The North American– boundary in –Guatemala–Honduras

LOTHAR RATSCHBACHER1*, LEANDER FRANZ1, MYO MIN1, RAIK BACHMANN1, UWE MARTENS2, KLAUS STANEK1, KONSTANZE STU¨ BNER1, BRUCE K. NELSON3, UWE HERRMANN3, BODO WEBER4, MARGARITA LO´ PEZ-MARTI´NEZ4, RAYMOND JONCKHEERE1, BLANKA SPERNER1, MARION TICHOMIROWA1, MICHAEL O. MCWILLIAMS2, MARK GORDON5, MARTIN MESCHEDE6 & PETER BOCK1 1Geowissenschaften, Technische Universita¨t Bergakademie Freiberg, 09599 Freiberg, Germany 2Geological and Environmental Sciences, Stanford University, Stanford, CA 94305, USA 3Earth and Space Sciences, University of Washington, Seattle, WA 98195, USA 4CICESE, 22860 Ensenada, B.C., Mexico 5Department of Geology and Geophysics, Rice University, Houston, TX 77251-1892, USA 6Geowissenschaften, Universita¨t Greifswald, 17487 Greifswald, Germany *Corresponding author (e-mail: [email protected])

Abstract: New structural, geochronological, and petrological data highlight which crustal sec- tions of the North American–Caribbean Plate boundary in Guatemala and Honduras accommo- dated the large-scale sinistral offset. We develop the chronological and kinematic framework for these interactions and test for Palaeozoic to Recent geological correlations among the Maya Block, the Chortı´s Block, and the terranes of southern Mexico and the northern Caribbean. Our principal findings relate to how the North American–Caribbean Plate boundary partitioned defor- mation; whereas the southern Maya Block and the southern Chortı´s Block record the Late Cretac- eous–Early Cenozoic collision and eastward sinistral translation of the Greater Antilles arc, the northern Chortı´s Block preserves evidence for northward stepping of the plate boundary with the translation of this block to its present position since the Late Eocene. Collision and translation are recorded in the ophiolite and subduction–accretion complex (North El Tambor complex), the continental margin (Rabinal and Chuacu´s complexes), and the Laramide foreland fold–thrust belt of the Maya Block as well as the overriding Greater Antilles arc complex. The Las Ovejas complex of the northern Chortı´s Block contains a significant part of the history of the eastward migration of the Chortı´s Block; it constitutes the southern part of the arc that facilitated the breakaway of the Chortı´s Block from the Xolapa complex of southern Mexico. While the Late Cretaceous collision is spectacularly sinistral transpressional, the Eocene–Recent translation of the Chortı´s Block is by sinistral wrenching with transtensional and transpressional episodes. Our reconstruction of the Late Mesozoic–Cenozoic evolution of the North American–Caribbean Plate boundary identified Proterozoic to Mesozoic connections among the southern Maya Block, the Chortı´s Block, and the terranes of southern Mexico: (i) in the Early–Middle Palaeozoic, the Acatla´n complex of the southern Mexican Mixteca terrane, the Rabinal complex of the southern Maya Block, the Chuacu´s complex, and the Chortı´s Block were part of the Taconic–Acadian orogen along the northern margin of South America; (ii) after final amalgamation of Pangaea, an arc developed along its western margin, causing magmatism and regional amphibolite–facies metamorphism in southern Mexico, the Maya Block (including Rabinal complex), the Chuacu´s complex and the Chortı´s Block. The separation of North and South America also rifted the Chortı´s Block from southern Mexico. Rifting ultimately resulted in the formation of the Late Jurassic–Early Cre- taceous oceanic crust of the South El Tambor complex; rifting and spreading terminated before the Hauterivian (c. 135 Ma). Remnants of the southwestern Mexican Guerrero complex, which also rifted from southern Mexico, remain in the Chortı´s Block (Sanarate complex); these complexes share Jurassic metamorphism. The South El Tambor subduction–accretion complex was emplaced onto the Chortı´s Block probably in the late Early Cretaceous and the Chortı´s Block collided with

From:JAMES, K. H., LORENTE,M.A.&PINDELL, J. L. (eds) The Origin and Evolution of the Caribbean Plate. Geological Society, London, Special Publications, 328, 219–293. DOI: 10.1144/SP328.11 0305-8719/09/$15.00 # The Geological Society of London 2009. 220 L. RATSCHBACHER ET AL.

southern Mexico. Related arc magmatism and high-T/low-P metamorphism (Taxco–Viejo– Xolapa arc) of the Mixteca terrane spans all of southern Mexico. The Chortı´s Block shows continu- ous Early Cretaceous–Recent arc magmatism.

Supplementary material: Analytical methods and data, and sample description are available at http://www.geolsoc.org.uk/SUP18360.

Structural, geochronological, and petrological data domains (e.g. Dickinson & Lawton 2001; Keppie document the Palaeozoic–Cenozoic pressure– 2004). Most tectonic models suggest that collision temperature–deformation–time (P–T–d–t) his- of the Maya and Chortı´s Blocks occurred along tory of the southern part of the Maya and northern south- or north-dipping subduction zones during part of the Chortı´s Blocks, and the Polochic, the Cretaceous, and caused emplacement of Motagua and Jocota´n–Chameleco´n fault zones Jurassic–Cretaceous Proto-Caribbean oceanic crust along the North American–Caribbean Plate bound- onto the blocks; thrusting of metamorphic and ary in southern Mexico, Guatemala and northern sedimentary strata was bivergent north and south Honduras. The principal questions addressed are: of the zone (e.g. Meschede & which crustal sections accommodated the large- Frisch 1998; Beccaluva et al. 1995 and references scale sinistral offset along the plate boundary, therein). Since Late Cretaceous–Early Cenozoic, what was the timing of offset and deformation, the MSZ developed into a sinistral wrench zone and in what kinematic framework did it occur? due to the different velocities of the North American What Palaeozoic to Cenozoic geological corre- and Caribbean plates (e.g. DeMets 2001). In lations exist between the Maya Block, the Chortı´s southern Mexico, Guatemala and Honduras this Block, the tectono-stratigraphic complexes (‘ter- still active fault system has dismembered the colli- ranes’) of southern Mexico, and the arc/subduction sional blocks as well as the allochthonous remnants complexes of the northern Caribbean. of the former oceanic crust; the latter comprises shaly me´lange, containing blocks of serpentinized ophiolite fragments, and metamorphic rocks such Northern Caribbean Plate boundary as amphibolite, eclogite, albitite and jadeitite (e.g. Tsujimori et al. 2005). Inferred Cenozoic displace- The Caribbean Plate represents a lithospheric unit ment accommodated along the MSZ is controver- between the North and South American plates. Its sial, but magnetic anomalies within the Cayman western and eastern margins consist of subduction trough suggest a minimum of c. 1100 km; spreading zones with magmatic arcs (Central America rates in the trough were c.3cm/annum from 44 to Isthmus, Lesser Antilles), whereas the northern 30 Ma and c. 1.5 cm/annum thereafter (Rosen- and southern margins correspond to strike–slip crantz et al. 1988). Active faulting and volcanism with subsidiary transpression or transtension zones related to eastward subduction of the Pacific plate (Polochic–Motagua, Cayman, Greater Antilles, overprinted the ophiolite emplacement and intra- North Andean and South Caribbean, e.g. Pindell continental translation features within the crust of et al. 2006). The Motagua Suture Zone (MSZ) of Mexico, Guatemala and Honduras (e.g. Guzma´n- southeastern Mexico, Guatemala, and Honduras Speziale 2001). extends from the Pacific Ocean to the Caribbean Sea for c. 400 km east–west and c. 80 km north– south (e.g. Beccaluva et al. 1995), separating the Methods Maya Block from the Chortı´s Block (Fig. 1a, Dengo 1969). This belt actually represents a compo- Methodical and technical aspects of our radiometric site structure, encompassing suture zone(s) with dating that includes U–Pb, Ar–Ar, Rb–Sr, and relics of Mesozoic ocean floor that characterize fission-track geochronology are provided as the northern Caribbean Plate boundary, and Ceno- Supplementary material. Also included in the Sup- zoic fault zones (mainly the Polochic, Motagua, plementary materials is our approach to structural Jocota´n–Chameleco´n faults) that link the Cayman analysis of ductile deformation and kinematics, trough pull-apart basin in the east with the subduc- to fault–slip analysis and definition of stress-tensor tion zone of the Pacific plate in the west (Fig. 1). groups in the brittle crust, and a review of the The Maya and Chortı´s Blocks showcase a geo- applied calculation techniques. Methodical and logical history that encompasses Middle Proterozoic technical aspects of our petrology and geothermo- to Quaternary and have long been recognized as barometry are included in the Supplementary key elements in understanding the interaction material along with locations of studied outcrops of Laurentia, Gondwana and the (Proto-)Pacific (‘stops’) and descriptions of analysed samples. Fig. 1. (a) Plate tectonic framework of northern Central America and crustal complexes (‘terranes’) of southern Mexico. Numbers in the Cayman trough give age of transitional to oceanic lithosphere (from Rosencrantz et al. 1988). (b) SRTM digital elevation model of the Motagua suture zone in southern Mexico, Guatemala, and Honduras, highlighting neotectonic structures such as the Polochic, Motagua and Tonala strike–slip faults and NNE-trending grabens (e.g. Sula graben). The Jalpatagua fault zone comprises distributed neotectonic dextral strike–slip deformation within the Cenozoic arc. (c) Basement geology of southern Mexico and northern Central America and available reliable geochronological data surrounding the Motagua suture zone. References for Belize, Guatemala and Honduras: 1, Ortega-Gutie´rrez et al. (2004); 2, Harlow et al. (2004); 3, Horne et al. (1976a); 4, Manton (1996); 5, Venable (1994); 6, Steiner & Walker (1996); 7, recalculated from Gomberg et al. (1968); 8, unpublished, referenced in Donnelly et al. (1990); 9, Bertrand et al. (1978); 10, Manton & Manton (1984); 11, Drobe & Oliver (1998); 12, Donnelly et al. (1990); 13, Emmet (1983); 14, Italian Hydrothermal (1987; reported in Rogers 2003); 15, Horne et al. (1976b); 16, Horne et al. (1974); 17, MMAJ (1980; reported in Rogers 2003); 18, Manton & Manton (1999); 19, Weiland et al. (1992); 20, Bargar (1991); 21, Ortega-Obrego´n et al. (2008); 22, Solari et al. (2008); 23, Martens et al. (2008). References for Mexico: A, Ortega-Gutie´rrez et al. (1999); B, Talavera-Mendoza et al. (2005); C, Sa´nchez-Zavala et al. (2004); D, Yan˜ez et al. (1991); E, Weber et al. (2006a); F, Ducea et al. (2004a); G, Elı´as-Herrera & Ortega-Gutie´rrez (2002); H, Solari et al. (2001); I, Vega-Granillo et al. (2007); J, Keppie et al. (2004); K, Mora´n-Zenteno (1992); L, Solari et al. (2007); M, Herrmann et al. (1994); N, Elı´as-Herrera et al. (2005); O, Tolson (2005); P, Robinson et al. (1989); Q, Wawrzyniec et al. (2005); R, Weber et al. (2005); S, Alaniz-Alvarez et al. (1996); T, Solari et al. (2004); U, Schaaf et al. (2002); V, Weber et al. (2008). Ages without label are our own data from southern Mexico and the central Chortı´s Block. N. AMERICA–CARIBBEAN PLATE BOUNDARY 221

Basement and cover of the Maya and within these ‘southern’ ophiolites show mid-ocean Chortı´s Blocks ridge basalt (MORB) affinity (Beccaluva et al. 1995). In the north-central Chortı´s Block (Fig. 2j), Figure 2 presents lithostratigraphic columns of the the clastic, partly turbiditic, rift-related Agua Frı´a southern Maya Block, the MSZ, and the north- Formation was deformed during Late Jurassic central Chortı´s Block, based mainly on the compi- (Viland et al. 1996). It is overlain by the shallow- lations given in Donnelly et al. (1990), Talavera- marine, shelf-type Yojoa Group that contains arc Mendoza et al. (2005, 2007), Vega-Granillo et al. and intra-arc rift deposits (Rogers et al. 2007a). (2007) and Rogers (2003). We modified the columns Possibly two molasse sequences are present: the according to our new data and the recent literature mainly red beds of the Albian–Cenomanian Valle (discussed below) and, where available, added strati- de Angeles Group and the ?Eocene–Oligocene graphic range (converted into Ma), thickness, reli- Subinal Formation; the latter may have been depos- able radiometric ages, PT estimates and first-order ited in pull-apart basins along the Motagua fault tectono-stratigraphic interpretations. We also show zone. Thus, molasse-type deposits occur on the columns of units that we will use in the discussion Chortı´s during late Early to Late Cretaceous times for correlation and refinement of tectonic models. (Valle de Angeles Group), whereas marine platform carbonates developed on the Maya Block (Coban Cover sequences and Campur Groups). Top-to-north thrust imbrica- tion affected the Jurassic–Cretaceous sequence of In general, the Maya Block (Fig. 2b) records shelf the Chortı´s Block during Late Cretaceous (Donnelly evolutions during Late Palaeozoic and Cretaceous et al. 1990; Colon fold belt of Rogers et al. 2007b). that were followed by terrestrial–shallow marine rift (Middle–Late Jurassic) and submarine-fan Basement complexes (Late Cretaceous) developments. Along the MSZ, ophiolites are imbricated with probably Upper Clear definition and subdivision of the rock units Jurassic to Upper Cretaceous submarine-fan and that comprise the MSZ (Fig. 1c) remain elusive prin- ocean-basin deposits (North El Tambor complex cipally because Cenozoic faults delineate parts of and El Pilar Formation, Fig. 2a) and crystalline these units (see also Ortega-Gutie´rrez et al. 2007). basement of the Chuacu´s complex; they are uncon- This section briefly describes the major units of the formably overlain by continental red beds (Subinal MSZ from north (Maya Block) to south (Chortı´s Formation). In the Sierra de Santa Cruz (Fig. 1c), Block). massive ophiolitic rocks that contain basalts of island-arc affinity are imbricated with Maastrichtian Maya Mountains and Altos Cuchumatanes (Maya to Eocene (c. 80–50 Ma) carbonate–arenaceous Block). Metamorphic rocks in the Maya Mountains pre-flysch and flysch (Sepur Formation; Wilson of Belize (Fig. 1c) comprise low-grade meta- 1974; Beccaluva et al. 1995). Aptian and Cenoma- sedimentary rocks with minor rhyolite–dacite inter- nian (c. 120–95 Ma) sedimentary rocks of the calations (Steiner & Walker 1996; Martens et al. Pete´n basin of northern Guatemala (Maya Block) 2010). Lower Palaeozoic beds are intruded by gran- record a pelagic facies, reflecting maximum trans- itoids; Upper Palaeozoic deposits were correlated gression and an open marine connection to the Gulf with the Santa Rosa Group based on lithology and of Mexico to the north, and the Proto-Caribbean Carboniferous–Permian fossils (Bateson 1972). seaway to the south (Fourcade et al. 1994, 1999). The Altos Cuchumatanes area (Fig. 1c) exposes low- During the Late Maastrichtian and Danian (c.70– to high-grade basement, intruded by granitoids, and 60 Ma), the southern Pete´n basin (Fig. 1b) was a mantled by Santa Rosa Group metasedimentary deep siliciclastic sink, the Sepur foredeep basin. rocks (Anderson et al. 1973). The depocentre of this basin shifted from south during the Late Campanian to north during the Rabinal complex: Salama´ schists, San Gabriel unit, Late Maastrichtian and Early Danian; this led to and Rabinal granitoids (Maya Block). Van den the demise of the pre-existing carbonate platform Boom (1972) called low-grade metasedimentary over more than 100 km from south to north. rocks south of the Polochic fault zone in central The ophiolites south of the Motagua fault (South Guatemala the Salama´ schists. These rocks were El Tambor complex, Fig. 2k) are limited to the north grouped into an older clastic and volcanic succes- by the Caban˜as fault and are overlain by Subinal red sion, the San Gabriel unit, which is intruded by beds and in the south by the Cenozoic–Recent the Rabinal granitoids, and a younger clastic and volcanic arc; they are cut by felsic intrusives. calcareous succession, probably, part of the Santa Together with Late Jurassic–Cretaceous ‘me´lange’, Rosa Group (Fig. 1c; Donnelly et al. 1990; they cover basement of the Las Ovejas complex and Ortega-Obrego´n et al. 2008). The basal conglomer- the San Diego Group of the Chortı´s Block. Basalts ate beds of the Santa Rosa Group comprise the 222 L. RATSCHBACHER ET AL.

Fig. 2. (Litho)stratigraphic and pressure–temperature–time–deformation evolution and tectono-stratigraphic interpretation of the Maya Block, the Motagua suture zone, and the northern Chortı´s Block and correlative units in southern Mexico and the southern Chortı´s Block (see text for discussion of correlations). Ages in italics are detrital U–Pb zircon ages or from inherited zircons in granitoids; other ages are from magmatic rocks. All ages are those reported in the captions of Fig. 1c and the text. Pressure–time conditions from this paper and literature are reviewed in the text.

Sacapulas Formation of van den Boom (1972). Carboniferous crinoids (van den Boom 1972) and Ortega-Obrego´n et al. (2004, 2008) proposed an up Mississippian (Tournaisian) conodonts (Ortega- to 10 km wide, south-dipping shear zone, the Baja Obrego´n et al. 2008). The Rabinal granitoids are Verapaz shear zone, thrusting (with a minor sinistral spatially poorly defined, as their sheared margins strike–slip component) Chuacu´s gneiss onto San and the surrounding sediments have similar appear- Gabriel rocks; deformation is most pronounced ance; they are predominately fault-bounded and within 300 m of the Chuacu´s–San Gabriel bound- widely tectonized. Here we use the term Rabinal ary. The lateral continuation of the shear zone is complex to describe the in general low-grade rock unknown. The San Gabriel unit has a continental assemblage that comprises pre-Middle Devonian depositional environment and is pre-Silurian, given volcanoclastic rocks and Ordovician–Early Devo- by the age of the Rabinal granitoids (see below). It nian granitoids; it is widely associated with its resembles low-grade metasedimentary rocks in clastic–calcareous cover, the Santa Rosa Group. We western Guatemala, which are post-Middle Protero- note that there is no obvious difference in Palaeo- zoic based on detrital zircon geochronology (Solari zoic lithostratigraphy and magmatism between the et al. 2008). The Santa Rosa Group yielded Early Maya Block (Maya mountains, Alto Cuchumatanes) N. AMERICA–CARIBBEAN PLATE BOUNDARY 223

Fig. 2. (Continued). and the Rabinal complex, suggesting that the calcsilicate and marble are uncommon. Local quan- Polochic fault is not a fundamental plate boundary titative petrology suggests upper greenschist- to fault (see Ortega-Gutie´rrez et al. 2007). amphibolite-facies conditions up to 650 8Cat c. 1.1 GPa; retrogression to greenschist facies is Chuacu´s complex. In Guatemala, Ortega-Gutie´rrez poorly documented (van den Boom 1972; Machorro et al. (2004) redefined the Chuacu´s complex as 1993). Both foliation concordant and discordant schists and gneisses outcropping between the quartz–albite–white mica pegmatites occur. Eclo- Motagua fault and the Baja Verapaz shear zone of gitic relicts contain possible ultra-high pressure the Salama´ –Rabinal area (Fig. 1c). Chuacu´s schists metamorphism, and pyroxene with up to 45 vol% and gneisses contain quartz þ albite þ white mica þ omphacite and garnet together with phengite þ epidote/zoisite + garnet + biotite + amphibole + rutile þ zoisite þ epidote. Thermobarometric ana- omphacite + allanite + chlorite + calcite + titanite + lyses yielded c. 740 8Catc. 2.3 GPa with retrogres- rutile (Martens et al. 2005). Banded migmatite sion at c. 590 8C and c. 1.4 GPa. Decompression and (garnet) amphibolite (locally relict eclogite), seems associated with local partial melting 224 L. RATSCHBACHER ET AL.

Fig. 2. (Continued).

(Ortega-Gutie´rrez et al. 2004). In central-eastern three Rabinal granites and one pegmatite are discor- Guatemala (San Agustı´n Acasaguastla´n), ortho- dant (Ortega-Obrego´n et al. 2008); they yielded gneiss (quartz–monzonite, granodiorite) outcrops lower intercepts of 496 + 26 and 417 + 23 Ma together with high-grade, typical Chuacu´s para- (granite), 462 + 11 Ma (pegmatite), and an upper gneiss; the latter locally shows sillimanite and abun- intercept of 483 + 23 Ma (granite). Muscovite dant K-feldspar. The volcanosedimentary protolith from two pegmatites cutting the San Gabriel unit of the Chuacu´s complex is probably younger than yielded K–Ar cooling ages of 453 + 4 and c. 440 Ma based on detrital zircon geochronology 445 + 5 Ma. Ortega-Obrego´n et al. (2008) sug- (Solari et al. 2008). gested intrusion of the Rabinal granitoids between 462 and 453 Ma. The Matanzas granite of eastern Existing geochronology: Maya Block, Rabinal and central Guatemala, possibly intruding the Rabinal Chuacu´s complexes. U–Pb zircon and monazite complex, yielded a two feldspar–whole rock Rb– ages obtained from two batholiths in the Maya Sr isochron of c. 227 Ma, and Ar/Ar white mica mountains (Steiner & Walker 1996) yielded and biotite ages of c. 212 and 161 Ma, respectively 404–420 Ma (the best constrained samples are (Donnelly et al. 1990). Four zircon fractions com- 418 + 3.6 and 404 + 3.3 Ma); K–Ar biotite ages bining two Chuacu´s gneiss samples south of of these igneous rocks are 230 + 9 Ma (weighted Salama´ in central Guatemala gave a lower intercept mean of seven dates compiled in Steiner & Walker U–Pb age of 326 + 85 Ma (recalculated from 1996). Martens et al. (2010) discriminated a Gomberg et al. 1968). Three single zircon and two post-520 Ma (detrital zircon geochronology) volca- small populations from refolded leucosome inter- nosedimentary sequence (Baldy unit) that is inter- bedded with relict eclogite (El Chol area) yielded bedded with 404 + 7 Ma rhyolite in its upper intercept U–Pb ages of 302 + 6 and 1048 + section (Bladen Formation). Granitoids of the 10 Ma (Ortega-Gutie´rrez et al. 2004). K–Ar and Altos Cuchumatanes area yielded 269 + 29 and Ar/Ar ages from Chuacu´s-complex rocks are 391.2 + 7.4 Ma (Solari et al. 2008). Weber et al. mostly poorly located and documented. Hornblende, (2008) reported 482 + 5 Ma U–Pb zircon and white mica and biotite gave Ar/Ar ages of 78–66, 473 + 39 Ma Sm–Nd garnet–whole rocks ages 76–67 and 67–64 Ma, respectively (Sutter 1979). from S-type granite from the Chiapas Massif of Amphibole from garnet amphibolite and white the southwestern Maya Block. Three zircon frac- mica from retrograde phengite-bearing gneiss tions of a Rabinal orthogneiss yielded a concordant yielded K–Ar ages of c. 73 and c. 70 Ma, respect- 396 + 10 Ma age (recalculated from the data given ively. Coarse- and finer-grained white mica from by Gomberg et al. 1968). All zircon fractions of foliated metapegmatite gave c. 73 and c. 62 Ma, N. AMERICA–CARIBBEAN PLATE BOUNDARY 225 respectively (weighted mean ages from several frac- þ8.6 and þ9.2). One northern eclogite has a signi- tions given by Ortega-Gutie´rrez et al. 2004). White ficantly more enriched signature (87Sr/86Sr ¼ mica from Chuacu´s-complex paragneiss and peg- 0.70536; 1Nd ¼ 22.1). Harlow et al. (2004) matite 4 km south of the main trace of the Baja obtained Ar/Ar white mica ages from jadeitite, albi- Verapaz shear zone gave ages between 74.1 + 1.3 tite, mica-bearing metamorphic rocks and altered and 65.3 + 1 Ma; they were suggested to date eclogite within the MSZ, ranging between shearing along the Baja Verapaz zone (Ortega- 77–65 Ma for the northern high-P belt and Obrego´n et al. 2008). The oldest Ar/Ar age from 125–113 Ma for the southern belt. In the northern the Chuacu´s complex is c. 238 Ma (amphibole; belt, three phengite-only samples are indistinguish- Donnelly et al. 1990). able within error at 76 Ma, whereas two phengite– phlogopite–paragonite mixtures are younger at 71 El Tambor complexes: metamorphic–metasomatic and 65 Ma. The 600–650 8C peak T of the northern rocks of the Motagua suture zone. The El Tambor high-P rocks suggests that the phengite (closing T complexes (Fig. 1c) comprise metasedimentary c. 400 8C, von Blanckenburg et al. 1989) dates retro- rocks of a subduction–accretionary complex, sheets gression under blueschist-facies conditions when of serpentinized peridotite, serpentinite me´lange Na–Al–Si-rich fluids produced jadeitite and albitite and exotic blocks, containing eclogite, jadeitite, in serpentinite (Harlow 1994). All dated minerals mylonitized gabbro, (garnet) amphibolite, pillow from the southern belt are phengite; four ages are lavas and radiolarian cherts; as far as is known, within error at 120 Ma, one is older, two younger. the basal contacts are tectonic (e.g. Bertrand & These ages probably date fluid infiltration during Vuagnat 1975; Harlow 1994; Beccaluva et al. blueschist-facies retrograde metamorphism. Ber- 1995; Tsujimori et al. 2005; Chiari et al. 2006). trand et al. (1978) reported K–Ar mineral and Regional prehnite–pumpellyite-facies metamorph- whole-rock ages from poorly located samples ism affected the volcanic rocks, forming albite þ north of the Motagua fault zone. We divided their chlorite, predominantly pumpellyite, and locally samples regionally and recalculated the ages: a prehnite. Centimetre- to metre-scale interlayering metadolerite from the Baja Verapaz ophiolite of mafic rocks, siliceous marble and quartzite (prob- sheet (Fig. 1c) north of Salama´ yielded c. 76 Ma, ably metamorphosed chert) are reported in the La three metadiabase samples from pillow basalt over- Pita complex of the eastern MSZ (Martens et al. lying the Chuacu´s complex gave a weighted mean 2007); assemblages with barroisite þ garnet þ whole-rock age of c. 74 Ma (the 40Ar/36Ar v. epidote þ albite þ white mica are characteristic. 40K/36Ar isochron gives atmospheric 40Ar/36Ar), Jadeitite formed sporadically within the serpentinite and six amphibole separates from garnet amphibo- matrix of the El Tambor me´lange; it consists of lites of the same unit yielded a weighted mean age jadeitic pyroxene with minor paragonite, phengite, of 57.2 + 5.4 Ma (again, atmospheric 40Ar/36Ar phlogopite, albite, titanite, zircon, apatite and with an ‘isochron age’ of c. 56 Ma). All ages prob- garnet and formed at 100–400 8C and 0.5– ably date metamorphism. 1.1 GPa (Harlow 1994). Two contrasting high-P The available petrology (lawsonite and zoisite belts contain lawsonite eclogite and zoisite eclogite eclogites south and north of the Motagua fault south and north of the Motagua valley, respectively. zone, respectively), geochronology (c. 120 and The southern lawsonite eclogites record unusually 76 Ma for blueschist-facies retrogression south low-T–high-P with prograde eclogite facies at and north of the Motagua fault zone, respectively) c. 450 8C and c. 1.8–2.4 GPa, followed by retro- and isotope geochemistry (isotopically depleted gression with infiltration of fluids at ,300 8C and v. enriched south and north of the Motagua c. 0.7 GPa (Tsujimori et al. 2004, 2005, 2006). fault zone, respectively) demonstrate the existence Amphibolitized, paragonite-bearing zoisite eclo- of two distinct subduction–accretion belts along gites in antigorite schist north of the Caban˜as fault the MSZ, even though ages of crust formation formed at significantly higher T, 600–650 8Cat and eclogite–facies metamorphism may appear 2.0–2.3 GPa (Tsujimori et al. 2004). similar. Herein, we use the terms South and North El Tambor complexes for these two distinct Existing geochronology: El Tambor complexes. subduction–accretion belts and their ophiolitic Two Sm–Nd mineral isochrons from eclogites sheets, with the South El Tambor complex being south of the Motagua fault zone gave ages of located south of the Caban˜as strand of the 131.7 + 1.7 and 132.0 + 4.6 Ma, which are identi- Motagua fault zone, and the North El Tambor cal to the 130.7 + 6.3 Ma age from eclogite complex located to its north. sampled north of the fault (Brueckner et al. 2005). The southern eclogites show MORB major and Las Ovejas complex, San Diego phyllite, and Caca- trace element patterns and are isotopically depleted guapa schist (Chortı´s Block). The Las Ovejas 87 86 ( Sr/ Sr ratios of 0.70374 and 0.70489, 1Nd of complex (Fig. 1c) comprises biotite schist, gneiss, 226 L. RATSCHBACHER ET AL. migmatite, amphibolite, marble and deformed over South El Tambor complex metaclastic rocks granitoids. Specifically, amphibolite (15% of the that show distinctly lower metamorphic grade. Las Ovejas complex in Guatemala), garnetiferous gneiss, staurolite schist, garnet–two-mica schist Existing geochronology: Las Ovejas complex, San and marble with calcsilicate layers are exposed toge- Diego phyllite and Cacaguapa schist, igneous ther with metaigneous rocks, diorite, tonalite, gran- rocks. Most of the available geochronology is, in a odiorite and minor gabbro. Mineral assemblages modern sense, poorly documented and samples are define amphibolite-facies conditions; metamorphic only approximately located (Fig. 1c). U–Pb zircon grade of rocks containing staurolite increases ages on igneous rocks are: c. 93 Ma (Rio Julian toward the NW where a sillimanite zone was ident- granite), c. 80 Ma (sheared Rio Cangrejal orthog- ified (IGN 1978). Both basement and cover of the neiss), 81.0 + 0.1 (El Carbon granodiorite), northern Chortı´s Block continue as the Bonacca c. 29 Ma (Banderos granodiorite; all Manton & Ridge, forming a series of islands south of the Manton 1984), 124 + 2 Ma (dacite; Drobe & Swan Island Fault (Fig. 1c), the southern boundary Oliver 1998), and 1017 Ma (orthogneiss; Manton fault to the Cayman Trough. Roata´n Island, which 1996). We recalculated the zircon data of the El is the best studied (Ave´ Lallemant & Gordon Carbon granite of Manton & Manton (1999) and 1999), exposes amphibolite-facies basement with obtained lower and upper intercepts of 30 + 19 (augen)gneiss, schist, amphibolite (partly metagab- and 126 + 42 Ma, respectively. Deformed granitic bro), marble and rare pegmatite. plutons gave a 300 + 6 Ma Rb–Sr whole-rock The San Diego metamorphic rocks consist of isochron for the Quebrada Seca complex (Horne phyllite and minor thin quartzite and slate layers et al. 1976a). Granitoids gave the following (Schwartz 1976). Their stratigraphic age is post- Rb–Sr ages: 150 + 13 Ma (San Marcos granodior- Early Cambrian (youngest detrital zircons are ite, whole rock; Horne et al. 1976a), c.69Ma c. 520 Ma; Solari et al. 2008). The Chiquimula bath- (K-feldspar and plagioclase, cooling age of sheared olith thermal overprint produced cordierite and El Carbon granodiorite; Manton & Manton 1984), andalusite (Clemons 1966). Greenschist-facies c. 80 Ma (whole-rock isochron, Trujillo grano- and locally higher-grade metamorphic rocks in diorite; Manton & Manton 1984), 140 + 15 Ma Honduras include calcareous phyllite, white mica– (Dipilto granite, whole rock; Donnelly et al. 1990), chlorite and graphitic schists and slate (‘younger 118 + 2 Ma (Carrizal diorite, K-feldspar and plagi- sequence’ of Horne et al. 1976a). The San Diego oclase; Manton & Manton 1984), c. 54 Ma (Rı´o phyllite of southern Guatemala and northern Cangrejal granodiorite), 39 Ma (Sula valley (Sula Honduras has been correlated with the Cacaguapa graben) granodiorite), and 35 Ma (Confadia granite; schists south and east of the Jocota´n–Chameleco´n all biotite cooling ages; Manton & Manton 1984). fault system (Fig. 1c; Burkart 1994). The K–Ar cooling ages of igneous rocks comprise greenschist-facies lithology at Roata´n Island com- 93.3 + 1.9, 80.5 + 1.5 and 73.9 + 1.5 Ma (Tela prises (quartz) phyllite, chlorite schist, serpentinite, augite-tonalite, one amphibole and two biotite chert and marble to limestone; locally granite por- ages from different samples; Horne et al. 1974), phyry is present (Ave´ Lallemant & Gordon 1999). 72.2 + 1.5 and 56.8 + 1.1 Ma (Piedras Negras tonalite, amphibole and biotite; Horne et al. 1974), Sanarate complex. We introduce the Sanarate 35.9 + 0.7 Ma (San Pedro Sula granodiorite, complex as a distinct unit separate of the South El biotite; Horne et al. 1974), 122.7 + 2.5 and Tambor and the Las Ovejas complexes due to its dis- 117.0 + 1.9 Ma (San Ignacio granodiorite, hornble- tinct lithology and P–T–d–t parameters; however, nde and biotite; Emmet 1983), 114.4 + 1.7 Ma (San its boundaries and lithostratigraphy are mostly Ignacio adamellite, biotite; Horne et al. 1974), unknown and we may have exaggerated its extend 79.6 + 1.6 Ma (gabbro dike, amphibole; Emmet in the geological maps (e.g. Fig. 1c). The complex 1983), 62.2 + 2.6 Ma (diorite, whole rock; Italian constitutes the westernmost basement outcrop Hydrothermal 1987, reported in Rogers 2003), south of the MSZ and comprises partly retrograde 60.6 + 1.3 and 59.3 + 0.9 Ma (Minas de Oro gran- amphibolite, garnet- and quartz-bearing amphi- odiorite, biotite; Horne et al. 1976b; Emmet 1983), bolite, retrograde leucogabbro, chlorite–epidote– 58.6 + 0.7 Ma (San Francisco dacite pluton, biotite; hornblende schist, micaschist, phyllite and chlorite Horne et al. 1976b). Ar/Ar ages are from ande- phyllite. Massive amphibolite is characteristic and site and diorite of the eastern Chortı´s Block most widespread. These lithologies probably rep- (75.6 + 1.3 and 59.9 + 0.5 Ma, amphibole and resent an oceanic environment. The complex is in biotite; Venable 1994). A K–Ar biotite age of fault contact with Cenozoic andesite and red beds 222 + 8 Ma from micaschist probably reflects a and is cut by andesite dykes. Along its eastern Triassic thermal event (MMAJ 1980). Hornblende margin, the Sanarate-complex rocks shows west- from the amphibolite-facies basement of Roata´n dipping foliation and apparently thrusted top-to-east Islandyielded36.0 + 1.2 Ma,andagraniteporphyry N. AMERICA–CARIBBEAN PLATE BOUNDARY 227 from Barbareta Island, east of Roata´n Island, yielded diagrams: ages from sample 5C-36 of the Chuacu´s a 39.4 + 2.8 Ma zircon fission-track age (closing T complex immediately below North El Tambor c. 250 8C; Ave´ Lallemant & Gordon 1999). complex serpentinite, as these rocks show extreme, late-stage metasomatism; ages from very coarse- New geochronology grained, pegmatitic white mica. In the following section, we present the data; we relate the ages to ID-TIMS and ion-probe U–Pb geochronology can the main temperature–deformation (T–d) events be found in the Supplementary materials, and in the petrology and structure sections. Figs 3 & 4 show the U–Pb data in Concordia dia- grams and present representative zircon CL pictures, U–Pb zircon geochronology respectively. Our new Ar/Ar and Rb–Sr mineral ages are summarized in Table 1 and the Supple- Rabinal complex. Sample G18s (stop G1; Figs 3a & mentary material, respectively, and are plotted in 7a) is a migmatitic gneiss from western Guatemala Fig. 5. In this paper, we discuss Phanerozoic U– overprinted by lower greenschist-facies defor- Pb ages; late Middle Proterozoic ages, which we mation. Four fractions with identical 75–125 mm find in both the Maya and Chortı´s Blocks, will be grains but different magnetic properties were uti- reported elsewhere. Our apatite fission-track lized for ID-TIMS analysis. A discordia fit to three (AFT) and titanite fission-track (TFT) data are fractions yields intercepts of 220.0 + 3.5 and listed in the Supplementary material; confined 1292 + 39 Ma (MSWD ¼ 0.78); one of these frac- track-length data appear in the Supplementary tions is concordant at 215.92 + 0.18 Ma. The fourth material. Figure 6 depicts AFT age v. elevation dia- fraction is concordant at 268.04 + 0.55 Ma. CL grams for the Rabinal–Chuacu´s and Las Ovejas images (Fig. 4) depict well-rounded cores with complexes of Guatemala/Honduras and the Chiapas variable but mostly high U-content, and low-U (on Massif of southeastern Mexico (Fig. 1c), and mod- average 63 ppm), oscillatory zoned, idiomorphic elled T–t paths for selected samples (Chuacu´s and rims with Th–U ratios (on average 0.7) that suggest Las Ovejas complexes, granitoids). We excluded magmatic origin. The weighted average of six the samples from western Guatemala from the age grains analysed with the SHRIMP gives a v. elevation data (too few readings and widely 207-corrected 206Pb– 238U age of 217.7 + 7.3 Ma, spaced) and also emphasize that the plotted data within error of the concordant TIMS fraction. We do not strictly fulfill the criteria for exhumation interpret this age as the age of leucosome crystalli- estimates from age v. elevation data; for example, zation within the western Rabinal complex; the the samples cover considerable horizontal distances. well-rounded cores suggest a paragneiss protolith We assigned errors to the elevation data, as several and the concordant fraction at c. 268 Ma may indi- samples were collected in the early-1990s when cate an earlier phase of crystallization. Sample hand-held GPS-receiver precision was still inferior G27s (stop G57, western Guatemala, Figs 3a & and large-scale topographic maps were locally 7a) is orthogneiss that shows low-T mylonitization unavailable. We emphasize that the regression along discrete sinistral shear zones. Five fractions data that we present are first-order estimates. varying in grain size or magnetic properties are all Figure 7 shows the regional distribution of ages, discordant but define a chord with lower and upper highlights age groups, and depicts T–t fields for intercept ages at 284 + 58 and 958.7 + 6.8 Ma, selected regions. We employed standard closing respectively (MSWD ¼ 0.63); this rock is probably temperatures for the T–t diagrams: Ar hornblende, a Permian intrusion. Sample GM8s (stop G8; high-T steps, 525 + 25 8C, low-T steps, 400 and Figs 3a & 7a) is granite overprinted by greenschist- 300 + 100 8C, depending on grain size; Ar white facies mylonitization from western Guatemala. Four mica, 350 + 25 8C; Ar biotite, 300 + 25 8C; Ar K- zircon fractions with variable grain size and mag- feldspar, 400 + 100 8C for high-T steps and netic properties define lower and upper intercepts 200 + 25 8C for low-T steps, depending on the at 165.2 + 5.4 and c. 990 Ma, respectively (ID- likely presence of single or multiple domains; Rb TIMS); one fraction is weakly reverse discordant white mica, 500 + 25 8C, except for centimetre- and two fractions define a concordant age of sized, pegmatitic white mica, 600 + 100 8C; Rb 163.80 + 0.38 Ma (Fig. 3a). CL images show oscil- biotite, 275 + 25 8C; TFT, 275 + 50 8C; AFT, latory zoned rims around Proterozoic cores (Fig. 4) 100 + 25 8C; exceptionally, errors were set and oscillatory zoned coreless grains with marked higher, imposed by variable grain size and strong changes in luminescence from centre to rims; the deformation. Zircon crystallization was taken as ages of these grains are identical within error. Con- 800 + 50 8C, except where zircon growth occurred sequently, the Th–U ratios obtained from SHRIMP during peak-T metamorphism and related migmati- spots are variable and are with one exception .0.2, zation, which is constrained by our petrology (see implying magmatic origin. The 207-corrected ages below). We excluded a few samples from the T–t for a group of six rims and coreless grains yield 228 .RATSCHBACHER L. TAL. ET

