IV CONGRESO DE GEOLOGIA Y MINERIA

GEOLOGÍA REGIONAL
GREG .07

P-T EVOLUTION AND HYDROUS MELTING OF METAPELITES IN THE

PINOS TERRANE (CUBA): TECTONIC IMPLICATIONS FOR THE EARLY
DEVELOPMENT OF THE CUBAN COLLISIONAL BELT

Antonio García Casco (1), Rafael Luís Torres Roldán(1), Guillermo Millán Trujillo(2),
Patrick Monié(3) y Faouziya Haissen(4)
(1) Departamento de Mineralogía y Petrología, Universidad de Granada, Fuentenueva s/n, 18002-
Granada, España, C. Eléct.: agcasco@ugr.es
(2) Instituto de Geología y Paleontología, Via Blanca y Carretera Central, La Habana, Cuba, C. Eléct.:
igpcnig@ceniai.inf.cu
(3) Laboratoire de Tectonique et Géochronologie, URA CNRS 1371, USTL, Place E. Bataillon, 34095
Montpellier Cédex, Francia, C. Eléct.: monie@dstu.univ-montp2.fr
(4) Département de Géologie, Faculté des Sciences, Université Chouaib Doukkali, B.P. 20, El Jadida,
Marruecos

ABSTRACT

Detailed assessment of phase relations in high grade rocks (metapelites and leucocratic segregates) from
the Pinos terrane indicate that peak temperature was attained at relatively high pressure (750±50 ºC at
11±1 kbar), which was followed by strong decompression (down to 3±1 kbar at 575-600 ºC), concomitant
with the development of the main synmetamorphic deformation (D2 and D3) of the rocks. Initially,
decompression was near isothermal and was accompanied by the formation of sillimanite and the
segregation of partial melts of trondhjemitic composition. At lower pressure cooling was progressively
more important until metamorphism terminated during the Uppermost Cretaceous (68±2 Ma at ca. 400 ºC;
40
Ar/39Ar dates on biotite and muscovite). Overall, the P-T- path suggest the occurrence of crustal
thickening, followed by extensional collapse of the thickened crust.
Formation of trondhjemitic melts during decompression of the Pinos terrane points towards the occurrence
of fluid infiltration, which agrees with the known occurrence of ca. 68-60 Ma old subvolcanic activity and
fluid-related ore mobilization and with the early opening of the Yucatan basin in late Cretaceous times.
The extensional event affected a thickened crust which originated during the collision of the southeastern
margin of the Yucatan block with the Cretaceous volcanic island arc of Cuba. The new Ar/Ar dates
indicate that the Pinos terrane was implicated early in that collision, implying that it had been located
southwest along the Yucatan continental margin. Once sutured, the metamorphic-arc complex is likely to
have achieved its present location at the time the collision front was moving N-NE and left-lateral tectonics
were increasingly characterizing the Yucatan/Caribbean plate boundary during the early Paleocene.
INTRODUCTION

The metamorphic basement of the Isle of Pines (or Isle of Youth in the current political division of Cuba),
herein called the Pinos terrane, is one of the so-called Southwestern Cuban Terranes (SWCT) that also
include the Guaniguanico and Escambray terranes. Same as other SWCT, the Pinos terrane has a
continental foundation made of stratigraphic sequences that span Lower-Middle Jurassic to Upper
Cretaceous age, which are thought to have formed along the southern margin of Yucatan block (e.g.,
Millán, 1997; Iturralde-Vinent, 1997; Marton and Buffler, 1999; Pszczólkowski, 1999). This margin was
involved in the Late Cretaceous to Eocene built up of the Cuban Orogenic Belt, accomplished by collision

MEMORIAS GEOMIN 2001, LA HABANA, 19-23 DE MARZO. ISBN 959-7117-10-X GREG- 33

IV CONGRESO DE GEOLOGIA Y MINERIA
GEOLOGÍA REGIONAL
GREG .07

of the Yucatan and North America plate margins with the Great Arc of the Caribbean (Gordon et al., 1997;
Mann, 1999).

Within the framework of other metamorphic terranes in the Greater Antilles, the Pinos terrane is distinctive
in that it underwent intermediate to low pressure metamorphism, eventually reaching high-grade
conditions (Millán, 1997; Babushkin et al., 1990), suggesting that it represents a different tectonic scenario
within this part of the ancient Caribbean plate boundary. The current contribution is a study of high-grade
rocks in the Pinos terrane, aimed at determining their reaction history and P-T path and its implications
with regard to tectonic development at the northern margin of the Caribbean plate.

