Central Cuba and Regional Tectonic Implications

Precambrian Research, 42 (1989) 325-341 325 Elsevier Science Publishers B.V., Amsterdam -- Printed in The Netherlands

4°Ar/agAr AND U-Pb EVIDENCE FOR LATE PROTEROZOIC (GRENVILLE-AGE) CONTINENTAL CRUST IN NORTH- CENTRAL CUBA AND REGIONAL TECTONIC IMPLICATIONS

PAUL R. RENNE 1, JAMES M. MATTINSON 2, CHARLES W. HATTEN 3, MARK SOMIN 4, TULLIS C. ONSTOTT 5, GUILLERMO MILLAN 6 and EVELIO LINARES 7

~Department of Geological and Geophysical Sciences, Princeton University, Princeton, NJ 08544 (U.S.A.) eDepartment of Geological Sciences, University of California, Santa Barbara, CA 93106 (U.S.A.) :~San Carlos Oil and Gas Corporation, 1515 Hope Street, Suite 204, South Pasadena, CA 91030 (U.S.A.) 4Institute of Physics of the Earth, Soviet Academy of Sciences, Bolschaya Gruzinskaya str. 10, 123810 Moscow D-242 (U.S.S.R.) "SDepartment of Geological and Geophysical Sciences, Princeton University, Princeton, NJ 08544 (U.S.A.) 6Instituto de Geologia, Academia de Ciencias de Cuba, La Habana (Cuba) 7Centro de Investigaciones Geologicas, Oficios I54, e/Teniente Rey y Amargura, La Habana I (Cuba)

(Received February 1, 1988; revision accepted March 30, 1988 )

Abstract

Renne, P.R., Mattinson, J.M., Hatten, C.W., Somin, M., Onstott, T.C., Millfin, G. and Linares, E., 1989. 4°Ar/39Ar and U-Pb evidence for Late Proterozoic (Grenville-age) continental crust in north-central Cuba and regional tectonic implications. Precambrian Res., 42: 325-341.

Central Cuba is composed of fault-bounded tectonostratigraphic terranes juxtaposed and deformed during plate collision and subsequent transform motion between the Caribbean and North American plates in the Late Mesozoic- Cenozoic. One of these, the Las Villas terrane, contains crystalline basement rocks thought to be pre-Upper Jurassic on stratigraphic grounds. The Socorro Complex occurs in the northwestern Las Villas terrane, and consists of marbles and siliciclastic metasedimentary rocks, and the Rio Carla Granite. An 4°Ar/39Ar plateau date of 903.5 _+ 7.1 Ma for phlogopite from a marble corroborates previous K-Ar dates from this unit, and establishes unambiguously a Late Proterozoic age for high-grade metamorphism. Discordance of the 4°Ar/agAr age spectrum can be reliably attributed to diffusive Ar loss, which if modeled as an episodic thermal event implies a reheating age that closely coincides with the Late-Cretaceous-Paleogene collision between the Caribbean and North American plates. U-Pb zircon data indicate an intrusive age of 172.4 Ma for the Rio Carla Granite, and reveal an inherited zircon component with an age of ~ 900 Ma. Radio-isotopic data from the Socorro Complex display no evidence of Pan-African age thermal overprinting. These observations, combined with constraints provided by published results from nearly Pangean landmasses, suggest that the complex lay substantially to the southwest (present-day co-ordinates) during the Early Paleozoic. Following its genesis in the mid-Mesozoic, the Caribbean plate evidently transported fragments of an extensive Gren- ville-age belt that spanned the Americas during the Late Proterozoic.

Introduction ied extensively in the last decade (e.g., Pindell and Dewey, 1982; Sykes et al., 1982; Wadge and The tectonic evolution of the Caribbean plate Burke, 1983; Mattson, 1984; Duncan and Har- and its northern boundary zone has been stud- graves, 1984; Burke et al., 1984; Pindell, 1985),

0301-9268/89/$03.50 © 1989 Elsevier Science Publishers B.V. 326 during which time considerable progress has Regional geology of central Cuba been made in understanding the Cretaceous and younger geologic history of the region in terms Central Cuba comprises several distinct tec- of plate tectonic processes. The pre-Late Jur- tonostratigraphic terranes {used in the non-ge- assic history of the northern Caribbean region netic sense of Howell et al., 1985), shown in Fig. remains much more poorly understood because 1, that are imbricated along WNW-ESE trend- rocks known to be this old are relatively rare ing faults. As discussed by Somin and Mill~in and are generally not well dated. Collision be- (1977, 1981), Mill~in and Somin (1981, 1985) tween the North American and Caribbean and Hatten et al. (1988), these terranes con- plates during the Late Cretaceous-Eocene has tain lithologic suites characteristic of conver- produced intense deformation that obscures gent plate margins, including volcanic and plu- earlier stratigraphic and structural relation- tonic arc-related rocks (Zaza and Manicaragua ships. Post-Eocene sinistral transform motion terranes), ortho-amphibolites (Manicaragua between these plates has produced major rota- terrane), high pressure-low temperature tions (e.g., Gose and Eby, 1987) which further (blueschist and eclogite facies) tectonites (Es- complicate geologic relations. The effect of this cambray terrane), ophiolite fragments (Zaza motion in Cuba is largely unknown, but prelim- terrane), and serpentinite melanges {espe- inary paleomagnetic data {P.R. Renne, unpub- cially along thrusts within and between ter- lished data) suggest that parts of central Cuba ranes). The imbrication of these terranes, and have undergone ~80 ° of post-Early Creta- much of the deformation and metamorphism of ceous counterclockwise rotation with respect to rocks composing them, took place during the the North American plate. Late Cretaceous-Paleogene collision of the The origin of the Caribbean plate as a tec- Caribbean and North American plates (Hatten tonic entity, as envisaged by Pindell and Dewey et al., 1988). Evidence reviewed by Mattson (1982) and most subsequent investigators, fol- (1984), Burke et al. (1984) and Pindell (1985) lowed the rifting apart of Pangea in the Middle implies that these terranes originated on the Jurassic. Shortly thereafter, the northeastern Caribbean plate. The active plate boundary in margin of the Caribbean plate was the site of this area since ~ 36 Ma ago has been the Cay- the Cretaceous magmatic arc (s) of the Greater man trough, south of Cuba (Pindell and Dewey, Antilles, marking the zone of convergence be- 1982; Burke et al., 1984), suggesting that Cuba tween the Caribbean and North American was sutured to the North American plate dur- plates. The nature and age of the basement upon ing the collision. which this arc was constructed are poorly constrained, as is the location of its origin. The Las Villas terrane The pre-rift tectonic history of the Carib- bean region has obvious implications for the re- The Las Villas terrane (Hatten et al., 1988) lationship of the North and South American comprises a belt of highly deformed Late Jur- continents prior to the Middle Jurassic. This assic to Early Eocene sedimentary rocks un- paper presents data that provide constraints on conformably overlying diverse types of crystal- the nature, age and original location of crustal line basement. In the southeastern Las Villas fragments that eventually were transported by terrane defined by Hatten et al. (1958, 1988), the Caribbean plate. Some applications of these basement complexes include (from oldest to constraints to the paleotectonic relationship of youngest) the Perea metabasites, the San Mar- the Laurentian and Amazonas cratons of North cos Troctolite and the Tres Guanos granitoids. and South America respectively, during the Late Cross-cutting relationships discussed by Hat- Proterozoic are presented in the following. ten et al. (1958, 1988) permit the delineation 327