Fig. 3. U–Pb zircon data and age interpretation. See the Supplementary material for sample location and data listings. ID-TIMS results are shown with 2s error ellipses and discordia are given as lines. SHRIMP results show 2s error ellipses and are corrected for common Pb. Selected age groups are displayed as weighted average and relative probability diagrams (206 Pb– 238 U ages are 207-corrected). Data evaluation and plotting was supported by the program package of Ludwig (2003). MSWD, mean square weighted deviation. The upper intercept ages and ages older than Early Palaeozoic will be discussed elsewhere. (a) Data from the Rabinal complex of the Maya Block. .AEIACRBENPAEBOUNDARY PLATE AMERICA–CARIBBEAN N.

Fig. 3. (Continued)(b) Data from the Chuacu´s complex. 229 230 .RATSCHBACHER L. TAL. ET

Fig. 3. (Continued)(c) Data from the Chortı´s Block; all SHRIMP ages are from the northeastern edge of the Las Ovejas complex and granitoids intruding it. Th–U diagrams aid discrimination of metamorphic and magmatic zircons (or rims). .AEIACRBENPAEBOUNDARY PLATE AMERICA–CARIBBEAN N.

Fig. 3. (Continued)(d) Data from the Chorti´s Block, TIMS U–Pb zircon ages. 231 232 L. RATSCHBACHER ET AL.

Fig. 4. Cathodoluminescence images of representative zircons. Numbers give sample, spot and spot age (in Ma, below or above the sample number). a weighted average of 171.0 + 2.1 Ma. Granite low-T overprint; it defines the ‘Rabinal’ granitoids emplacement is either given by this date or is in the area of San Miguel Chicaj, central Guatemala slightly younger (concordant TIMS fractions). (Fig. 7a). Rounded, often mottled, Proterozoic Sample 5CO-1 is two-mica, K-feldspar– cores are surrounded by in general thin, structure- porphyroblastic orthogneiss, mylonitized during less, often high-U rims that have variable but Table 1. 40 Ar/39 Ar data

Sample Stop Mineral J Weight Grain size Total fusion Weighted Isochron age MSWD 40Ar/36 Ar %39 Ar Steps (mg) (mm) age (Ma) mean age (Ma) used (Ma)

Rabinal complex G19s G1 Phe 0.003192 10.1 10–20 66.66 + 0.33 66.53 + 0.33 66.50 + 0.37 4.1/2.0 296 + 4 61 6–16/21 Phe 0.003185 10.1 20–40 67.92 + 0.34 67.38 + 0.33 67.00 + 0.37 4.1/1.9 307 + 5 79 7–19/23 BOUNDARY PLATE AMERICA–CARIBBEAN N. G27s G57 Bt 0.004555 0.3 250 37.38 + 1.37 35.06 + 1.29 37.21 + 10.06 0.6/2.4 350 + 74 48 5–10/15 Chuacu´s complex G10s G29 WM 0.004564 1.2 100–200 66.04 + 0.65 66.00 + 0.65 65.96 + 0.66 14.6/1.7 299 + 9 93 3–21/26 G12s G31 WM 0.0007854 n.d. n.d. 71.80 + 0.89 72.59 + 0.65 67.8 + 5.0 31 297 + 64 100 1–19/19 5CO-4b 5CO-4 Bt 0.004480 4.3 250 64.31 + 0.63 64.34 + 0.63 64.80 + 0.68 26.6/1.8 264 + 11 98 2–17/19 WM 0.004590 2.9 250 67.77 + 0.67 67.51 + 0.66 67.56 + 0.73 174/1.7 286 + 17 100 1–17/17 5CO-5a 5CO-5 WM 0.004586 2.1 250 234.4 + 2.2 225.1 + 2.1 225.2 + 2.2 0.9/2.6 294 + 3 5 1–5/23 Hbl 0.004581 8.9 250 670.61 + 5.60 — — — — 100 1–20/20 5PB-9b 5PB-9 WM 0.004570 5.1 250 60.27 + 0.59 — — — — 100 1–27/27 5PB-13b 5PB-13 Hbl 0.0007852 n.d. n.d. 94 + 18 69.3 + 1.4 64.5 + 5.4 0.31 1283 + 270 35 5 of 12 5PB-14 5PB-14 WM 0.004568 2.9 250 64.67 + 0.63 65.01 + 0.64 64.93 + 0.64 1.2/2.3 297 + 16814–20/20 854F 854 Bt 0.014640 28 250 125.3 + 3.5 — — — — 100 10/10 880C 880 WM 0.014640 16.8 250 102.4 + 2.7 101.0 + 2.5 100.4 + 1.5 0.9/2.6 335 + 32 88 5–9/10 883B 883 WM 0.014640 17.1 250 75.2 + 1.8 74.7 + 1.5 74.2 + 0.8 0.6/2.4 338 + 21 93 5–10/10 883B 883 Bt 0.014640 14.9 250 68.2 + 2.0 68.6 + 2.0 71.3 + 1.8 9.0/3.0 146 + 90 100 1–8/8 887C 887 WM 0.014376 5.4 250 70.5 + 1.2 69.8 + 1.2 70.9 + 3.8 13 321 + 120 78 1–7/10 Metasomatized southern Chuacu´s complex G6s G26 WM 0.004565 0.6 c. 63 71.40 + 0.80 62.01 + 0.69 59.55 + 2.37 2.2/2.3 497 + 367 48 6–12/16 G7s G27 WR rhyolite 0.004446 20.0 n.a. 68.37 + 0.94 69.91 + 1.06 63.66 + 9.19 9.5/3.0 305 + 15 59 2–5/7 5C-36a 5C-36 WM 0.004579 1.9 250 48.04 + 0.47 48.32 + 0.48 48.20 + 0.50 16.7/1.8 298 + 3 86 9–22/22 Sanarate complex G3s G23 WM 0.004534 2.2 c. 63 168.05 + 1.61 170.56 + 1.63 151.2 + 13.22 80/2.4 1339 + 918 69 3–8/17 G5s G24 Hbl 0.004519 9.5 250 128.08 + 1.29 155.85 + 1.50 155.92 + 1.91 75/2.7 291 + 83613–17/17 (Continued) 233 234

Table 1. Continued

Sample Stop Mineral J Weight Grain size Total fusion Weighted Isochron age MSWD 40Ar/36 Ar %39 Ar Steps (mg) (mm) age (Ma) mean age (Ma) used (Ma)

Western Las Ovejas complex, Guatemala 5C-2c 5C-2 Hbl 0.000786 n.d. n.d. 22.1 + 6.6 19.2 + 1.5 20.1 + 6.4 0.32 280 + 36 73 1–4/6 5C-5 5C-5 Bt 0.004574 0.5 c. 63 29.21 + 0.29 29.18 + 0.29 29.90 + 0.33 3.4/2.6 255 + 6 88 2–6/7 5C-23d 5C-23 Hbl 0.004609 23.4 250 35.99 + 0.37 35.43 + 0.36 35.81 + 0.40 1.6/2.0 276 + 9 64 6–15/17 5C-23c 5C-23 Bt 0.004612 0.8 250 25.57 + 0.26 25.24 + 0.25 25.34 + 0.26 0.3/3.0 291 + 2 60 3–6/9 5C-26c 5C-26 Hbl low-T 0.004607 18.9 250 27.84 + 0.28 17.76 + 0.18 18.21 + 0.45 48.3/2.6 272 + 18 45 1–5/19 medium-T 21.73 + 0.24 21.72 + 0.35 2.4/2.4 384 + 31310–15/19 high-T 38.01 + 0.38 35.78 + 2.28 43.8/3.8 314 + 18 17 15–17/19 RATSCHBACHER L. Bt 0.004605 2.9 250 13.84 + 0.62 20.49 + 0.24 21.08 + 0.84 18.1/2.0 294 + 2 76 4–9/11 5C-26d 5C-26 Kfs 0.004600 12.7 400 30.25 + 0.59 30.42 + 0.32 29.94 + 0.54 6.4/2.3 297 + 18910–16/17 5C-33 5C-33 Bt 0.004597 2.5 250 22.42 + 0.24 23.04 + 0.23 22.80 + 0.26 2.2/2.1 297 + 1 58 6–13/17 Kfs low-T 0.004524 26.8 400 129.9 + 2.0 7.4 + 1.1 6.7 + 3.2 1.7/2.6 298 + 25 23 6–10/21 medium-T 31.38 + 2.78 32.98 + 10.47 0.1/3.8 292 + 7 7 11–13/21 5C-37b 5C-37 Hbl high-T 0.004566 8.6 250 41.06 + 0.43 39.11 + 0.55 39.10 + 3.93 1.2/2.1 512 + 117 37 8–11/12 low-T 18.23 + 0.53 18.27 + 1.63 1.7/2.3 362 + 20 20 1–3/12 5C-37d 5C-37 Bt 0.004593 2.9 250 17.81 + 0.18 18.15 + 0.19 19.49 + 1.18 60/2.3 289 + 6 74 4–10/12 AL. ET 5PB-5a 5PB-5 Hbl 0.0007864 n.d. 250 30.4 + 1.0 31.03 + 0.70 29.9 + 1.4 67/– 243 + 77 96 2–13/15 Bt 0.004572 1.6 250 28.22 + 0.28 28.20 + 0.28 28.29 + 0.32 18.2/2.1 289 + 6 100 1–8/8 Eastern Las Ovejas complex, Honduras 5H-3a 5H-3 Bt (laser) 0.004420 s.g. 250 25.58 + 0.26 18.98 + 0.22 19.00 + 2.78 17.4/2.1 661 + 412 100 1–10/10 5H-4b 5H-4 Bt (laser) 0.004424 s.g. 250 25.58 + 0.34 25.70 + 0.34 23.05 + 1.86 1.6/3.0 314 + 13 100 1–4/4 Hbl 0.004532 5.3 250 36.87 + 0.40 38.81 + 0.40 38.86 + 1.78 93.7/2.6 291 + 135 92 5–9/10 5H-4c 5H-4 Bt 0.004577 2.3 250 24.77 + 0.25 25.75 + 0.26 24.67 + 0.44 0.6/3.0 311 + 4 54 5–8/9 Hbl 0.004527 7.5 250 32.57 + 0.40 31.53 + 0.32 31.44 + 0.65 52.3/2.1 297 + 6 98 4–11/11 5H-4d 5H-4 Hbl 0.0007881 n.d. n.d. 28.6 + 2.0 30.0 + 1.5 26.1 + 3.6 6.1/– 334 + 16 100 1–17/17 Granitoids of the arc complex 5C-11 5C-11 Bt þ Chl 0.004422 s.g. n.a. 84.51 + 0.89 89.98 + 0.89 — — — 100 1–10/10 5C-21 5C-21 Kfs (laser) 0.004416 s.g. 400 20.03 + 0.23 19.09 + 0.25 18.80 + 0.45 0.3/3.0 297 + 1 58 5–9/9 5H-5 5H-5 Kfs low-T 0.004529 12.3 400 32.98 + 0.33 28.29 + 0.28 28.31 + 0.36 2.4/2.6 291 + 59 27 10–14/33 5H-13 5H-13 Bt (laser) 0.004425 3.0 250 57.71 + 0.70 60.89 + 0.61 61.95 + 0.86 9.7/2.4 265 + 14 100 1–6/6 Southern Mexico ML-18 ML-18 Bt 0.004540 1.1 250 148.82 + 1.43 162.2 + 0.8 — — — 10 19–21/21 ML-37 ML-37 WM high-T 0.004542 1.7 250 302.76 + 2.76 315.09 + 2.87 316.3 + 0.4 5/3 289 + 23716–20/20 ML-37 ML-37 WM low-T 0.004542 1.0 250 302.76 + 2.76 246.70 + 2.76 255.39 + 22.70 68/3.8 289 + 14 2 1–3/20 MU-12 MU-12 Bt 0.0015255 1.0 250 29.69 + 0.35 28.60 + 0.29 28.62 + 0.40 0.02/3.8 295 + 25813–16/17 MU-13 MU-13 Bt 0.0015249 1.2 250 35.06 + 1.24 34.38 + 0.39 35.12 + 0.70 0.5/3.0 275 + 12 56 11–15/16 ML-39 ML-39 Bt 0.0015307 1.5 250 15.22 + 0.46 14.57 + 0.59 13.81 + 1.90 1.6/2.6 300 + 10 51 1–5/15 PA3-4 PA3-4 Hbl 0.0042502 s.g. — 8.1 + 0.6 8.8 + 1.5 1.9 290 + 27 100 1–5/5 PA3-4 PA3-4 Bt 0.0042502 203/s.g. 3 exp. — 8.5 + 0.3 2.1 Forced 100 1–9/9 CB35 CB35 Hbl 0.0036469 198/320 2 exp. — 28.4 + 1.3 3.0 293 + 25 — 1–7/9 CB35 CB35 Bt 0.0035870 35/203 2 exp. — 8.2 + 0.1 0.57 297 + 3 100 1–12/12

CB31 CB31 Hbl 0.0049746 207/305 2 exp. — 15.5 + 1.2 0.89 305 + 19 — 1–6/8 BOUNDARY PLATE AMERICA–CARIBBEAN N. CB34 CB34 Hbl 0.0035870 195/306 2 exp. — 11.0 + 1.8 0.33 307 + 6 100 1–9/9 PI28-1 PI28-1 Hbl 0.0042502 s.g. — 8.1 + 0.3 — — — 100 1–4/4 PI28-1 PI28-1 Bt 0.0042502 191/s.g. 3 exp. — 8.0 + 0.5 1.3 303 + 3 100 1–8/8

J is the irradiation parameter; MSWD is the mean square weighted deviation, which expresses the goodness of fit of the isochron; isochron and weighted mean ages are based on fraction of 39Ar and steps listed in the last two columns. WM(P)A is the weighted mean (plateau) age. Abbreviations: sg, single-grains (laser total-fusion ages); Phe, phengite; WR, whole rock; WM, white mica; Hbl, amphibole; Kfs, potassium feldspar; Bt, biotite; Chl, chlorite; low-T, low-temperature steps; medium-T, intermediate-temperature steps; high-T, high-temperature steps; exp., heating or laser degassing experiments on different fractions of minerals from the same sample; n.a., not applicable; n.d., not determined. Errors are 2s. Age interpretation – Rabinal complex: G19s fine-grained phengite: unified age for two grain-size fractions: 67 + 1 Ma; G27s biotite: WMA calculated for 40Ar/36Ar ¼ 350; 37 + 5 Ma. Central Chuacu´s complex: 5CO-4b biotite: 64.5 + 1.0 Ma; 5CO-4b white mica: 67.5 + 2.0 Ma. G10s white mica: 66 + 1 Ma; G12s white mica: 70 + 3 Ma; 5CO-5a white mica: hump-shaped spectrum indicates 40Ar excess, 219–243 Ma; low-T plateau at 225 + 5 Ma; 5CO-5a hornblende: loss profile from c. 720 to c. 153 Ma; 5PB-9b sericitic white mica: 52–62 Ma hump-shaped spectrum indicates 40Ar excess; 5PB-13b hornblende: mixing of different age components; minimal age of old component 137 Ma, maximal age of young component 20 Ma; major age component at 65 + 6 Ma; 5PB-14 white mica: loss profile from c.65toc. 50 Ma; mid and high-T ‘plateau’ at 65.2 + 0.8 Ma; 854F biotite: age range from 148 to 94 Ma; 880C white mica: age range from 127 to 100 Ma; likely Cretaceous metamorphic overprint on Early Mesozoic or older metamorphic rock; 883B white mica: 75 + 2 Ma; 883B biotite: 68.6 + 2.0 Ma; 887C white mica: 70.0 + 1.5 Ma. Metasomatized southern Chuacu´s complex: G6s white mica: U-shaped spectrum indicates 40Ar excess; 60 + 5 Ma; G7s rhyolite gash filling: 67 + 5 Ma; 5C-36a white mica: 48 + 1 Ma. Sanarate complex: G3s white mica: loss profile from c. 195 to c. 137 Ma; mid-T ‘plateau’ at 155 + 10 Ma; G5s hornblende: loss profile from c. 157 to c. 45 Ma; high-T ‘plateau’ at 156 + 5 Ma, low-T ‘plateau’ at 60 + 20 Ma. Western Las Ovejas complex, Guatemala: 5C-2c hornblende: 20 + 2 Ma; 5C-5 biotite: loss profile from c.31toc. 26 Ma; mid-T ‘plateau’ at 29 + 2 Ma; 5C-23c biotite: 25.3 + 0.5 Ma; 5C-23d hornblende: 35.5 + 1.0 Ma; 5C-26c hornblende: loss profile from c. 64 to 17 Ma; low-T ‘plateau’ at 18 + 2 Ma, intermediate-T ‘plateau’ at 21 + 2 Ma and high-T ‘plateau’ at 36 + 3 Ma; intermediate and low-T ‘plateau’ and chlorite-biotite ages are identical within error at 20 + 2 Ma; 5C-26c chlorite-biotite mixture: loss profile from c.28toc. 3 Ma; mid-T ‘plateau’ at 20 + 2 Ma; 5C-26d potassium-feldspar: 30 + 1 Ma; 5C-33 biotite: loss profile from c.26toc. 15 Ma; mid-T ‘plateau’ at 23 + 2 Ma; 5C-33 potassium-feldspar: loss profile from c. 370 to c. 6 Ma, low-T 40Ar excess, fast cooling at 32+4 and 7 + 2 Ma; 5C-37b hornblende: high-T WMA calculated for 40Ar/36Ar ¼ 553; 39 + 3 Ma; low-T WMA calculated for 40Ar/36Ar ¼ 362; 18 + 3 Ma; 5C-37d biotite: loss profile from c.26toc. 9 Ma; mid-T ‘plateau’ at 19 + 2 Ma; 5PB-5a hornblende: 30 + 2 Ma; 5PB-5a biotite: 28.2 + 0.4 Ma. Eastern Las Ovejas complex, Honduras: 5H-3a biotite: single-grain laser-fusion age (10 grains); WMA calculated from 40Ar/36Ar ¼ 661; 22 + 4 Ma; 5H-4b biotite: single-grain laser-fusion age (4 grains) at 25 + 3 Ma; 5H-4c biotite: loss profile from c.28toc. 18 Ma; minor 40Ar excess; mid-T ‘plateau’ at 25 + 2 Ma; Integrated biotite age of stations 5H-3 and 5H-4 at 24 + 2 Ma; 5H-4b hornblende: 38 + 2 Ma; 5H-4c hornblende: 31 + 5 Ma; 5H-4d hornblende: 29 + 3 Ma; Integrated hornblende age of stations 5H-4 at 34 + 4 Ma. Granitoids: 5C-11 chloritized biotite: single-grain laser-fusion age (10 grains) at 90 + 10 Ma; 5C-21 potassium-feldspar: single-grain laser-fusion age (8 grains) at 20 + 1 Ma; 5H-5 K-feldspar: loss profile from c.37toc. 28 Ma, low-T ‘plateau’ indicates fast cooling at 28.3 + 0.5 Ma; 5H-13 biotite: single-grain laser-fusion age (6 grains) at 59 + 3 Ma. Southern Mexico: ML-18 biotite: loss profile from 163 to 88 Ma, high-T ‘plateau’ at 162 + 2 Ma, low-T ‘steps’ at 90 + 2 Ma; ML-37 white mica: loss profile from 316 to 238 Ma, high-T ‘plateau’ at 315 + 3 Ma, low-T ‘plateau’ at 247 + 3 Ma; MU-12 biotite: 29 + 1 Ma; MU-13 biotite: 35 + 1 Ma; ML-39 biotite: 15 + 2 Ma; PA3–4 hornblende: 8.5 + 1.5 Ma; PA3–4 biotite: 8.5 + 1.0 Ma; CB35 hornblende: 28.5 + 1.5 Ma; CB35 biotite: 8.2 + 0.5 Ma; CB31 hornblende: 15.5 + 1.5 Ma; CB34 hornblende: 11.2 Ma; PA28-1 hornblende: 8.1 + 0.5 Ma; PI28-1 biotite: 8.0 + 0.5 Ma. 235 236 L. RATSCHBACHER ET AL.

Fig. 5. New 40Ar/39Ar mineral spectra and Rb–Sr whole rock–feldspar–mica ‘isochrons’. See the Supplementary material for sample locations, age data and interpretations. Weighted mean ages (WMA) were calculated using black steps of the given temperature range; TFA, total fusion age; IA, isochron age. Uncertainties are +1s. N. AMERICA–CARIBBEAN PLATE BOUNDARY 237

Fig. 5. (Continued) ‘Atm.’ in the isochron diagrams is the 40Ar/36Ar ratio of the atmosphere (295.5). MSWD, mean square weighted deviation. Abbreviations: wr, whole rock; Pl, plagioclase; Kfs, K-feldspar; WM, white mica; Bt, biotite. 238 L. RATSCHBACHER ET AL.

Fig. 5. (Continued). N. AMERICA–CARIBBEAN PLATE BOUNDARY 239

Fig. 5. (Continued). 240 L. RATSCHBACHER ET AL.

Fig. 5. (Continued). N. AMERICA–CARIBBEAN PLATE BOUNDARY 241

Fig. 5. (Continued). 242 L. RATSCHBACHER ET AL.