GEOLOGICAL SETTING, SAMPLES AND METHODS

The main geological features of the Isle of Pines have been described by Millán (1997 and references
therein), Pardo (1990), Pardo and Moya (1988) and Babushkin et al. (1990). The Pinos metamorphic
terrane occupies most of the crystalline core of the island, except to the NW where it is tectonically
covered by the Sabana Grande volcanics, a fragment of the Cretaceous Volcanic Arc. Four main phases
of synmetamorphic deformation have been identified. The main phase D2 transposed early D1 structures,
and developed tight folds (F2), the main foliation (S2) and a stretching NW-trending lineation (L2). The
deformation phases D3 and D4 are late- to post-metamorphic, and developed open folds and crenulations
of S2. Previously available radiometric data from metamorphic rocks (mostly K-Ar) distribute with no
statistical maxima in the range of 78±4 to 49.3±3.8 Ma (see Iturralde-Vinent et al., 1996, for review). Late-
to post-metamorphic acid subvolcanic rocks have K-Ar ages of 60-68 Ma (Buguelski et al., 1985).

High-grade rocks are restricted to the NW area of the island and consist predominantly of well foliated
pelitic gneisses and, to a lesser extent, less severely deformed cm- to m-sized leucocratic pods of
trondhjemitic (s.l.) composition with microgranitic to pegmatitic textures. The leucocratic material
concentrates into diffuse pods that are mingled with the pelitic material, and into well defined layers
parallel to the main foliation (S2) of the host metapelite.

Four samples of pelitic gneiss (IJ9, CM10, IP5, and PIB), and two leucocratic rocks (IJ11 and IJ17) were
selected for detailed textural and mineral composition study. All samples of pelitic gneiss contain
Grt+Bt+Fib+Pl+Qtz+Rt+Ilm+Po+graphite, with kyanite being an important constituent in IJ9, PIB and
CM10, and include barely deformed leucocratic material rich in quartz and plagioclase that forms diffuse
patches or layers following the main foliation. Muscovite occurs in these diffuse pods in samples IJ9 and
PIB, and andalusite is a late reaction product after muscovite in sample IJ9 and after kyanite in sample
CM10. Sample IJ11 is a medium grained texturally igneous-looking rock formed by Qtz+Pl+Ms+Bt, with
minor K-feldspar, apatite and zircon. Sample IJ17 is a pegmatitic rock formed by Qtz+Pl+Ms, with minor
K-feldspar replacing muscovite.

MEMORIAS GEOMIN 2001, LA HABANA, 19-23 DE MARZO. ISBN 959-7117-10-X GREG- 34

IV CONGRESO DE GEOLOGIA Y MINERIA
GEOLOGÍA REGIONAL
GREG .07

All elemental analyses were done at the CIC Labs of the University of Granada. Whole-rock element
determinations were carried out by XRF and ICP-MS. Mineral compositions and electronic images were
obtained using a CAMECA SX-50 machine.

MINERAL ASSEMBLAGES AND PHASE RELATIONSHIPS

Bulk compositions

The samples of pelitic gneiss have bulk compositions typical of ordinary pelites, with K2O > Na2O and
relatively low CaO, otherwise comparable to those of lower grade staurolite-bearing metapelites in the
same terrane. The absence of K-feldspar in these samples can be explained through hydrated melting of
pelite, or through subsolidus, fluid-assisted metamorphic differentiation or metasomatism (Ashworth,
1985), yet the former explanation is more consistent with the REE pattern of trondhjemite sample IJ11
(Figure 1), which lacks of an Eu anomaly and suggests garnet fractionation and a significant participation
of plagioclase in melt formation. Accordingly, we infer that melts rich in Na, analogous compositionally to
sample IJ11, formed through partial melting of the pelitic gneisses.

Sequence of mineral assemblages

Pre-D2 assemblages. The main pre-D2 assemblage in the pelitic gneisses is formed by Grt+Ky+Bt+Rt. An
earlier assemblage is recorded as inclusions of Qtz+Bt+Rt+Po±Ms within garnet and kyanite
porphyroblasts, indicating that these minerals grew in part before muscovite was wasted up during
prograde metamorphism. In samples CM10 and PIB, smaller corroded garnets (0.2-0.5 mm in diameter)
are found in the matrix and as inclusions within kyanite, indicating that at least part of the kyanite had
developed after garnet.