80 ° 79 ° I I

23 °

.22 °

~ LASVILLAS ZAZA o.~ so KM

MANICARAGUA/ESCAMBRAY ! ! Fig. 1. Tect~nostratigraphic te~anes of central Cuba, modified from Hatten eta]. (1988) and discussed in text. Contacts between te~anes are faults. Cross-hatched area is shown in Fig. 2. of relative ages of these units. K-Ar dates of (1) The San Marcos 'Troctolite' (as ob- 79+2 Ma on whole rock (Somin and Mill~in, served in the Jarahueca Fenster) is actually a 1981) and 61 + 1 Ma on biotite (Meyerhoff et leucogabbro with anorthositic affinity, much al., 1969) samples have been reported for the more reminiscent of anorthosite complexes or Tres Guanos quartz monzonite, but Hatten et layered basic intrusions than of ophiolites; al. (1958, 1988) presented stratigraphic argu- (2) The Tres Guanos granitoids are compo- ments that this granitoid has a pre-Tithonian sitionally distinct from the typical plagiogran- (pre-152 Ma; Palmer, 1983) intrusive age, and ites of ophiolite suites. Further research (in speculated that the K-Ar dates reflect a tecton- progress) on these rock units is needed before othermal overprint, presumably related to plate their age and tectonic significance can be collision. addressed. Somin and Mill~in (1981) considered the Basement of the northwestern Las Villas ter- Perea metabasites, San Marcos Troctolite and rane, the focus of this paper, is herein referred Tres Guanos quartz monzonite to be part of an to as the Socorro Complex, consisting of de- Upper Jurassic (?)-Cretaceous ophiolite asso- formed granitic and metasedimentary rocks de- ciation, but several factors weigh against this scribed below. interpretation: 328

Socorro Complex ologic relationships are often obscured. Rela- tionships among the various lithologic units The Socorro Complex is exposed in the area shown in Fig. 2 are poorly exposed and thus may shown in Fig. 2, where it crops out poorly in a be subject to slightly different interpretations. region of subdued topography. This complex Our observations and mapping strongly suggest comprises the Rfo Carla Granite and Sierra that the Socorro Complex is allochthonous, its Morena Marbles discussed by Somin and Mil- northeastern border resting with tectonic dis- l~n (1977, 1981). Best exposures of the com- cordance on highly deformed thin-bedded Late plex occur in streambeds of the area {particu- Jurassic carbonates of the Veloz Formation and larly the Rio Carla and its tributaries), but even Early Cretaceous cherts of the Amaro Forma- in these the exposure is discontinuous and lith- tion. To the southwest, the crystalline complex

[-9 Tertiary Corralillo Amaro Cherts (lower K) ~ 0 cherts w/shaly interbeds ~] Veloz Fm. (upper J): thin - bedded limestones Constancia Fro. (upper J[?}): arkosic conglomerates

\ Rio Ca6a Granite (middle J )~ mylonitized granite [] Metasedimentary Rocks (pre-900 Ma)= marbles and quartzites

acts

ion

Fig. 2. Geologic map of the Rio Carla Area showing the Socorro Complex. 329

A A

,M, ,h' ' '~ I ......

, , ,,,, ,, ',' ' ,,,,,,i,,,,,,L,, ...... 1 - ,,,,,,::i!ii:,ii!i!i:',iil;!i:!i!:,!i i

...... '.'".'.',;','i"i ..... ',2" " "~" "." "

0 5km | i i I I I

Fig. 3. Interpretive cross-section along A-A' in Fig. 2. appears to be overlain unconformably by the CaO, 4.63% K20, and 3.90% Na20 by weight. Veloz and Amaro Formations (Hatten et al., Locally, the granite is highly deformed and 1958; Pszczolkowski, 1986). Arkosic conglom- commonly possesses a cataclastic texture fea- erates of the Constancia Formation, derived turing porphyroclasts of feldspars and strained from the complex, occur discontinuously be- quartz in a comminuted matrix of feldspars, tween the Rio Carla Granite and the Veloz For- quartz, and biotite (see Plate l(a)). Plagio- mation along this unconformity (Hatten et al., clase porphyroclasts are often kinked or bro- 1958; Pszczolkowski, 1986). Pardo (1975) also ken. Biotite is frequently chloritized, and the considered the Socorro Complex (Sierra Mor- feldspars contain sericitic alteration products. ena) to represent allochthonous pre-Neocom- Irregular pods of greenish breccia, possibly ian basement outcrops. Pszczolkowski (1986) brecciated mafic xenoliths, occur within the interprets the complex to comprise pre-Tithon- granite. ian basement fragments between thrust sheets. Contact relations of the granite with the Me- Our structural interpretation portrays the So- sozoic sedimentary rocks are not well exposed, corro Complex as a basement uplift, thrust from but it appears to be overlain unconformably by southwest to northeast, as shown in Fig. 3. arkosic conglomerates of the Constancia For- The distribution of the Socorro Complex in mation, which is in probable depositional con- surface outcrops is limited, but similar rocks tact with the overlying Veloz Formation in the occur elsewhere in evaporite diapirs (discussed southern portion of the complex (Fig. 2 ). These below) and as clastic components in arkosic inferred relationships suggest a pre-Late Jur- sediments of the Upper Cretaceous (Campan- assic age for the Rio Carla Granite. Somin and ian-Maastrichtian) Bacanayagua Formation Millfin (1981) reported whole-rock K-Ar dates along the Gulf coast in Habana and Matanzas of 150 + 5, 139 + 5 and 140 + 2 Ma for two sam- provinces (Somin and Mill~n, 1981 ). Thus the ples of the granite. Interpretation of these data Socorro Complex may have been much more must accommodate the observations that: widely exposed in the Cretaceous. (1) the granite contains 30-60 vol.% microcline, whose high K content implies that a Rio Ca~a Granite whole-rock K-Ar date is probably in effect a microcline date; The Rio Carla Granite (Hatten et al., 1988) (2) the microcline is generally sericitized, of the Socorro Complex is a coarse-grained, quite strongly in many cases. pink-colored granite, consisting of perthitic mi- The K-Ar dates thus may reflect the most re- crocline, plagioclase, quartz, biotite, apatite and cent cooling of the granite through the blocking zircon. A chemical analysis of a sample of the temperature ( ~ 100-200 oC; Berger and York, granite reported by Somin and Milldn (1981) 1981; Harrison and McDougall, 1982) of Ar in includes 68.32% SiO2, 14.94% A1203, 1.89% K-feldspar. The K-Ar dates may also reflect the 330