Fig. 5. (Continued). magmatic Th–U ratios (mean 0.4) (Fig. 4). The Santa Rosa Group. Zircons are rounded but a sig- 207-corrected spot ages, mostly from rims but also nificant portion is sub- to euhedral. Besides ages from a few coreless grains with relatively flat CL .1 Ga, we obtained four concordant ages at response, range from 410 to 485 Ma (Fig. 3a). 416.5 +1.6, 282.2 +1.7, 260.4 + 2.9 and 245.1 + There may be two age groups with weighted 2.9 Ma. The Devonian spot has low Th–U (0.012), average ages of 464.6 + 5.3 and 412.9 + 3.7 Ma; the other zircons (Th–U on average 0.64) probably Th–U ratios are similar. We suggest that these crystallized from a magma (note difference in lumi- ages indicate prolonged Ordovician–Early Devo- nescence of 416 and 257 Ma grains in Fig. 4). Our nian magmatism (see below). ages imply post-Early Triassic deposition for these Sample GM1 (stop GM15, central Guatemala; metasediments. Figs 3a & 7a) is quartz phyllite close to Salama´ Samples GM13s (stop GM42) and GM14s (stop (central Guatemala); these rocks were mapped as GM43) (Figs 3a & 7a) are from western Guatemala. N. AMERICA–CARIBBEAN PLATE BOUNDARY 243

GM13s is (?ortho)gneiss and GM14s granite intrud- 74.24 + 0.52 Ma and five younger spots that may ing gneisses along the Polochic fault. Seven zircon constitute a cluster at 70.3 + 1.5 Ma; the latter fractions of GM13s with different magnetic proper- group may be geologically meaningless, if it reflects ties and, in part, grain size yielded lower and upper progressive Pb loss of grains from the former group. intercepts at 20 + 15 and 941 + 7.9 Ma, respect- Sample 5CO-6b is most straightforwardly inter- ively. We analysed six fractions of GM14s differing preted as a Triassic magmatite that inherited Proter- in grain size and magnetic properties. All are discor- ozoic and Devonian zircons and experienced high-T dant and yield lower and upper intercepts at 21 + 33 metamorphism during Late Cretaceous, probably and c. 894 Ma; five fractions yield a lower intercept coeval with pegmatite injection; similar very at 13 + 29 Ma. Phanerozoic spot ages of zircons coarse-grained pegmatites also yielded Late Cretac- from GM14s gave one concordia age at 425.5 + eous K–Ar (Ortega-Gutie´rrez et al. 2004) and 6.0 Ma and 14 spots with a 14.65 + 0.42 Ma Rb–Sr white mica cooling ages (see below). The weighted average age. These ages demonstrate regional significance of magmatism is supported Middle Miocene Central American arc magmatism. by Triassic hornblende and white-mica cooling ages from this region (238–225 Ma, Donnelly Chuacu´s complex. Sample 5CO-4b, high-P meta- et al. 1990; see below). As we interpret the ortho- morphic (see below) garnet orthogneiss from the gneiss as a sill, the high-P event is probably pre- northern Chuacu´s complex, central Guatemala Middle Triassic. (Fig. 7a), comprises complex, mostly Proterozoic zircons. Zircon grain shapes include both rounded Chortı´s Block. Sample 5H-3 is syntectonic, syn- and prismatic–idiomorphic. In general, low-U migmatitic, plagioclase-rich granite gneiss associ- cores are surrounded by variable, higher-U rims ated with migmatite, leucogranite and gabbro– that both have Proterozoic ages (1180–1017 Ma diorite; it is typical of the Las Ovejas complex and concordant spots; Fig. 3a). Spots that include the crops out at the western flank of the Sula graben in former grain portions and tiny, very luminescent northwestern Honduras (Fig. 7a). The 207-corrected low-U outer rims, irresolvable with the spot size spot ages on typically magmatic zircons (Figs 3c & we employed, yielded the youngest ages. The 4) range from 40.0 + 0.4 to 36.5 + 0.9 Ma with lower intercepts of discordia fitted to two groups three grains defining a concordia age at of grains indicate crystallization at 638–477 Ma 37.0 + 0.52 Ma. We interpret this age as the time (numbers include errors), predating the high-P of magmatism and deformation. Sample 5H-9a metamorphic event. Sample 5CO-6b (south-central comprises orthogneiss and migmatite south of the Chuacu´s complex; Fig. 7a), a two-mica orthogneiss, Jocota´n fault zone (Fig. 7a) that are distinctly is intruded by very coarse-grained pegmatite and different in grain-size (coarse) and deformation interlayered with paragneiss and garnet amphibo- (weak) from the gneisses and migmatites of the lite; the latter contains relict eclogite (see petrology Las Ovejas complex along the MSZ. Besides Proter- for analysis of equivalent eclogite sample 5CO-5a). ozoic ages, there are concordant spot ages at All rock types are ductilely deformed. The amphi- 400.3 + 3.5 and 328.9 + 4.0 Ma. There are two bolite forms boudins in the orthogneiss and is main age groups: 272.8 + 2.8 Ma (eight spots) more strongly deformed. We interpret the massive and 244.8 + 2.3 Ma (four spots). These groups orthogneiss as a sill and the eclogite-facies meta- have distinctly different Th–U ratios and morphism pre-dating its emplacement. The zircons CL-response (Figs 3c & 4). The younger group, of the orthogneiss are complex (Figs 3a & 4): we probably metamorphic, has mean Th–U at 0.04 detected a few well-rounded Proterozoic cores, and appears dark and unzoned in CL. The older one core that has flat CL response, low, likely group has mean Th–U of 0.7 and clear oscillatory metamorphic Th–U (0.020) and a 207-corrected zoning. We suggest Early Permian protolith crystal- age of 402.9 + 4.0 Ma, and many oscillatory zoned lization and Early Triassic metamorphism. Sample grains with magmatic Th–U (on average 0.82); 5H-11a is low-T, mylonitic orthogneiss (phyllonite), three of the latter spots give a concordia age of probably metarhyolite intercalated with paragneiss 238.0 + 2.5 Ma, and eight spots yield a well- and marble; the sample is also from the Jocota´n defined group with a weighted average of 218.2 + fault zone (Fig. 7a). Six out of seven 207-corrected 2.3 Ma. Almost all grains have mostly thin spot ages yield a 130.5 + 2.7 Ma concordia age (5–40 mm) overgrowths with flat CL response and (Fig. 3c); Th–U ratios are on average 0.41. We low Th–U (on average 0.01) that span 89 to 69 Ma. interpret this date as reflecting Early Cretaceous The oldest grain has the highest Th–U and probably magmatism. Sample 5PB-5c is a small orthogneiss samples different age domains (core and rim). pod in massive migmatite that is enclosed in large Selected omission of grains from the remaining granite in easternmost Guatemala (Fig. 7a); it is group yields one statistically significant group of associated with garnet amphibolite (see petrology). six grains with a weighted average age of Three spots define a 36.2 + 2.1 Ma concordia age 244 L. RATSCHBACHER ET AL.

Fig. 6. Low-temperature thermochronology data. (a) Apatite fission-track (AFT) age v. elevation plots for the central Chuacu´s complex (top), the Las Ovejas complex, and undeformed arc granitoids. Locations and fission-track parameters are given in the Supplementary material. (b) AFT temperature–time (T–t) paths for rocks of the Rabinal (western Guatemala), Chuacu´s (central Guatemala) and Las Ovejas (western area) complexes, and three arc granitoids; the latter are: 5C-6, Chiquimula batholith in the Las Ovejas complex; 5C-32, small granite east of the Chiquimula batholith; GM15s, augengneiss of the westernmost Rabinal complex overprinted by c. 15 Ma granite. Good-fit solutions N. AMERICA–CARIBBEAN PLATE BOUNDARY 245

(Fig. 3c), probably the age of migmatization. One of Chortı´s Block in central Honduras (Fig. 1c). SF1D the spots is from a relatively high-U, oscillatory is weakly deformed granite gneiss surrounded by zoned central grain portion that is surrounded by a Jurassic metasedimentary rocks of the Agua Frı´a rim poorer in U (Fig. 4); this rim is c.5Ma Formation; the contact relationships are unclear. (31.2 + 1.2 Ma) younger. The major group of All five fractions are discordant and lower intercept Phanerozoic ages spans 140–191 Ma with two, ages between 275 and 234 Ma can be defined probably significant, age groups: 158.8 + 0.52 Ma (Fig. 3c). 92QO404 is mega-crystic, K-feldspar (n ¼ 4) and 165.79 + 0.87 Ma (n ¼ 13); the con- metagranite surrounded by phyllite. Five out of six sistently lower Th–U ratios (on average 0.12) of fractions define intercepts at 396 + 11 and c. the former (the latter has on average 0.36) support 884 Ma (MSWD ¼ 0.33; Fig. 3c). 86TA100 is K- the grouping. There are other Phanerozoic con- feldspar augen schist with white mica, quartz, cordant spots at 139.8 + 1.0 Ma (n ¼ 2), epidote, microcline, biotite and garnet. Four out of 190.6 + 1.4 Ma (n ¼ 1) and 240.6 + 3.5 Ma five fractions define 404 + 13 and 1149 + 23 Ma (n ¼ 1). We suggest a Middle Jurassic granite proto- lower and upper intercept ages, respectively lith that inherited Proterozoic to Early Jurassic (MSWD ¼ 1.9; Fig. 3c). Besides Grenville inheri- zircons. Sample 92RN412 is granitic gneiss sur- tance, these samples indicate Early Devonian mag- rounded by amphibolite-facies rocks from the Las matism in the Chortı´s Block. Ovejas complex of northwestern Honduras (Fig. 7a); it is grossly the same rock as 5H-3. Rb– Ar/Ar and Rb–Sr geochronology Sr and K–Ar biotite cooling ages are c. 39 and c. 36 Ma, respectively (Horne et al. 1974, 1976a). We determined ages of deformation and meta- Three of four zircon fractions with different grain morphic cooling of the Maya and Chortı´s Block size and magnetic properties plot close to concordia; rocks by analysis of syn- to post-tectonic minerals. one fraction is concordant at 38.3 + 0.53 Ma. All We also dated cooling in pre- to post-tectonic fractions define intercepts at 36.6 + 3.1 and c. magmatic rocks, again providing age limits for 1014 Ma (Fig. 3c); three fractions are at deformation. 37.49 + 0.53 and c. 1136 Ma. The c. 38 Ma age is within error identical to sample 5H-3. The age con- Rabinal complex. Sample G19s (stop G1, Figs 5a & strains deformation within the eastern Las Ovejas 7a) is weakly migmatitic gneiss from western complex to be younger than 39 Ma (see below). Guatemala retrogressed to hornblende–epidote Sample 94SA602 is amphibolite within the meta- schist during localized but high-strain, low-T morphic sequence that contains the granite gneiss (400 8C) sinistral shear. Two sericite fractions of of sample 92RN412; the protolith was probably different grain size yielded nearly identical weighted gabbro. We obtained three zircon fractions that mean ages (WMA). The pooled 67 + 1 Ma age is plot close to and on concordia, defining a lower interpreted to slightly post-date low-T plasticity in intercept at 88 + 51 Ma (Fig. 3c); the concordant the shear zones. Sample G27s (stop G57, western fraction is at 88.27 + 0.56 Ma, probably dating Guatemala, Figs 5a & 7a) is orthogneiss associated gabbro crystallization. Sample 93LP522 is granite with amphibolite and shows low-T mylonitization directly south of the Jocota´n–Chameleco´n fault along discrete sinistral shear zones; undeformed, zone (Fig. 7a). Four zircon fractions define a lower probably Miocene (c. 15 Ma, see above) granite intercept at 167.6 + 2.6 Ma (MSWD ¼ 10.9), on occurs in the vicinity. The age spectrum is complex concordant fraction is at 167.99 + 0.46 Ma and the c. 37 Ma WMA may reflect partially reset (Fig. 3c); the c. 168 Ma age is within error identical Late Cretaceous–Early Cenozoic biotite. to sample 5PB-5c and demonstrates Middle Jurassic magmatism in the Chortı´s Block. Samples SF1D, Chuacu´s complex. Sample 5CO-4b is garnet gneiss 92QO404 and 86TA100 typify orthogneisses associated with K-feldspar–porphyroblastic gneiss embedded in the Cacaguapa schists of the central and schist from the north-central Chuacu´s complex

Fig. 6. (Continued) (all T–t paths with a merit function value of at least 0.5, Ketcham et al. 2000) are dark grey or black; acceptable-fit solutions (all T–t paths with a merit-function value of at least 0.05) are light grey. (c) AFT age v. elevation plot and elevation–age–distance from the Pacific-coast diagram for a vertical profile and regional samples of the south-central Chiapas Massif, southern Mexico. The Tonala shear zone, associated granitoids, metamorphic rocks and mylonites and their ages are indicated (Hbl, hornblende; Bt, biotite); note their influence on one AFT sample. (d) AFT and apatite (U–Th)/He data for two coast-normal sections in southern Mexico from Ducea et al. (2004b) plotted in age v. elevation diagrams; these data are reference sections for our own data and are discussed in the text. Dark grey vertical bars give age range (U–Pb zircon) and intrusion depth of granitoids from about the same area as the thermochronology data. Light grey bars give interpreted closure through 70–100 8Corc. 4 km depth. 246 L. RATSCHBACHER ET AL.

(Fig. 7a); metamorphism evolved through a high-P indicating 40Ar excess, the Rb–Sr white mica– stage and amphibolite– to epidote–amphibolite- whole rock isochron is at 76.5 + 0.59 Ma (Fig. 5b). facies retrogression (see below). Low-Si phengite Samples 5CO-5a,b, 5CO-6b, 880C, 883B, 885A and biotite are part of the recrystallized groundmass and 887C are from the central Chuacu´s complex (510 8C, c. 0.7 GPa). The phengite spectrum is and typical for the garnet amphibolite–micaschist– weakly hump-shaped but the total-fusion (TFA), biotite–(?ortho)gneiss–metapegmatite intercalati- WMA, and isochron (IA) ages correspond within ons of this region (Fig. 7a). The pegmatites are error at 67 + 2 Ma; the biotite is c. 3 Ma younger. variably deformed but massive stocks are weakly A Rb–Sr phengite–plagioclase–whole rock iso- strained. All rocks share various stages of post- chron is within error of the Ar/Ar ages (Fig. 5b). tectonic annealing. The Ar-release spectrum of Epidote–amphibolite-facies retrogression is thus phengite of garnet amphibolite 5CO-5a is hump- Late Cretaceous. Sample G10s (stop G29, Fig. 7a) shaped, indicating 40Ar excess; a low-T plateau is from the southern Chuacu´s complex and com- is at 225 + 5 Ma (Fig. 5b). Hornblende shows prises marble interlayered with garnet micaschist a classic loss-profile spectrum (c. 720–153 Ma). and gneiss just north of El Tambor complex serpen- The 70.0 + 1.5 Ma Ar/Ar white mica age from tinites. The Ar/Ar age of the coarse white mica in intercalated quartzofeldspathic paragneiss sample the marble overlaps a Rb–Sr biotite age (G11s; 887C expresses Late Cretaceous reheating, also Fig. 7a) of nearby two-mica orthogneiss within highlighted by our U–Pb zircon geochronology error at c. 67 Ma (Fig. 5b). Sample G12s (stop (see above). White mica of metapelite sample G31, Fig. 7a) is fine-grained, mylonitic orthogneiss 880C defines a shear zone related to late stage (c. 1 Ga, unpublished U–Pb zircon) north of G10s. folding (F2); the individual heating steps cover The grain size is due to dynamic recrystallization 127–100 Ma. White mica and biotite from para- of quartz at the transition between subgrain rotation gneiss sample 883B define a s-c fabric, charac- and grain-boundary migration (c. 425 8C, e.g. Stipp teristic of late-stage deformation; sinistral flow et al. 2002). White mica is related to the recrystalli- is constrained at 74 + 2 Ma (white mica) to zation fabric, includes albite and is cogenetic with a 68.6 + 2.0 Ma (biotite). Sample 854F is migmatitic second generation of garnet and chlorite; the ubiqui- paragneiss weakly overprinted by late-stage (retro- tous albite may be due to Na-rich fluid exchange grade) shear zones from the Rio Hondo area of the between the Chuacu´s complex and the El Tambor southeastern Chuacu´s complex; individual heating complex rocks in the hanging wall. Mylonitization steps cover 148–94 Ma, probably reflecting a and fluid exchange occurred at or slightly after mixture of age components. Rb–Sr metapegmatite 70 + 3 Ma (Ar/Ar white mica, Fig. 5b). Samples white mica ages are at 76.1 + 1.5 (5CO-5b), 5PB-13b and 5PB-14 typify rocks of the east-central 57+8 (5CO-6b) and 67 + 11 Ma (885A, extremely Chuacu´s complex (Fig. 7b): 5PB-13b is coarse- coarse grained mica from the Guatemarmol mine grained garnet amphibolite associated with garnet west of Granados), demonstrating Late Cretaceous micaschist metamorphosed at c. 540 8C and pegmatite emplacement; our ages confirm those c. 0.7 GPa (see below). 5PB-14 is hornblende– from foliated metapegmatite (K–Ar 73–62 Ma) of mica–chlorite schist with quartzitic layers formed Ortega-Gutie´rrez et al. (2004). We interpret our at identical PT conditions. Quartz recrystallized ages as Triassic cooling from granite emplacement dynamically by subgrain rotation and grain- and associated metamorphism and Late Cretaceous boundary migration. The hornblende spectrum of reheating associated with regional metamorphism 5PB-13b is complex with an age component older and local magmatism. The two samples that than 137 Ma; several steps are at 65 + 6 Ma. The suggest cooling in the Early Cretaceous character- latter are within error of the Ar/Ar (5PB-14) and istically are from areas with relative weak late-stage Rb–Sr (5PB-13b) white mica ages (Fig. 5b). As tectonic overprint. On one hand, they (880C) the closing temperatures for these systems are support a pre-Late Cretaceous metamorphism in close to the peak-T conditions represented by the the core of the central Chuacu´s complex, where synkinematic minerals, deformation is dated at sinistral flow is localized and latest Cretaceous 65 Ma. Sample 5PB-9b is retrograde biotite– (883B). On the other hand, they (854F) demonstrate chlorite gneiss from ortho–paragneiss intercala- pre-Cretaceous migmatization in the southern part tions at the eastern end of the Chuacu´s-complex of the eastern Chuacu´s. outcrop in the Sierra de Las Minas area (below the Samples G6s, G7s, 860B and 5C-36a (Figs 5c & Juan de Paz serpentinite unit; Fig. 7a). Mylonitiza- 7a) are Chuacu´s-complex rocks immediately below tion is upper greenschist facies with basal ,a. the North El Tambor complex rocks. These slip and grain-boundary migration recrystallization localities show imbrication of basement (now retro- in quartz and coexisting fracturing and incipient grade phyllonite, chlorite phyllite, epidotized biotite recrystallization in feldspar. The Ar/Ar white mica gneiss, carbonate-bearing garnet–actinolite mica- spectrum is classic hump-shaped (52–62 Ma) schist), marble (San Lorenzo marble), serpentinite Fig. 7. (a) Major geological units of the Motagua suture zone, distribution of Palaeozoic–Cenozoic igneous rocks and new geochronological ages. (b) Definition of mineral cooling and (c) zircon crystallization-age groups in relative probability plots; (d) temperature–time diagrams of all data and (e) selected localities/samples (closing temperatures see text). N. AMERICA–CARIBBEAN PLATE BOUNDARY 247 and low-grade metasiltstones (Jones Formation). (c. 35 Ma); it remains unclear whether the late They share a history of extreme fluid flow that cooling of these samples through biotite closure resulted in strong retrogression, local carbonate (c. 23 Ma) may indicate long-lasting deformation, and albite ‘metasomatism’ and post-tectonic reactivation, or slow cooling. Samples 5C-26c,d graben formation, and locally intense quartz comprise garnet amphibolite and garnet-bearing veining. White mica of G6s (station G26), retro- biotite gneiss of the lithological unit described gressed basement, has a U-shaped spectrum indicat- above; distinct rocks are migmatitic gneiss and ing 40Ar excess with a flat-portion at 60 + 5Ma garnet-bearing leucosome. 5C-26c hornblende is (Fig. 5c), comprising most of the released Ar. Retro- aligned in the foliation (low strain) and exhibits grade biotite gneiss G7s (station G27) exhibits a loss profile from c. 64 to 17 Ma. Multiple-step centimetre-sized pull-apart-type tension gashes degassing allowed deconvolution into sub-plateaus locally filled with pseudotachylite that spectacularly that were analysed with isochrons (Fig. 5d). dates brittle–ductile, sinistral transtension at c. A low-T ‘plateau’ at 18 + 2 Ma has approxi- 70 Ma (Fig. 5c). 5C-36a shows coarse white mica mately atmospheric 40Ar/36Ar, an intermediate-T from relict gneiss in overall fine-grained, locally ‘plateau’ at 21 + 2 Ma is non-atmospheric but exhi- biotite-rich phyllonite; its 48 + 1MaAr/Ar and bits a good-fit isochron and a high-T ‘plateau’ at the 35 + 14 Ma Rb–Sr biotite ages are the youngest 36 + 3 Ma is only weakly non-atmospheric. The ages encountered in the Chuacu´s complex rocks and intermediate and low-T ‘plateaus’ are within error may reflect long-lasting fluid activity. of the 20 + 2 Ma WMA for the mid-T steps of a biotite–minor chlorite mixture; this biotite has a Chortı´s Block. The following samples characterize loss profile from c.28toc. 3 Ma. The high-T steps metamorphism and deformation in the Las Ovejas of K-feldspar give 30 + 1 Ma. Sample 5C-37b, complex and cooling of associated granitoids again typical Las Ovejas complex of eastern Guate- (Figs 5d & 7a). Sample 5C-2c is partly mylonitic mala, is amphibolite associated with migmatitic amphibolite intercalated with paragneiss and quart- biotite gneiss and metre-sized bodies of leuco- zite and crosscut by diabase dykes. Hornblende is granite. The hornblende spectrum is complex but clearly synkinematic to sinistral shear deformation contains a high-T ‘plateau’ with a WMA of and shows two high-T steps at .30 Ma and a 39 + 3 Ma, calculated for 40Ar/36Ar of 553, and a low-T ‘plateau’ at 20 + 2 Ma; we attribute limited low-T ‘plateau’ with a WMA of 18 + 3 Ma, calcu- significance to these ages, as only few steps are lated for 40Ar/36Ar of 362; the K–Ca ratios of the available due to a furnace problem. Sample 5C-5 latter suggests a contribution of gas released from is phyllite faulted against Cenozoic–Quaternary biotite micro-inclusions. 5C-37d biotite reveals volcanic rocks along the Jocota´n fault zone. Synki- Ar loss c. 26–9 Ma and a mid-T ‘plateau’ at nematic biotite shows Ar-loss from c.31toc.26Ma 19 + 2 Ma. The leucogranite gives a white mica– with a mid-T ‘plateau’ at 29 + 2 Ma (Fig. 5d). This K-feldspar–plagioclase–whole rock Rb–Sr iso- age indicates Early Oligocene Jocota´n fault activity, chron of 33.6 + 1.6 Ma. Located in the far-field of coeval with mylonitization in the Las Ovejas unit the large granitoids that were mapped crosscutting (see below). Samples 5C-23c,d comprise biotite the Las Ovejas complex, stations 5C-26 and 5C-37 gneiss and amphibolite in a unit of partly migmatitic typify cooling from amphibolite-facies metamorph- gneiss, metagabbro and granitic gneiss that is ism at c. 37 Ma and, possibly, relatively slow typical Las Ovejas complex lacking retrogression. cooling throughout the Oligocene–Miocene. The Synkinematic, fibrous hornblende gives a WMA of Rb–Sr leucogranite white mica age supports the 35.5 + 1.0 Ma; biotite is at 23.3 + 0.5 Ma and field observation that these small melt bodies are thus .10 Ma younger. Biotite of nearby 5C-33 cogenetic with Late Eocene metamorphism and gneiss depicts a loss profile from c.26toc.15Ma migmatization. Sample 5C-28b is pegmatite gneiss with a mid-T ‘plateau’ at 23 + 2 Ma, identical to crosscutting typical Las Ovejas biotite gneiss; its biotite of 5C-23c. K-feldspar shows a loss profile Rb–Sr whole rock–plagioclase–K-feldspar iso- from c. 370 to c. 6 Ma; the low-T steps, forming a chron is at 34.2 + 1.0 Ma and a whole-rock– good-fit IA with atmospheric 40Ar/36Ar at plagioclase–white mica isochron at c. 21 Ma. 7 + 2 Ma, are difficult to interpret geologically. Sample 5PB-5a is weakly deformed garnet amphi- The high-T steps indicate cooling at 32 + 4Ma bolite associated with massive migmatite and (Fig. 5d). Although from the vicinity of large post- small orthogneiss bodies and represents a huge tectonic granite, we interpret our 5C-23 and 5C-33 xenolith in granite. Peak-PT estimates using core- station ages to date amphibolite-facies metamorph- sections of minerals such as garnet that lack defor- ism and associated migmatization, the minerals lack mation are c. 630 8C at 0.7 GPa (see petrology); post-tectonic annealing and are in part synkinematic the analysed hornblende (WMA of 30 + 2 Ma) to sinistral strike–slip deformation. The amphibole and biotite (WMA at 28.2 + 0.4 Ma) are synkine- ages closely approximate the age of deformation matic matrix minerals that date deformation and 248 L. RATSCHBACHER ET AL. rapid cooling during weak, still amphibolite-facies event or Middle Jurassic reheating not reflected in retrograde overprint, also detected in the garnet the zircon geochronology in this sample but rims; peak-T conditions are at c.36Ma(U–Pb detected in the Las Ovejas complex of Guatemala zircon, see above). and in the central Chortı´s Block of Honduras. The next sample group comprises Las Ovejas We prefer the latter interpretation, as we consider complex of northern Honduras and structurally the extremely slow cooling unlikely in the active tec- footwall block of the Sula graben (Figs 5e & 7a); tonic setting of Early Mesozoic northern Central our dating aimed establishing potential along-strike America. changes in deformation/metamorphism/magma- tism and Sula-graben formation. Sample 5H-2a is Sanarate complex. Samples G3s and G5s (stops G23 post-tectonic, weakly deformed garnet–white mica- and G24; Figs 5f & 7a) are phyllite and quartz- bearing pegmatite in migmatitic biotite gneiss and bearing amphibolite from the Sanarate complex, rare leucogranite sills, and 5H-2c is a more strongly respectively. Quartz shows low-T plasticity with deformed, finer-grained variety of the pegmatite grain-growth inhibition and some post-tectonic (augen gneiss); these magmatites are outcrop-scale annealing; together with hornblende and white intrusives. The coarse white mica of 5H-2a is at mica it is synkinematic in discrete mylonite zones. 23.68 + 0.27 Ma (Rb–Sr whole rock–plagio- The white mica spectrum of G3s shows Ar-loss clase–white mica; the biotite isochron is c.7Ma from c. 195 to c. 137 Ma with a mid-T ‘plateau’ at younger) and provides a younger bound for the 155 + 10 Ma, and G5s hornblende from c. 157 to intrusion of the pegmatite protolith. The white c. 40 Ma; a high-T ‘plateau’ is at 156 + 5 Ma and mica–K-feldspar–plagioclase–whole rock iso- low-T steps at 40 Ma. We interpret these data as chron of 5H-2c is c. 2 Ma younger (Fig. 5e), prob- indicating Middle–Late Jurassic deformation and ably reflecting the higher strain. Sample 5H-3a is Cenozoic reactivation. small, weakly deformed but syntectonic granite in strongly migmatitic biotite gneiss; the 22 + 4Ma Granitoids of the Chortı´s Block. We dated cooling isochron age of 10 biotite crystals analysed by in a few of the mostly undeformed granitoids of single-grain laser fusion is .10 Ma younger than the northern Chortı´s Block (Figs 5g & 7a). Sample the intrusion age of the granite (c. 37 Ma, U–Pb 5C-11, granodiorite of the Chiquimula batholith, zircon, see above). Nearby station 5H-4 typifies shows biotite partly altered to chlorite; 10 grains the heterogeneous lithology of the Las Ovejas dated by laser fusion give a 90 + 10 Ma WMA. complex in this area. Sample 5H-4b is biotite and Eight out of 12 K-feldspar grains of sample hornblende-bearing pegmatite at the rim of a 5C-21, sub-volcanic granite also from the Chiqui- small, weakly deformed leucogranite body; all mag- mula batholith, have a 20 + 1 Ma laser-fusion matites at this locality are deformed and in several WMA; this demonstrates a composite nature of cases the strain is concentrated in leucocratic this batholith. K-feldspar of sample 5H-5, granite dikes. The 38 + 2 Ma hornblende is identical to from the rift flank of the Sula graben, shows Ar the age of the concordant zircon fraction of nearby loss from c.37toc. 28 Ma; the low-T ‘plateau’ at granite gneiss 92RN412 (see above); biotite is 28.3 + 0.5 Ma indicates rapid cooling (Fig. 5g). In .10 Ma younger at 25 + 3 Ma (laser single-grain line with the synkinematic hornblende and biotite ages). Sample 5H-4c shows synkinematic horn- ages reported from this area (see above), this age blende and biotite from a sinistral transtensive, suggests initiation of formation of the Sula graben greenschist-facies shear zone in amphibolite: horn- at 28 + 3 Ma. Six biotite grains of sample 5H-13, blende is at 30 + 2 Ma, biotite show a 25 + 2Ma granite from south of the Las Ovejas unit, gave a mid-T ‘plateau’ (Fig. 5e); these ages best date for- laser-fusion age of 59 + 3 Ma. The lack of mation of the Sula graben (see also below). Large modern U–Pb geochronology still impedes clear hornblende crystals of sample 5H-4d, the weakly age assignments of plutonic events in the Chortı´s deformed host amphibolite to the above shear Block. Because of the often sub-volcanic nature of zone, contain small hornblende crystals and are these granitoids, we assume that the available geo- within error (29 + 3 Ma) of the synkinematic chronological data reflect, to first-order, these sample 5H-4c amphiboles. Sample 5H-9a, ortho- events. Figure 5h pools all granitoid data and gneiss and migmatite that yielded Permo-Triassic gives first-order groupings: there is significant mag- U–Pb zircon ages (see above) and are distinctly matism at c. 36 Ma starting 40 Ma, at c. 59 Ma, different from the Las Ovejas complex rocks, and less significant at c. 81, c. 120, c. 161 and gives Rb–Sr whole rock–plagioclase–white mica c. 260 Ma. and biotite isochrons of c. 162 and c. 135 Ma, res- pectively. These dates have two possible inter- Southern Mexico. For correlative purposes (see pretations: either they reflect very slow cooling discussion), we analysed samples from southern from the Permo-Triassic metamorphic/magmatic Mexico with the Ar/Ar method (Figs 5i & 1c). N. AMERICA–CARIBBEAN PLATE BOUNDARY 249