Syn-D2 assemblages. In the pelitic gneisses, ubiquitous replacement of garnet by Bt±Fib±Pl, of kyanite
by fibrolite and of matrix rutile by ilmenite, defines an overall overprinting of the main Grt+Ky+Bt+Rt
assemblage by Fib+Bt+Pl+Ilm. This overprinting was concomitant with the formation of the main external
foliation S2. Oriented trails of syn-S2 Fib+Bt also occur in cm-sized trondhjemitic layers within sample IP5,
indicating that the growth of syn-D2 fibrolite accompanied segregation of the leucocratic material.

Post-D2 assemblages. The trondhjemitic samples IJ11 and IJ17 bear the primary assemblages
Qtz+Pl+Bt+Ms and Qtz+Pl+Ms, respectively. These samples are only slightly deformed, which indicates
that crystallization took place after the major ductile deformation event (D2). In sample IJ11, plagioclase
and quartz form phenocrysts of up to 3 mm in length, and their implications in symplectitic intergrowths
further suggests their late crystallization at the eutectic of the system. K-feldspar, very scarce (<5 % vol),
appears as fine xenomorphic interstitial crystals that are also interpreted to have crystallized late, near or
at the solidus of the rock. Biotite forms medium grained crystals dispersed in the matrix and included
within porphyritic plagioclase. Muscovite is often associated with matrix biotite forming composite Ms-Bt
clots where muscovite overgrows biotite. Fine-grained muscovite (plus Al-silicate, probably andalusite) is

MEMORIAS GEOMIN 2001, LA HABANA, 19-23 DE MARZO. ISBN 959-7117-10-X GREG- 35

IV CONGRESO DE GEOLOGIA Y MINERIA
GEOLOGÍA REGIONAL
GREG .07

also seen to replace plagioclase phenocrysts along preferred crystallographic directions. In the pelitic
gneisses (samples IJ9 and PIB), undeformed (post-D2) muscovite laths appear within the leucocratic-rich
material. These muscovites are medium grained, euhedral in shape and free of inclusions, all of which
points to their late crystallization from interstitial melt during the cooling history of the rocks. These primary
muscovites (plus garnet) reacted later to form post-D2 And+Bt+Pl+Ilm assemblages. Late overprinting
within the andalusite stability field is also denoted by partial replacement (polymorphic inversion) of
kyanite by andalusite.

In summary, replacement textures and fabrics of high grade rocks from the Pinos terrane consistently
record reaction processes that progressed from the kyanite (pre-S2 Ky+Grt+Bt) to the sillimanite (syn-D2
Fib+Bt) to the andalusite (post-D2 And+Bt) stability fields, with syn- to post-kinematic segregation and
crystallization of leucocratic material.

Mineral chemistry

Garnet displays compositional features typical of high grade, with a flat spatial distribution of all the
elements within the cores and steep changes at the rims where Mg# (=Mg/(Mg+Fe) decreases and Mn
increases, the latter resulting from net transfer reactions involving garnet dissolution during decreasing
temperature. Of major petrogenetic importance is that the corroded garnets included within kyanite
porphyroblasts also bear retrograde rims, implying that the retrograde reactions already started within the
kyanite stability field, and thereafter continued in the sillimanite and andalusite stability fields
(replacements by Fib+Bt and And+Bt). Reverse zoning also develops adjacent to biotite inclusions within
garnet althougn in this case the decrease of Mg# is not coupled with an increase in Mn, as expected from
Fe-Mg exchange alone. The significance of all these differences in the retrograde zoning of garnet is
discussed below.

Biotite. In the pelitic gneisses, no clear chemical distinction can be made between matrix grains and
those appearing in the Bt+Als replacements after garnet. Both display high Ti contents (up to 0.4 atoms
per 22 O), typical of high grade biotites from metapelites, and a remarkably constant Mg# notwithstanding
larger variations in absolute Fe or Mg contents. The same trend is apparent in zoned biotite grains that
replace garnet, where Ti decreases and Fe and Mg increase toward the Bt-Grt interface but Mg# is
constant. These observations do otherwise indicate that matrix biotites in these rocks do not retain their
peak-T composition, given that their Mg/(Fe+Mg) ratio is similar to that of replacement biotite. Biotite
inclusions within garnet have intermediate Ti (0.2-0.3 atoms pfu) and distinctively higher Mg# (up to 0.5). It
is likely that these biotites still retain their original (moderate) Ti content, at the time of their trapping within
garnet at medium grade conditions, notwithstanding that their Mg/(Fe+Mg) ratios reflect re-equilibration at
higher grade and during the ensuing cooling history of the rocks through continued Fe-Mg exchange with
host garnet.