J~ (.9

0 331 age of sericitization. Considering the evidence by the intrusion of the Rio Carla Granite, the for widespread overprinting of K-Ar system- geochronologic data presented below are incon- atics (implying greenschist grade or higher sistent with this. Shearing and cataclasis of both metamorphism) in basement rocks elsewhere rock units apparently occurred under brittle in Cuba during the Late Cretaceous-Paleogene conditions, and thus are probably unrelated to (Meyerhoff et al., 1969; Somin and Mill~n, the intrusive event. 1977, 1981; Hatten et al., 1988), it is somewhat The biostratigraphic ages of the Veloz and surprising that microcline and/or sericite were Amaro Formations (Hatten et al., 1958; apparently not reset during this time interval, Pszczolkowski, 1986), which appear to be in but more detailed analyses of these rocks with unconformable depositional contact with the 4°Ar/39Ar techniques (in progress) are needed Socorro Complex to the southwest, suggest a to evaluate better their thermal histories. pre-Late Jurassic metamorphism of these rocks. This constraint cannot be considered rigorous, Metasedimentary rocks however, as the nature of the contact (depositional or tectonic) is ambiguous. K-Ar anal- Associated with the Rio Carla Granite are yses of different portions of a single coarse metasedimentary rocks, dominantly marbles phlogopite crystal from a marble have yielded (Sierra Morena Marble of Somin and Mill~n, apparent dates of 945+25 and 910+25 Ma 1977, 1981 ) and quartzites, whose contacts with (Somin and Milldn, 1977). Somin and Millfin the granite are obscured by poor exposure. The (1977) concluded on this basis that the marbles quartzites are generally fine grained, but the were substantially older than the Rio Carla marbles range from fine to very coarse grained; Granite, and related their genesis to the Gren- in the latter case the rock may resemble skarn. ville orogeny. The significance of these K-Ar Textures of the metasedimentary rocks are typ- dates is not entirely clear. Phlogopite closes to ically granoblastic or porphyroblastic, but are thermally activated Ar diffusion at ~350- locally schistose in thin, discrete shear zones 400°C (Giletti and Tullis, 1977), and thus is which may be genetically related to cataclastic more resistant to resetting than feldspars and textures developed in the granite. Traces of other micas, and might not be outgassed by low molybdenite occur locally associated with greenschist facies metamorphism. However, the quartz in irregular veins in the clastic rocks. K-At technique does not permit the recogni- Samples of marble collected for this study tion of excess Ar, particularly problematic contain variable proportions of calcite + diop- in metamorphic rocks, which can produce side + phlogopite. Equant patches (up to 5 mm spuriously old K-Ar dates (e.g., Lanphere and diameter) of serpentine are locally abundant, Dalrymple, 1976). resembling pseudomorphs after forsterite. Ve- suvianite was also reported in some of these Geochronologic data rocks by Somin and Mill~in (1977). These as- semblages are indicative of high-grade meta- Despite some uncertainties about their sig- morphism in either the pyroxene-hornfels or nificance, published geochronologic data sug- upper amphibolite/granulite facies of contact gest that the Socorro Complex has a substan- or regional metamorphism, respectively tial pre-Cretaceous history, and its study is (Turner, 1981 ). Their bulk compositions indi- therefore relevant to understanding the early cate that dolomitic carbonate rocks were the tectonic evolution of the Caribbean region. In protoliths for the marbles. Although field and order to resolve some ambiguities posed by ex- petrologic data are consistent with the contact isting K-Ar dates, the following 4°Ar/39Ar and metamorphism of the metasedimentary rocks U-Pb data have been obtained as part of our 332 continuing geochronologic, petrologic and tec- lector instrument, allowing simultaneous col- tonic studies of Cuba. lection of all the relevant Pb or U isotopes.

Analytical procedures 4°Ar/39Ar data on phlogopite (metasedimentary rocks) A phlogopite-bearing marble (shown in Plate l(d) ) collected ~ 1 km from the contact with Data from the step heating experiment are the Rio Carla Granite was crushed and sieved. summarized in Table I, and the corresponding Phlogopite crystals were hand picked from the apparent age spectrum is shown in Fig. 4a. The 1-2-mm size fraction, which is the approximate age spectrum (corrected for atmospheric Ar size range of phlogopite porphyroblasts in this having 4°Ar/36Ar = 295.5 ) shows young appar- sample. Crystals were rinsed in distilled water ent ages at low extraction temperatures, in- for 30 min in an ultrasonic cleaner, then dried, creasing monotonically from 153.4 Ma at 600 °C and 0.040 g was packaged in A1 foil for irradia- to a plateau which begins with the 950 ° C fraction in the McMaster University nuclear reac- tion at 898.6 Ma. The plateau spans ~ 67% of tor. After irradiation, 0.035 g was degassed in the total 39Ar released, and the volume-weighted 15 temperature steps and the purified Ar from age calculated from these steps is 903.5 ___7.1 Ma. each extraction analyzed with a Varian-MAT The pattern of discordance shown in Fig. 4a is GD150 mass spectrometer at Princeton Uni- suggestive of either ( 1 ) partial diffusive Ar loss, versity. Details of instrumentation, tempera- such as may occur during a reheating event ture control, gas purification, data collection, (Turner, 1968) or (2) a mixture of two phases background and mass discrimination correc- having different ages and degassing behavior. tions, corrections for interfering nucleogenic An isotope correlation diagram (Fig. 5) aids isotopes and decay of 3TAr and 39At, decay con- the characterization of several components of stants, age and error calculations were similar Ar. Low-temperature fractions (600-700°C) to those reported by Onstott and Peacock define a parabolic trajectory whose intercept on ( 1987 ). K/Ca and K/C1 atomic ratios were cal- the 36Ar/4°Ar axis is indistinguishable from culated from corrected 39Ar/~TAr and 39Ar/38Ar present-day atmospheric Ar. This strongly sug- ratios using the data reported by Onstott and gests that all the gas fractions are composed of Peacock (1987) for the McMaster reactor. a mixture between an atmospheric component A sample of the Rio Carla Granite was and a radiogenic component. Non-linearity of crushed, and zircons separated by conventional the array indicates a variable 39Ar/4°Ar for the methods. Microscopic examination of the zir- radiogenic component of each fraction. As the con population reveals a high proportion of non-radiogenic component appears to be of clear, euhedral grains, plus some possibly uniform atmospheric composition, the 39Ar/ rounded or corroded grains (Plate 1 (c)), some 4°At* ratio (4°Ar* is radiogenic 4°At) of any of which appear to have partial, euhedral over- fraction is thus the intercept on the 39Ar/4°Ar growths (Plate 1 (b)). After acid washing in hot axis of a line connecting each datum point with HCI, followed by hot HN03, non-magnetic the atmospheric 36Ar/4°Ar value. coarse and fine size fractions were purified by Examination of ~TAr/39Ar and 3SAr/39Ar ra- hand picking and analyzed by methods modi- tios (Fig. 4b and c ) shows that the low-temper- fied after Krogh (1973). Details of the analyt- ature extractions are correlated with relatively ical procedures and error analysis are described high Ca/K and C1/K, which monotonically de- by Mattinson (1987). Mass spectrometric crease to uniform values as this component is analysis was on the UCSB MAT-261 multicol- exhausted (by 950 ° C). The plateau values of 333

TABLE I

4°Ar/39Ar analytical data: CUH-4 phlogopite

T f(39) [1 /2 Atmospheric 4°Ar* 37Ar/39Ar 38Ar/39Ar Date (Ma) (°C) (X10 -3) (×10 -2 ) (%) (E-6 cm3 STP) (_+ 1S.D.)