Sample ML-18 is Magdalena migmatite of the respectively; the former and latter are similar to eastern Acatla´n complex (Fig. 2g). Biotite shows 14.65 + 0.42 granite GM14c along the Polochic Ar loss from c. 163 to c. 88 Ma with a high-T fault zone of westernmost Guatemala (see above) ‘plateau’ at 162 + 2 Ma; the latter traces cooling and a 10.3 + 0.3 Ma, pervasively sheared pluton from migmatization and San Miguel dyke intrusion (U–Pb zircon, Wawrzyniec et al. 2005) of the (175–171 Ma U–Pb ages; e.g. Keppie et al. 2004; Tonala shear zone, respectively. Hornblende and Talavera-Mendoza et al. 2005). White mica from biotite from three mylonitic gneisses (PA3-4, sample ML-37, mylonitic gneiss of the southern- CA35, PI28-1) gave within error identical ages most Oaxaca complex (Fig. 1c), shows Ar loss at c. 8.2 Ma, interpreted to date sinistral shear from a high-T ‘plateau’ at 315 + 3 Ma to a low-T deformation. One hornblende age (CB35) at ‘plateau’ at 247 + 3 Ma; the latter corresponds to 28.5 + 1.5 Ma is speculatively interpreted as ages of Permo-Triassic arc magmatism/meta- approximating the age of the higher-T flow/meta- morphism, exemplified by the Chiapas Massif of morphism, in line with Oligocene deformation southeastern Mexico (e.g. Weber et al. 2006), the ages along the northern margin of the Xolapa former corresponds to a widespread tectonothermal complex (see above). event in the Acatla´n complex of southern Mexico (e.g. Vega-Granillo et al. 2007). Biotite from Fission-track thermochronology sample MU-12, migmatitic gneiss at Cruz Grande (Xolapa complex, Fig. 1c) yielded 29 + 1Ma Our TFT and AFT thermochronology is only recon- (WMA), in line with cooling from widespread c. naissance in nature (Figs 6 & 7a). The two TFT ages 35–30 Ma Xolapa plutonism (Herrmann et al. support the available Ar/Ar and Rb–Sr geochronol- 1994; Ducea et al. 2004a). Sample MU-13 is Las ogy, being close to biotite cooling ages. The AFT Pin˜as mylonitic orthogneiss, sampled east of La ages from Guatemala and Honduras show positive Palma, which forms the structural top of the age v. elevation correlations both in the central Xolapa complex. Solari et al. (2007) reported a Chuacu´s complex and the central and eastern parts 54.2 + 5.8 Ma intrusion age (U–Pb zircon), and of the northern Chortı´s Block (Fig. 6a); exhumation 50.5 + 1.2 and 45.3 + 1.9 Ma cooling ages (K– rates, if significant, are similar and slow for the areas Ar and Rb–Sr biotite, respectively) from a moder- north and south of the Motagua fault zone (c. 0.038 ately deformed variety. Our sample was specifically and 0.035 mm/annum, respectively), and possibly taken to date low-T, high-strain mylonitic flow (c. slightly more rapid along the western Sula-graben 300 8C) along the Tierra Colorada (Riller et al. flank of northern Honduras (c. 0.046 mm/annum). 1992) or La Venta (Solari et al. 2007) top-to-NW, The Chuacu´s-complex ages depict a cluster at sinistral-transtensional shear zone; synkinematic c. 30 Ma, the low elevation ages from the Las biotite is at 35 + 1 Ma. Sample ML-39, mylonitic Ovejas complex at c. 12 Ma (Fig. 6a). Despite gneiss at Pochutla, is from a narrow, late-stage high elevation, the youngest ages north of the shear zone in the regional Chacalapa shear zone Motagua fault zone occur in western Guatemala along the northern edge of the eastern Xolapa along the Polochic fault in a small, c.15Ma complex; its major sinistral strike–slip stage is con- granite and the mylonitic augengneiss it intrudes strained at 29–24 Ma (Tolson 2005; Nieto- (samples GM13s, GM14s, Fig. 7a). The c.2Ma Samaniego et al. 2006). Our 15 + 2 Ma biotite age difference between crystallization and cooling age suggests reactivation, also indicated by faulting through c. 100 8C of the granite reflects quenching in the 29 + 1 Ma, post-mylonitic Pochutla granite at high crustal levels. The c. 11.5 Ma TFT and the (Herrmann et al. 1994; Meschede et al. 1996). c. 4.8 Ma AFT ages from low-T mylonite approxi- Samples PA3-4, PI28-1, CB31, CB35 and CA34 mate the age of ductile–brittle to brittle Polochic are from the Chiapas Massif of southeastern Mexico faulting. The outcrop conditions prevent assessment (Figs 1c & 5i). A belt of small plutons along the of whether the apparently undeformed granite is southwestern margin of the massif is associated locally affected by mylonitization but both mylonite with metamorphic rocks that are partly mylonitized and granite show east-trending, sinistral faults. The and faulted along a NW-trending, .100 km long oldest ages at high elevation in the Chortı´s Block zone, named Tonala shear zone by Wawrzyniec cluster at c. 22.5 Ma. The trend in the northern et al. (2005). The late ductile and brittle kinematics Chortı´s Block for AFT ages to become younger is sinistral strike–slip, overprinting an older, toward major fault zones (Motagua fault, Sula higher-T fabric; the zone probably connects to graben), with ages as young as 7.8–3.8 Ma, similar sinistral strike–slip fabrics and plutons (Figs 6a & 7a), may be apparent, as these samples along the Polochic fault (Fig. 1c; see below). Our also occupy the lowest elevations and thus are cer- goal was to obtain ages for the intrusions and tainly also an expression of the age v. elevation sinistral shear. Hornblende of diorite CB31 and trend. The presence of these ages at low elevations tonalite CA34 yielded 15.5 + 1.5 and 11 + 2 Ma, along the Motagua fault suggests that its current 250 L. RATSCHBACHER ET AL. activity (e.g. M7.5 1976 Guatemala earthquake, is about coeval with pegmatite emplacement at incorporating strike–slip and normal components; c. 74 Ma (Rb–Sr and K–Ar on very coarse-grained Plafker 1976) extended at least into the latest white-mica). Local Early Cretaceous cooling ages Miocene–Pliocene. (two samples) are probably geologically meaning- The track-length distributions of our samples are less, possibly reflecting pre-Cretaceous events unimodal, narrow, and symmetric. Modelled T–t overprinted by Late Cretaceous deformation. Mig- paths give monotonous continuous-cooling type matization in the Rı´o Hondo area is pre-Cretaceous solutions (Fig. 6b). All T–t models (except the and it remains to be demonstrated that it is Triassic, c. 15 Ma granite from western Guatemala, see as in the Rabinal complex of western Guatemala; above) show a phase of rapid cooling in the last lithology, structure (see below) and our Ar/Ar few Ma that is an artefact caused by annealing at data suggest that this region may constitute a ambient temperatures acting over geological time. distinct unit. Low-T track-length reduction has been described In the Las Ovejas complex of the northern for fossil tracks in age standards (e.g. Donelick Chortı´s Block and in the central Chortı´s Block, we et al. 1990) and borehole samples (Jonckheere & resolved Early Permian (c. 264–283 Ma, main age Wagner 2000). This reduction is not incorporated group at 273 Ma) magmatism followed by Early into the annealing equations derived from labora- Triassic (c. 242–253 Ma, main age group at tory experiments on induced fission tracks, which 245 Ma) high-grade metamorphism (including mig- account for annealing processes that take place matite) that is also reflected by a K–Ar biotite within the partial annealing zone (Jonckheere cooling age (222 + 8 Ma; MMAJ 1980), and 2003a, b). Middle to Late Jurassic (c. 140–191 Ma, age Our samples from the Chiapas Massif, southeast- groups at 159 and 166 Ma) magmatism and sub- ern Mexico, comprise a 1.5 km elevation profile sequent cooling (Rb–Sr and Ar/Ar ages at c. (Fig. 6c) and two swaths separated by ,40 km 155 Ma). Overwhelmingly, the northern margin of with little elevation difference; no apparent major the Las Ovejas complex is dominated by Late fault separates these samples. The youngest Eocene high-grade metamorphism, migmatization, sample is at c. 15.7 Ma, close to the Tonala shear and plutonism at 40–35 Ma (age group at 37 Ma). zone and the c. 11 Ma tonalite (Fig. 6c); it is most Cooling ages (Rb–Sr, Ar/Ar) related to this per- likely thermally influenced, supporting the vasive event cover c. 35–20 Ma and also come Miocene magmatism and deformation along this from Roata´n Island (c. 36 Ma). Our geochronology shear zone. The remaining samples range from also supports Silurian (c. 400 Ma) and Jurassic c.25toc. 39 Ma without a clear age v. elevation (c. 167 Ma) magmatism, and likely a Permo-Triassic trend and a weighted average age of 30.4 + event both in the northern and central Chortı´s Block. 2.5 Ma; we suggest that these samples indicate rela- Based on U–Pb zircon geochronology, the Maya tively rapid cooling at c. 30 Ma. and Chortı´s Blocks share Ordovician to Jurassic events (Fig. 7c; the evident Proterozoic similarity Interpretation of new geochronology is not addressed here). The northern and central Chortı´s Blocks are distinctly different from the Our U–Pb zircon geochronology outlines Early Rabinal and Chuacu´s complexes in the Late Cretac- Palaeozoic to Late Cenozoic high-T metamorphism, eous and Early Cenozoic: Late Cretaceous high- migmatization, and magmatism (Figs 3 & 7c). In the grade metamorphism only occurs in the Chuacu´s Rabinal complex of the Maya Block, we resolved complex and Eocene magmatism and metamorph- prolonged Ordovician–Silurian (c. 410–485 Ma, ism only can be found in the northern Chortı´s Block. age groups at 413 and 465 Ma), Permo-Triassic The most striking, first-order result showcased (c. 215–270 Ma, main age group at 218 Ma), and by our Ar/Ar and Rb–Sr geochronology is the Middle Jurassic (c. 163–177 Ma, main age group clear-cut separation of the Rabinal and Chuacu´s at 171 Ma) events (Fig. 3a); the Triassic event complexes from northern Chortı´s Block reflected includes migmatites. Miocene (c. 15 Ma) granite in all cooling ages (Fig. 7a & b): the Rabinal and intruded into shallow crustal levels (nearly identical Chuacu´s complexes, rimmed by the Polochic and U–Pb zircon and AFT ages) in western Guatemala. the Motagua fault zones, cooled through c. 500 8C Chuacu´s-complex rocks inherited Devonian zircons and through 275–350 8C c. 40 Ma earlier than the (c. 403 Ma) and experienced Triassic (c. 239– Las Ovejas complex south of the Motagua fault 211 Ma, age groups at 238 and 218 Ma) magmatism zone; cooling through c. 100 8C occurred c.15Ma and Cretaceous high-grade metamorphism (c.70– earlier in the north than the south. Obviously, the 78 Ma, main age group at 74 Ma; Th–U , 0.03). rock units north and south of the Motagua fault These events are partly supported by Ar/Ar and zone experienced distinctly different thermal Rb–Sr cooling ages (c. 238–212; c. 161; c.75– evolutions during the Late Cretaceous and Ceno- 65 Ma). The Late Cretaceous metamorphic event zoic: the Chuacu´s complex was heated to high N. AMERICA–CARIBBEAN PLATE BOUNDARY 251 amphibolite-facies metamorphism (see below) and suggests that the change from relatively rapid to cooled relatively rapidly (c.288C/Ma) to upper relatively slow rates occurred relatively early in crustal temperatures during the Late Cretaceous the exhumation process (c. 60 Ma in the Chuacu´s (Fig. 7d); subsequently, cooling (c.68C/Ma) and complex; c. 20 Ma in the Las Ovejas complex). exhumation was apparently slow (c. 0.038 mm/ The fact that the late-stage exhumation rates (AFT annum; see Figs 6a & 7d). Multi-mineral T–t age v. elevation data) are grossly similar in the paths from central and east-central Chuacu´s sam- Chuacu´s and Las Ovejas complexes suggests, ples (Fig. 7e) support this two-stage history. The given the strongly different cooling rates, much Las Ovejas complex lacks Cretaceous regional hotter crust in the Las Ovejas complex. metamorphism; its cooling from 40 Ma appears The difference between the southern Maya (here steady-state at c.168C/Ma (Fig. 7d & e). Our geo- including the Chuacu´s complex) and northern chronology does not prove whether slow cooling Chortı´s Blocks during the Cretaceous–Cenozoic is actually occurred, thus potentially reflecting long- emphasized by the distinctly different P–T–d–t term wrenching, or represents deformation and conditions encountered in the North and South El cooling at 40–35 Ma and reactivation at 25– Tambor complex allochtons (Harlow et al. 2004; 18 Ma. As we found clear synkinematic mineral Tsujimori et al. 2004): extremely low-T lawsonite growth defining both age groups (see below), we eclogites experienced blueschist-facies retrogres- favour the latter scenario. Along the southern sion at c. 120 Ma south of the Motagua fault, margin of the Chiapas Massif, U–Pb and Ar/Ar whereas higher-T zoisite eclogites show blueschist- geochronology demonstrates Cenozoic plutonism facies retrogression at c. 76 Ma north of the fault. (c. 29, c. 16, c. 11 Ma) and sinistral strike–slip at These P–T–d–t conditions illustrate differences c. 8 Ma along the Tonala shear zone. in tectonic evolution and time of the emplacement Most AFT ages within the Rabinal and Chuacu´s of these subduction–accretion complexes. Signifi- complexes group around c. 30 Ma and thus corre- cantly however, in the high-P/low-T case, the spond to the c. 30 Ma age cluster in the Chiapas region south of the Motagua fault zone cooled Massif. We suggest that these ages indicate a earlier and cooling in the North El Tambor thermal disturbance at 30 Ma induced by Ceno- complex is coeval with the onset of cooling in the zoic arc magmatism and sinistral displacement Chuacu´s complex. between the Maya and Chortı´s Blocks along and across this arc (see below). The young TFT New petrology and AFT ages along the Polochic fault zone relate its sinistral strike–slip deformation to the Late We selected 12 representative thin sections of mag- Miocene–Recent deformation along the Tonala matic and metamorphic rocks from both sides of the shear zone, thus supporting their contemporaneous MSZ for electron microprobe analysis. Mineral activity (Fig. 1c). These ages also support Burkart’s abbrevations taken from Kretz (1983). Mineral (1983) estimate of major displacement (c. 130 km) assemblages and summary of the P–T–t estimates along the Polochic fault between 10 and 3 Ma. can be found in the Supplementry materials, and The presence of very young AFT ages along the Figure8 showsthePT space ofourandliterature data. Motagua fault zone extends its neotectonic activity into the latest Miocene–Pliocene. The cluster of Rabinal complex, western Guatemala AFT ages at c. 22.5 Ma in the northern Chortı´s Block probably reflects continuation of the cooling Strongly foliated hornblende–epidote schist G19s from Late Eocene–Early Oligocene magmatism contains phengite (Si-content c. 3.3 p.f.u.), green and metamorphism; the c. 12 Ma cluster is attribu- amphibole (edenite to ferro-edenite associated ted to a thermal event that is connected with the with magnesio- to ferro-hornblende), epidote flare-up of magmatism (ignimbrite event) along (XPs 0.15; cf. Dahl & Friberg 1980), and saus- Central America (see below). suritized plagioclase with an An-content of Taking the peak-P conditions of c. 0.7 GPa (see 0–2 mole%. Biotite forms post-tectonic, topotactic petrology below) reached in the Chuacu´s and Las flakes on phengite. Accessory minerals are opaques Ovejas complexes during their Cretaceous and titanite. Phengite geobarometry yielded (Chuacu´s complex, c. 75 Ma) and Cenozoic (Las .0.7 GPa at 500 8C (Massonne & Szpurka 1997). Ovejas complex, c. 40 Ma) metamorphism, it The calibrations of Spear (1981) and Plyusnina appears that (i) early average exhumation rates (1982) gave 0.7–0.8 GPa at 450–500 8C, which were at least one order of magnitude more rapid are reproduced by Colombi’s (1988) Ti-in- than those derived from the AFT age v. elevation hornblende geothermometer. Our 67 + 1Ma relationship (Fig. 6a); (ii) the presence of AFT Ar/Ar phengite age demonstrates latest Cretaceous ages at high elevations, which approach those of exhumation through the middle crust with an the Ar/Ar and Rb–Sr biotite geochronometers, average rate c. 0.37 mm/annum. Epidote schist 252 L. RATSCHBACHER ET AL.

Fig. 8. Pressure–Temperature–time diagrams: (a) Pressure–temperature–time space of our new petrology and geochronology, and literature data of Ortega-Gutie´rrez et al. (2004). A few Chuacu´s-complex samples plot in the eclogite field. They follow an isothermal decompression path to amphibolite facies. The eclogite-facies metamorphism is at best constrained to between c. 440 and 302 Ma (see text). Most of the complex passed through epidote–amphibolite facies metamorphism and subsequent retrogression to greenschist facies; these events are well-dated as Late Cretaceous (75 Ma). The Sanarate complex of the northwestern Chortı´s complex gives Jurassic (c. 155 Ma) epidote– amphibolite-facies metamorphism that evolved prograde out of greenschist facies. The Las Ovejas complex (Chortı´s Block) shows Cenozoic (40 Ma) epidote–amphibolite to amphibolite facies. (b) Motagua suture zone pressure– temperature–time evolution from this study and Tsujimori et al. (2004, 2005, 2006), Harlow (1994), Harlow et al. (2004) and Brueckner et al. (2005). The Late Cretaceous Chuacu´s complex and Mid-Cenozoic Las Ovejas complex pressure–temperature fields are from this study. The northern zoisite eclogites (on Maya Block) formed at c. 131 Ma and reached mid-crustal conditions at about the same time as the Chuacu´s rocks experienced reheating at epidote– amphibolite-facies conditions accompanied by local magmatism. We suggest that the two are linked to the same oceanic-to-continental subduction zone (see text). The southern lawsonite eclogites (on Chortı´s Block) formed at c. 132 Ma and were at mid-crustal conditions already by c. 120 Ma. Their emplacement is probably not genetically linked to the high-grade metamorphism and magmatism of the Las Ovejas rocks, which is a result of the Caribbean Plate translation and the formation of the Cayman Trough (see text).

G21s is mylonitic with numerous porphyroclasts phengite (c. 3.3 Si p.f.u.) are common; opaques, of epidote (XPs core ¼ 0.295, XPs rim ¼ 0.215), titanite, apatite and zircon are accessories. albite (An0.3 – 0.9) and microcline (Kfs98Ab1–2 Strong mineral chemical disequilibria between Cel1–2An,0.5). The feldspars lack recrystallization the phases preclude PT estimates; the Si-content but quartz recrystallized dynamically by subgrain of phengite suggests c. 0.8 GPa (Massonne & rotation. Greenish biotite (XMg c. 0.47) and Szpurka 1997). Fig. 9. Overview map (top left) locates geological–structural map (bottom) in the Maya Block of western Guatemala. The latter compiles and reinterprets available 1 : 50 000 geological maps and plots the main structural features both on map and in stereonets (lower hemisphere, equal area); it also serves as legend for the following structural data maps. LPO (lattice preferred orientation) of quartz c-axis; D, deformation event; X, Y, Z, principal axes of finite strain. Fault-slip data and principal stress orientations 1–3: faults are drawn as great circles and striae are drawn as arrows pointing in the direction of displacement of the hanging wall. Confidence levels of slip-sense determination are expressed in the arrowhead style: solid, certain; open, reliable; half, unreliable; without head, poor. Arrows around the plots give calculated local orientation of subhorizontal principal compression and tension. N. AMERICA–CARIBBEAN PLATE BOUNDARY 253

Chuacu´s complex, central Guatemala (pargasite to edenite). Omphacite with 35 mole% jadeite at the rim to 40 mole% in the core is Garnet gneiss 5CO-4b is dominated by orthoclase present as relicts in garnet and hornblende cores (Kfs93Ab5Cel1An,0.5) and minor, partly saussuri- and less commonly in the matrix. Well-aligned tized plagioclase (An0.2 – 1.4). Quartz shows phengite is the dominant sheet silicate; Si-content subgrain-rotation recrystallization. Chloritized decreases from core (3.32 p.f.u.) to rim (3.27 biotite (XMg ¼ 0.17) and phengite have epidote p.f.u.). Garnet displays weak zoning with decreasing (XPs ¼ 0.23) and plagioclase inclusions. Phengite grossular and pyrope contents and increasing has elevated Si-contents of up to 3.50 p.f.u. in the almandine contents from core (Alm60.5GAU23- core, indicating high-P conditions, whereas Sps0.5Prp16) to rim (Alm63GAU20.5Sps0.5Prp15.5). Si-contents as low as 3.27 p.f.u. occur in the rim sec- Plagioclase (An2–4) is secondary, forming along tions. Garnet forms porphyroblasts with inclusions the rims of garnet and hornblende; irregular spots of plagioclase, quartz, minor ferro-hornblende/ of An10 – 14 are also present. Late-stage biotite ferro-tschermakite, omphacite (10 mole% jadeite) (XMg c. 0.53) grew at the expense of phengite. and rutile. It displays compositional zoning with a Apatite, rutile, ilmenite and pyrrhotite are acces- bell-shaped spessartine curve (core Alm43GAU44- sories. PT conditions of the eclogite facies event Sps10Prp0.5; rim Alm47GAU48Sps5Prp0.5). Carbon- were calculated applying the garnet–phengite– ate, zircon and titanite form accessory minerals in omphacite geobarometer of Waters & Martin the matrix. Relict phases highlight the polymeta- (1993) (activity models see sample 5CO-4b) in morphic evolution of 5CO-4b. Assuming equili- combination with the garnet–clinopyroxene brium between the omphacite inclusion in garnet, geothermometer of Krogh (2000). Using the the adjacent garnet core section and the core of method of Ryburn et al. (1976) to estimate Fe3þ the matrix phengite, PT conditions of c. 583 8C in clinopyroxene, PT conditions of c. 675 8Cat and c. 2.04 GPa are calculated using the garnet– c. 2.4 GPa were calculated using the core compo- 2 phengite–omphacite geobarometer of Waters & sition of garnet (highest aGrs *aPrp), the omphacite Martin (1993) (phengite activity model of Holland inclusion in garnet with the highest jadeite & Powell 1998), clinopyroxene activity model of content, and the phengite with the highest Si Holland (1980, 1990), and garnet activity model of content (Carswell et al. 2000). The application of Ganguly et al. (1996) in combination with the other methods to estimate the ferric iron content in garnet–clinopyroxene geothermometer of Krogh omphacite (e.g. Carswell et al. 2000; Schmid et al. (2000) (calculation of ferric iron in clinopyroxene 2000; Krogh 2000; Krogh & Terry 2004) results in according to Ryburn et al. 1976). Similar tempera- an uncertainty of +50 8C for the geothermometry. tures are calculated for high-P phengite and garnet Omphacite in the matrix probably was affected by core analyses using the thermometer of Green & retrograde processes and consequently shows a Hellman (1982). These PT conditions, however, wide range of temperatures (480–690 8C). Temp- only indicate a point on the prograde PT path as the eratures for retrograde overprint of the eclogite 2 highest pressure is indicated by the highest aGrs *aPrp were estimated at 630–690 8C using the garnet– (Schmid et al. 2000) at an intermediate position hornblende thermometer of Graham & Powell between garnet core and rim. The strong retrograde (1986) for garnet rims and adjacent hornblende. overprint is mirrored by pervasive growth of matrix The GRIPS geobarometer of Bohlen & Liotta biotite and low-Si phengite. Outermost garnet rims (1986) calculates pressures of c. 1.1 GPa for these (highest XPrp) and adjacent biotite yielded c. 510 8C minerals and associated plagioclase, rutile and (geothermometer of Hodges & Spear 1982) at ilmenite. Slightly higher pressures of 1.2–1.4 GPa c. 0.7 GPa (barometry of Massonne & Szpurka are calculated using the garnet–amphibole– 1997) using paragenetic phengite; this demonstrates plagioclase geobarometer of Kohn & Spear (1990). re-equilibration under epidote–amphibolite-facies These estimates indicate isothermal decompression conditions. Because of this overprint, feldspar ther- during exhumation of the eclogite. Ortega-Gutie´rrez mometry (Fuhrman & Lindsley 1988) on coexisting et al.’s (2004) c. 302 Ma lower intercept zircon age plagioclase (An0.2 – 1.4) and K-feldspar (integrated from a leucosome interbedded with relict eclogite analysis with a 50 mm beam diameter) indicates from the area of 5CO-5 was interpreted to provide cooling temperatures of 400–440 8C. The lower a minimum age for the Chuacu´s high-P event. intercept U–Pb zircon age indicates that high-P The strong foliation of garnet-bearing micaschist metamorphism is ,638–477 Ma and our Ar/Ar 5PB-13a is formed by alternating mica- and quartz- and Rb–Sr phengite and biotite ages date the latest rich layers; quartz shows subgrain-rotation and portion of the retrograde path at 68–64 Ma. grain-boundary-migration recrystallization. Small, Garnet amphibolite (eclogite) 5CO-5a contains almandine-rich and spessartine-poor garnet grains minor quartz showing subgrain-rotation recrystalli- (Alm70GAU10Sps1Prp18) are included in large zation and foliation-defining, large calcic amphibole albite porphyroblasts (An0.2 – 0.7), and larger garnet 254 L. RATSCHBACHER ET AL. porphyroblasts with Alm74GAU2Sps13Prp8 crystal- 500–550 8Catc. 0.8 GPa are suggested by the lized within the matrix; the garnets are slightly geothermobarometer of Plyusnina (1982). This esti- inhomogeneous without regular zonation patterns. mate is reproduced by the phengite geobarometry Green to brownish biotite (XMg c. 0.58) is partly (Massonne & Szpurka 1997) and the amphibole– altered to oxychlorite. Si-content in phengite plagioclase geothermometers of Spear (1981), decreases from core (3.3 p.f.u.) to rim (,3.2 Blundy & Holland (1990), and Holland & Blundy p.f.u.). This correlates with decreasing Ti-content (1994). A temperature of 600–630 8C is indicated and indicates a retrograde evolution. Because of by the Ti-in-hornblende geothermometer of the lack of critical mineral assemblages, only a Colombi (1988), probably due to the high bulk few thermobarometers can be used to evaluate the Ti-content of the rock. This may, however, also PT conditions of these schists. The Si-content of indicate an early stage of medium- to high-grade, phengite yields 0.6–0.8 GPa; the garnet–biotite amphibolite-facies metamorphism. Our 65 + 1Ma geothermometer (Hodges & Spear 1982) and the Ar/Ar white mica age dates cooling during garnet–phengite geothermometer (Green & greenschist-facies retrogression. Hellman 1982) point to 560–580 8C for the garnet Gneiss G14s is inhomogeneous with alternating inclusions in albite and 460–475 8C for the matrix plagioclase–quartz and mica–hornblende layers. garnets. Garnet amphibolite 5PB-13b is well foli- Plagioclase is albite (An0.5 – 2); epidote (XPs ¼ 0.2) ated, medium- to coarse-grained, and comprises and chlorite are present in the mafic layers. Phengite large, pale-green calcic amphibole (magnesio- has Si-contents of 3.27–3.34 p.f.u. Biotite yields hornblende to edenite) with numerous garnet and XMg of c. 0.51. Green amphibole is magnesio- and quartz inclusions; the matrix has also phengite and ferro-hornblende. Small garnet within the mafic clinozoisite. Phengite Si-content (3.3–3.15 p.f.u.) layers and as tiny inclusions in plagioclase has and XMg (0.76–0.66) decrease toward the rims. a bell-shaped spessartine curve and rimward Clinozoisite (XPs ¼ 0.08–0.09) forms large decreasing XFe and XMn; cores and rims are Alm40- prisms. Quartz shows both dynamic subgrain- GAU31Sps25Prp1 and Alm48GAU37Sps13Prp2, res- rotation and grain-boundary-migration recrystalli- pectively. Accessories are carbonate, opaques and zation. Albite (An2–6) has numerous quartz titanite. Triboulet’s (1992) geothermobarometer inclusions. Small, euhedral garnet yields Alm55- calculates c. 610 8Catc. 0.62 GPa for the rim com- GAU35Sps5Prp4 in the core and Alm54GAU30Sps5- position of the minerals, supported by phengite geo- Prp11 in the rim and has increasing XMg values barometry at 0.6–0.7 GPa at 500–600 8C. About from core to the rim, indicative of prograde 550 8C is obtained by the garnet–hornblende growth. One garnet differs compositionally, with (Graham & Powell 1986) and the garnet–chlorite high XMg in the core and abruptly increasing alman- geothermometer (Ghent et al. 1987). These PT dine content at the expense of grossular and spessar- estimates indicate peak conditions at the transition tine towards the rim; this may indicate an earlier, from epidote-amphibolite to amphibolite-facies prograde metamorphic evolution. Retrograde chlor- metamorphism. ite is widespread; apatite, zircon, and titanite are accessories. Pressures of 0.7 GPa (Massonne & Szpurka 1997) at 530–550 8C (garnet–hornblende Intrusive sequence, Juan de Paz ophiolite, geothermometer of Graham & Powell 1986) are North El Tambor complex obtained; the Ti-in-hornblende thermometer of Colombi (1988) provides c. 590 8C. Our Ar/Ar Fine- to medium-grained microdiorite 5PB-8 shows amphibole and Rb–Sr phengite ages suggest meta- agglomeration of hornblende and plagioclase. morphism at c. 65 Ma at this locality. Plagioclase shows normal zoning with An-content Hornblende–mica–chlorite schist 5PB-14 decreasing from 85 to 55 mol% from core to rim. shows distinct layering with amphibole–mica and Calcic amphibole (magnesiohornblende) is slightly quartz that is dynamically recrystallized by subgrain brownish and short prismatic. A few clinopyroxene rotation and grain-boundary migration. Pseudo- relicts yield 95–98 mol% augite (end member morphs of chlorite (ripidolite after Hey 1954) after calculation after Banno 1959) and are mantled relict garnet and lamellae of oligoclase (An12) by hornblende. Opaque minerals and apatite within albite (An0.5 – 2) demonstrate retrograde are common accessories. Anderson & Smith’s overprint. Green amphibole (tschermakite to mag- (1995) Al-in-hornblende geobarometer indicates nesiohastingsite) and clinozoisite (XPs c. 0.13) are c. 140 MPa at 730–860 8C, calculated with the subordinate. White mica shows increasing Ti-in-hornblende and the plagioclase–hornblende Si-content from the core to the rim (3.12–3.20 geothermometers. Bertrand et al. (1978) dated a p.f.u.). Pale brown biotite is rare. Carbonate is likely metamorphic overprint in related dolerites Fe-bearing dolomite. Opaque minerals, rutile and from the Baja Verapaz (Fig. 1c) at c. 75 Ma (see titanite form accessory phases. PT conditions of above). N. AMERICA–CARIBBEAN PLATE BOUNDARY 255

Sanarate complex Prismatic amphibole is magnesiohornblende (classi- fication of Leake et al. 1997) with relicts of augite in Chlorite–epidote–amphibole schist 5G-1a contains the cores. Reddish-brown biotite contains up to albite (An ), clinozoisite (X 0.1), chlorite 1 – 3.5 Ps 4.6 wt% TiO2, has an XMg of c. 0.45 and is partly (ripidolite), amphibole (edenite) and phengite transformed to chlorite. Accessories are opaque (3.2–3.3 Si p.f.u.). The application of the geother- minerals, apatite, and zircon. Anderson & Smith’s mobarometer of Triboulet (1992) yields peak PT (1995) Al-in-hornblende geobarometer yields conditions of c. 550 8Catc. 0.65 GPa for the rims 100–180 MPa at 740–860 8C, calculated with the of the minerals. The cores indicate c. 380 8Cat Ti-in-hornblende and plagioclase–hornblende c. 0.55 GPa, outlining a prograde metamorphic geothermometers (Colombi 1988; Blundy & loop. These data highlight the transition from Holland 1990; Holland & Blundy 1994). Our lower greenschist to epidote–amphibolite facies Ar/Ar geochronology demonstrates that the Chiqui- and are supported by the geothermometry of mula granite is a composite batholith (c.90Ma Colombi (1988) and Holland & Blundy (1994), biotite–chlorite age, 5C-11; c. 20 Ma K-feldspar, which provide c. 415 8C for the core and c. 560 8C 5C-13); assuming that the 20 Ma K-feldspar age for the rim of the amphibole. Si-contents of syntec- of 5C-11, adjacent to 5C-13, is close to the intrusion tonic phengite accordingly yield 0.6–0.65 GPa age of this shallow granodiorite (3–5.5 km), the (Massonne & Szpurka 1997). A likely age of this average exhumation rate is 0.15–0.27 mm/annum. metamorphism is Middle Jurassic, given by the This supports our inference (see above) that rela- hornblende and white mica ages from adjacent tively rapid exhumation in the Las Ovejas com- locations G23, G24, which comprise similar rocks. plex is pre-20 Ma.