Plagioclase. In the pelitic gneisses, plagioclase appears as porphyroblasts, matrix grains and within
replacements after muscovite and garnet. The porphyroblasts are unzoned or bear slight concentric

MEMORIAS GEOMIN 2001, LA HABANA, 19-23 DE MARZO. ISBN 959-7117-10-X GREG- 36

IV CONGRESO DE GEOLOGIA Y MINERIA
GEOLOGÍA REGIONAL
GREG .07

reverse zoning that includes homogeneous cores (e.g., Xan=0.18-0.20 in IJ9) and Ca-richer rims
(Xan=0.20-0.25). Local patches of Na-rich plagioclase (Xan down to 0.03) may overprint the concentric
zoning. We interpret the Na-richer cores represent equilibrium conditions at peak (or near-peak)
temperature. The composition of plagioclase from the matrix and the replacements after muscovite and
garnet is similar to that to the of Ca-rich rims of the porphyroblasts, suggesting that increasing Ca
contents developed as garnet decomposed during decreasing temperature. Plagioclase phenocrysts in
the trondhjemitic samples IJ11 and IJ17 are more albitic than those of the pelitic gneisses. In IJ11, the
phenocrysts bear normal zoning with Ca decreasing toward the rims (Xan=0.28-0.12), and these are
compositionally similar to the matrix crystals (Xan=0.12-0.15. In IJ17, plagioclase phenocrysts are
unzoned, but late plagioclase grown after muscovite is nearly pure albite (Xan=0.05). The contrasting
trends of plagioclase in the pelitic gneisses and trondhjemites reflect their distinct relation to reactions
involving garnet dissolution and evolution of the melt during crystallization, respectively.

Muscovite. Primary muscovites from samples IJ9 and PIB have low Si (6.151-6.125 atoms/22 O) and
high Ti contents (0.094-0.123), typical of high grade and magmatic muscovites. Matrix muscovites from
trondhjemite IJ11 are also Si-poor (6.079-6.167) and Ti-rich (0.095-0.170), indicative of igneous
crystallization. The fine-grained crystals from the Ms+Als replacements after plagioclase phenocrysts are
similar in composition to primary crystals, but with somewhat lower Ti contents (0.095-0.097), and are
though to have also formed at relatively high temperature, perhaps near the solidus when fluids exsolved
from the crystallizing melt. The pegmatitic muscovites of sample IJ17 have slightly higher Si (<6.25) and
Mg (up to 0.25) contents than matrix muscovite from IJ11, and much lower Ti contents (0.008-0.025). This
would suggest a lower temperature of crystallization, although the low Ti content could also derive from
the subsaturated nature of the pegmatitic fluid. No chemical variations were observed in relation with the
late replacements by Kfs+Ab.

Phase relationships

The phase relationships for observed assemblages in sample IJ9 are illustrated in Figure 2. The relations
correspond to early (higher P-T) and late (lower P-T) stages of evolution. At lower P-T the compositions
are better constrained, and muscovite is assumed to have coexisted with Fe-rich garnet and biotite (i.e.,
retrogressed garnet rim composition and biotite replacements after garnet), plus an Fe-rich melt
represented by the composition of trondhjemite IJ11 (note that this melt composition plots within the tie-
triangle Grt-Bt-Ms). For the earlier stage at higher P-T, muscovite was absent and garnet with the
composition of the relict cores (high Mg#) coexisted with biotite and melt of unknown composition. Yet,
Figure 2a shows two possible topologies for this stage, corresponding to Mg#Grt < Mg#melt < Mg#Bt and
Mg#Grt < Mg#Bt < Mg#melt, both of which could have been feasible at peak pressure. Although the
composition of biotite at high P is not known, it can be assumed to have had an Mg# value intermediate
between that of matrix/replacement biotite and that of biotite inclusions within garnet, as done in Fig. 2a.
This is because Mg# changes of matrix/replacement and inclusion biotites are respectively driven in

MEMORIAS GEOMIN 2001, LA HABANA, 19-23 DE MARZO. ISBN 959-7117-10-X GREG- 37

IV CONGRESO DE GEOLOGIA Y MINERIA
GEOLOGÍA REGIONAL
GREG .07

opposite directions by their contrasting reaction histories. Fe-Mg changes in dissolving garnet during
retrogression correlate well with observed plagioclase changes, yet disequilibrium is suspect because the
amounts of Ca across retrograde garnet rims are constant, no matter of the extent of retrogression. Given
that the chances that a P-T path follows exactly the traces of Xgrs surfaces in every rock are insignificant,
this reflects that dissolving garnet rims failed to equilibrate with growing plagioclase.