600 0.0093 -4.173 1.053 89,59 0.356 6.4103 0.443 153.4+24.4 650 0.0154 -0.955 0.826 65,75 0.385 1.4677 0.329 247.7_+ 17.3 700 0.0233 - 0.568 1.006 39.39 0.759 0.8729 0.270 364.2 -+ 14.6 750 0.0359 -0.303 1.557 11.87 2.056 0.4657 0.263 579.1_+8.0 800 0.0729 -0.108 2.307 2.21 8.795 0.1652 0.246 791.2 + 1.5 850 0.1737 -0.052 2.565 0.87 26.599 0.0792 0.238 861.9-+2.9 900 0.3342 -0.036 5.104 0.29 44.256 0.0558 0.217 892.5 _+ 1.0 950 0.5474 - 0.028 3.743 0.31 59.306 0.0432 0.225 898.6 ___ 1.0 975 0.6574 -0.026 3.424 0.31 30.718 0.0399 0.225 901.3 -+ 1.1 1000 0.7303 - 0.029 4.169 0.28 20.513 0.0452 0.225 906.4 -+ 1.1 1025 0.7963 -0.029 2.383 0.50 18.375 0.0445 0.210 898.8_+ 1.6 1050 0.8603 - 0.019 0.792 0.98 17.889 0.0286 0.223 902.6 _+ 1.2 1075 0.9154 - 0.017 0.889 0.79 16.701 0.0262 0.222 907.5 _+ 1.4 1100 0.9718 -0.020 3.543 0.23 14.798 0.0306 0.222 911.0_+1.4 1180 1.0000 -0.032 1.143 1.10 8.211 0.0488 0.197 931.5_+ 1.3

Sample mass = 0.035 g. Average J-value = 0.023833 _+ 0.000031. Integrated date = 870.1 ___ 1.0 Ma. Plateau date = 903.5 _+ 7.1 Ma. Uncertaintyin extraction temperatureis -+2°C.f1=1/[1 - ( 37 Ar /39 Ar ) ca/ ( 3VAr / 39Ar ) M ] , and f 2 = [l [1 - (36Ar/39Ar)ca/(36At/ 39Ar)M], where ()ca = Isotope ratio of Ar extracted from irradiated Ca salts {from Onstott and Peacock, 1987), and ( ) M= isotope ratio of Ar extracted from irradiated unknown. Atmospheric (%) = % of 4°At that is of atmospheric composition. 4°At* is radiogenic At, corrected for atmospheric Ar (4°Ar/36Ar = 295.5), with 36At corrected for nucleogenic 3~Arca. f(39) is the cumulative fraction of 39At released in each step. 3VAr/39Ar and 38Ar/39Ar reflect interference- and decay-corrected values of nucleogenic 3VArca, 38Arcl, and 39ARK. Date calculated using decay constants and isotopic abundances of Steiger and Jager (1977). Error calculations according to Onstott and Peacock (1987).

37Ar/39Ar and 38Ar/39Ar correspond to Ca/ moved during mineral separation, e.g., possibly K=0.060 and C1/K=0.048 (using the data of present along cleavages or cracks. Onstott and Peacock, 1987), indistinguishable The release patterns of individual isotopes, within error from ratios (Ca/K = 0.047 ___0.018 shown as a function of extraction temperature and C1/K=0.037_+0.015) determined from in Fig. 6, are consistent with a single mineral- electron microprobe analysis of a phlogopite ogic source for 39ARK and 4°At*. In contrast, the crystal from the same hand sample. Micro- bimodal release pattern for 37Arca supports a probe data indicate that K, Ca and C1 are ho- low-temperature contribution from a high-Ca mogeneous over all but the outermost 10-20 pm phase which dominates 37Arca in the first three of the crystal, where marginal serpentinization or four gas fractions. This phase does not con- results in K-depletion while Ca and C1 are un- tain significant 39ARK or 4°Ar*; thus calcite is changed. Initial high Ca/K and C1/K may thus further implicated as a contaminant. The re- reflect the contribution of marginal serpentine, lease pattern of 38Arcl reveals a minor C1 anom- although the volume fraction of serpentine cal- aly in the 600°C, and possibly in the 700°C culated from the above is <0.2%, much less fractions, whose source is unknown. Thus, de- than the amount required to generate the 39ARK spite evidence for additional phases besides associated with Ca/K and C1/K anomalies. The phlogopite reflected in the low-temperature gas anomalously high initial Ca/K may also reflect fractions, these phases do not appear to con- contamination from calcite incompletely re- tribute to the 39ARKor 4°Ar* budget and the age 334

temperature fractions has atmospheric com- 0.6 C position, hence the apparent ages of these frac-

0.4 tions are meaningful and can be used to model c the pattern of discordance. If this Ar-loss was 0.2 due to a single reheating event, the algorithm of Hall and York (1982) can be applied to the "--T- theory of Turner (1968) to calculate the age of 0.4 this event for a given diffusion geometry and "c fraction of gas lost. In this model, the age cal- < culated from the first increment of gas released, 153.4 ___ 24.4 Ma, is an upper limit on the possible age of the secondary event. For the present 10] ' case, with 3.7% 4°Ar* loss (per cent difference between the total gas age and the plateau age ) 0 0 and cylindrical diffusion geometry, a model age 1,1,1 o (tm) of 59.5+9.4 Ma is calculated for the re- i!, A heating event and 905.8 + 9.8 Ma for the phlogopite's primary cooling age. For small frac- 0 015 1.0 tional Ar loss (L), as in this case, model CUMULATIVE FRACTION 39Ar RELEASED reheating ages calculated by this method are Fig. 4. Results of incrementalheating experiment, showing very sensitive to small perturbations in the the followingvalues as a function of cumulativefraction of value of L: for L=0.036, tm=24.9_9.9 Ma; for total 39Ar released: (A) apparent age, calculated from corrected 4°At*/agArratios; (B ) (3SAt)Cl/(a9Ar) K, the nucleo- L=0.038, tm=91.5___9.0 Ma. The present lim- genic ratio of Cl-derived 3BAr to K-derived agAr; (C) ited data do not warrant detailed interpretation (3TAr)ca/(39At) K, the nucleogenicratio of Ca-derived 37At of this model overprint age, but we note that the to K-derived39Ar. results cited above are consistent with partial outgassing during the pervasive Cretaceous- 0.004" Paleogene tectonothermal activity previously r 600 1 [] discussed. 0.003' 650 The K-Ar dates (910 + 25 and 945 + 25 Ma) El reported by Somin and Milldn (1977, 1981) 0.002' from a Sierra Morena Marble (discussed above) [] 700 are slightly older than the 4°Ar/39Ar plateau 0.001' date reported herein. This may reflect the much 800 a 750 larger grain size (hence higher blocking tem- 0.000 , ~, [] , 0.00 0.02 0.04 0.06 0.08 perature if the scale of Ar diffusion approaches