Las Ovejas complex Interpretation of new petrology Garnet amphibolite 5PB-5a displays a fine- to Garnet amphibolite and felsic gneiss of the central medium-grained matrix with synkinematic biotite, Chuacu´s complex in the El Chol area preserve, in amphibole and large, poikiloblastic garnet. Matrix accordance with Ortega-Gutie´rrez et al. (2004), a quartz shows numerous subgrains. Biotite is Ti-rich high-P event. Garnet gneiss 5CO-4b probably (up to 4.1 wt% TiO2) and has an XMg of about 0.24. records part of a prograde path (c. 580 8Cat The green amphibole is hastingsite. Large plagio- c. 2 GPa) that culminated at c. 675 8C and clase grains are chemically homogeneous with c. 2.4 GPa (5CO-5a). Isothermal decompression An-content of c. 35 mol%. Garnet cores show to c. 660 8Catc. 1.3 GPa followed the near ultra- Alm69GAU19Sps8Prp4; XFe and XMn increase high pressure conditions and the retrograde path towards the rims, indicating retrograde overprint. can possibly be traced to c. 520 8Catc. 0.7 GPa. Accessories are opaques and apatite. Garnet– Only restricted areas of the Chuacu´s complex hornblende geothermometry (Graham & Powell escaped the regional overprint at 450–550 8C and 1986) in combination with garnet–amphibole– 0.7–0.8 GPa. The age of the high-P event is plagioclase–quartz geobarometry (Kohn & Spear pre-238–218 Ma, the age of two-mica orthogneiss 1990) yields 627–655 8C at 0.7–0.8 GPa. Com- intrusion (U–Pb ages) and high-T metamorphism parable temperatures are calculated with garnet– (Ar/Ar ages) in the El Chol region. The high-P biotite thermometers of Hodges & Spear (1982), event is probably post-638–477 Ma, the lower Indares & Martignole (1985) and Perchuk & Lavren- intercept U–Pb zircon age (including errors) of t’eva (1983). Ar/Ar hornblende and biotite ages out high-P garnet–K-feldspar gneiss at location of shear bands date deformation and rapid cooling 5CO-4. It is possibly older than the c. 302 Ma through the middle crust at 30–28 Ma. lower intercept U–Pb age of a leucosome inter- preted to be associated with eclogite decompression Chiquimula batholith within Las Ovejas (Ortega-Gutie´rrez et al. 2004); we have so far not complex been able to reproduce this age at the same locality using SHRIMP geochronology (unpublished Coarse-grained granodiorite 5C-13 is part of results). The high-P event should also be younger the Chiquimula batholith with the mineral assemb- than c. 440 Ma, the age of the youngest zircons in lage quartz þ K-feldspar þ plagioclase þ biotite þ Chuacu´s paragneiss (Solari et al. 2008). The hornblende. Subhedral and anhedral plagioclase c. 403 Ma age of likely metamorphic zircon incorpo- crystals are typically altered to sericite and have rated in the Triassic two-mica orthogneiss indicates An-contents of 35–55 mol% with normal zoning Devonian metamorphism in the Chuacu´s complex. patterns. Twinning is in part deformational. Sub- The regional 450 8C and 0.7–0.8 GPa overprint hedral orthoclase has a K-feldspar component is Late Cretaceous and the most outstanding feature of 80–90 mol% and displays perthitic unmixing. of most of the Chuacu´s complex. This regional 256 L. RATSCHBACHER ET AL. pressure-dominated metamorphism is associated imbricate stack that initiated contemporaneously with local magmatism. The PT estimates of the (c. 75 Ma). Jurassic–Cretaceous South El Tambor Cretaceous event are grossly similar in the Rabinal complex oceanic rocks incorporated c. 132 Ma eclo- complex of western Guatemala and the Chuacu´s gite into an accretionary wedge, reached crustal complex of central Guatemala, but somewhat levels at c. 120 Ma, and were emplaced onto higher temperatures and pressures occur in central Chortı´s-Block crust, whose last thermal overprint Guatemala. The microdiorite from the Juan de Paz was during the Middle Jurassic. The South El ultramafic sheet of the North El Tambor complex Tambor complex was variable affected by the probably records seafloor hydrothermal overprint Cenozoic deformation and metamorphism that followed by low-grade Cretaceous metamorphism. characterizes the Las Ovejas complex (see below). Our reconnaissance petrology of the northern Characteristically, the South El Tambor complex Chortı´s Block identifies Cenozoic amphibolite- is preserved outside the area of strong Cenozoic facies metamorphism, migmatization, local magma- tectono-metamorphic overprint, i.e. outside the tism and associated deformation for the Las Ovejas central and eastern Las Ovejas complex; we may complex starting at 40 Ma. Peak PT conditions speculate that parts of the latter may actually be are c. 650 8C at 0.7–0.8 GPa, veiling earlier high- South El Tambor complex subjected to Cenozoic grade metamorphism associated with magmatism high-grade metamorphism and deformation. and migmatization that is Permo-Triassic and Juras- sic (273–245, 166–158 Ma). Distinctly different, Structure and kinematics Middle Jurassic (c. 155 Ma) prograde, lower green- schist- to epidote–amphibolite-facies metamorph- Rabinal complex and Late Palaeozoic cover, ism is preserved in the non-migmatitic Sanarate western Guatemala complex of southern Guatemala, west and outside of the Cenozoic-shaped Las Ovejas complex. The Figures 9 and 10 compile structural and kinematic latest Cretaceous–Cenozoic Chiquimula batholith field and laboratory data and Figure 11 shows repre- is a composite, shallow crustal level intrusion sentative deformation features and dated synkine- (3–5.5 km). Sub-volcanic c. 20 Ma intrusives in matic minerals. The Rabinal complex in western this batholith demonstrate exhumation from Guatemala shows regional, heterogeneously distrib- c. 25 km during 40–20 Ma, supporting a two-stage uted, low-grade deformation (D2) that locally over- deformation–exhumation history within the Las prints high-grade, partly migmatitic flow structures Ovejas complex; that is relative rapid pre-20 Ma (D1). Most pronounced is a deflection of the first followed by slower post-20 Ma exhumation (see foliation, s1, into ubiquitous, sub-vertical, sinistral, above). D2 shear zones. In this study, we concentrated on Figure 8a summarizes the available P–T–t data the Late Cretaceous (see below) greenschist-facies from the Rabinal, Chuacu´s, Sanarate and Las Ovejas event D2; its low grade is distinct, as the same complexes. The intricate metamorphic evolution event reaches amphibolite facies in the central in the Chuacu´s complex apparently is related Chuacu´s complex of southern Guatemala and is to Silurian–Carboniferous near ultra-high pressure coeval with magmatism. Another distinct map-scale metamorphism and its retrogression into the feature is a NW structural trend (foliation, stretching epidote–amphibolite-facies field; it is overprinted lineation, shear planes), oblique to the c. east trend by Triassic magmatism and metamorphism, and of the Polochic fault zone; the latter is clearly regional Cretaceous epidote–amphibolite to younger (see above). amphibolite-facies reheating. The northern Chortı´s The basement mainly comprises locally migma- Block contains crust dominated by Jurassic titic paragneiss and orthogneiss, amphibolite and epidote–amphibolite metamorphism, only weakly (hornblende)–epidote schist, augen gneiss of mag- overprinted in the Cenozoic; this Cenozoic meta- matic protolith, pegmatite and variably strained morphism defines the Las Ovejas complex, which granitoids. The metagranitoids also intrude possibly is characterized by epidote–amphibolite to Precambrian, low-grade sedimentary rocks (San amphibolite-facies reheating. Figure 8b compares Gabriel unit of Solari et al. 2009). We also studied the P–T–t evolutions of the me´lange complexes localities of Late Palaeozoic metasediments (often (North and South El Tambor) and the continental conglomerate; Chicol and Sacapulas Formations). blocks onto which they were emplaced. Jurassic– D1 shows high-T plasticity, dominant coaxial flat- Early Cretaceous North El Tambor complex tening (s fabrics), and intrafolial and isoclinal oceanic and immature arc rocks incorporated folds, which refold an even earlier, relict foliation. c. 131 Ma eclogite into an accretionary wedge that D1 is accompanied by local migmatization, traced reached crustal levels in the Late Cretaceous by feldspar layers. D1 quartz LPO shows medium- (c. 76 Ma) and was incorporated as the roof nappe to high-T prism ,a. glide (G20s of stop G45; into the Chuacu´s complex–southern Maya Block Fig. 9, bottom right) that contrasts with D2 low-T N. AMERICA–CARIBBEAN PLATE BOUNDARY 257

Fig. 10. Additional structural data from the Maya Block of western Guatemala. Legend for structural symbols as in Fig. 9. ccw, counterclockwise. 258 L. RATSCHBACHER ET AL.

Fig. 10. (Continued) . Fig. 11. Representative dated minerals and deformation-geometry and kinematic structures. (a) Chuacu´s complex and Rabinal complex rocks. 1, Stop G27, sample G7s: weathered retrograde biotite gneiss with sinistral brittle–ductile faults and pull-apart-type tension gashes that were locally filled with pseudotachylite (c.70MaAr/Ar whole rock). 2, Western Guatemala: stop G46, sample G21s, partly mylonitic epidote schist; epidote and feldspar are brittle, quartz is ductile with subgrain-rotation recrystallization; quartz layers show synrecrystallization s–c fabric and strain concentration in shear bands (sb); phengite (c. 0.8 GPa) dates deformation at 67 + 1 Ma (stop G1, Ar/Ar); inset (top-left, stop G8, sample GM8s) shows bulging recrystallization in quartz of mid-Jurassic granitoid that is locally mylonitized along discrete Cretaceous low-temperature shear zones. 3, Stop 5CO-6: Triassic orthogneiss with pegmatite dykes and pods deformed top-to-NE during amphibolite-facies, c. 74 Ma (U–Pb zircon) flow. 4, Stop 854: second deformation, likely to be Cretaceous, sinistral shear zones anastomose around relict garnet and white mica microlithons. 5, Stop 856: undeformed pegmatite dykelet (quartz þ plagioclase, plagioclase mostly converted to epidote) intrudes parallel to c planes of second deformation. Recrystallized s–c fabric in quartz-rich layers of orthogneiss; pegmatite is Late Cretaceous (e.g. samples 885, 5CO-6b). Feldspar in orthogneiss is broken but locally recrystallized. 6, Stop 5PB-9, eastern Guatemala: high-strain, sinistral shear zone dated by synkinematic white mica at c. 76 Ma (Ar/Ar and Rb–Sr). (b) Chortı´s Block rocks. 1, Stop 5PB-5, Las Ovejas complex: synkinematic amphibole (hastingsite) and biotite in c. 650 8C, 0.75 GPa (garnet) amphibolite date rapid cooling at c. 30 Ma. 2, Stop G24, sample G5s, Sanarate complex: c. 155 Ma synkinematic amphibole and chlorite in amphibolite. 3, Stop 5C-26, Las Ovejas complex: weakly deformed, migmatitic amphibolite, hornblende is synkinematic at c. 35 Ma (Ar/Ar) with slow cooling to or reheating at c. 20 Ma (Ar/Ar hornblende and biotite). 4, Stop 5C-2, Las Ovejas-complex amphibolite: c. 20 Ma synkinematic hornblende in mylonite. 5, Stop 5C-37, Las Ovejas complex: retrograde amphibolite–biotite gneiss sequence with mostly foliation-parallel leucogranite dykes and quartz-segregation veins. Part of the high strain zone (bottom) and details (insets top) shows sub-vertical, greenschist-facies, second deformation under sinistral wrenching. Hornblende dates beginning of deformation at c. 39 Ma and biotite retrogression at c. 20 Ma (Ar/Ar). Hammer and coins for scale. 6, Stop 5H-4, Las Ovejas complex, Honduras: greenschist-facies, sinistral shear in diorite; hornblende and biotite date deformation between 38 and 25 Ma (Ar/Ar). (c) Northern foreland fold–thrust belt. Stop G60: Cretaceous limestone; NE–SW shortening and NW–SE extension indicated by pressure-solution surfaces, fold-axis parallel tension gashes, and faults. Ductile strain is recorded by elliptical shape of crinoid stems; foliation dips steeply into the picture, bedding is sub-parallel to the outcrop surface. Hammer, compass, and coin for scale. Fig. 12. Overview map (top left) locates geological–structural map (bottom right) in the Maya Block of central Guatemala. Main structural features are plotted both on map and in stereonets (lower hemisphere, equal area). See Figs 9 & 10 for legend; B, fold axis. N. AMERICA–CARIBBEAN PLATE BOUNDARY 259 quartz LPO (see below). Quartz is coarse-grained metagreywacke (coarse-grained feldspar and white and recrystallized by grain-boundary migration. mica) and K-feldspar porphyroblastic granitoid D1 is Late Triassic (c. 215 Ma), suggested by our (U–Pb zircon ages at 485–410 Ma, stop 5CO-1) U–Pb zircon age of phyllonitic (due to D2) migma- that were transformed to quartz phyllite and augen titic gneiss (G18s of stop G1). D2 formed 10–80 cm gneiss. The distinction between ortho- and para- wide, chlorite- and carbonate-bearing shear zones gneiss is problematic at many locations and makes that boudinaged s1 (e.g. G1), imposed locally well- mapped lithological boundaries questionable (e.g. developed augen- to ultra-mylonites that evolved BGR 1971, Fig. 12). Locally marble and amphibo- during decreasing temperature into ductile–brittle lite intercalations occur. The major deformation, mylonite with clinozoisite-, epidote- and quartz- attributed to the WNW-trending Baja Verapaz bearing faults (e.g. G3) and late fracturing (e.g. shear zone by Ortega-Gutie´rrez et al. (2004), is G45). These discrete shear zones widen into belts locally mylonitic and low grade. Foliation dips of mylonite, transforming the basement gneisses intermediate to steeply south(west) and the stretch- into biotite–chlorite schist and epidote–hornblende ing lineation (str1) plunges intermediate to the SW phyllite. Strain is locally prolate (dominant (Figs 12 & 13). A continuous kinematic evolution l-fabrics; e.g. G56/57). D2 quartz LPO indicates from ductile shear bands to ductile–brittle faults basal ,a. slip (G1) with a strong coaxial flow and quartz-filled, en-echelon tension gashes (e.g. component (c-axis cross-girdles). Quartz recrystal- stops 5CO-1, 5CO-3, 867) indicates sinistral trans- lization mechanisms are variable but typically pression. At station 5CO-1, typical Rabinal augen low-T and dynamic; subgrain rotation is dominant gneiss, shear bands evolve into ductile–brittle (e.g. G21s of stop G46; Fig. 11a-2) but evolves faults (quartz ductile), and faults with large out of bulging recrystallization in quartz ribbons quartz–white mica fibres that are conjugate to (e.g. G25s of stop G56, Figs. 11a-2, inset). Horn- kink bands that are partly faulted and indicate blende and feldspar are broken. Plagioclase shows north–south shortening and east–west extension rare incipient stages of recrystallization along twin along str1. Quartz shows typical low-T plasticity planes; thus, deformation occurred mostly below with subgrain-rotation recrystallization. K-feldspar c. 450 8C, evolving to very low grade (300 8C). is brittle. Local ultramylonite (,1 m thick, stop Sinistral shear criteria are ubiquitous: flow folds, 867) has cogenetic chlorite þ sericite and quartz s clasts, shear bands and s-c- fabrics in layers ribbons that show weak bulging- and dominant of quartz recrystallization (Fig. 9). Epidote subgrain-rotation recrystallization. Late kink folds (+hornblende) schists G19s (stop G1) and G21s are interpreted as flow heterogeneities at the end (stop G3) provide quantitative PT estimates for the of deformation. Quartz fibres in the strain shadow initial, higher-T stages of D2 at 450–500 8C and of pyrite are un-recrystallized. The post-245 Ma 0.7–0.8 GPa. D2 is precisely dated at 67 + 1Ma (see above) phyllite and conglomerate (mostly by synkinematic white mica (G19s of stop G1, intermediate volcanic rock pebbles, stops Fig. 11a-2). D2 also developed in the Palaeozoic 865–866) records identical, low-T structural geo- metasediments and is more pronounced (sinistral) metries and kinematics. Paragneiss (stop 867) pre- transpressional (e.g. G4, Fig. 9) than in the under- serves pre-Cretaceous deformation with large lying basement. Location GM36 shows stretched white mica and pre-existing foliation and felsic conglomerate (Fig. 9) cut by aplitic veins with the veins that are cut by orthogneiss that is typical same c. east–west extension that stretched the Rabinal gneiss. pebbles; low-T plasticity with dynamic subgrain rotation occurred in quartz clasts. The basement Chuacu´s complex, central Guatemala rocks are variably overprinted by faults that are related to the Polochic fault zone (Fig. 10); the prin- Central Guatemala exposes the deepest part of the cipal faults strike WNW (Polochic-parallel) and Chuacu´s complex studied by us (Figs 12 & 13) and north. Faulting is 20 Ma, given by the c.20Ma allows insight into pre-Cretaceous deformation– U–Pb zircon lower intercept of samples GM13s metamorphism–magmatism. The Cretaceous struc- and GM14s (stops GM42, GM43) and the c. tural overprint increases both toward south and 23 Ma AFT age of stop GM35 (sample GM9s); it north, i.e. the North El Tambor complex of the is best constrained by the TFT and AFT ages hanging wall. Typical Chuacu´s-complex rocks are along the Polochic fault (11.5–4.8 Ma, see above). garnet-bearing para- and orthogneiss, garnet amphi- bolite that is partly retrograde eclogite (see petrol- Rabinal complex, Rabinal granitoids and ogy), and marble (e.g. stops 880–884, 886, low-grade metasediments, central Guatemala 5CO-4–6). Felsic and mafic rocks are often concor- dantly interlayered and resemble a bimodal volcanic The Rabinal complex in central Guatemala sequence dominated by metasediments. The layer- comprises paragneiss that typically resembles ing is accentuated by felsic sills (dominant) and 260 L. RATSCHBACHER ET AL.

Fig. 13. Additional structural data from the Maya Block and the Early Cenozoic Subinal Formation in central Guatemala. Legend for structural symbols see Figs 9, 10 & 12. dykes that crop out as partly pegmatitic orthogneiss garnet (+kyanite in quartz veins) growth. Local, and locally comprise 50% of the rock volume. Our large-scale fluid flow is indicated by up to 1 m new geochronology assigns them to the Triassic thick quartz-segregation veins and ubiquitous (see above). This rock association is sheared, albite growth. Locally, but increasingly abundant folded, and thermally overprinted by syn- to post- to the south, a second type of pegmatite is observed tectonic feldspar þ hornblende þ white mica þ (e.g. stop 884). It is muscovite-rich, crosscuts N. AMERICA–CARIBBEAN PLATE BOUNDARY 261

Fig. 13. (Continued) . the sills/dykes, and is associated with lower (U–Pb zircon, stop 5CO-6) and their 74–62 Ma amphibolite- to upper greenschist-facies meta- white mica ages (see above). morphism. These pegmatites are undeformed Where preserved, Triassic deformation is dis- where massive but sheared along quartz-rich tinct. Foliation is sub-vertical and folded tight- layers. They are Late Cretaceous as indicated by isoclinally to ptygmatically with axes parallel to c. 74 Ma metamorphic overprint in the host rocks the mineral stretching lineation (F1 ,a. type 262 L. RATSCHBACHER ET AL.

Fig. 13. (Continued) . N. AMERICA–CARIBBEAN PLATE BOUNDARY 263 folds) that trends NNW (Fig. 12, stops 5CO-5, sinistral, often anastomosing shear zones 880–886) and thus is clearly different from the (Fig. 11a-4) show subgrain-rotation recrystalliza- overall trend of the Late Cretaceous deformation tion and reaction softening by transfer to sericite (c. ENE). F1 verge mostly (north)east and are and epidote, respectively; locally feldspar is recrys- locally sheath folds. At these locations, overprinting tallized (c. 500 8C). At 854 and 856, undeformed Cretaceous F2 folds are steeply plunging, open, pegmatite and aplite dykes/dykelets cut migmatitic, northeast-vergent, and associated with dextral sillimanite- and white mica-bearing ortho- and para- and sinistral (conjugate) shear bands (Fig. 12). gneiss and are younger than the D2 fabric Garnet–K-feldspar gneiss (5CO-4), resembling (Fig. 11a-5). Pure quartz layers show prism ,a. Rabinal augen gneiss, has old garnet with an internal slip and subgrain-rotation with transitions to grain- foliation that is 908 to the external (Cretaceous) foli- boundary migration recrystallization. Common F2 ation. Overall, Triassic deformation appears coaxial refold both s1 and D2 shear zones, imposing and the structures are annealed (widespread albite prolate strain. Radiometric age control on these overgrowth, see above). deformations is weak (854F); migmatization and Structural geometry and kinematics of the D1 are pre-149 Ma, however, structural geometry, regionally dominant Cretaceous deformation kinematics and associated metamorphism suggest appear simple (Figs 11a-3, 12 & 13): the stretching a Late Cretaceous age for D2. lineation consistently trends c. ENE and parallels We also studied the Chuacu´s complex along a the ubiquitous B2 axes; where s1 is sub-vertical, it section at its easternmost outcrop at the base of the is transposed by sinistral shear zones/bands; Juan de Paz ophiolite unit (Fig. 12). The structural where s1 is sub-horizontal, shear zones/bands trend changes to ENE, from WNW in western and record sinistral transpressive, top-to-NE flow. Late east in central Guatemala. Deformation is continu- faults have huge fibres and indicate north–south ous from greenschist facies to brittle (boudinaged shortening (e.g. stop G29). This Cretaceous high- quartz rods) and heterogeneous. At station 5PB-9, strain, NE–SW stretching is suggested by various deformation is high-strain, mostly coaxial with sinis- features. Stop G31, for example, shows relatively tral and dextral shear bands and ,5 cm thick mylo- idiomorphic, inclusion-free, pre-Cretaceous garnet nite zones, shows coaxial, basal ,a. slip in quartz, that is fractured and connected with chlorite; a and open ,a. folds. The whole rock– second generation of garnet and chlorite is co- synkinematic white mica (Fig. 11a-6), Rb–Sr iso- genetic. Quartz shows transition from subgrain- chron at 76.5 + 0.6 Ma precisely dates deformation. rotation to grain-boundary-migration recrystalliza- At station 5PB-10 (Fig. 12), ubiquitous vertical, tion (e.g. stop G31). At stops 880–886 late, sinistral shear bands, a single- to type-I cross-girdle mostly unfoliated/unfolded dykes/veins are quartz c-axis orientation and strong preferred orien- exposed; some of these quartz þ feldspar þ tation of a-axis and m-planes indicate dominant muscovite þ biotite pegmatites show high-strain rhomb and prism ,a. slip under somewhat higher boudinage. At stop 887, a several metres thick, deformation temperatures than at station 5PB-9 weakly foliated pegmatite crosscuts the major and dominant non-coaxial flow. (Triassic) foliation but deflects it sinistrally. F3 refold F2 openly. Where F3 is not developed l-tectonites are common. Our interpretation is that Metasomatized southern Chuacu´s complex folding is represented by prolate strain (e.g. stop rocks below the North El Tambor complex, 887). Our geochronology constrains cooling central Guatemala from c. 550 to 275 8C and thus deformation at 75–65 Ma (see above). The immediate footwall of the North El Tambor Stations 854 and 856 reveal Chuacu´s complex serpentinites comprises imbricates that stack that is dominated by orthogneiss (Fig. 12). At 856, Chuacu´s-complex rocks, Chuacu´s rocks with low- ,0.5 m thick, D1 mylonite–ultramylonite shear grade sedimentary rocks (Jones Formation and San zones cut augen gneiss, biotite granite and aplite Lorenzo marble), and these rocks with serpentinite dykes. The host rock is weakly deformed at high-T (Figs 12 & 13). Rocks that are imbricated are phyl- preserving NW trending str1. Discrete ultramylonite lite (retrograde biotite gneiss), epidotized biotite shear zones that differ in degree of localization – the gneiss, (+garnet)–actinolite–clinozoisite–chlorite younger are more strongly localized – overprint a phyllite, marble and low-grade siltstone. These first, east striking, dextral, high-T shear zone rocks show neocrystallization of calcite, quartz, fabric. D1 is syn-migmatitic, F1 are isoclinal. D2 Mg-chlorite and albite (e.g. stops 5PB-15, 860) imprints ENE-trending ultra-mylonitic–mylonitic, depending on the surrounding lithology. The bound- Ramsay–Graham-type shear zones with late chlor- ary between serpentinite and Chuacu´s-complex ite–quartz veins, in which quartz is fibrous tracing rocks is, where exposed, a shallow-dipping breccia the NW–SE str2. Quartz and plagioclase in the zone with serpentinite and marble clasts (stop 264 L. RATSCHBACHER ET AL.

G26, Fig. 13). Nearly all of the stations (Fig. 13) the ductile–brittle regime with a dominant northeast show relict, mostly sub-horizontal, high-grade foli- striking set. At station G27 pseudotachylite (Figs ation (s1) that is folded (e.g. G26, G27, 5C-36). 11a-1 & 13, sample G7s) fills pull-apart tension During D2 sinistral ductile–brittle shear bands fractures. These are again involved in continuous developed where s1 was steeply dipping (e.g. deformation, as east-trending quartz fibres formed 5PB-15, 860). Stretching is WSW–ENE and exem- as strain shadows. The pseudotachylite dates D2 at plified by s1 boudinage. Quartz-filled tension gashes c. 67 Ma. are ubiquitous (e.g. G26, 5PB-15) and testify to fluid flow. F2 flexural-glide folds are north(east) vergent North El Tambor complex on Chuacu´s and show lengthening along B2 (e.g. G26, 5C-36). complex, central Guatemala Quartz deformed by subgrain-rotation recrystalliza- tion (e.g. G26) and basal ,a. glide (5C-36). The We analysed imbrications of serpentinite, ophical- LPO indicates dominantly coaxial flow. White cite, and felsic volcanic rocks (stop G34) and mica at stop G26 is sericite and its Ar/Ar age serpentinites (stops G28, G35, G38, GM11-13, dates deformation at 60 Ma. This corresponds to 5PB-7; Figs 13 & 14). All stations show a low- the Rb–Sr white mica–whole rock isochron at c. grade, ductile–brittle fabric with shear bands 63 Ma from a chlorite phyllite. Faulting started in and faults that slipped sinistrally, coeval with