METAMORPHIC CONDITIONS AND P-T-t PATH

Equilibrium problems certainly apply to the studied samples, yet meaningful approximate results can be
obtained by careful consideration of the observed phase relations within the context of the equilibria being
used. For that purpose we have used the T-sensitive Fe-Mg exchange between garnet and biotite (GARB
thermometer) and the P-sensitive net-transfer reaction among garnet, Al-silicate, quartz and plagioclase
(GASP barometer), to estimate P-T conditions at apparent peak-T. Additionally, a number of P-sensitive
reactions involving muscovite were used for the Grt-Bt-Als-Ms retrograde assemblages. For the
calculations we used the TWQ software (v. 2.02) of Berman (1991).

Peak P-T conditions

A precise estimate of peak-T conditions is precluded in the studied samples because biotite composition
continuously readjusted during retrogression. A “best” estimate can be calculated, nonetheless, by taking
into account that core compositions of the largest garnet grains and of the those included within kyanite
should most closely resemble the composition at peak-T, and that the Mg# value of the original peak-T
biotite must lie within the bracket defined by the maximum Mg# of matrix/replacement biotite and the
minimum Mg# of included biotite, as explained earlier. The corresponding temperature bracket so
calculated is in the range of 700-800 ºC (Fig. 3a) for a total calculated range of pressures of 8-12 kbar,
using same compositions of garnet cores and the homogeneous cores of plagioclase porphyroblasts. Our
preferred pressure estimate, extracted from samples IJ9 and PIB at a peak temperature of 750±50 ºC, is
11±1 kbar.

Conditions during retrogression

To estimate post peak P-T conditions we have used compositions of adjacent biotite-garnet pairs, Ca-rich
plagioclase (from grains produced after garnet and rims of porphyroblasts) and igneous muscovite (when
present). Yet, the use of these compositions translates into large uncertainties (575-700 ºC). These
temperatures are not likely to represent a sequence of local equilibrium along the retrogression path, but
rather problems with the spatial resolution of the microprobe when adjacent Bt-Grt pairs are to be
analyzed. Because of this, we judge the lower limit of 575-600 ºC as a best maximum estimate for this
stage. A large uncertainty is also associated to pressure estimates for retrogression (3-7 kbar). The
calculations based on the GASP barometer yield pressures more than 2 kbar higher than Ms-bearing
equilibria (Figure 3). This difference is thought to represent mostly disequilibrium among garnet rims and
replacing plagioclase, as inferred above, rather than a sequence of blocking pressures of the different net

MEMORIAS GEOMIN 2001, LA HABANA, 19-23 DE MARZO. ISBN 959-7117-10-X GREG- 38

IV CONGRESO DE GEOLOGIA Y MINERIA
GEOLOGÍA REGIONAL
GREG .07

transfer reactions during decompression. As a consequence, the pressure estimates based on equilibria
involving the anorthite component of plagioclase and the grossular component of garnet are maximum
estimates. Indeed, the P-sensitive equilibrium Alm + Ms = Als + Ann + Qtz, unaffected by these problems,
yields a tighter pressure estimate of 3±0.5 kbar (Figure 3) that should correspond to the late replacement
of garnet and muscovite by Pl+Bt+And.