39Ar/40Ar the grain dimensions) of the phlogopite used in Fig. 5. 36Ar/*°Ar versus 39Ar/4°Ar correlation diagram. the K-Ar analyses, or may be a result of excess Numbers refer to extraction temperatures. Star corre- Ar contamination*. sponds to present-dayatmospheric Ar. Symbolsare > 1S.D. errors. spectrum may be confidently interpreted as the * Note addedin proofi Seventeen 4°Ar/39Ar laser probe spot analyses made on the {001 } faceof a singlephlogopite crys- product of volume-diffusive loss of Ar. tal ( ~ 0.4 cm diameter) yield dates ranging from 952_ 12 The correlation diagram, Fig. 5, shows that Ma to 718_ 9 Ma. Older dates are more common near the the non-radiogenic Ar component of the low- center of the grain. 335

1.20 1.20 " [] 4OAr* [] 38Arc~ 1.00 1.00 "

0.80 0.80 '

0.60 0.60 ' [] [] B 0.40 • 0.40 - BBQ•[] ••mB• 0.20 - [] 0.20 • [] B [] [] [] O.OC --..m i , i , 0.00 i 600 700 800 900 1000 1100 1200 6OO 700 8OO 900 1000 1100 1200

Temp (°C) Temp (°C)

1.20 " 1.20 " . [] 37Arca 1.00 ' [] 39Ar K 1.00 " m 0.80 • 0.80 " B [] 0.60 ' 0.60 • []

0.40 - 0.40 • BB []BBD[] 0.20 - [] 0.20 " ~BB [] m o [] 0.00 i , i i 0.00 600 700 800 900 1000 9100 3200 6OO 700 800 900 1000 1100 1200

Temp (°C) Temp (°C)

Fig. 6. Release patterns of individual Ar isotopes as a function of laboratory extraction temperature. Values plotted on vertical axes are atomic abundances normalized to the abundance in the 950 ° C extraction.

TABLE II o. 032 200 ,~#~::

Rio Carla Granite zircon data O. 031 ~#~ ....

Sample" F201 Zir f F201 Zir c 0.030 ~ 190 ~#~:* / 2asu (ppm) 320.6 286.6 2°8"Pb (ppm) 8.231 7.711 0.029 ~ ~ " _ 2°Spb/2°6pbb 0.1955 0.1826 180 ~:i~ 2°Tpb/2°6pbb 0.05879 0.05609 0.028 J (Lower Inl, = 17~5 Ma 2°4pb/2°~pbb 0.000495 _+ 3 0.000244 _+ 2 • ca. OO a ) 2°~'pb/2asU (Ma)~ 188.4 +_ 0.4 197.3 _+ 0.4 O. 027 ,,ri >: 2°7"pb/2asU (Ma)~ 194.2 +_ 0.7 206.2 _+ 0.6 207Pb/235U 2°7"Pb/2°S'Pb (Ma) c 265 _+ 5 309 ___3 0.026 , I , I , I , I , 0.18 0.19 0.20 0.21 0.22 0.23 % and f designate coarse and fine size fractions, respec- Fig. 7. Expanded portion of U-Pb concordia, with data from tively. Zircon analytical methods after Mattinson (1987). bObserved isotopic ratios corrected for mass fractionation the Rio Carla Granite. The concordia curve (virtually a straight line for this small segment) is shown for ~ 170- of 0.125% per mass unit, based on replicate analyses of NBS 200 Ma by the heavy stippled line. Fine ( F ) and coarse (C) Pb standards. Two-sigma errors are shown for last digit of fractions of zircon define a discordia line with a lower in- the 204/206 ratio. Errors on the 208/206 and 207/206 ratios are <0.1%. tercept of 172.5 Ma, and an upper intercept (not shown) ~Ages calculated using decay constants of Jaffey et al. of ~ 900 Ma. ( 1971 ). Error analysis after Mattinson (1987). a strong pattern of 'normal' discordance. On a U-Pb data on zircon (Rio Carla Granite) concordia plot {Fig. 7), thetwo fractions define a discordia trajectory with a lower intercept of The two zircon fractions have moderately low ~ 172.5 Ma, and an upper intercept of ~ 900 U contents ( ~ 300 ppm-see Table II), and show Ma. Both points plot close to the lower inter- 336