Fig. 14. Geological map (bottom) of the Motagua suture zone with structural data from the northern Chortı´s Block, the North and South El Tambor complexes, and the Sanarate complex; the outline of the latter is conjectural. In addition structural data of Cretaceous to Late Cenozoic sedimentary and magmatic rocks straddling the suture are shown. Main structural features are plotted both on map (this figure) and in stereonets (lower hemisphere, equal area, Fig. 15). See Figs 9 & 10 for legend. Adapted from Bonis et al. (1970). N. AMERICA–CARIBBEAN PLATE BOUNDARY 265 shortening indicated by folding and oblique-slip material is partly deposited in veins that comprise faulting. White mica ages (c. 65 Ma) of greenschist- up to 20 vol% of the rock. Fold axis-parallel exten- facies rocks (Chuacu´s complex, stops G26, G29, see sion is accentuated by a second set of tension gashes above) in the immediate footwall of these El (tgyoung) and conjugate faults that represent tilted Tambor rocks suggest latest Cretaceous defor- horst-graben structures; the latter exactly constrain mation. At several stations, tectonic me´lange extension at 1288 (inset stop G61–G62, Fig. 10). occurs along major north dipping detachments that Abundant crinoids in the limestone record up to contain foliated, moderately serpentinized ultra- 50% shortening due to volume reduction that is mafic rocks (G34, G38) and pure serpentinite mostly accommodated by bedding-parallel stylo- (G35; Figs 13 & 14) with brittle–ductile s–c lites (Fig. 11c-1). Other stations, both in limestone fabrics. The ultramafic bodies are surrounded by and serpentinite, show less ductile flow but more an anastomosing, lenticular, scaly foliation that pronounced faulting. Folding is mostly gliding in records overall coaxial deformation. Shortening is bedding; the slip direction is oblique to the fold NNE–SSW and extension is WNW–ESE. Most axes (e.g. stop G55). Common to most stations is stations contain a younger fabric with unclear rotation of s1 from east–west to north–south in age relations: sinistral shear and normal faulting time and sinistral transpression. both with NW–SE extension. Late-stage, flattening- In central Guatemala, Permian limestone and the type extension is possibly caused by topographic contact zone between Cretaceous limestone and ser- collapse. pentinite provide a similar albeit less complete deformation history. Sinistral transpression and Northern foreland fold–thrust belt, rotation of s1 imply increased shortening and less Guatemala strike–slip over time (e.g. G43, Fig. 13). The weak structural record of serpentinite emplacement The ‘Laramide’ foreland fold–thrust belt north of and the overall paucity of shallowly dipping faults the Polochic fault zone and the overlying North El are notable. Early fault sets indicate c. northeast– Tambor complex allow comparisons of structural southwest shortening along conjugate strike–slip geometries and kinematics with those recorded in faults (e.g. G39). At station G40, c. north–south the Chuacu´s basement in the south. The strati- shortening along mostly strike–slip faults formed graphic range observed at our structural stations anastomosing fault zones with boudinage and ranges from Permian to Upper Cretaceous, that is east–west extension. Normal faults may be topogra- Permian Chochal and Tactic Formation limestone phically induced, and late cataclastites likely record and shale, and Lower and Upper Cretaceous active sinistral wrenching along a strand of the Coba´n and Campur Formation dolomite and lime- Polochic fault zone. The normal faulting observed stone (BGR 1971). In the following, we give an at station G41 could also be gravitationally overview but emphasize that the stereoplots induced or, alternatively, a transtensional continu- (Figs 10 & 13) contain a cornucopia of structural ation of sinistral wrenching. detail, for example, geometries, deformation evolution and changes in the palaeostress field. Cenozoic volcanic rocks and Early Cenozoic Deformation age is difficult to constrain directly. Subinal-Formation red beds, western and Structural and kinematic compatibility with the central Guatemala well-dated deformation in the Chuacu´s complex (see above) and that in the Sierra Madre Oriental– In western Guatemala, Cenozoic pyroclastic rocks, Yucata´n fold belt (e.g. Gray et al. 2001) suggest welded tuffs and, locally, sedimentary rocks cover the major deformation, sinistral transpression, is the Rabinal complex and are exposed along the latest Cretaceous–Early Cenozoic. Polochic fault zone. Our data (Figs 9 & 10) highlight In western Guatemala, stations G59–G61 within three strike–slip dominated events. From older to Cretaceous limestone and minor shale record younger these are: (i) c. north–south shortening; folding with well developed ductile–brittle, NW– (ii) c. NE–SW shortening, mostly along c. east SE extension along the regional fold trend trending sinistral strike–slip faults. This is by far (Figs 10 & 11c-1). Deformation starts during layer- the dominant set and traces the Polochic fault-zone parallel shortening before buckling with vertical deformation. The strata dip up to 808, indicating extension by tension gashes (tgold) and continues strong local shortening. A similar stress field but with buckling and layer-internal ductile flow with dominated by normal faults also occurs in subhori- foliation development, foliation boudinage and cre- zontal pyroclastic rocks that likely are part of the ation of a fold axis-parallel stretching lineation. Cenozoic ignimbrite province of Central America Later stages develop conjugate but mostly sinistral (c. 15 Ma; e.g. Jordan et al. 2007); we thus attribute faults and oblique stylolites that show strong this major event to the currently active stress field. volume-loss by pressure solution; dissolved (iii) Normal faulting with widely dispersed slip 266 L. RATSCHBACHER ET AL. directions. This event probably records topographic the Las Ovejas complex south of the central collapse. Motagua valley and north(east) of serpentinite and Deformation of the Subinal-Formation red beds me´lange rocks of the South El Tambor complex started with c. north–south shortening that tilted (Figs 14 & 15). This area comprises biotite granite beds (open to tight folds) and produced a fracture with diorite xenoliths, gabbro-amphibolite, migma- cleavage (Figs 13 & 14). At stops 872–873, two titic biotite gneiss and garnet-bearing leucosome. sets of older faults have constant ENE–WSW The main deformation is of variable intensity, extension (s3) but s1 and s2 permutated. The from weakly foliated to locally mylonitic–ultra- strike–slip faults have oblique lineations so that mylonitic but constant in orientation. Foliation has faulting probably started during late stages of intermediate dips toward the NW, kinematically folding. Faulting recorded at all stations has a related shear bands trend NE and dip, in general, clear sinistral-transpressive component. Late steeper than the foliation, and the stretching linea- normal faulting is interpreted as prolongation of tion is sub-horizontal NE–SW, parallel to axes of folding-related deformation with extension along ,a. type folds. These structures overprint relict the fold axes becoming dominant. earlier deformation fabrics and began syn-migmatic (Fig. 11b-3) with melt locally concentrated along Sanarate complex, central Guatemala shear bands; shear is overwhelmingly sinistral. Dykes intrude parallel to the foliation (5C-22, sub- The Sanarate complex, the westernmost basement volcanic granite), mostly normal to the stretching outcrop south of the MSZ, comprises retrograde lineation (5C-2, diorite), thus indicating syntectonic amphibolite, chlorite–epidote–hornblende schist, magmatism or crosscut irregularly (5C-28, pegma- micaschist and chlorite phyllite with peak PT con- tite). Synkinematic hornblende and white mica ditions at c. 550 8C and c. 0.65 GPa. The basement date deformation as pre-to syn-35 Ma; a cluster of is in fault contact with Cenozoic andesite and red c. 10 Ma younger ages (mostly cooling through beds (stops G23, G24, 5G-1; Figs 14 & 15). Foli- 300 8C) indicates later structural reactivation, as ation and relict isoclinal folds constitute D1.D2 suggested by locally well-defined, synkinematic shows low-T, partly mylonitic quartz–white mica mineral fabrics (e.g. stop 5C-2, Fig. 11b-4). fabrics along an east trending, vertical foliation; Paragneiss and micaschist at stop 5C-3 bound shear sense is sinistral. Quartz is post-tectonically South El Tambor serpentinite and show syntectonic, annealed. D3 occurs both in the red beds/volcanic greenschist-facies retrograde metamorphism. Late rocks and the basement. ENE–WSW shortening sinistral faults juxtapose these rocks with the ser- thrust the basement in a sinistral-oblique sense pentinite. Stations 5C-37 and 38 at the southeastern onto the red beds. Two generations of hornblende edge of the same South El Tambor allochton com- exist in the studied amphibolite. The dominant syn- prise migmatitic biotite gneiss, amphibolite and kinematic younger generation (Fig. 11b-2) and also leucogranite and their retrograde products. D1 synkinematic white mica in the micaschist/phyllite synamphibolite-facies deformation with local mig- date D2 as Middle Jurassic (samples G3s, G5s, matites is overprinted by localized but pervasive stations G23, G24, Fig. 5e). The reheating (quartz greenschist-facies D2 (Fig. 11b-5). The biotite weakly annealed) is probably Cenozoic (low-T gneiss and the amphibolite are in part extremely steps in G5s). Adjacent stations 853 and 870–871 deformed and often l-tectonites; s1 and sinistral- comprise massive amphibolite and rare garnet transtensional shear bands dip intermediate to NW amphibolite and leucogabbro close to the boundary and str1 plunges shallowly. The tourmaline– with the South El Tambor me´lange rocks. Their garnet–white mica leucogranites postdate the mig- apparently lower-grade metamorphism disting- matitic gneiss, intrude mostly as sills, and are uishes them from the Las Ovejas complex amphibo- locally deformed synintrusive with high strain. D2 lites. Asymmetric boudinage of hornblende–quartz induced vertical sinistral shear zones, ,a. folds, veins, s clasts and late biotite-rich shear bands foliation boudinage and syn- and mostly antithetic associated with kink bands record top-to-NE flow faulting of leucogranite dykes and quartz with a slight dextral component along west segregation veins (Fig. 11b-5). The transition from dipping foliations. Tight to isoclinal flow folds are shallowly dipping D1 shear bands to vertical, present in high-strain zones (Fig. 15). This defor- ductile–brittle D2 ones is abrupt (within c. 10 m). mation remains to be dated. Hornblende (amphibolite) and white mica (leuco- granite) date the initiation of deformation and Las Ovejas complexes and San Diego phyllite, syntectonic intrusion at 40–35 Ma. The likely kin- Guatemala and western Honduras ematic evolution from ductile (40 Ma) to ductile–brittle (c. 20 Ma, biotite) could be inter- Stations 5C-2, 5C-28, 5C-26, 27 and 5C-22–24 are preted in this outcrop as long-lasting progressive in basement lacking retrograde metamorphism of deformation instead of structural reactivation (see N. AMERICA–CARIBBEAN PLATE BOUNDARY 267

Fig. 15. Structural data from the Chortı´s Block and Cretaceous to Late Cenozoic sedimentary and magmatic rocks straddling the Motagua suture zone. Legend for structural symbols as in Figs 9, 10 & 12. above). Stations 5PB-1–5 and 861, 862 expose are post-tectonically annealed due to their proximity unretrogressed Las Ovejas complex migmatitic to granitoids (Fig. 14). At location 5PB-5 massive biotite gneiss, intercalations of (garnet) amphibolite, migmatite is enclosed in granite. Garnet amphibolite biotite–white mica–garnet–tourmaline pegmatite, contained in the migmatite yielded c. 650 8C at 0.7– marble and calcsilicate rocks. Locally these rocks 0.8 GPa peak metamorphic conditions and preserves 268 L. RATSCHBACHER ET AL.

Fig. 15. (Continued) .

pre-migmatitic garnet relicts. Deformation started Migmatization was at c. 36 Ma (U–Pb zircon and prior to migmatization, is locally high-strain Ar/Ar hornblende, see above). Station 5C-15 with migmatitic biotite–gneiss mylonite, and is shows the Jocota´n fault as a ductile–brittle sinistral sinistral-transpressive with local s–c mylonites. It fault zone. North–south ductile shortening with continued with sinistral ductile–brittle faulting. well-developed sinistral shear bands changed to N. AMERICA–CARIBBEAN PLATE BOUNDARY 269

Fig. 15. (Continued) .

NE–SW compression in the brittle field. Low-T sinistral transpression that started during the mig- plasticity is recorded by bulging to dominant matitic stage. Deformation continued into the subgrain-rotation recrystallization in quartz. ductile–brittle field with sinistral-transtensive, SE-dipping shear zones/normal faults that we Las Ovejas complex of northern Honduras relate to opening of the Sula graben. The pegmatites either show strain concentration with well- Stations 5H-1–6 characterize deformation of the developed mylonite and ultramylonite or are vari- western footwall of the Sula graben in northern ably but mostly weakly deformed, forming irregular Honduras (Figs 14 & 15). The Las Ovejas pods. In comparison with Guatemala, foliation is complex rocks and their deformation are similar to rotated, dipping moderately to steeply north with those in Guatemala. Partly migmatitic biotite the stretching lineation NE–SW; shear bands dip gneiss, biotite granite, diorite, micaschist and NW. Our 40–36 Ma U–Pb zircon ages from mig- small white mica- and garnet- or hornblende- matitic biotite gneiss and the c.38MaAr/Ar horn- bearing pegmatite pods and sills deformed under blende age from hornblende pegmatite constrain 270 L. RATSCHBACHER ET AL. small-scale magmatism and deformation (see formation dated by rapid cooling at c. 28 Ma (Ar/ above). Synkinematic hornblende and biotite Ar K-feldspar). Conjugate strike–slip faulting (Fig. 11b-6) and rapid cooling of K-feldspar in with north–south compression was followed by greenschist-facies, ductile–brittle shear zones normal faulting (s3 of both events trends east) in related to the Sula graben suggest its initiation at red beds and volcanic rocks at station 5H-14 c. 28 Ma. At stop 5H-7, a top-to-east shear/normal within the Jocota´n fault zone. fault zone separates orthogneiss from phyllite and is likely related to the Sula graben. The orthogneiss South El Tambor complex on Las Ovejas shows a relict older fabric (D1) with NE dipping foliation. Greenschist facies with up to mylonitic complex and San Diego phyllite, central deformation is present both in phyllite and ortho- Guatemala gneiss. Shear bands show sinistral transpression (NE–SW shortening) with transition to brittle– Stations 852, 855, 864 and 869 comprise rare un- ductile fabrics. Quartz shows subgrain-rotation foliated blueschist, serpentinite, graphite-bearing recrystallization. Stations 5H-8 and 9 comprise metasiltstone to phyllite with quartzite layers, and orthogneiss and migmatite with mostly Permo- greenschist of the southern South El Tambor Triassic U–Pb zircon ages and Jurassic–Cretaceous me´lange. The metasiltstones show ductile–brittle, cooling ages. These rocks are distinctly different heterogeneous fracture cleavage formed by from the Las Ovejas complex rocks along the pressure solution (diffusion creep), and quartz MSZ. Migmatization is pervasive and thus different veins in metasandstone layers. D1 shows open– from the c. 40 Ma event. D1 is coeval with migma- tight, c. south vergent folds and thrust imbricates tization. Again, NE-dipping s1 is overprinted by with associated fault-bend folds. We do not have discrete ductile–brittle shear bands/zones that age information on this me´lange-forming event. probably are Cenozoic. Stations 5H-11 and 12 These rocks are cut by sub-volcanic andesite comprise low-T, mylonitic orthogneiss (probably a dykes and sub-volcanic granite stocks, which we rhyolitic protolith) interlayered with thin paragneiss interpret as apophyses of the Chiquimula batho- and marble (?metavolcanic sequence). A relict lith. D2 affects the me´lange rocks and these high-T deformation, again with the characteristic dykes/granitoids by c. east-trending (transten- NE-dipping foliation, is cut by greenschist-facies sional) strike–slip and normal faults. Their low-T mylonite (locally chlorite out of biotite and sericite nature and the presence of c. 20 Ma sub-volcanic on mylonitic shear bands) with ,3 cm chlorite- granite in the Chiquimula batholith (see above) bearing ultramylonite layers. Overall deformation suggest Cenozoic faulting. is top-to-SW sinistral transpression and probably related to the Jocota´n fault zone. Interpretation of new structural geology Our work along the MSZ focused on the Late Cretaceous and Cenozoic–Quaternary Cretaceous–Cenozoic deformation. Clearly, how- sedimentary and magmatic rocks south of the ever, older events occurred. In the Rabinal and Motagua suture zone, central Guatemala Chuacu´s complexes, Triassic deformation, studied superficially, shows high-T ductile flow and is dis- Conjugate strike–slip fault sets that indicate c. tinct in structural geometry and kinematics: mostly east–west shortening dominate Upper Cretaceous (north)west striking foliation, NW-trending stretch- limestones (stops GM4, 5 and 7; Figs 14 & 15). ing directions, tight–isoclinal ,a. folds. Shorten- The quality of the field exposures does not allow ing appears coaxial north(east)–south(west) with us to discriminate among subgroups representing a strong along-strike lengthening. Associated high-T deformation sequence or to assess block rotations. metamorphism, migmatization, and magmatism The lack of a dominant east trending set of sinistral are Late Triassic (238–215 Ma, major group at strike–slip faults, which we attribute to the active c. 220 Ma). Foliation trajectories of the Late Creta- stress field along the Motagua fault, distinguishes ceous, Maya-Block (for this discussion it includes these stations from the Cenozoic–Quaternary the Rabinal and Chuacu´s complexes and their rocks (stops GM1, 2 and 10; Figs 14 & 15) but cover) deformation change from WNW in ties them to the deformation recorded in Subinal- western, to west in central, and SW in eastern Formation red beds. The latest increment of defor- Guatemala. Using the (west-)northwest strike of mation in all these stations is east–west extension the Late Cretaceous–Paleocene Mexican thrust by normal faulting. Biotite granite (37 Ma) at belt as a reference frame (e.g. Nieto-Samaniego stations 5H-5 and 6 within the footwall block of et al. 2006), this implies large-scale sinistral shear. the Sula graben shows brittle–ductile normal The southern Maya Block in Guatemala apparently faults and fracture cleavage related to graben comprises a ductile–brittle nappe stack, whose N. AMERICA–CARIBBEAN PLATE BOUNDARY 271 detailed geometry remains to be established. The be younger than c. 15 Ma (apparent post-tectonic North El Tambor allochton in the hanging wall, pluton), although our only Cenozoic age itself an accretionary stack of rocks of variable (c. 37 Ma) from western Guatemala (G27s, stop tectono-stratigraphic origin and subduction depth, G57), from a discrete low-T mylonite zone in is underlain by an up to several hundred metres- gneiss close to the Polochic fault zone, may hint at thick carapace of imbricated rocks, containing a Late Eocene slip history. Ductile to brittle slip Chuacu´s basement and cover, altered by fluid infil- along the Polochic is dated as Late Neogene– tration and metasomatism. The central Chuacu´s Recent in western Guatemala (TFT and AFT ages) complex may either constitute a Late Cretaceous and at c. 8 Ma (Ar/Ar chronology) along the, prob- antiformal stack or, our preferred model, an anti- ably kinematically related, Tonala shear zone in form above a blind thrust, exposing pre-Cretaceous southeastern Mexico. metamorphic and magmatic products in its core. Cenozoic volcanic rocks record a phase of pro- The Baja Verapaz shear zone is the sole thrust to nounced north–south shortening along the western this stack. The northern El Tambor sheets (Baja Polochic fault zone. Its active deformation is by a Verapaz, Sierra des Santa Cruz) are cut by out-of- kinematically related combination of strike–slip sequence thrusts that are part of the foreland fold– and normal faulting. We interpret the Subinal- thrust belt. The relatively high-P (0.7–0.8 GPa) Formation red beds as intra-continental molasse and relatively low-T (450 8C) metamorphism in related to late-stage, erosional unroofing of mostly the Chuacu´s complex suggests subduction of the Chuacu´s complex. Its deformation was devel- oceanic crust transitioning to continental crust (the oped during north–south shortening by folding Chuacu´s geothermal gradient is 18 8C/km from and fold-axis parallel extension, accompanied by our data) and steep continental subduction; this sinistral shear. The latest stage indicates north trend- is supported by the nearly coeval blueschist ing graben formation. (c. 76 Ma, North El Tambor complex) and epidote– Our study indicates that the Las Ovejas complex amphibolite- to amphibolite-facies (75 Ma, of the northern Chortı´s Block is less a lithologically Chuacu´s complex) metamorphism in the accretion defined unit than distinguished from the other base- stack and the sinistral–transpressive deformation. ment units by its Cenozoic metamorphism, magma- Penetration of deformation and overburden (PT tism and deformation. It forms a westward conditions) appear to increase from western to narrowing belt of overall sinistral transtension eastern Guatemala. Although in detail variable, along c. NW dipping foliations. Deformation pene- meso-scale Late Cretaceous deformation is simple: tration apparently increases toward north, the MSZ, north–south shortening by reverse shear zones and and seems to be more localized and to die out south- up to two generations of folds are accompanied by ward and westward. The belt of Cenozoic shear thus sinistral strike–slip shearing and major east–west appears to trend oblique to the Motagua fault. lengthening. Tangential stretching is dominant in Deformation interacts with amphibolite-facies most cases, as deduced from the mostly shallowly metamorphism, widespread migmatization and plunging stretching lineation. In detail, overall sinis- local magmatism, demonstrated most clearly by tral transpression is typically partitioned between small syn- to post-tectonic pegmatites. The close oblique stretching within the foliation and strike– spatial and temporal interaction between high-strain slip along steeper dipping shear zones/bands tran- deformation and Cenozoic metamorphism, migma- secting the foliation. Flow has a dominant non- tization and magmatism is confined to the northern coaxial component, most clearly expressed by Chortı´s Block. It appears that arc magmatism/ quartz LPOs, developed during dynamic, subgrain- metamorphism provided the heat input to localize rotation dominated recrystallization. There is a deformation. Deformation started pre-migmatitic clear kinematic continuity from ductile to brittle and occurred in a kinematic continuity from deformation. Although dominated by diffusion ductile to brittle. Batholiths are post-tectonic. Ceno- creep, deformation is first-order identical within zoic sinistral-transtensional wrenching appears, the northern foreland fold–thrust belt, the sole like the Cretaceous deformation of the Chuacu´s thrust carapace below the North El Tambor complex, partitioned between oblique stretching complex sheet, and within it. Emplacement of the along a shallower dipping foliation and strike–slip El Tambor allochton, poorly constrained by our along steeper dipping shear zones/bands. Defor- data, appears to change from tangential to frontal mation began prior to 40 Ma and we suggest over time (strike–slip before thrusting). that there were two major phases, 40–35 and Structural trends in the northern foreland fold– 25–18 Ma. This reactivation interpretation still thrust belt and the Rabinal complex of western must be tested. The Sula graben started in temporal Guatemala are NW and oblique to the Polochic and kinematic continuity with wrenching along fault zone, testifying to the younger age of this the Las Ovejas complex at c. 28 Ma. The topo- fault. Polochic mylonitic deformation appears to graphically well developed active grabens, 272 L. RATSCHBACHER ET AL. however, are a neotectonic feature. Also the Ortega-Obrego´n et al. (2008) suggested intrusion Jocota´n–Chameleco´n fault zone was part of this of the Rabinal granitoids between 462 and deformation system, albeit mostly ductile–brittle. 453 Ma. Combining all available U–Pb zircon The Cenozoic event veils pre-Cenozoic tectono- ages from the Maya Mountains (Belize), the Altos metamorphic–magmatic events, with distinct Cuchumatanes (western Guatemala), the Chiapas structural orientation and kinematics. The latest Massif (southeastern Mexico) and the Rabinal preserved is Jurassic and has a NW structural complex (see above; Figs 1c & 16a), we suggest trend. The South El Tambor complex was emplaced that the Maya Block experienced prolonged by south vergent thrusting and associated folding. Ordovician–Early Devonian magmatism. We It is involved into the Cenozoic sinistral wrench consider the c. 396 Ma age of a Rabinal orthogneiss deformation. reliable, as 404–391 Ma magmatites (both granite The Sanarate complex with its distinct lithology, and rhyolite) occur north of the Rabinal complex. metamorphism and age of metamorphism is Our Rabinal augen–gneiss sample is dominated affected by the Cenozoic deformation but has a by 485–400 Ma zircons and shows two age groups characteristic west dipping foliation and top-to-east at 465 and 413 Ma. Similar ages occur in the flow in its dominant amphibolite lithology. This Maya Mountains (c. 418 Ma; Steiner & Walker deformation is Jurassic or younger. Late Cenozoic 1996) and in other Rabinal granitoids (c. 462, deformation, preserved in Late Cretaceous and c. 417 Ma; Ortega-Obrego´n et al. 2008). The Cenozoic–Quaternary rocks on the Chortı´s Block, Chiapas Massif and the Rabinal area of central is by distributed sinistral strike–slip. The latest Guatemala may also host c. 482 Ma granitoids event we recognized in the Chortı´s Block and the (Weber et al. 2008; Ortega-Obrego´n et al. 2008; Subinal Formation is east–west extension by this study). normal faulting. The entire MSZ area shows gravi- The Chuacu´s-complex rocks apparently recycle tationally induced faulting due to the large topo- these Early Palaeozoic magmatites in their sediments graphic relief. and granitoids: Solari et al. (2009) reported, among others, 480–402 Ma (main age group at 436 Ma) detrital zircons in a Chuacu´s-complex paragneiss, Discussion and we found c. 403 Ma, probably metamorphic zircon in Triassic orthogneiss. The Chuacu´s com- Here we return to the principal questions this paper plex shows high-P, perhaps ultrahigh-P (Ortega- aims to address: (i) which crustal sections of the Gutie´rrez et al. 2004) metamorphism that is, at MSZ accommodated the large-scale sinistral best, constrained in age to between c. 440 Ma, the offset along the northern Caribbean Plate boundary, youngest significant zircon-age group of the when did it occur, and in which kinematic frame- Chuacu´s metasediments and possibly 302 Ma, the work, and (ii) what Palaeozoic–Cenozoic geologi- age of post-high-P migmatite leucosome, and cal correlations exist between the Maya Block, the conservatively to between c. 440 (see above) and ´ Chortıs Block, the tectono-stratigraphic complexes 238–218 Ma, the age of Triassic orthogneiss / (‘terranes’) of southern Mexico, and the arc (Figs 2d & 16a; see above). The Rabinal complex subduction complexes of the northern Caribbean. apparently lacks high-P metamorphism but this has First we focus on geological correlations proceeding to be confirmed by quantitative petrology. Lower from Palaeozoic to Recent, and then we refine Carboniferous–Permian Santa Rosa Group sedi- Caribbean tectonic models for the Cretaceous and ments cover the Rabinal complex unconformably Cenozoic. (Fig. 2c). Recent studies within and along the southern Taconic–Acadian (Early–Middle margin of the Maya Block suggest a subdivision Palaeozoic) orogeny of its lithostratigraphy into pre-Early Ordovician clastic rocks (e.g. Jocote complex, intruded by The Chuacu´s complex contains a paragneiss- c. 482 Ma granite in the southern Chiapas Massif, dominated sequence, in which metabasalts prevail Weber et al. 2008; post-c. 520 Ma Baldy unit of among the magmatic rocks, and quartzite and the Maya mountains, Martens et al. 2010), and a calcsilicate/marble are minor constituents (Fig. 2d). younger, again mostly clastic sequence that has In contrast, the hanging Rabinal complex is rhyolite intercalations (c. 404 Ma; Bladen For- orthogneiss-dominated with minor metaclastic and mation of the Maya mountains, Martens et al. rare metavolcanic rocks (Fig. 2c). The Chuacu´s– 2010) and is intruded by granitoids as young as Rabinal boundary is a Cretaceous shear zone but c. 404 Ma (Maya mountains, Steiner & Walker this may represent reactivation of an older structure 1996). Ortega-Obrego´n et al. (2008) suggested the (e.g. Ortega-Gutie´rrez et al. 2007). Both complexes presence of pre-Middle Ordovician basement in have inherited Proterozoic crust (U–Pb zircon). the Rabinal area of central Guatemala (San N. AMERICA–CARIBBEAN PLATE BOUNDARY 273

Fig. 16. Tectonic–palaeogeographic evolution models for the North American–Caribbean Plate boundary and Palaeozoic–Cenozoic correlations between southern Mexico and the Chortı´s Block. (a) Overview of our new and published geochronology from the Maya Block, the Chortı´s Block, and southern Mexico. Data are referenced in the captions of Fig. 1 and in the text.

Gabriel unit intruded by the Rabinal granitoids hinterland and was subducted into the mantle at c. 462 Ma); Solari et al. (2009) extended the likely in the Devonian. San Gabriel unit to western Guatemala. Our infer- Next, we discuss possible correlations of the ence that intrusions in the Rabinal area span Rabinal and Chuacu´s complexes with the Acatla´n 482–396 Ma suggests similar prolonged volcano- complex of the Mixteca terrane of southern clastic deposition and intrusion for the Rabinal Mexico (Fig. 1a & c). The Acatla´n complex com- complex and firmly ties this complex to the prises a nappe stack containing Middle Proterozoic Maya Block. All areas of the Maya Block exposing to Palaeozoic units imbricated during the Palaeo- Ordovician–Early Devonian granitoids are cove- zoic. To date, seven nappes containing distinctive red by the Santa Rosa Group; characteristically stratigraphy have been discriminated (Fig. 2i; these sediments contain Early Palaeozoic detrital Yan˜ez et al. 1991; Ortega-Gutie´rrez et al. 1999; zircons. Tectono-stratigraphically, the Rabinal com- Talavera-Mendoza et al. 2005; Keppie et al. 2006; plex probably represents continental crust over- Vega-Granillo et al. 2007). The stack shares a printed by an arc to back-arc setting (see also c. 325 Ma regional metamorphism. The hanging geochemical data reviewed in Ortega-Obrego´n Esperanza suite is dominated by Early Silurian et al. 2008). Detrital zircon geochronology ties the (442 + 5 to 440+14 Ma) mega-crystic, K-feldspar Maya Block to the South American side of the augen gneiss and contains metasediments and Iapetus–Rheic oceans (e.g. Weber et al. 2008 and metabasalts transformed to eclogite. The entire references therein). The Chuacu´s complex may rep- suite records peak PT at 730–830 8C and 1.5– resent a marginal basin that received detritus from 1.7 GPa attained at c. 430–420 Ma. Partial over- the Ordovician–Silurian arc, and Pan-African printing occurred at 640–690 8C and 1.0–1.4 GPa (c. 0.5–0.7 Ga), and Grenvillian (c. 0.9–1.2 Ga) prior to c. 375 Ma. The Tecolapa suite consists of 274 L. RATSCHBACHER ET AL.

Fig. 16. (Continued)(b) Permo-Triassic ages of southern Mexico and northern Central America. Ages in italics refer to detrital zircons. (c) Palaeogeographic model for the Permian modified from Weber et al. (2006) and Vega-Granillo et al. (2007) emphasizing arc magmatism along the margins of Laurentia and Gondwana and a common position of the Acatla´n complex and the Maya Block in the inner orogenic zone of the Acadian orogen. (d) Jurassic ages along the northern Caribbean Plate boundary. Subdivision of the Chortı´s Block follows Rogers et al. (2007c). Mineral abbreviations as in Fig. 1. (e) Middle to Late Jurassic tectonics of the Caribbean emphasizing rifting of the Chortı´s Block from southern Mexico and its connection with the Arteaga complex of the Zihuatanejo terrane of southwestern Mexico. N. AMERICA–CARIBBEAN PLATE BOUNDARY 275

Fig. 16. (Continued)(f) Early Cretaceous ages of southern Mexico and northern Central America. (g) Late Early Cretaceous tectonics of the Caribbean underlining emplacement of the South El Tambor complex rocks onto the Chortı´s Block and the existence of continuous arcs off southwestern and southern Mexico: Chortı´s–Xolapa–Taxco–Viejo; South El Tambor complex–Teloloapan; Arteaga (Zihuatanejo)–Sanarate (Chortı´s).

Middle Proterozoic mega-crystic granitoids and is the most widespread unit of the Acatla´n complex tonalitic gneisses and Early Ordovician leucogra- and consists of monotonous, greenschist-facies nites (478–471 Ma). The greenschist-facies El clastic rocks that are younger than c. 410 Ma Rodeo suite contains Cambrian–Early Ordovician and basalts. This passive margin sequence was clastics and basalts of continental-rift affinity and imbricated with the other nappes before the Early is intruded by Ordovician (476–471 Ma), Devonian Carboniferous (Figs 2i & 16a, Acatla´n). (371 + 34 Ma), and Early Permian (287 + 2 Ma) We see several gross similarities between the granitoids. The Xayacatla´n suite comprises eclogite- Acatla´n-complex suites and the Rabinal and facies, mafic and pelitic rocks and is intruded by Chuacu´s complexes (Fig. 2c, d & i): (1) the upper Ordovician leucogranites (478+5 to 461 + 9 Ma), nappe stack, from the Xayacatla´n suite at the post-dating high-P metamorphism. Peak PT con- bottom to Esperanza suite at the top, contains exclu- ditions were at 490–610 8C and 1.2–1.3 GPa, sively Proterozoic detrital zircons, the lower stack reached during Early Ordovician (c. 490–477 Ma). also Ordovician and younger zircons (Magdalena, The suite re-equilibrated in greenschist facies 275 Ma; Cosoltepec, 410 Ma; Ixcamilpa, during the Mississippian (c. 332–318 Ma). The 477 Ma). The Maya-Block rocks (Baldy unit, Ixcamilpa suite consists of greenschists and blue- Maya Mountains; Jocote complex, southeastern schists (200–580 8C, 0.65–0.9 GPa) intercalated Chiapas), including the Rabinal complex (San with metaclastic rocks that are younger than Gabriel unit, Guatemala), contain exclusively Pro- c. 470 Ma. The high-P event is loosely constrained terozoic detrital zircons as well, the Chuacu´s between 450 and 330 Ma; the absence of Silurian– complex has a major component of Early Ordovi- Devonian detrital zircons and the existence of a cian–Early Devonian detrital zircons (480– major magmatic event of Silurian age within 402 Ma; c. 436 Ma) besides Proterozoic ages. (2) several units of the Acatla´n complex, which may Although elucidating the Proterozoic history of be genetically linked to subduction, suggest that northern Central America is beyond the topic of high-P metamorphism was Late Ordovician–Early this paper, we note that the major Proterozoic detri- Devonian. The greenschist-facies Cosoltepec suite tal zircon age clusters in both the upper stack suites 276 L. RATSCHBACHER ET AL.