P-T-t relations

Thermobarometric calculations suggest that the studied rocks reached a peak temperature of 750±50 ºC
at the pressure of 11±1 kbar, then decompressed down to ca. 3 kbar without much cooling (i.e., down to
575-600 ºC). Yet, the maximum stability limit of muscovite plus quartz was not intersected during
decompression, as denoted by the absence of Kfs+Als replacing muscovite in all types of rock, and this
implies that some cooling did indeed accompany the decompression, in all likelihood along a decreasing
dP/dT slope as tentatively shown in Figure 3. The time at which peak conditions were achieved is not
radiometrically constrained. Yet, biotite from pelitic gneiss IJ9, and biotite and muscovite from
40
trondhjemite IJ11 give Ar/39Ar plateau ages that cluster at 68±2 Ma. These ages, which are thought to
reflect a range of closure temperatures between 370-430 ºC, indicate that final cooling and exhumation of
the rocks took place in the Uppermost Cretaceous. The overlap of cooling ages for muscovite and biotite
points towards high rates of cooling along the low pressure section of the P-T path, which is consistent
with the occurrence of post-metamorphic subvolcanic dikes with whole-rock K-Ar ages of 68-60 Ma
(Buguelski et al., 1985). High rates of cooling at low pressure, plus the general form of the retrograde path
as shown in Figure 3, make conceivable that ending of metamorphism in the Pinos terrane had been
related to extensional collapse of earlier thickened crust.

DISCUSSION AND CONCLUSIONS

The main result of our analysis of phase relations indicates that the studied rocks experienced strong
decompression (from 11 down to 3 kbar) at relatively high temperature (> 600 ºC) in the presence of
partial melts of trondhjemitic composition. It is also inferred that peak temperature was attained at high
pressure (ca. 750 ºC and 11 kbar), and that the growth of Al-silicates (including part of the kyanite) took
place upon decompression and cooling, rather than during prograde development. Further,
decompression and cooling from the kyanite down the andalusite stability fields are consistent with
decompression dominating at high to intermediate pressure, while cooling was progressively more
important at low P until metamorphism terminated during the Uppermost Cretaceous (68±2 Ma at ca. 400
ºC). Paths with steeper dP/dT slopes at high P and low dP/dT slopes and high rates of cooling at low P,
are deemed typical of metamorphic regions affected by tectonic extension. Thus, we propose that the
deformations D2 and D3 (defined by orientation of fibrolite) of the Pinos terrane are the mechanical results
of tectonic extension rather than compression. The occurrence of trondhjemitic melts do further point that
fluids were available during tectonic extension. The source of any such fluids is a matter of speculation,
although the Pinos terrane has been known for some time to have underwent a “tectonometamorphic

MEMORIAS GEOMIN 2001, LA HABANA, 19-23 DE MARZO. ISBN 959-7117-10-X GREG- 39

IV CONGRESO DE GEOLOGIA Y MINERIA
GEOLOGÍA REGIONAL
GREG .07

reactivation” which, as described by Pardo (1990), included subvolcanic activity and fluid-related ore
mobilization in a distensive regime dated as being ca. 68-60 Ma old. In our view, such
“tectonometamorphic reactivation” is likely to have been related to the extensional event recorded in the
studied rocks, which started at high temperature and continued during cooling of Pinos terrane.
Extensional tectonics are otherwise consistent with the metamorphic break that occurs at the tectonic
contact between the Pinos terrane and the overlying unmetamorphosed Sabana Grande volcanics, which
could be regarded as a low-angle normal fault, analogous to those bounding metamorphic core-
complexes.

At present there is a large latitude in the geotectonic interpretation of the inferred Upper Cretaceous
extensional collapse of the Pinos terrane. Interestingly, however, the nearby Yucatan basin opened during
a brief period of NE extension sometime between the Late Cretaceous and the Eocene (Rosencrantz,
1996), which allows envisaging that these two events had been somehow connected. As a working
hypothesis, we suggest that the retrograde metamorphism in the Pinos terrane records an early stage of
development of the Yucatan basin in a context of tectonic extension. Further, the Pinos terrane is likely to
represent the collapsed remains of a thickened crust that originated from the collision of the southeastern
margin of the Yucatan block with the Cretaceous volcanic island arc of Cuba, represented here by the
Sabana Grande volcanics. The collision of the Great Arc of the Caribbean with the North- and South-
America plate margins is diachronous, and commenced in the Late Cretaceous in northern Central
America (e.g., Rosenfeld, 1990; Mann, 1999). In the Guaniguanico terrane in Western Cuba, the collision
is estimated to have taken place in Paleocene-Early Eocene times (e.g., Bralower and Iturralde-Vinent,
1997; Gordon et al., 1997). Our Ar/Ar dates in the Pinos terrane would hence suggest that during the
Mesozoic the Isle of Pines was located farther southwest in the Yucatan Block, relative to the
Guaniguanico terrane, so that it was implicated at an early stage in the collision with the Great Arc. This
early sutured metamorphic-arc terrane could have thereafter migrated northeastwards, and achieved its
present location at the time the collision front was moving N-NE and left-lateral tectonics were increasingly
characterizing the Yucatan/Caribbean plate boundary during the early Paleocene (Pindell and Barrett,
1990; Lara, 1993).