cept, which is therefore tightly constrained. Be- Conclusions from geochronologic data cause of the long projection, and the fact that only two data points define the chord, the upper 4°Ar/39Ar and U-Pb data summarized herein intercept is much less well constrained. Addi- clearly show evidence of Late Proterozoic con- tional fractions, in progress, will allow more tinental crust in the Las Villas terrane. 4°At/ rigorous statistical analysis, with meaningful 39Ar data indicate that phlogopite from a So- errors on the upper and lower intercepts. The corro Complex marble cooled through its Ar coarser fraction has a lower U concentration, closure temperature at ~ 903.5 Ma ago, follow- and plots higher on the discordia trajectory than ing high-grade metamorphism. Additionally, a does the finer fraction. In general, the overall minor secondary thermal event at ~ 60 Ma ago pattern of discordance of the zircons (i.e., plot- is suggested by modeling of the step heating ting close to the lower intercept) is typical of data. The U-Pb data indicate that the Rio Carla that for igneous rocks containing older inher- Granite was emplaced and crystallized in the ited zircons (for example, see Fig. 4 of Mattin- Early-Middle Jurassic, at ~ 172.4 Ma, but that son and James, 1985). The finer fraction con- the source of the granite was, at least in part, tains a larger proportion of newly crystallized continental crustal rocks containing 900-Ma- zircons (plus possibly a greater proportion of old zircons. The clastic metasedimentary rocks Pb loss from the inherited component) and the associated with the marble, or similar rocks at coarser fraction contains a greater fraction of greater depth, are a likely source for these in- inherited zircons, possibly as cores (seed crys- herited zircons. The whole-rock K-Ar ages of tals ) upon which igneous zircon nucleated. 150-139 Ma reported for the granite by Somin The most straightforward interpretation of and Millan (1981) evidently reflect either sim- the zircon data is that the lower intercept on ple cooling of the granite (through ~ 100-200 ° C concordia at ~ 172.4 Ma represents the time of as discussed earlier) or a separate, later ther- magmatic emplacement of the granite, and that mal disturbance. the upper intercept on concordia at ~ 900 Ma Partial outgassing of the phlogopite, indi- represents the minimum age of the inherited cated by the low-temperature part of the 4°Ar/ zircons. Thus, the granite was emplaced in the 39Ar age spectrum, may also reflect the em- Jurassic, and derived in part by anatexis or as- placement and cooling of the granite, although similation of (probably) continental > 900 Ma this outgassing is more reasonably modeled as source rocks, or sedimentary or metasedimen- a later event. An additional problem with the tary rocks derived from a > 900 Ma source. latter interpretation is that heat flow consid- A possible alternate interpretation is that the erations (Jaeger, 1957; discussed by Turner, granite was emplaced and crystallized at ~ 900 1981, pp. 18-24) suggest that a pluton of this Ma, and that the observed discordance is the size and general composition should have pro- result of a Jurassic thermal event. This would duced a temperature >_ 400 ° C at a distance of 1 require the loss of almost 98% of the Pb from km from the contact. This almost certainly the fine-grained zircons during the Jurassic. In should have been sufficient to outgas > 3.7% of light of the evidence for chiefly cataclastic the Ar from phlogopite, whose Ar blocking tern- (cold) deformation of the granite (see earlier perature is < 372°C, the value calculated for a description), and the relatively low U concen- 1.0 mm diffusion radius and a cooling rate of trations in the zircons (and thus low accumu- 1000 °C Ma-1 from the diffusion data of Giletti lation of radiation damage, an important factor and Tullis (1977) using the formulation of in susceptibility to Pb loss -- see Silver, 1963 ), Dodson (1973). Because of the poorly exposed this second interpretation seems highly contact relationships, however, we cannot be unlikely. sure that the present relative positions of the 337 marble and granite reflect their relative posi- The Socorro Complex contains a signifi- tions when the granite was intruded. A tectonic cantly older geologic record than is known from contact somewhere between the two units is anywhere else in the Greater Antilles. Its tec- possible, and the marble sampled for this study tonothermal history is markedly discordant may have originally been further from the gran- with respect to those of the oldest known nearby ite, possibly explaining the apparent absence of rock units to the north and east, including those thermal effects from the intrusion on Ar iso- from DSDP holes 537 and 538A and peninsular topic systematics of the phlogopite. Florida, which preserve a record of pervasive Somin and Mill~in (1981) reported frag- Pan-African ( ~ 650-500 Ma) age intrusion and ments of rocks resembling some Socorro Com- metamorphism (Dallmeyer, 1984, 1987a, 1989; plex lithologies from the San Adrian gypsum Dallmeyer et al., 1987). Neither the a°Ar/39Ar diapir (Meyerhoff and Hatten, 1968), located nor the U-Pb data presented herein show any ~ 40 km east of the city of La Habana. Among evidence for Pan-African tectonothermal ac- these are arkosic sandstones (resembling those tivity superimposed on rocks that were meta- of the Constancia Formation) and phlogopitic morphosed at ~ 900 Ma, which is remarkably marbles. Fine-grained phlogopite separated similar in age to biotite K-Ar cooling dates from one of the latter yields a K-At date of from the Grenville belt (e.g., Berger and York, 123+5 Ma (Somin and Mill~in, 1981). If this 1981; Dallmeyer, 1987b) and Llano Province rock was indeed derived from the Socorro Com- plex, then the K-Ar date reflects a more com- plete outgassing of this phlogopite than observed in the sample reported above, probably a result of either its smaller grain size or variable thermal conditions during the reheating event, or both.

Regional tectonic implications

The Socorro Complex apparently represents a fragment of continental crust, whose history includes: (1) pre-900 Ma deposition of dolomitic carbonates and siliciclastic sedimentary rocks; (2) high-grade metamorphism of these rocks, probably at mid-crustal depth, at ~ 900 Ma; (3) intrusion of the Sierra Morena Granite at ~ 172 Ma; (4) uplift and erosion of the complex (rep- resented by the Constancia Formation) prior to the Late Jurassic deposition of the Veloz Fig. 8. Present-day distributionsof Grenville ( ~ 1200-800 Formation; Ma) and Pan-African ( ~ 650-500 Ma) age orogenic com- (5) a minor Late Cretaceous-Paleogene re- plexes. Sources in text. G = Grenville Province, LL = Llano heating event; Province, D=DSDP holes 537 and 538A, OC=Oaxaca (6) cataclastic deformation of both the gran- Complex, YB=Yucatan Block, CB=Chortis Block, SC = Socorro Complex, GSMB = Garzdn-Santa Marta Belt, ite and the metasedimentary rocks, which is SB = Sunsas Belt, SS = Suwanee Suture, CBS = Caribbean- constrained only to be post-172 Ma. Bahamas Suture. 338

(Denison et al., 1984) of North America. Fig- ure 8 shows the distribution of rocks strongly NA affected by the Grenville and Pan-African \ ..... events, illustrating the apparent discordance

across the Caribbean-Bahamas suture. \ .,,- ~ / \ : :: C \ I Grenville-age high-grade metamorphic rocks ? 11 seemingly unaffected by Pan-African tecton- \ othermal events occur in a discontinuous belt (Fig. 8) that extends from the Llano Province \l, southward through the Yucatan block (Gom- berg et al., 1968), Chortis block (Home et al.,

1976), and Oaxaca Complex (Ruiz et al., 1988; Fig. 9. Diagrammatic reconstruction of Pangea at ~ 180 Ma, Ballard et al., 1989) of Central America; and showing Pan-African Orogens and postulated Grenville the Garzdn-Santa Marta belt of the Colombian (sensu lato ) Belt (queriedwhere existenceuncertain ) dis- Andes (Alvarez, 1981; Kroonenberg, 1982; cussed in text. NA= North America, AF=Africa, SA = South America, C = 'Caribbeana'. Priem et al., 1989). The Sunsas belt of the western Amazonian craton may also be genetically Mojave Megashear and to the southeast by the related to the foregoing, although 1280-950 Ma Colombian Andes. In this scenario, Caribbeana thermal activity in the Sunsas orogen was dom- was subsequently fragmented and dispersed inated by voluminous granitoid intrusion and during the birth and motion of the Caribbean lower-grade metamorphism (Priem et al., 1971; plate, beginning in the Middle Jurassic and Litherland and others, 1986) than amphibol- continuing to the present. It must be empha- ite-granulite-facies metamorphism common to sized that Fig. 9 does not attempt to reconstruct the other complexes mentioned above. The the Grenville (sensu lato) Belt at the time of Sunsas belt further differs from these in being its origin 1.2-0.8 Ga ago, but rather depicts its truncated to the southeast by a Pan-African remnants just prior to the Mesozoic rifting of (Brasiliano) mobile belt. Pangea. The Socorro Complex thus may represent The potassic Rio Carla Granite is composi- part of an extensive Grenville orogen, charac- tionally akin to intra-cratonic rift-associated terized by high-grade metamorphism + pluton- granites, and its intrusion in the Middle Jur- ism, which spanned the Americas during the assic may have accompanied the rifting apart Late Proterozoic. Correlations between the of Pangea. Uplift and erosion of the Socorro Grenville Province and the GarzSn-Santa Complex by the Late Jurassic is not consistent Marta belt of Colombia have been proposed be- with a simple post-rifting subsidence history fore (Kroonenberg, 1982, and references beyond the Late Jurassic, but may instead re- therein ), but the possible role of Caribbean and flect an earlier episode of plate collision prior Central American terranes in such correlations to the widely recognized Late Cretaceous-Pa- have yet to be considered. A speculative recon- leogene event. The age of subsequent cataclas- struction of part of the expanded Grenville or- tic deformation in the Socorro Complex is ogenic belt, immediately prior to the Mesozoic poorly constrained, but this may have coin- rifting apart of Pangea, is shown in Fig. 9. This cided with the thermal event indicated by 4°At/ reconstruction places all the Central Ameri- ~gAr data discussed previously. The rather can/Caribbean crustal blocks (Chortis, Yuca- poorly constrained age ( ~ 60 Ma) of this event tan, Oaxaca and Socorro Complex) together in is consistent with its occurrence during the Late a landmass herein termed 'Caribbeana', Cretaceous-Paleogene collision of the Carib- bounded to the northeast by the Sonora- bean and North American plates. 339