Fig. 16. (Continued)(h) Block diagrams illustrating north–south swath across the Motagua suture zone at c. 70 Ma. Left: the southern Maya Block is imbricated by sinistral transpression. Centre: truncation of the continental margin by sinistral transpression. Roll-back and break-off of the North El Tambor complex slab may have caused magmatism in the southern Maya Block. Right: hanging wall arc magmatism terminated contemporaneously to slightly after (reflecting transfer to the Bahamas continental margin) their collision with the southern Maya and southern Chortı´s Blocks (see text for further discussion). Hanging wall arc diagram modified from the Cordillera Central, Hispaniola, situation (Escuder-Viruete et al. 2006a). (i) Geochronology of the footwall and hanging wall of the Motagua suture zone. Left: ages of imbrication and associated metamorphism in the southern Maya Block (this paper). Centre: ages in the North El Tambor complex (see references in the text), and the subduction–accretion complexes in Cuba (Garcia-Casco et al. 2002, 2007; Schneider et al. 2004; Krebs et al. 2007). Right: ages from the Greater Antilles arc, Cuba, and Cordillera Central arc, Hispaniola (Grafe et al. 2001; Hall et al. 2004; Stanek et al. 2005; Kesler et al. 2005; Rojas-Agramonte et al. 2006; Escuder Viruete et al. 2006b). (j) Subduction–accretion and arc complexes of Cuba and Hispaniola and major age groups emphasizing the evolution of the hanging wall of the Motagua suture zone. (k) Late Cretaceous tectonics of the northern Caribbean highlighting the two Late Cretaceous suture zones on the Chortı´s Block. of the Acatla´n complex and the Maya Block/ Middle Devonian intrusions, whereas the lower Rabinal complex are Grenvillian (c. 0.95– stack suites (Cosoltepec to Ixcamilpa) received the 1.25 Ga) and South American (c. 1.3–1.8 Ga; erosional detritus of these granitoids. Similarly, Rondoˆnia–San Ignacio and Rı´o Negro–Juruena the Maya Block/Rabinal complex carries Early provinces; see Weber et al. 2008 for locations; Palaeozoic magmatism, whereas the Chuacu´s Fig. 2c & i). (3) All upper stack suites (Xayacatla´n complex collected its detritus. Items (1)–(3) link to Esperanza) have massive Early Ordovician to the Maya Block/Rabinal complex to the Xayacatla´n N. AMERICA–CARIBBEAN PLATE BOUNDARY 277

Fig. 16. (Continued)(l) Late Cretaceous and Early Cenozoic ages along the northern Caribbean Plate boundary bringing the Chortı´s Block to its pre-c. 40 Ma position off southern Mexico. Geochronological data compiled from Herrmann et al. (1994), Mora´n-Zenteno et al. (1999), Ducea et al. (2004a), Cerca et al. (2007) and references therein. Deformation zones (numbers 1–9) and their ages are discussed in the text. Convergence direction at c. 45 Ma is from Pindell et al. (1988). (m) Middle Miocene ages along the northern Caribbean Plate boundary. Geochronology is from this paper, Mora´n-Zenteno et al. (1999), and Jordan et al. (2007) and references therein. Deformation zones (numbers 1–5) and their ages are discussed in the text. 278 L. RATSCHBACHER ET AL. to Esperanza stack, and the Chuacu´s complex to the complex that is rich in mafic rocks best correlates Cosoltepec and Ixcamilpa stack of the Acatla´n with the Ixcamilpa suite. These units formed a mar- complex. In both areas, southern Mexico and Guate- ginal basin that was transferred into a Silurian– mala, Permian, Triassic and Jurassic magmatism Carboniferous subduction–accretion complex; the stitches the various units together (Figs 1c, 2c, d & Chuacu´s complex may have represented the lead- i). In contrast, eclogite-facies metamorphism is con- ing edge of this subduction complex. The clastics- fined to the upper stack in the Acatla´n complex, and dominated Chuacu´s complex may correspond to the Chuacu´s complex in Guatemala; the lower to the Cosoltepec suite of the Acatla´n complex. stack in Mexico shows blueschist-facies meta- (3) Although data from the Chortı´s Block are still morphism only. As outlined above, upper stack sparse, our U–Pb analysis identified c. 400 Ma crys- high-P metamorphism is Ordovician (Xayacatla´n tallization ages in the basement rocks, suggesting suite) and Silurian (Esperanza suite), and the that this block also experienced Early Devonian lower stack (blueschist-facies) high-P metamorph- magmatism, identical to the youngest ages from ism is 450–330 Ma (best guess of Vega-Granillo the Maya Block/Rabinal complex. et al. 2007 is 458–420 Ma). Our loose constrain We propose that the southern Maya Block/ for the Chuacu´s complex eclogite-facies event is Rabinal complex and the Chuacu´s complex were 440–302 Ma (see above). These age constrains tie parts of the Taconic–Acadian orogen. The Chua- the Chuacu´s complex high-P event either to the cu´s complex due to its high-P metamorphism and Esperanza or the Ixcamilpa suite; disparate litho- Rabinal complex due to its extensive Ordovician– stratigraphy and age of metaclastic sequence make Silurian magmatism were located together with the Esperanza suite–Chuacu´s complex correlation the Acatla´n complex within the orogenic interior unlikely (Fig. 2c, d & i). along the northern margin of South America. It is Can we go further with the correlations? (1) The likely that the Chortı´s Block also belonged to this most obvious correlation, based on lithostratigraphy orogen (Fig. 16c). Subduction produced an Early and type and age of magmatism, compares the Ordovician arc magmatism on the Tecolapa and El Rabinal complex with the Esperanza suite of the Rodeo suites, which was followed by high-P meta- Acatla´n complex. In this scenario, at least the pre- morphism and accretion of the Xayacatla´n suite Silurian parts of the Rabinal complex should to the Tecolapa and El Rodeo suites. Subduction contain high-P relicts. However, placing the continues beneath the Xayacatla´n producing the Rabinal complex (and the Maya Block) further post-high pressure granitoids within the Xayacatla´n into the hinterland of the Ordovician orogeny than suite. Subsequently the Maya Block/Rabinal com- the Esperanza suite would allow variable meta- plex and the Esperanza suite were accreted and morphic overprint. (2) We predict that the the Esperanza suite involved in continental sub- Chuacu´s complex will see separation into several duction. Subduction again migrated outward, pro- units. For example, several samples of Chuacu´s- ducing the Ixcamilpa and Chuacu´s complexes and complex orthogneiss (data not shown in this involved them in subduction. In the Carboniferous paper) yielded Proterozoic ages (0.98–1.2 Ga) that Laurentia and Gondwana collided producing the are not or are little rejuvenated by Palaeozoic– nappe stack of the Acatla´n complex and Chuacu´s Mesozoic events; it is possible that these gneisses complex. define a unit that corresponds to the Tecolapa suite of the Acatla´n complex. If our correlations are Permo-Triassic arc on Maya and correct, where are the El Rodeo (?arc and continen- Chortı´s Blocks tal rift) and Xayacatla´n (oceanic basin) suites of southern Mexico in Guatemala? First, these suites Gneiss of the Rabinal complex of western may represent relative small-scale, regionally con- Guatemala was migmatized at c. 218 Ma and inher- fined successions (as several units of the Cretaceous ited c. 268 Ma zircon; thus the high-grade rocks Mixteca–terrane arc succession; see below). that accompany the low-grade metasedimentary Second, only few parts of the Chuacu´s complex rocks of the San Gabriel unit are probably a result and the Maya Block/Rabinal complex have been of local Triassic (and Jurassic magmatism, see studied in detail. The abundance of mafic rocks in below) migmatization. In central Guatemala, the some parts of Chuacu´s complex and the strong Chuacu´s-complex orthogneiss, intruding high-P variability of lithology within it (compare, for paragneiss and garnet amphibolite (relict eclogite), example, the amphibolite–paragneiss-dominated records two zircon-crystallization events at 238 El Chol and orthogneiss and migmatite-dominated and 219 Ma. Regional magmatism and metamorph- Rı´o Hondo areas described in the Structure and ism is demonstrated by similar hornblende and Kinematics section) suggests that other suites, white mica cooling ages (238–212 Ma; Fig. 16a & reminiscent to those of the Acatla´n complex, may b). Phyllite interlayered with conglomerate attribu- be discovered. Here we propose that the Chuacu´s ted to the Santa Rosa Group of central Guatemala N. AMERICA–CARIBBEAN PLATE BOUNDARY 279

(Rabinal–Salama´ area) contains detrital zircons at Chuacu´s complexes are c. 30 Ma younger than c. 282, c. 260 and c. 245 Ma. Volcanic pebbles dom- those in the Chiapas Massif and also younger than inate the conglomerate. Triassic deformation in the most Permo-Triassic ages in southern Mexico. We Rabinal and Chuacu´s complex shows high-T flow speculate that the Chuacu´s complex occupied, with north(east)–south(west) shortening and both together with the Maya Mountains of Belize (c. sub-vertical and along-strike, NW–SE extension. 230 Ma reheating), an easterly position along the In the Chortı´s Block, migmatitic orthogneiss crys- main Chiapas–Chortı´s arc (Fig. 16c). The younger tallized at c. 273 Ma and underwent high-T meta- ages may reflect an episode of flat-slab subduction morphism at c. 245 Ma. Inherited c. 241 Ma that shortened the main arc (Weber et al. 2006) zircon in Jurassic orthogneiss and biotite cooling and caused a younger episode of magmatism and ages support regional Triassic magmatism and metamorphism in its hinterland. Shortening metamorphism (Fig. 16a & b). directions are similar in the Rabinal and Chuacu´s Permo-Triassic granitoids are documented along complexes and the Chiapas Massif. Associated the length of Mexico and probably extend into with the shortening there is probably a dextral- northwestern South America (e.g. Centeno-Garcı´a transpressional component, widespread in southern et al. 1993). Their apparent linear arrangement prob- Mexico (e.g. Malone et al. 2002), which is consist- ably traces an active continental margin that origi- ent with tangential stretching in the Triassic nated after final amalgamation of Pangaea along Chuacu´s complex. Finally, the conglomerate clasts the Ouachita–Marathon suture at c. 290 Ma that are predominantly sourced from intermediate (Fig. 16c; Torres et al. 1999). Isotopic ages of volcanic rocks, and c. 245, c. 260 and c. 280 Ma det- these granitoids span 287–210 Ma with reliable rital zircons within phyllite demonstrate that parts of U–Pb zircon ages at 287–270 Ma (Figs 1c & 16a, the clastic cover sequence of the Rabinal complex in Ttoletec/Magdalena/Oaxaca, 16b; Yan˜ez et al. central Guatemala is not Lower Carboniferous– 1991; Solari et al. 2001; Elı´as-Herrera & Ortega- Permian Santa Rosa Group. We suggest that a part Gutie´rrez 2002; Ducea et al. 2004a). In contrast to of this Group comprises either a new unit or is these mostly undeformed granitoids, the Chiapas part of the undated basal conglomeratic sequence Massif batholith rocks (Fig. 16b) are deformed and of the Jurassic–Cretaceous Todos Santos Formation metamorphosed (Weber et al. 2006). Rock types (Fig. 2b & d). Detritus and age of detrital zircons include orthogneisses that are intruded by calc- make the Chiapas–Chortı´s-arc Massif the likely alkaline rocks varying from granite to gabbro, source of these molasse-type, basal deposits. metasedimentary rocks (metapelite/psammite, calc- silicate; Sepultura unit) and mafic rocks (hornblende Jurassic rifting and arc magmatism gneiss; Custepec unit). Zircons from orthogneiss yielded a concordia age at 271.9 + 2.7 Ma, and Granite intruded into the Rabinal complex of three samples from migmatitic paragneiss and leuco- western Guatemala at c. 171 Ma (164–177 Ma some yielded 254.0 + 2.3 to 251.8 + 3.8 Ma ages zircons; Fig. 16a). Most of the Maya Block was (Fig. 16a, Chiapas). The latter ages date the most probably emergent during the Triassic–Jurassic prominent event in the Chiapas Massif – partial with the likely Middle Jurassic–Early Cretaceous anatexis in pelite/psammite and, locally, ortho- Todos Santos Formation reflecting rifting during gneiss. High-grade metamorphism is contempora- Jurassic separation of North and South America neous with c. north–south shortening (Weber et al. (Fig. 2b; e.g. Donnelly et al. 1990). Granites also 2006). intruded the northern Chortı´s Block during the Our new ages indicate that during the Permo- Middle Jurassic (c. 159 and 166 Ma, zircons at Triassic the Rabinal and Chuacu´s complexes and 140–191 Ma; 168 Ma; Fig. 16d). The Middle Juras- the northern Chortı´s Blocks were part of a conti- sic, clastic, partly greenschist-facies Agua Frı´a nuous magmatic arc from northeastern Mexico to Formation (Fig. 2j) thickens from the central to northern South America (e.g. Dickinson & Lawton the eastern Chortı´s Block and is interpreted to be 2001), which initiated after the formation of related to transtension and rifting during the Jurassic Pangaea along the western continental margin of opening between North and South America; e.g. Gondwana (Fig. 16c). Sample 5H-9a zircons of mig- the Guayape fault zone may have originated as a matitic orthogneiss in the north-central Chortı´s Jurassic normal fault (Fig. 16d; e.g. James 2006; Block have within error the same intrusion and high- Gordon 1993). The rocks of the Agua Frı´a For- grade metamorphic ages as the Chiapas Massif rocks mation were deformed during the Late Jurassic (Fig. 16a). This identical geological history suggests (Viland et al. 1996). a direct correlation between the Chiapas Massif and Early–Middle Jurassic magmatism–metamor- the northern Chortı´s Block along the Permo-Triassic phism affected the Chazumba and Cosoltepec arc prior to the opening of the Gulf of Mexico units of the Acatla´n complex and produced the Mag- (Fig. 16b & c). The ages in the Rabinal and dalena migmatites of southern Mexico (Figs 2i & 280 L. RATSCHBACHER ET AL.

16a, Totoletec/Magdalena/Oaxaca; e.g. Talavera- eclogite formation in the South El Tambor Mendoza et al. 2005; our new c. 162 Ma cooling complex and is probably pre-Late Jurassic, the prob- age); inferred metamorphic conditions were low-P/ able age of the oldest radiolarians in the South El high-T (c. 730 8Catc. 0.55 GPa; Keppie et al. Tambor cherts (Chiari et al. 2006). Rifting affected 2004). Anatectic granitic dykes of the San Miguel also the interior of the Chortı´s Block (Agua Frı´a For- unit (175–172 Ma; e.g. Yan˜ez et al. 1991) cut the mation; Fig. 16d). These rift structures trend c.NE former units. These events have been interpreted as (present coordinates), sub-parallel to the northwes- the result of a plume breaking up Pangaea, opening tern and southeastern margins of the Chortı´s Block of the Gulf of Mexico and Triassic–Middle Jurassic and the Guayape fault. This rifting is distinctly magmatic arc activity (e.g. Keppie et al. 2004). different from the WNW-trending Mid-Cretaceous Approximately at 165 Ma, NNW-trending, dextral intra-arc rifting (Rogers et al. 2007a; see below). transtension shear, kinematically compatible with The Late Jurassic regional metamorphism (165– opening of the Gulf of Mexico, occurred along 150 Ma; Fig. 16a) and deformation (Viland et al. the Oaxaca fault (Alaniz-Alvarez et al. 1996), and 1996) may be rift and strike–slip related. NW-trending, Jurassic–earliest Cretaceous dextral Several likely connections exist between the transpression took place within the northern Chortı´s Block, the Xolapa complex, and the Guer- Oaxaca complex (Solari et al. 2004). The Xolapa rero composite terrane. The Sanarate complex complex (Fig. 2l) also shows 165–158 Ma (Fig. 2e), the westernmost basement outcrop of the magmatic–metamorphic ages (Fig. 16a, Xolapa; Chortı´s Block south of the MSZ, was included Herrmann et al. 1994; Ducea et al. 2004a). The with the South El Tambor complex because of its Zihuatanejo and Huetamo mature arc sections of metamorphism (lower metamorphic grade than the Guerrero terrane rest unconformably on the that of the Las Ovejas unit) and rock assemblage Arteaga–Las Ollas basement, which comprises (dominated by mafic and clastic rocks; e.g. ocean-floor–continental-slope–subduction com- Tsujimori et al. 2004, 2006); however, the Jurassic plexes that were deposited in the Late Triassic– age for the metamorphism precludes such a corre- Early Jurassic, intruded by 187 + 7 Ma granite, lation. Here we suggest a correlation of this com- and deformed in the Middle Jurassic (Fig. 2m). plex with the Arteaga–Las Ollas units (Figs 2m & Major late Middle–Late Proterozoic–Phanerozoic 16e) based on lithology, likely proximity to the detrital zircon populations from these complexes Jurassic continental margin of Mexico and distinct are c. 247, c. 375, c. 475, c. 575 and c. 975 Ma. The Jurassic metamorphism (Fig. 2e & m). The Sanarate overlying Cretaceous arc has major zircon popu- complex probably also includes parts of the Cretac- lations at c. 160, c. 258, c. 578 and c. 983 Ma eous Zihuatanejo and Huetamo volcanic and clastic (Fig. 2m; Talavera-Mendoza et al. 2007). units (Fig. 2m). Both the Arteaga–Las Ollas units Our tectonic model (Fig. 16e) integrating the and the overlying Cretaceous arc contain zircons Jurassic evolution of the Maya and Chortı´s Blocks that may have originated from the Chortı´s Block. with the magmatic–tectonic evolution of southern The c. 247, c. 257 and 475 Ma clusters, particularly Mexico is modified from Mann (2007). Onset of in the 85 Ma arc rocks, are difficult to derive rifting in the Gulf of Mexico and between North from North America because of intervening spread- and South America reduced Triassic (see above) to ing ridges and subduction zones (Figs 2m & 16e, see Jurassic volcanic eruption rate in southern Mexico below). The pre-migmatitic basement of the Xolapa and the Maya and the Chortı´s Blocks. Late intrusion complex, containing evidence for a Grenville into active dextral fault zones (Mexico) may explain crustal component (Herrmann et al. 1994) and the concentration of magmatism at 175–166 Ma Permian magmatism (see above) in an alternation (Fig. 16d). Rifting of the Chortı´s Block from of metagreywacke and metabasic rocks, probably southern Mexico probably initiated along the correlate to the Permian and Jurassic section of the arc and developed either as back-arc extension Mixteca terrane (Fig. 2m, Olinala´ unit). Finally, above the proto-Pacific subduction zone, a failed connecting the Jurassic magmatic–metamorphic arm of the Proto-Caribbean rift, or a combination. rocks of southern Mexico and the northern Chortı´s Additionally, the Arteaga–Las Ollas basement Block implies 1000 km post-Jurassic offset. units rifted from southern Mexico, as they likely share the Late Triassic–Jurassic history of southern Mexico (Fig. 2m; e.g. Talavera-Mendoza et al. Early Cretaceous subduction, top-to-south 2007). Our model requires rifting the Chortı´s emplacement of the South El Tambor complex, Block from southern Mexico and creation of and proto-Pacific arc formation oceanic crust, because MORB oceanic crust and me´lange of the South El Tambor complex is Spreading between the Chortı´s Block and southern present between the Chortı´s Block and southern Mexico must have reverted to subduction before Mexico. This must predate 132 Ma, the age of 132 Ma (see above). The c. 120 Ma phengite ages N. AMERICA–CARIBBEAN PLATE BOUNDARY 281 of the South El Tambor complex blueschists are Chortı´s Block (120 Ma, see above). Laterally, interpreted as marking fluid infiltration during retro- Xolapa-complex magmatism is correlated with gression from eclogite-facies conditions (see above) that in the Taxco–Viejo unit of the Mixteca and could date exhumation during emplacement of terrane, also an arc on continental crust (e.g. the South El Tambor complex me´lange onto the Centeno-Garcı´a et al. 1993) and dated at Chortı´s Block (Fig. 8b). The El Tambor rocks are 130.0 + 2.6 and 131.0 + 0.85 Ma (Campa & emplaced top-to-south (present coordinates). Iridono 2004). This correlation is supported by Aptian–Albian calc-alkaline, arc volcanic rocks of the, within error, identical ages of the Chapolapa the Manto Formation and volcanoclastic rocks of Formation meta-andesite/rhyolite (structural top the Tayaco Formation, sandwiched between of the Xolapa complex) and the volcanic flows of shallow water carbonates, were deposited in a the Taxco–Viejo unit (Fig. 2l & m). Magmatism zone of intra-arc extension in the central Chortı´s in the Teloloapan intra-oceanic arc ended at c. Block, the Agua Blanca rift (Rogers et al. 2007a; 120 Ma. Magmatism in the oceanic arc and Figs 2j & 16f). The strike of the rift is WNW back-arc region of the Arcelia units, outboard (present coordinates), oblique to the northern (southwesterly) of the Teloloapan unit continued margin of the Chortı´s Block and the Jurassic exten- into the Cenomanian (Fig. 2m). The youngest detri- sional structures (see above). Cretaceous (peaks at c. tal zircon fractions in the clastic rocks above the vol- 128, c. 118 and c. 81 Ma; Figs 16a, f & 5h) mag- canic succession in the Teloloapan and Taxco– matic rocks are widespread on the Chortı´s Block Viejo units, comprising magmatic zircons from (Fig. 16f) and probably mark continuous arc for- local sources, are c. 124 and c. 132 Ma, respectively mation above the proto-Pacific subduction zone (Talavera-Mendoza et al. 2007), thus contempora- (Fig. 16g). The Xolapa complex, the largest plutonic neous with the Xolapa-complex zircons. Again, and metamorphic mid-crustal basement unit in these ages indicate that magmatism terminated at southern Mexico, preserves abundant migmatites that time and suggest a close spatial relationship that were produced during a single metamorphic of the Taxco–Viejo–Xolapa–Teloloapan arcs event with peak prograde conditions at 830– (Figs 2m & 16g). The NE-dipping subduction 900 8C and 0.63–0.95 GPa. This occurred within a zone of the South El Tambor complex may have continental magmatic arc and overprinted a pre- ended at a transform fault at the eastern end of the migmatitic metamorphic assemblage (Corona- Maya Block (Fig. 16g). Cha´vez et al. 2006). Deformation, likely (north-) The Teloloapan, Taxco–Viejo, and Olinala´ units northeast–(south-)southwest shortening (Corona- and the Xolapa complex are covered by Albian– Cha´vez et al. 2006), associated with low-P Lower Cenomanian Morelos shelf limestones metamorphism and migmatization predates 129 + (Fig. 2l & m). These limestones may correspond 0.5 Ma in the Tierra Colorada area of the western to the massive La Virgen limestone atop the Xolapa complex (Solari et al. 2007). The migma- South El Tambor complex (Fig. 2k). Carbonate tites have been directly dated at 131.8 + 2.2 Ma passive margin strata (Coba´n–Campur Formations; (Herrmann et al. 1994) and 130–140 Ma (Ducea Fig. 2b) were deposited on the Maya Block. The et al. 2004a). The structural top of the migmatized post-suturing deposits on the Chortı´s Block are sequences is the 126–133 Ma Chapolapa Formation probably the red beds and foreland-basin deposits (Campa & Iriondo 2004; Herna´ndez-Trevin˜o et al. of the Valle de Angeles and Gualaco Formations. 2004) that mostly comprises andesite/rhyolite; its The latter is typically intercalated with Upper hanging wall, Albian–Cenomanian Morelos For- Cretaceous volcanic rocks indicating establishment mation shelf limestone, is again in tectonic contact of the Sierra Madre Occidental arc in Late Cretac- (Figs 2l & 16a, Xolapa; e.g. Riller et al. 1992; eous, also on the Chortı´s Block (see below). Charac- Solari et al. 2007). teristically, proto-Pacific subduction continued Rogers et al. (2007a) suggested several features beneath the Chortı´s Block and the Zihuatanejo and piercing lines that tie the Chortı´s Block to unit far into the Late Cretaceous (Fig. 2j & m). In southern Mexico; here we elaborate on the tectonic our interpretation the Agua Blanca rift and the model explaining the correlation (Fig. 16g). We Manto volcanic rock intercalations are related to suggest that the 141–126 Ma migmatization– proto-Pacific subduction and intra-arc extension. magmatism of the Xolapa complex represents arc We attribute the inversion of the Agua Blanca magmatism related to subduction of the South El intra-arc rift to the accretion of the Zihuatanejo– Tambor complex lithosphere. Characteristically, Huetamo units–Sanarate complex to southern no Cretaceous ages younger than c. 126 Ma have Mexico and the western Chortı´s Block. Top-to-east been found in Xolapa complex, so the end of arc shear in the greenschist facies, metabasaltic/clastic magmatism in the South El Tambor subduction Sanarate complex of the western Chortı´s Block zone is likely to be contemporaneous with emplace- reflects this accretion of the Guerrero terrane/ ment of the South El Tambor Complex onto the Sanarate complex. 282 L. RATSCHBACHER ET AL.