Acknowledgements

We wish to thank José Tomás Cordova (Empresa Geólogo-Minera Isla de la Juventud) for his help during
field work in the Pinos Island. This contribution has received financial support from the Spanish DGES-
MEC (Project PB96-1426) and the Cuban MINBAS.

REFERENCES

Ashworth, J.R. (1985): Introduction. In: Migmatites (J. R. Ashworth, Ed.). Blackie and Sons Ltd. Glasgow,
1-35.
Babushkin, V. & others (1990): Informe de los trabajos de levantamiento geológico-geofísicos a escala
1:50000 y búsquedas acompañantes en el municipio especial Isla de la Juventud en colaboración

MEMORIAS GEOMIN 2001, LA HABANA, 19-23 DE MARZO. ISBN 959-7117-10-X GREG- 40

IV CONGRESO DE GEOLOGIA Y MINERIA
GEOLOGÍA REGIONAL
GREG .07

con la URSS. Oficina Nacional de Recursos Minerales. MINBAS. La Habana, Cuba (unpublished
report).
Berman, R.G. (1991). Thermobarometry using multi-equilibrium calculations: A new technique, with
petrological applications. Canadian Mineralogist, 29, 833-855.
Bralower, T.J. & Iturralde-Vinent, M.A. (1997): Micropaleontological dating of the collision between the
North American plate and the Greater Antilles arc in western Cuba. Palaios, 12, 133-150.
Buguelski, Y., Vázquez, O. & Grigorieva, I. (1985): Ore deposits of Cuba (in Russian). Ed. Nauka,
Moscow, 246 p.
Gordon, M.B., Mann, P., Cáceres, D., & Flores, R. (1997): Cenozoic tectonic history of the North America-
Caribbean plate boundary in western Cuba. Journal of Geophysical Research, 102, 10055-10082
Iturralde-Vinent, M.A. (1996): Introduction to Cuban Geology and Geophysics. In: Ofiolitas y Arcos
Volcánicos de Cuba (M. A, Iturralde-Vinent, Ed.), IGCP Project 364 Spec. Contrib. 1, 3-35. Miami,
USA.
Iturralde-Vinent, M.A. (1997): Introducción a la Geología de Cuba. In: Estudios sobre Geología de Cuba
(G. F. Furrazola Bermúdez & K. E. Núñez Cambra, Eds.), 35-68. Centro Nacional de Información
Geológica, La Habana, Cuba.
Iturralde-Vinent, M.A., Millán, G., Korkas, L., Nagy, E. & Pajón, J. (1996): Geological interpretation of the
Cuban K-Ar data base. In: Ofiolitas y Arcos Volcánicos de Cuba (M. A, Iturralde-Vinent, Ed.), IGCP
Project 364 Spec. Contrib. 1, 48-69. Miami, USA.
Lara, M.E. (1993): Divergent wrench faulting in the Belize Southern Lagoon: Implications for Tertiary
Caribbean plate movements and Quaternary reef distribution. American Association of Petroleum
Geologists Bulletin, 77, 1041-1063.
Mann, P. (1999): Caribbean sedimentary basins: Classification and tectonic setting. In: Caribbean Basins
(P. Mann, Ed.), Sedimentary Basins of the World, 4, 3-31. Elsevier Science B.V., Amsterdam.
Marton, G.L. & Buffler, R.T. (1999): Jurassic-Early Cretaceous tectono-paleogeographic evolution of the
Southeastern Gulf of Mexico basin. In: Caribbean Basins (P. Mann, Ed.), Sedimentary Basins of the
World, 4, 63-91. Elsevier Science B.V., Amsterdam.
McDonough, W.F. & Sun, S.S. (1995): The composition of the Earth. Chemical Geology, 120, 223-253.
Millán G. (1997): Geología del macizo metamórfico de la Isla de la Juventud. In: Estudios sobre Geología
de Cuba (G. F. Furrazola Bermúdez & K. E. Núñez Cambra, Eds.), 259-270. Centro Nacional de
Información Geológica, La Habana, Cuba.
Pardo, M. (1990): La constitución geológica del macizo de la Isla de la Juventud y metalogenia endógena
vinculada al magmatismo ácido. Transactions of the 12th Caribbean Geological Conference, St-
Croix U.S. Virgin Islands. Miami Geol. Soc. p. 68-91.
Pardo, M. & Moya, I. De (1988): Nuevos datos sobre la estructura geológica de la Isla de la Juventud.
Parte II: Estructura interna del macizo metamórfico. Revista Tecnológica 18 (3), 3-12.
Pindell, J. L. & Barrett, S. F. (1990): Geologic evolution of the Caribbean region; A plate-tectonic
perspective. In: The Caribbean Region (G. Dengo & J. E. Case, Eds.). The Geology of North
America, v. H, 405-432. The Geological Society of America, Boulder, Colorado, USA.
Pszczólkowski, A. (1999): The exposed passive margin of North America in Western Cuba. In: Caribbean
Basins (P. Mann, Ed.). Sedimentary Basins of the World, 4, 93-121. Elsevier Science B.V.,
Amsterdam.
Rosencrantz, E. (1996): Basement structure and tectonics in the Yucatan basin. In: Ofiolitas y Arcos
Volcánicos de Cuba (M.A, Iturralde-Vinent, Ed.), IGCP Project 364, Spec. Contrib. 1, Miami, USA,
pp. 36-47.
Rosenfeld, J. H. (1990): Sedimentary rocks of the Santa Cruz ophiolite – a proto-Caribbean history.
Transactions of the 12th Caribbean Geological Conference, U.S. Virgin Islands, pp. 513-519.