The reconstruction depicted in Fig. 9 differs of the Deep Sea Drilling Project Leg 77, southeastern from most (e.g., those of Pindell and Dewey, Gulf of Mexico: tectonic implication. In: R.T. Buffler 1982; Burke et al., 1984; Pindell, 1985) in that and W. Schlager (Editors), Initial Reports of the Deep Sea Drilling Project, LXXVII: 496-504. a much tighter fit between North and South Dallmeyer, R.D., 1987a. 4°Ar/39Arage of detrital muscovite America is obtained by removing the interven- within Lower Ordovician sandstone in the coastal plain ing crustal blocks which we place in Carib- basement of Florida: implications for west African ter- beana. Details of the distribution of these blocks rane linkages. Geology, 15: 998-1001. Dallmeyer, R.D., 1987b. 4°Ar/39Ar mineral age record of in a Pangea reconstruction is a major source of variably superimposed Proterozoic tectonothermal discrepancy between various published tec- events in the Grenville Orogen, central Labrador. Can. tonic models, and their transport history dur- J. Earth Sci., 24: 314-333. ing the early evolution of the Caribbean plate Dallmeyer, R.D., 1989. Contrasting accreted terranes in the southern Appalachian orogen, basement beneath the is not well understood. More quantitative data, Atlantic and Gulf Coastal Plains, and Western Africa. particularly from geochronologic and paleo- Precambrian Res., 42: 387-409. magnetic studies, are needed to help delineate DaUmeyer, R.D., Caen-Vachette, M. and Villeneuve, M., the individual tectonothermal histories and 1987. Emplacement age of post-tectonic granites in tectonic motions of these fragments in order to southern Guinea (west Africa) and the peninsular Flor- ida subsurface: implications for origins of southern Ap- help evaluate the foregoing speculations. palachian exotic terranes. Geol. Soc. Am. Bull., 99: 87- 93. Acknowledgments Denison, R.E., Lidiak, E.G., Bickford, M.E. and Kisvar- sanyi, E.B., 1984. Geology and geochronology of Pre- We are grateful to the Centro de Investiga- cambrian rocks in the central interior region of the ciones Geologicas de Cuba and officials of the United States. U.S. Geol. Surv. Prof. Pap., 1241-C: 1- International Geological Correlation Program 20. Dodson, M.H., 1973. Closure temperature in cooling geo- Project 165 for logistical support of fieldwork. chronological and petrological systems. Contrib. Min- We are particularly indebted to G. Echevarria, eral Petrol., 40: 259-274. J. Sanchez, and J. Yparraguirre for their assis- Duncan, R.A. and Hargraves, R.B., 1984. Plate tectonic tance. Microprobe analyses were provided by S. evolution of the Caribbean region in the mantle refer- Swapp. Critical comments on this manuscript ence frame. In: W.E. Bonini, R.B. Hargraves and R. Shagam (Editors), The Caribbean-South American by H. Gaudette, P. Pindell, and W. Teixeira are Plate Boundary and Regional Tectonics. Geol. Soc. Am. appreciated. Mem., 162: 81-92. Giletti, B.J. and Tullis, J., 1977. Studies in diffusion, IV. Pressure dependence of Ar diffusion in phlogopite mica. References Earth Planet. Sci. Lett., 35: 180-183. Gomberg, D.N., Banks, P.D. and McBirney, A.R., 1968. Alvarez, J.A., 1981. Determinacion de la edad Rb/Sr en Guatemala: preliminary zircon ages from the Central rocas del Macizo de Garzon, Cordillera Oriental de Col- Cordillera. Science, 162: 121-122. ombia. Geologia Norandina, 4: 31-38. Gose, W. and Eby, H., 1987. Relative displacement be- Ballard, M.M., Van der Voo, R. and Urrntia-Fucugauchi, tween the North American and Caribbean Plates In- J., 1989. Paleomagnetic results from Grenvillian-aged ferred from Paleomagnetic Data. EOS Trans. Am. Geo- rocks from Oaxaca, Mexico: evidence for a displaced phys. Union, 68: 291. terrane. Precambrian Res., 42: 343-352. Hall, C.M. and York, D., 1982. Least-squares estimation of Berger, G.W. and York, D., 1981. Geothermometry from formation and metamorphic ages from 4°Ar/3~Ar age 4°Ar/39Ar dating experiments. Geochim. Cosmochim. spectra. EOS Trans. Am. Geophys. Union, 63: 453-454. Acta, 45: 795-811. Harrison, T.M. and McDougall, I., 1982. The thermal sig- Burke, K., Cooper, C., Dewey, J.F., Mann, P. and Pindell, nificance of potassium feldspar K-Ar ages inferred from J.A., 1984. Caribbean tectonics and relative plate mo- 4°Ar/39Arage spectrum results. Geochim. Cosmochim. tions. In: W.E. Bonini, R.B. Hargraves and R. Shagam Acta, 46: 1811-1820. (Editors), The Caribbean-South American Plate Hatten, C.W., Schooler, O.E., Giedt, N. and Meyerhoff, Boundary and Regional Tectonics. Geol. Soc. Am. Mem., A.A., 1958. Geology of central Cuba, eastern Las Villas 162: 31-63. and western Camaguey provinces, Cuba. La Habana, Dallmeyer, R.D., 1984. 4°Ar/39Ar ages from pre-Mesozoic Ministerio de Industrias archives, 250 pp. (unpublished crystalline basement penetrated at holes 537 and 538A report). 340

