al. (2002) and Gerya and Yuen (2003). This Press, p. 97 108.

789

UHP GARNET ULTRAMAFIC ROCKS

Abbott, R. N., Jr., Draper, G., and Keshav, S., 2001a, Gar- Draper, G., and Nagle, F., 1991, Geology, structure, and

net peridotite found in the Greater Antilles: EOS tectonic development of the Rio San Juan Complex,

(Transactions of the American Geophysical Union), v. northern Dominican Republic, in Mann, P., Draper,

82, no. 35, p. 381 388. G., and Lewis, J. F., eds., Geologic and Tectonic devel-

opment of the North American Caribbean plate

Abbott, R. N., Jr., Draper, G., and Keshav, S., 2001b,

boundary in Hispaniola: Geological Society of Amer-

Alpine-type garnet peridotite in the Caribbean [abs.]:

ica Special Paper 262, p. 77 95.

Geological Society of America Abstracts with Pro-

grams, v. 33, no. 6, Boston. Eberle, W., Hirdes, W., Muff, R., and Pelaez, M., 1982,

The geology of the Cordillera Septentrional (Domini-

Abbott, R. N., Jr., Draper, G., and Keshav, S., 2003a, Tec-

can Republic), in Snow, W., Gil, N., Llinas, R., Rod-

tonic implications of UHP garnet peridotite in the

riguez-Torres, R., and Tavares, I., eds., Transactions,

Cuaba Unit, Rio San Juan Complex, Dominican

9th Caribbean Geological Conference, Santo Domingo,

Republic [abs.]: Geological Society of America

Dominican Republic, 1980, p. 619 632.

Abstracts with Programs, v. 35, no. 2, Puerta Vallarta.

Gerya, T. V., Stockert, B., and Perchuk, A. L., 2002, Exhu-

Abbott, R.N., Jr., Draper, G., and Keshav, S., 2003b, UHP

mation of high-pressure metamorphic rocks in a sub-

magmatic paragenesis, garnet peridotite, Cuaba Unit,

duction channel: A numerical simulation: Tectonics, v.

Rio San Juan Complex, Dominican Republic [abs.]:

21, 6-6, 1056 [doi:10.1029/2002TC001406].

Geological Society of America Abstracts with Pro-

grams, v. 35, no. 6, Seattle. Gerya, T. V., and Yuen, D. A., 2003, Rayleigh-Taylor

instabilities from hydration and melting propel cold

Abbott, R. N., Jr., Draper, G., and Keshav, S., 2005a, UHP

plumes at subduction zones: Earth and Planetary

magma paragenesis, garnet peridotite, and garnet

Science Letters, v. 212, p. 47 62.

clinopyroxenite: An example from the Dominican

Republic. International Geology Review, v. 47, 233 Gilbert, M. C., Heltz, R. T., Popp, R. K., and Spear, F. S.,

247. 1982, Experimental studies of amphibole stability, in

Veblen, D. R., and Ribbe, P. H., eds., Amphiboles:

Abbott, R. N., Jr., Draper, G., and Keshav, S., 2005b, UHP

Petrology and experimental phase relationships:

metamorphism in garnet peridotite, Cuaba unit, Rio

Reviews in Mineralogy, v. 9B, p. 229 354.

San Juan complex, Dominican Republic, in Draper,

G., ed., Transactions of the 16th Caribbean Geological Gordon, T., 1999, Generalized thermobarometry: WEBIN-

Conference, Barbados: Caribbean Journal of Earth VEQ w ith the T WQ 1.02 data base [http: //

Science, v. 39, p. 13 20. ichor.geo.ucalgary.ca/~tmg/Webinveq/rgb95.html].

Ackermand, D., Seifert, F., and Schreyer, W., 1975, Insta- Graham, C. M., and Powell, R., 1984, A garnet-horn-

bility of sapphirine at high pressures: Contributions to blende geothermometer: Calibration, testing, and

Mineralogy and Petrology, v. 50, p. 79 92. application to the Pelona Schist, Southern California:

Journal of Metamorphic Geology, v. 2, p. 13 21.

Ai, Y., 1994, A revision of the garnet-clinopyroxene Fe2+-

Mg exchange geothermometer: Contributions to Miner- Holland, T. J.B., and Powell, R., 2000, AX, Mineral activ-

alogy and Petrology, v. 115, p. 467 473. ity calculation for thermobarometry: Cambridge, UK,

Cambridge University, computer program AX2, v. 2.2.

Broman, B. N., Abbott, R. N., Jr., and Draper, G., 2005,

P-T path for UHP garnet ultramafic rocks, Cuaba Jansma, P. E., Mattioli, G. S., Lopez, A., DeMets, C.,

Gneiss, Rio San Juan Complex, Dominican Republic: Dixon, T. H., Mann, P., and Calais, E., 2000, Neotec-

Geological Society of America Abstracts with Pro- tonics of Puerto Rico and the Virgin Islands, northeast-

grams, v. 37, no. 6, Salt Lake City. ern Caribbean, from GPS geodesy: Tectonics, v. 19, p.

1021 1037.

Carswell, D. A., and Zhang, R. G., 1999, Petrographic

characteristics and metamorphic evolution of ultra- Kretz, R., 1983, Symbols for rock-forming minerals:

high-pressure eclogites in plate-collision belts: Inter- American Mineralogist, v. 68, p. 277 279.

national Geology Review, v. 41, p. 781 798. Krogh, E. J., 1988, The garnet-clinopyroxene Fe-Mg geo-

Draper, G., Abbott, R. N., Jr., and Keshav, S., 2002, Indi- thermometer a reinterpretation of existing experi-

cation of UHP metamorphism in garnet peridotite, mental data: Contributions to Mineralogy and

Cuaba Unit, Rio San Juan Complex, Dominican Petrology, v. 99, p. 44 48.

Republic: Sixteenth Caribbean Geological Confer- Lewis, J. F., and Draper, G., 1990, Geology and tectonic

ence, Barbados. evolution of the northern Caribbean region, in Dengo,

Draper, G., Gutierrez, G., and Lewis, J. F., 1996, Thrust G., and Case, J. E., eds., The Caribbean region: Boul-

emplacement of the Hispaniola peridotite belt: Oro- der, CO, Geological Society of America, Geology of

genic expression of the mid-Cretaceous Caribbean arc North America, v. H, p. 77 140.

polarity reversal?: Geology, v. 24, p. 1143 1146.

Copyright © 2006 by V. H. Winston & Son, Inc. All rights reserved.

Contact this candidate