MEMORIAS GEOMIN 2001, LA HABANA, 19-23 DE MARZO. ISBN 959-7117-10-X GREG- 41

IV CONGRESO DE GEOLOGIA Y MINERIA
GEOLOGÍA REGIONAL
GREG .07

1000
IJ9G
IJ9P
IJ11
100

Rock/condrite
10

0.1
La* Ce* Pr* Nd* Sm*Eu* Gd* Tb* Dy* Ho* Er* Tm* Yb* Lu*

Figure 1: Chondrite-normalized REE patterns for pelitic gneiss IJ9P, Qtz+Pl rich zone IJ9G, and
trondhjemite IJ1. Chondrite composition after McDonough and Sun (1995).
KAlO 2 Al 2O3
+Qtz +Qtz
Ky/Fib/And
+Als +Ms
(a) Ms + Pl (core) (b) +Pl (core)
+Ilm +melt (IJ11)
+Ap +Ilm
+H2O +Ap
IJ11 melt +H2O

low P-T high P-T

CM10
matrix Bt Bt inclusions
Grt rim Grt core
hypothetical IJ9G
high P-T biotite
IJ9G, IJ9P,CM10 and melts IJ9P

FeO (rim) Grt (core) MgO FeO MgO

hypothetical
high P-T biotite
matrix/ Bt incluions
replacement Bt

Figure 2: KFM and AFM diagrams illustrating phase relationships in pelitic gneiss IJ9. In both diagrams,
early (higher P) and late (lower P) assemblages are tied with solid and dashed lines, respectively.

14
Bt matrix/replac. (max T) (a) 14
key to Grt-Bt-Pl-Ms equilibria
(b)
Bt inc. Grt (min T) ms
al s+p hl

rs+
prp+m

+Grt core +g an
+Pl core alm ann+
+qtz

12 12
s

Ky prp+ms+ grs Ky
qt z an+phl
l rs+ ) l
Si +g S P
als (GA Si
10 10 an
s
+m t z
Pressure (kbar)

alm+phl
Pressure (kbar)

alm nn+q
prp+ann

qt z + a
als
s+
8 +gr 8
n
als a
B
PI J9
I 10
alm+phl
prp+ann

6 6
CM
IJ9
5 CM10
4 IP 4 IP5
Ky Ky
A nd An
d
2 Sil 2 Si l
tz O tz O
Bt inclusions +
+Q H2 An +Q H 2 An Grt host pairs
Ms Al s+ d Ms Al s+ d Bt+Grt+Pl triplets
s+ +
0 Kf 0 Kfs
400 500 600 700 800 900 400 500 600 700 800 900
Temperature (ºC) Temperature (ºC)

Figure 3: P-T results for a) peak P-T and b) retrogression of pelitic gneisses IJ9, CM10, IP5 and PIB.

MEMORIAS GEOMIN 2001, LA HABANA, 19-23 DE MARZO. ISBN 959-7117-10-X GREG- 42