Hatten, C.W., Somin, M., Mill~n, G., Renne, P.R., Kistter, tectonica y metamorfismo del macizo de Escambray. R.W. and Mattinson, J.M., 1988. Tectonostratigraphic Edit. Academia, La Habana, 104 pp. units of central Cuba. Eleventh Carib. Geol. Conf. Symp. Mill~n, G. and Somin, M., 1985. Contribucfon al conoci- Vol., Barbados, (in press). miento geologico de las metamorfitas del Escambray y Home, G.S., Clarke, G.S. and Pushkar, P., 1976. Pre-Cre- del Purial. Reporte de Investigacfon del Institut6 de taceous rocks of northwestern Honduras: basement ter- Geologfa y Paleontologia No. 2, Academfa de Ciencias rane in Sierra de Omoa. Am. Ass. Petrol. Geol. Bull., 60: de Cuba, 74 pp. 566-583. Onstott, T.C. and Peacock, M.W., 1987. Argon retentivity Howell, D.G., Jones, D.L. and Schermer, E.R., 1985. Tec- of hornblendes: a field experiment in a slowly cooled tonostratigraphic terranes of the Circum-Pacific re- metamorphic terrane. Geochim. Cosmochim. Acta, 51: gion. In: D.G. Howell (Editor), Tectonic-Stratigraphic 2891-2904. Terranes of the Circum-Pacific Region. Circum-Pacific Palmer, A.R., 1983. The decade of North American geology Council for Energy and Mineral Resources, Houston, 1983 geologic time scale. Geology, 11: 503-504. TX, pp. 3-30. Pardo, G., 1975. Geology of Cuba. In: A.E.M. Nairn and Jaeger, J.C., 1957. Temperatures outside a cooling intru- F.G. Stehli (Editors), The Ocean Basins and Margins. sive sheet. Am. J. Sci., 257: 44-54. Vol. 3: The Gulf of Mexico and the Caribbean. Plenum, Jaffey, A.H., Flynn, K.F., Glendenin, L.E., Bentley, W.C. New York, pp. 553-615. and Essling, A.M., 1971. Precision measurement of the Pindeil, J.A., 1985. AUeghenian reconstruction and subse- half-lives and specific activities of U 235 and U 238. Phys. quent evolution of the Gulf of Mexico, Bahamas, and Rev., Ser. C, 4: 1899-1906. proto-Caribbean, Tectonics, 4: 1-39. Krogh, T.E., 1973. A low-contamination method for hydro- Pindell, J.A. and Dewey, J.F., 1982. Permo-Triassic recon- thermal dissolution of zircons and extraction of U and struction of western Pangea and the evolution of the Pb for isotopic age determinations. Geochim. Cosmo- Gulf of Mexico/Caribbean region. Tectonics, 1:179-211. chim. Acta, 37: 485-494. Priem, H.N.A., Boelrijk, N.A.I.M., Hebeda, E.H., Verdur- Kroonenberg, S.B., 1982. A Grenvillian Granulite Belt in men, E.A.Th., Verschure, R.H. and Bon, E.A., 1971. the Colombian Andes and its relation to the Guiana Granitic complexes and associated tin mineralizations Shield. Geol. Mijnbouw, 61: 325-333. of'Grenville' age in Rondonia, western Brazil. Geol. Soc. Lanphere, M.A. and Dalrymple, G.B., 1976. Identification Am. Bull., 82: 1095-1102. of excess 4°Ar by the 4°Ar/39Ar age spectrum technique. Priem, H.N.A., Kroonenberg, S.B., Boelrijk, N.A.I.M. and Earth Planet. Sci. Lett., 32: 141-148. Hebeda, E.H., 1989. Rb-Sr and K-Ar evidence for the Litherland, M. and others, 1986. The geology and mineral presence of a 1.6 Ga basement underlying the 1.2 Ga resources of the Bolivian Precambrian shield. Overseas GarzSn-Santa Marta granulite belt in the Colombian Mem. Br. Geol. Surv., 9, 153 pp. Andes. Precambrian Res., 42: 315-324. Mattinson, J.M., 1987. U-Pb ages of zircons: a basic ex- Pszczolkowski, A., 1986. Sequencia estratigrafica da las amination of error propagation. Chem. Geol. (Isotope provincias de Matanzas y Villa Clara. Bull. Polish Acad. Geosci. Sect.), 66: 151-162. Sci., 34: 67-78. Mattinson, J.M. and James, E.W., 1985. Salinian Block U/ Ruiz, J., Patchett, P.J. and Ortega-Gutierrez, F., 1988. Pro- Pb age and isotopic variations: implications for origin terozoic and Phanerozoic basement terranes of Mexico and emplacement of the Salinian terrane. In: D.G. How- from Nd isotopic studies. Geol. Soc. Am. Bull., 100: 274- ell (Editor), Tectonic-Stratigraphic Terranes of the 281. Circum-Pacific Region. Circum-Pacific Council for En- Silver, L.T., 1963. The relation between radioactivity and ergy and Mineral Resources, Houston, TX, pp. 215-226. discordance in zircons. Nuclear Geophys. Conf. Proc., Mattson, P.H., 1984. Caribbean structural breaks and plate 1962, Woods Hole, MA, NAS-NRC Publ. 1075, pp. 34- movements. Caribbean tectonics and relative plate mo- 42. tions. In: W.E. Bonini, R.B. Hargraves and R. Shagam Somin, M. and Mill~n, G., 1977. Sobre la edad de las rocas (Editors), The Caribbean-South American Plate metamorficas Cubanas. Informe cientffico-tecnico no. Boundary and Regional Tectonics. Geol. Soc. Am. Mem., 2, Instituto de Geologfa y Paleontologfa, Academia de 162: 81-92. Ciencias de Cuba, 11 pp. Meyerhoff, A.A. and Hatten, C.W., 1968. Diapiric struc- Somin, M. and Mill~n, G., 1981. Geology of the Metamor- tures in central Cuba. Am. Assoc. Petrol. Geol. Mem., 8: phic Complexes of Cuba. Nauka, Moscow, 218 pp. (in 315-357. Russian. ) Meyerhoff, A.A., Khudoley, K.M. and Hatten, C.W., 1969. Steiger, R. and Jager, E., 1977. Subcommission on geo- Geologic significance of radiometric ages from Cuba. Am. chronology: convention on the use of decay constants in Assoc. Petrol. Geol. Bull., 53: 2494-2500. geo- and cosmochronology. Earth Planet. Sci. Lett., 36: Mill~n, G. and Somin, M., 1981. Litologia, estratigrafia, 359-362. 341

Sykes, L.R., McCann, W.R. and Kafka, A.L., 1982. Motion rite Research. Reidel, Dordrecht, pp. 407-417. of the Caribbean plate during the last 7 million years Turner, F.J., 1981. Metamorphic Petrology. McGraw-Hill, and the implications for earlier Cenozoic movements. J. New York, 524 pp. Geophys. Res., 25: 929-955. Wadge, G. and Burke, K., 1983. Neogene Caribbean plate Turner, G., 1968. Thermal histories of meteorites by the rotation and associated Central American tectonic evo- 39Ar-4°Ar method. In: P.M. Millman (Editor), Meteo- lution. Tectonics, 2." 633-643.