Why were so many individual arcs and subduc- into the footwall of the Siuna terrane of the southern tion–accretion complexes developed along the Chortı´s Block and the Nicaragua Rise and to the east southwestern margin of Mexico (the Guerrero com- into the subduction–accretion and arc complexes of posite terrane) and the Chortı´s Block (the Sanarate Jamaica, Hispaniola (Dominican Republic) and complex) during the Early and Late Cretaceous Cuba (Fig. 16j & k). In comparison with the North (e.g. Talavera-Mendoza et al. 2007)? If rifting El Tambor complex (Fig. 2a), the Siuna me´lange between the Chortı´s Block–Arteaga unit and (Fig. 2f) contains similar lithology (ultramafic southern Mexico was mostly related to spreading rocks, radiolarite, meta-andesite, volcanic arenites) between the two Americas and formation of the of proven Late Jurassic age, high-P phengites at c. Gulf of Mexico (Fig. 16e), spreading was probably 139 Ma, blueschist and eclogite, and Aptian– north–south with c. north-trending transform Albian turbiditic sequence. These rocks are overlain faults. Transformation of this rift to convergence by the A´ tima Formation and intruded by diorite (c. along the c. NNW-trending proto-Pacific subduc- 60 Ma) and andesite dykes (c. 76 Ma; Fig. 2f; tion zone may have allowed activation of some Flores et al. 2007; Venable 1994). The poorly transform faults as subduction zones. dated (post-80 Ma) Colo´n fold belt of the eastern Chortı´s Block and the Nicaragua Rise probably reflects imbrication along the footwall of the Siuna Late Cretaceous North El Tambor complex terrane (Rogers et al. 2007b). The Siuna terrane subduction and Maya Block–Caribbean probably extends westward into the ocean igneous arc collision complexes of the Santa Elena and Nicoya peninsu- las of Costa Rica; these complexes contain c. Our structural, petrological, and geochronological 165–83 Ma pillow lavas, sills and flows, radiolarian data demonstrate sinistral transpressive collision chert, basalt breccias, and gabbro and plagiogranite. along the southern margin of the Maya Block They are unconformably capped by 75 Ma reef during Late Cretaceous. Collision transferred the limestones (Figs 2g & 16i; e.g. Hoernle & Hauff southern Maya Block into a nappe stack with the 2005). The Cuban subduction–accretion complex North El Tambor ophiolitic me´lange at top. Charac- contains c. 148–65 Ma high-P rocks (major peaks teristically, deformation involved north–south at c. 116 and c. 70 Ma; Fig. 16i & j; Krebs et al. shortening and east–west lengthening. Sinistral 2007, Schneider et al. 2004). The Greater Antilles transpression was partitioned on all scales between arc of Cuba was active at 133–66 Ma (Fig. 16i & oblique stretching along the foliation and sinistral j; major peaks at c. 72 and c. 80 Ma; Grafe et al. strike–slip along steeply dipping shear zones 2001; Hall et al. 2004; Rojas-Agramonte et al. (Fig. 16h). To a first order, deformation is identical 2005; Stanek et al. 2005), and the Cordillera in the North El Tambor complex, Chuacu´s complex, Central arc (Dominican Republic) spanned and the northern foreland fold–thrust belt; 450– c. 117–74 Ma (Fig. 16h–j; major peaks at c. 75, 550 8C and 0.7–0.8 GPa regional Cretaceous c. 84 and c. 116 Ma; Kesler et al. 2005; Escuder reheating within the Chuacu´s complex, associated Viruete et al. 2006b). Thus these subduction– with local magmatism, is synkinematic. Penetration accretion complexes and arcs cover time ranges of deformation and overburden (PT conditions) comparable to that of the North El Tambor increases from west to east along the collision complex, with subduction and arc formation termi- zone. The relatively high-P/low-T metamorphism nated c. 10 Ma after collision along the southern and nearly coeval blueschist- (North El Tambor Maya Block. This indicates the time span required complex) and amphibolite-facies metamorphism to transfer these subduction–accretion and arc (Chuacu´s complex) suggest a cold geotherm and complexes to the Bahamas-platform collision zone steep continental subduction (Fig. 16h). Southern when the southern Maya Block acted as a sinistral Maya-Block imbrication is dated at 75–65 Ma transform fault zone. Accordingly, termination (Fig. 16i) and cooling to upper crustal levels of pronounced magmatism in the Santa Elena– (c. 100 8C) was attained by Early Cenozoic (AFT Nicoya subduction-accretion complexes is approxi- results at high elevation, see above). The Jurassic– mately 10 Ma earlier (Fig. 16i) and reflects their Early Cretaceous North El Tambor oceanic and earlier collision with the southwestern Chortı´s immature arc rocks incorporated c. 131 Ma eclogite Block. Based on the termination of magmatism at into an accretionary wedge that reached crustal 74 Ma within the Cordillera Central arc and its levels by Late Cretaceous (c. 76 Ma). well-developed sinistral transpressive deformation What collided with the southern Maya Block in (Escuder Viruete et al. 2006a, b), we speculate the latest Cretaceous? We follow and broaden the this arc was the immediate hanging wall of the model of Mann (2007) and Rogers et al. (2007b) collision zone along the southern Maya Block. and suggest that the Maya Block–North El What was the origin of the 75 Ma magmatism Tambor complex suture extends toward the west (pegmatites) in the southern Maya Block? Causes N. AMERICA–CARIBBEAN PLATE BOUNDARY 283 might include Pacific-arc magmatism, local anatexis 1995). This magmatic arc continued into the due to crustal thickening, or break-off of the Siuna central Chortı´s Block. Lining up the northwest terrane–North El Tambor complex–Caribbean arc trend of these rocks in Mexico (the arc continues slab. We favour the latter (Fig. 16h), as entry of NW of the area shown in Fig. 16l) and on the the Nicoya–Siuna–North El Tambor–Caribbean Chortı´s Block yields 1100 km sinistral displace- slab into the Proto-Caribbean sea must have indu- ment and 408 anticlockwise rotation of the ced its rapid anticlockwise rotation, rapid along- Chortı´s Block (Figs 1c & 16l). Based on one data strike length reduction, and thus slab fracturing point from the central Chortı´s Block, the palaeo- and roll-back. In contrast, contemporaneous Pacific magnetically determined anticlockwise post-60 Ma arc subduction occurred far to the west on the rotation is c.328 (Gose 1985). During the Oligo- Chortı´s Block (Fig. 16k) and the thermal overprint cene, eastward migration of Chortı´s induced a north- due to crustal stacking was probably insufficient eastward migration of the trench (Herrmann et al. and too early in the Chuacu´s complex to cause 1994) and the associated Sierra Madre del Sur arc anatexis (see above). shifted inland and eastward into southeastern Cerca et al. (2007) outlined a mostly east Mexico. The arc trends ESE (Mora´n-Zenteno et al. vergent, partly sinistral transpressive, north- 1999), which is particularly evident on the Chortı´s trending fold-thrust belt along the western Mixteca Block (Fig. 16l). terrane, the Guerrero–Morelos platform that origi- When did the Chortı´s Block break away from nated during the Coniacian (90 Ma; Cenomanian, southern Mexico? In the Chortı´s Block, high-grade 100 Ma, after Talavera-Mendoza et al. 2005) and and high-strain deformation began at 40 Ma was active into the Cenozoic. They related this and we suggested above that there were major east–west shortening to the collision of the Carib- deformation phases at 40–35 Ma and possibly at bean Plate with the Chortı´s Block. Rogers et al. 25–18 Ma. Sinistral transtension in the La Venta– (2007b) and our data show, however, that this col- Tierra Colorada shear zone along the northwestern lision is c. north–south and 76 Ma on the southern margin of the Xolapa complex is low-grade at Chortı´s and Maya Blocks (see above). Probably, c. 300 8C (low-T quartz ductility; Riller et al. the Laramide fold–thrust belt orocline on the 1992) and started at 45 Ma (Solari et al. 2007). Guerrero–Morelos platform, along the Sierra de Local high-strain flow occurred at c. 35 Ma (see Jua´rez (Cerca et al. 2007), and within the southern above). Post-tectonic intrusions are 34 + 2Ma Maya Block (this study) is a composite structure, (Tierra Colorada granite, Herrmann et al. 1994); related to accretion of the most outboard Guerrero- however, brittle, sinistral transtension continued terrane units, Pacific subduction and collision of during cooling of this granite (Riller et al. 1992). the Siuna terrane–North El Tambor complex– The distinctly different exposure level of the con- Caribbean Plate. We suggest that shortening in the tinuous Early Cretaceous Taxco–Viejo (west; up Guerrero–Morelos platform and its continuation to greenschist facies) and Xolapa arc (east; amphi- into the Chortı´s Block (Agua Blanca rift inversion) bolite facies and migmatization) is ascribed to is mainly related to the 80 Ma Arteaga–western major exhumation of the latter. Large-scale sinistral Chortı´s (Sanarate complex) suture and Pacific sub- transtension along the northwestern Xolapa- duction and the Sierra de Jua´rez–southern Maya complex margin certainly contributed to this differ- Block shortening due to the 76 Ma Caribbean ence. Robinson et al. (1990) obtained a c. 40 Ma age Plate collision (Fig. 16k). Structural superpositions for synkinematic tonalite north of Acapulco. Sinis- must have occurred mostly in the Mixteca– tral shear occurred along the northern margin of Oaxaca basement terrane area. the eastern Xolapa complex between 29 and 24 Ma (Chacalapa shear zone, point 4 in Fig. 16l; Late Eocene–Recent eastward displacement Tolson 2005). Other sinistral transtensive defor- of the Chortı´s Block mation zones in southern Mexico outside the Xolapa complex are dated to between c. 37 and c. Most tectonic models place the Chortı´s Block oppo- 27 Ma (points 1 and 3 in Fig. 16l). Corona-Cha´vez site southern Mexico during Late Cretaceous–Early et al. (2006) estimated that about 70% of the Cenozoic (e.g. Mann 2007; see above). Rogers et al. eastern Xolapa complex was affected by discontinu- (2007a, b) summarized features and piercing lines ous mylonitic–cataclastic, post-migmatitic defor- that support this connection. Here we go over mation. We conclude that the migration of the additional evidence and trace the eastward displace- Chortı´s Block initiated 40 Ma and was fully ment of the Chortı´s Block to its present position. active at 40–35 Ma. This is compatible with the Approximately 80–45 Ma magmatic rocks of the chronology of spreading in the Cayman Trough Sierra Madre Occidental crop out along the coast pull-apart basin (see above). of southwestern Mexico and inland up to c.998W Using (U–Th)/He thermochronology, Ducea (Fig. 16l; Mora´n-Zenteno et al. 1999; Schaaf et al. et al. (2004b) estimated average exhumation rates 284 L. RATSCHBACHER ET AL. for the Xolapa complex. A western segment at weakness, the South El Tambor suture. Most impor- Acapulco yielded ages between 26 and 8.4 Ma and tantly, sinistral break away started in an area that a rate of 0.22 mm/annum; an eastern profile at had been an arc at least since Late Cretaceous (see Puerto Escondido gave 17.7–10.4 Ma ages and above) and thus was thermally weakened. This is a 0.18 mm/annum. This slow Neogene exhumation common process in oblique subduction settings. contrasts with petrological data that suggest exhu- Consequently, the largest volume (Mora´n-Zenteno mation from 25 km (migmatites; Corona-Cha´vez et al. 1999) and the concentration in time et al. 2006) and 20–13 km (granitoids; Mora´n- (c. 33 Ma Chortı´s Block, c. 34 Ma Xolapa com- Zenteno et al. 1996) since 34–27 Ma, the age of plex; Fig. 16a) of magmatism and metamorphism the granitoids; therefore, most exhumation must seems to have been controlled by the NW–SE have taken place before the Miocene. Pooling the shear zones associated with transtensional– (U–Th)/He and AFT data of Ducea et al. (2004b) transpressional tectonics at the boundary between and excluding outliers (e.g. ages older than cry- southern Mexico and the Chortı´s Block (Herrmann stallization and those with very large errors) indi- et al. 1994; Fig. 16l). cate that most samples of the western profile Early and Middle Miocene (c. 20–10 Ma) volca- cooled through 70–100 8C or were exhumed from nic rocks are distributed around the Isthmus of c. 4 km at c. 25 Ma. Except for three outliers most Tehuantepec (Fig. 16m). In this region, it is samples of the eastern profile have identical ages evident that volcanism in the inland regions was over 2 km elevation and thus suggest rapid cooling active after cessation of Oligocene plutonism in at c. 15 Ma (Fig. 6d). Using the three plutons for the Xolapa complex (Mora´n-Zenteno et al. 1999). which U–Pb zircon ages are available (Herrmann The earliest manifestations of magmatic activity et al. 1994), we obtained exhumation rates of along the eastern and central Trans-Mexican volca- c. 1.5 mm/annum between 31 and 25 Ma (12.9 km nic belt are as old as Middle Miocene (c. 17 Ma; depth at 31 Ma and c. 4 km at c. 25 Ma) and Ferrari et al. 1994) and thus this belt may form a c. 1.9 mm/annum between 32 and 25 Ma (17.9 km direct continuation of the volcanism around the depth at 32 Ma and c. 4 km at c. 25 Ma) for the Isthmus of Tehuantepec. To the east coeval magma- western section, and 1.2 mm/annum (20.4 km tism occurred along the western part of the Polochic depth at 29 Ma and c. 4 km at c. 15 Ma) for the fault zone in westernmost Guatemala (Fig. 16m), eastern profile. For the western section, Herrmann indicating 100 km offset along the Polochic fault et al. (1994) obtained two c. 47 Ma migmatization since then. This is similar to the c. 130 km offset ages; this yields c. 0.8 mm/annum between 47 and proposed along this fault between 10 and 3 Ma 31 Ma (c. 26 km depth at 47 Ma and 12.9 km (Burkart 1983; Burkart et al. 1987). Still further at 31 Ma). Along the northern rim of the Chortı´s east, the c. 16.9–13.4 Ma ignimbrite province Block our data indicate an exhumation rate of strikes northwest–southeast across the Chortı´s c.1mm/annum between 36 and 9 Ma (c.28km Block (Fig. 16m; e.g. Jordan et al. 2007). Aligning depth at 36 Ma and c. 3 km at 9 Ma). These data the southeastern Mexican and Chortı´s Block volca- indicate that the early stages of Chortı´s-Block nic provinces yields c. 400 km displacement since migration were associated with rapid exhumation c. 15 Ma and c. 300 km offset along the Motagua due to sinistral transtension and erosion. fault (Fig. 16m). Several elements trace the displa- The early stage of distributed sinistral displace- cement during this time (Fig. 16m): tectonic and ment and erosion that affected a broad zone in stratigraphic features NW of the Isthmus of Tehuan- southern Mexico and the northern Chortı´s Block tepec indicate sinistral transtension with c. north also must have affected the southern Maya Block, trending pull-apart basins coeval with volcanism as its eastern extension, the Nicaragua Rise, was (Mora´n-Zenteno et al. 1999); the easternmost part located off the western edge of the Maya Block of the Xolapa complex experienced rapid cooling (Fig. 16k). We interpret our AFT data from the at c. 15 Ma (Fig. 6d); the Chacalapa shear zone southern Chiapas Massif and the Chuacu´s com- was reactivated at c. 15 Ma; the Tonala and Polochic plex, both clustering at c. 30 Ma (Fig. 6a & c), to shear zones were active at 9–8 and 12–5 Ma, record this distributed transtension–transpression– respectively; and most of the AFT ages in the sinistral migration and erosion phase. Chortı´s Block closed at c. 12 Ma (all data from Why did the Chortı´s Block break away at this study). 40 Ma? Following Ratschbacher et al. (1991) and Riller et al. (1992), we propose that the contem- Variations in the Late Cenozoic stress field poraneous change in several plate tectonic boundary conditions triggered break away. During Eocene, North–south rifts that disrupt the ignimbrite pro- shear stresses across the arc were high due to rapid vince in western Honduras and southeastern convergence of relatively young lithosphere under Guatemala commenced after c. 10.5 Ma (Gordon an angle of c.458 to a pre-existing lithospheric & Muehlberger 1994). They are related to plateau N. AMERICA–CARIBBEAN PLATE BOUNDARY 285 uplift following slab break-off underneath Central southern Mexico and arc/subduction complexes America (Rogers et al. 2002). Our 11.2–7.9 Ma of the northwestern Caribbean constrain models AFT data from the western flank of the Sula for the Early Palaeozoic to Recent evolution of the graben support the Late Miocene onset of rifting, North American–Caribbean Plate boundary. although extension structures formed as early as During the Early Palaeozoic, southern Mexico c. 28 Ma, contemporaneous with and kinematically (Acatla´n complex of the Mixteca terrane), the related to sinistral wrenching along the northern southern Maya Block (Rabinal complex), the margin of the Chortı´s Block. Numerical modelling Chuacu´s complex and also the Chortı´s Block were suggests that presently the western edge of the part of the Taconic–Acadian orogen along the Chortı´s Block is pinned against North America, northern margin of South America. The most making the triple junction between the Cocos, obvious correlation, based on lithostratigraphy and North American and Caribbean plates a zone of type and age of magmatism, compares the Rabinal diffuse deformation (A´ lvarez-Go´mez et al. 2008). complex with the Esperanza suite of the Acatla´n Our ‘Late Cenozoic deformation’ structural event complex. In this scenario, at least the pre-Silurian (Figs 10, 13 & 15), reflecting sinistral slip along parts of the Rabinal complex should contain c. east-striking strike–slip faults locally interacting high-P relicts, which are not observed. Placing the with c. north-striking normal faults, is distributed Rabinal complex (and the Maya Block) further across the entire plate boundary from the northern into the hinterland of the Ordovician orogeny than foreland to south of the Jocota´n–Chameleco´n the Esperanza suite allows variable metamorphic fault zone. The transition from older dominant overprint. The Chuacu´s complex best correlates strike–slip to younger, prevailing normal faulting, with the Ixcamilpa suite and the Cosoltepec suite proposed by Gordon & Muehlberger (1994), of the Acatla´n complex. These units formed a mar- is also evident from our data. However, normal ginal basin that was transferred into a Silurian– faulting exemplified by the grabens in southern Carboniferous subduction–accretion complex; the Guatemala and western Honduras (Fig. 15), Chuacu´s complex may have represented the extends northward across the Motagua fault zone leading edge of this subduction complex. into the Cenozoic Subinal Formation (Fig. 13), the After the final amalgamation of Pangaea, an arc North El Tambor rocks on the central Chuacu´s developed along its western margin, causing mag- complex (Fig. 13), and the Rabinal complex and matism and regional amphibolite-facies meta- Cenozoic–Quaternary volcanic rocks of western morphism in southern Mexico (c. 290–250 Ma), Guatemala (Fig. 10). This may indicate that today the Rabinal complex of the southern Maya Block the Polochic fault constitutes the major plate bound- and the Chuacu´s complex (c. 270–215 Ma) and ary fault. This is compatible with termination of the the Chortı´s Block (c. 283–242 Ma). Identical intru- active strike–slip and reverse fault province of sion and metamorphic ages (273 and 254 Ma) in the eastern Chiapas north of the Polochic fault zone Chiapas Massif of the southwestern Maya Block and (Andreani et al. 2008). Furthermore, our data the northern Chortı´s Block suggest a similar pos- support additional salient features of the Late Ceno- ition within the arc. Younger ages in the Rabinal zoic stress distribution: shortening directions trend and Chuacu´s complexes are attributed to their pos- c. NW–SE in western and west-central Guatemala ition in the hinterland of the main arc and a speculat- and in particular in western Guatemala south of ive episode of flat-slab subduction. the Polochic fault (‘Late Cenozoic deformation’; The Jurassic opening of the Gulf of Mexico sep- Figs 10, 13 & 15) and outline the dextral, WNW- arated the Maya Block from North America. Granite trending Jalpatagua distributed fault zone along intruded the Rabinal complex (Maya Block) of the active volcanic arc (Gordon & Muehlberger western Guatemala at 164–177 Ma. Most of 1994; A´ lvarez-Go´mez et al. 2008). the Maya Block was probably emergent during Triassic–Jurassic with the likely Middle Jurassic– Conclusions Early Cretaceous Todos Santos Formation reflecting rifting during separation of North and South New geochronological, petrological, and structural America. Dextral shear along c. NNW trending data constrain the P–T–d–t history of the northern shear zones within southern Mexico is kinemati- Caribbean Plate boundary in southern Mexico, cally related to opening the Gulf of Mexico and central Guatemala and northwestern Honduras. In may account for the concentration of magmatism particular, we investigated which crustal sections at c. 175–166 Ma in southern Mexico. Granites of the Motagua suture zone accommodated the also intruded the northern Chortı´s Block during large-scale sinistral offset along the plate boundary, Middle Jurassic time (c. 168–159 Ma). The and the chronological and kinematic framework of Middle Jurassic Agua Frı´a Formation records this process. Correlations between the Maya and upper crustal transtension and rifting. Rifting of Chortı´s Blocks, tectono-stratigraphic complexes of the Chortı´s Block from southern Mexico probably 286 L. RATSCHBACHER ET AL. initiated within the thermally weakened proto- triggered breakaway of the Chortı´s Block from Pacific arc, possibly as a failed arm of the Proto- southern Mexico at 40 Ma, coeval with initiation Caribbean rift. This rifting separated the northern of the Cayman trough pull-apart basin. Sinistral Chortı´s Block from the Xolapa complex of southern break away started in an area, the Xolapa complex Mexico and also parts of the Guerrero terrane from and the northern Chortı´s Block, that had been an parts of southwestern Mexico that have Jurassic- arc at least since Late Cretaceous and thus was ther- aged basement. The Sanarate complex of the north- mally weakened. The Las Ovejas complex of the western Chortı´s Block correlates with the Arteaga– northern Chortı´s Block forms a westward narrowing Las Ollas units of the Guerrero terrane, sharing a belt of distributed sinistral transtension and rep- dominantly mafic lithology and Jurassic meta- resents the southern part of the high-strain displace- morphism. Rifting produced the Late Jurassic– ment zone along which the Chortı´s Block migrated Early Cretaceous oceanic crust of the South El eastward. Deformation interacted with amphibolite- Tambor complex and terminated before 132 Ma, facies metamorphism, widespread migmatization, the age of eclogite-facies metamorphism in that and local magmatism, illustrated most clearly by complex. small syn- to post-tectonic pegmatites. The close Rifting between Chortı´s Block and southern spatial and temporal interaction between high-strain Mexico reverted to subduction in the Early Cretac- deformation and Cenozoic metamorphism, migma- eous. Arc magmatism and high-T/low-P meta- tization and magmatism is confined to the northern morphism (141–126 Ma) along the Xolapa Chortı´s Block. It appears that arc magmatism pro- complex of southern Mexico is probably related to vided the heat input to localize deformation. Conse- subduction of South El Tambor complex litho- quently, the largest volume and the concentration in sphere, which may have been emplaced southward time (40–20 Ma) of magmatism and metamorphism onto the Chortı´s Block at or after c. 120 Ma. Later- seems to have been controlled by shear zones ally, Xolapa-complex magmatism is correlated with associated with transtensional–transpressional tec- that in the Taxco–Viejo unit of the Mixteca terrane. tonics at the boundary between southern Mexico While collision of the Chortı´s Block with southern and the Chortı´s Block. The Sula graben, and poss- Mexico terminated arc magmatism in the Xolapa ibly other north-trending grabens of southern Guate- complex, proto-Pacific subduction continued on mala and northern Honduras, started in temporal and the Chortı´s Block and the Zihuatanejo unit of the kinematic compatibility with wrenching along the Guerrero terrane to its NW. This unit and the Sana- Las Ovejas complex. The topographically well rate complex of the Chortı´s Block amalgamated to expressed active grabens, however, are neotectonic southern Mexico during Late Cretaceous. features. Also the Jocota´n–Chameleco´n fault Spectacular sinistral transpressive collision zone was part of this deformation system. Break- occurred along the southern margin of the Maya away and early eastward translation of the Chortı´s Block during Late Cretaceous (76 Ma). Collision Block induced exhumation rates of 0.8–1.9 mm/ transferred the southern Maya Block into a nappe annum in the Xolapa complex and the northern stack with the North El Tambor complex at top. Chortı´s Block. Initial eastward translation is also Sinistral transpression was partitioned on all scales recorded by rapid cooling within the southern/ between oblique stretching along the foliation and southwestern Maya Block at c. 30 Ma. Approxi- sinistral strike–slip along steeply dipping shear mately 175, c. 130 and c. 40 Ma magmatic belts zones. To first order, deformation is identical in both in southern Mexico and on the Chortı´s Block the North El Tambor complex, Chuacu´s complex, consistently argue for 1100 km offset since and the Laramide, northern foreland fold–thrust 40 Ma along the northern Caribbean Plate bound- belt. The relatively high-P/low-T metamorphism ary. Post-40 Ma structures and a c. 15 Ma magmatic and nearly coeval blueschist- (North El Tambor belt indicate that c. 700 km of that offset occurred complex) and lower amphibolite-facies metamorph- prior to c. 15 Ma. The Tonala shear zone, bounding ism (Chuacu´s complex) suggest a cold geotherm the Chiapas Massif to the SW, and Polochic shear and steep continental subduction. The Maya zones were major deformation zones at 9–8 and Block–North El Tambor complex suture extends 12–5 Ma, respectively. westward into the footwall of the Santa Elena– North–south rifts in western Honduras and Nicoya complexes, the Siuna terrane of the southern southeastern Guatemala are related to plateau Chortı´s Block, and the Nicaragua Rise and to the east uplift following slab break-off underneath Central into the subduction–accretion and arc complexes of America. Apatite–fission track ages of 11.2– Cuba and Hispaniola. Postulated break-off of the 7.9 Ma from the western flank of the Sula graben North El Tambor complex slab may account for support the Late Miocene onset of rifting. Presently, 75 Ma magmatism in the southern Maya Block. the triple junction between the Cocos, North High shear stress across the Caribbean–North American and Caribbean plates is a zone of American plate boundary during the Eocene diffuse deformation. Normal faulting, exemplified N. AMERICA–CARIBBEAN PLATE BOUNDARY 287 by the grabens in southern Guatemala and western origin of the Tela Basin (Honduras). In:MANN,P. Honduras, extends northward across the Motagua (ed.) Caribbean Basins. Sedimentary Basins of the fault zone into the Chuacu´s complex, indicating World, 4. Elsevier, Amsterdam, 197–218. that today the Polochic fault may constitute the BANNO, S. 1959. Aegirinaugites from crystalline schists major plate boundary fault. Distributed, dextral, in Sikoku. Journal Geological Society Japan, 65, 652–657. WNW trending fault zones are active along the BARGAR, K. E. 1991. Fluid inclusions and preliminary active volcanic arc. studies of hydrothermal alteration in core hole PLTG-1, Platanares geothermal area, Honduras. This work grew out of our research in southern Mexico in Journal of Volcanology and Geothermal Research, the early 1990s and was started by G. Dengo (Guatemala 45, 147–160. City) in 1993. Stimulating discussions with C. Rangin BATESON, J. H. 1972. New interpretation of geology of and his group at College de France, Aix-en-Provence, Maya Mountains, British Honduras. American Associ- and the inspiring work of our Mexican colleagues ation of Petroleum Geologists, Bulletin, 56, 956–963. prompted the first author to put together the data that BECCALUVA, L., BELLIA,S.ET AL. 1995. The northwes- always seemed to be less bright than those from Asia; we tern border of the Caribbean Plate in Guatemala: new hope that this paper proves otherwise. B. Wauschkuhn geological and petrological data on the Motagua ophio- (Freiberg) compiled Fig. 1b. G. Seward (UCSB) con- litic belt. Ofioliti, 20, 1–15. tributed the EBSD LPO measurement of sample BERTRAND,J.&VUAGNAT, M. 1975. Donne´es chimi- 5PB-10. E. Enkelmann (Lehigh) supported the AFT ques diverses sur des ophiolites du Guate´mala. Bulletin work. W. Frisch (Tu¨bingen), G. Michel (Bermuda) and Suisse de Mine´ralogie Pe´trographie, 57, 466–483. U. Riller (Toronto) helped in the field. L. Ratschbacher BERTRAND, J., DELALOYE, M., FONTIGNIE,D.& thanks B. Hacker (UCSB) for endless support in the VUAGNAT, M. 1978. Ages (K–Ar) sur diverses ophio- Ar-lab when both were at Stanford in the mid lites et roches associe´es de la Cordille`re centrale du 1990s. M. Kozuch collected sample SF1D, and Guate´mala. Bulletin Suisse de Mine´ralogie Pe´trogra- A. Andrews-Rodbell and J. M. Gutierrez sample phie, 58, 405–412. 86TA100. Funded by DFG grants RA442/6 and 25 (to BGR, BUNDESANSTALT FU¨ R BODENFORSCHUNG 1971. L. Ratschbacher), NSF grants EAR 0510325-02 (to Geologische U¨ bersichtskarte 1:125 000, Baja U. Martens), EAR 9219284 (to M. Gordon), CONACYT Verapaz und Su¨dteil der Alta Verapaz (Guatemala). project D41083-F (to B. Weber), and awards from D-3 Hannover-Buchholz. DAAD and Studienstiftung des Deutschen Volkes to BLUNDY,J.D.&HOLLAND, T. 1990. Calcic amphibole K. Stu¨bner. Reviewed by L. Kennan, J. L. Pindell and an equilibria and a new plagioclase-amphibole geo- anonymous colleague. Finally, we thank the editors to thermometer. Contributions to Mineralogy and Petro- allow a long and data-packed paper into their volume. logy, 104, 204–224. BOHLEN,S.&LIOTTA, J. 1986. A barometer for garnet amphibolites and garnet granulites. Journal of Petrol- References ogy, 27, 1025–1034. BONIS, S., BOHNENBERGER,O.H.&DENGO, G. 1970. ALANIZ-ALVAREZ, S., VAN DER HEYDEN, P., NIETO- Mapa geolo´gio de la Repu´blica de Guatemala. Insti- SAMANIEGO,A.F.&ORTEGA-GUTIE´ RREZ,F. tuto Geogra´fico Nacional, Guatemala. 1996. Radiometric and kinematic evidence for BRUECKNER, H. K., HEMMING, S., SORENSEN,S.& Middle Jurassic strike–slip faulting in southern HARLOW, G. E. 2005. Synchronous Sm-Nd mineral Mexico related to the opening of the Gulf of Mexico. ages from HP terranes on both sides of the Motagua Geology, 24, 443–446. Fault of Guatemala: convergent suture and strike– A´ LVAREZ-GO´ MEZ, J. A., MEIJER, P. T., MARTI´NEZ- slip fault? AGU Fall Meeting, Supplement, 86, 52. DIAZ,J.J.&CAPOTE, R. 2008. Constraints from BURKART, B. 1983. Neogene North America–Caribbean finite element modeling on the active tectonics of north- Plate boundary across northern Central America: ern Central America and the . offset along the Polochic fault. Tectonophysics, 99, Tectonics, 27, TC1008; doi: 10.1029/2007TC002162. 251–270. ANDERSON,J.L.&SMITH, D. R. 1995. The effects of BURKART, B. 1994. Northern Central America. In: temperature and f O2 on the Al-in-hornblende barom- DONOVAN,S.&JACKSON, T. (eds) Caribbean eter. American Mineralogist, 80, 549–559. Geology. UWI Publ. Association, Kingston, 265–283. ANDERSON, T. H., BURKART, B., CLEMONS, R. E., BURKART, B., DEATON, B. C., DENGO,C.&MORENO, BOHNENBERGER,O.H.&BLOUNT, D. N. 1973. G. 1987. Tectonic wedges and offset Laramide struc- Geology of the western Altos Cuchumatanes, north- tures along the Polochic fault of Guatemala and western Guatemala. Geological Society of America Chiapas, Mexico: reaffirmation and large Neogene dis- Bulletin, 84, 805–826. placement. Tectonics, 6, 411–422. ANDREANI, L., LE PICHON, X., RANGIN,C.& CAMPA,M.F.&IRIONDO, A. 2004. Significado de data- MARTI´NEZ-REYES, J. 2008. The southern Mexico ciones creta´cicas de los arcos volca´nicos de Taxco, Block, main boundaries and new estimation for its Taxco Viejo y Chacalapa, en la evolucio´n de la plata- Quaternary motion. Bulletin de la Socie´te´ Ge´ologique forma Guerrero Morelos. IV Reunio´n Nacional de de France, 179, 209–223. Ciencias de la Tierra, Juriquilla, Quere´taro, 338. AVE´ LALLEMANT,H.G.&GORDON, M. B. 1999. Defor- CARSWELL, D., O’BRIAN, P., WILSON,R.&ZHAI,M. mation history of Roata´n Island, Implications for the 2000. Metamorphic evolution, mineral chemistry and 288 L. RATSCHBACHER ET AL.

thermobarometry of schists and orthogneisses hosting CASE, J. (eds) The Caribbean Region. Geology of ultra-high pressure eclogites in the Dabie Shan of North America. Geological Society of America, H, Central China. Lithos, 52, 121–155. 37–76. CENTENO-GARCI´A, E. 2005. Review of Upper Paleozoic DROBE,J.&OLIVER, D. 1998. U–Pb age constraints on and Lower Mesozoic stratigraphy and depositional Early Cretaceous volcanism and stratigraphy in environments of central and west Mexico: constraints central Honduras, Geological Society of America. on terrane analysis and paleogeography. In:ANDER- GSA Cordilleran Section Meeting, 30, 12. SON, T. H., NOURSE, J. A., MCKEE,J.W.& DUCEA, M., GEHRELS, G. E., SHOEMAKER, S., RUI´Z,J. STEINER, M. B. (eds) The Mojave-Sonora Megashear &VALENCIA, V. A. 2004a. Geologic evolution of Hypothesis: Development, Assessment, and Alterna- the Xolapa Complex, southern Mexico: evidence tives. Geological Society of America Special Paper, from U–Pb zircon geochronology. Geological 393, 233–258. Society of America Bulletin, 116, 1016–1025. CENTENO-GARCI´A, E., RUIZ, J., CONEY, P. J., DUCEA, M., VALENCIA,V.A.ET AL. 2004b. Rates of PATCHETT,P.J.&ORTEGA, G. F. 1993. Guerrero sediment recycling beneath the Acapulco trench: terrane of Mexico: its role in the Southern Cordillera constraints from (U-Th)/He thermochronology. from new geochemical data. Geology, 21, 419–422. Journal of Geophysical Research, 109; doi: 10.1029/ CERCA, M., FERRARI, L., LO´ PEZ-MARTI´NEZ, M., 2004JB003112. MARTINY,B.&IRONDO, A. 2007. Late Cretaceous ELI´AS-HERRERA,M.&ORTEGA-GUTIE´ RREZ, F. 2002. shortening and Early Cenozoic shearing in the central Caltepec fault zone: an Early Permian dextral transpres- Sierra Madre des Sur, southern Mexico: insights into sional boundary between the Proterozoic Oaxacan and the evolution of the Caribbean–North American Palaeozoic Acatla´n complexes, southern Mexico and plate interaction. Tectonics, 26, TC3007; doi: regional tectonic implications. Tectonics, 21, 1–19. 10.1029/2006TC001981. ELI´AS-HERRERA, M., ORTEGA-GUTIE´ RREZ, F., CHIARI, M., DUMITRICA, P., MARRONI, M., PANDOLFI, SA´ NCHEZ-ZAVALA, J. L., MACI´AS-ROMO, C., L. & PRINCIPI, G. 2006. Radiolarian biostratigraphic ORTEGA-RIVERA,A.&IRIONDO, A. 2005. La falla evidence for a Late Jurassic age of the El Tambor de Caltepec: raı´ces expuestas de una frontera tecto´nica Group ophiolites (Guatemala). Ofioliti, 31, 173–182. de larga vida entre dos terrenos continentales del sur de CLEMONS, R. E. 1966. Geology of the Chiquimula Me´xico. Boletı´n de la Sociedad Geolo´gica Mexicana, quadrangle, Guatemala, Central America. PhD LVII, 83–109. thesis, University of Texas, Austin, TX. EMMET, P. 1983. Geology of the Agalteca Quadrangle, COLOMBI, A. 1988. Me´tamorphisme et ge´ochimie des Honduras, Central America. MS thesis, The University roches mafigues des Alpes onest centrals (ge´oprofil of Texas, Austin. Vie´ge–Domodossola–Lacarno. PhD thesis, Univer- ESCUDER VIRUETE, J., CONTRERAS,F.ET AL. 2006a. sity Lausanne, Switzerland. Transpression and strain partitioning in the Caribbean CORONA-CHA´ VEZ, P., POLI,S.&BIGIOGGERO, B. 2006. island-arc: fabric development, kinematics and Ar-Ar Syn-deformational migmatites and magmatic-arc ages of syntectonic emplacement of the Loma de metamorphism in the Xolapa Complex, southern Cabrera batholith, Dominican Republic. Journal of Mexico. Journal of Metamorphic Geology, 24, Structural Geology, 28, 1496–1519. 169–191. ESCUDER VIRUETE, J., DI´AZ DE NEIRA,A.ET AL. 2006b. DAHL,P.S.&FRIBERG, L. M. 1980. The occurrence and Magmatic relationships and ages of Caribbean Island chemistry of epidote-clinozoisites in mafic gneisses arc tholeiites, boninites and related felsic rocks, from the Ruby Range, southwestern Montana. Contri- Dominican Republic. Lithos, 90, 161–186. butions to Geology, University Wyoming, 18, 77–82. FERRARI, L., GARDUN˜ O, V. H., PASQUARE´ ,G.& DEMETS, C. 2001. A new estimate for present-day TIBALDI, A. 1994. Volcanic and tectonic evolution Cocos–Caribbean Plate motion: implications for slip of central Mexico: oligocene to present. Geofı´sica along the Central American volcanic arc. Geophysical Internacional, 33, 91–105. Research Letters, 28, 4043–4046. FLORES, K., BAUMGARTNER, P. O., SKORA, S., DENGO, G. 1969. Problems of tectonic relations between BAUMGARTNER, L., COSCA,M.&CRUZ, D. 2007. Central America and the Caribbean. Transactions of The Siuna serpentinite me´lange: an Early Cretaceous the Gulf Coast Association of Geological Societies, subduction/accretion of a Jurassic arc. EOS Trans- 19, 311–320. actions, AGU Fall Meeting, Supplement, 88/52, DICKINSON,W.R.&LAWTON, T. F. 2001. Carboniferous Abstract T11D-03. to Cretaceous assembly and fragmentation of Mexico. FOURCADE, E., ME´ NDEZ,J.ET AL. 1994. Dating of the Geological Society of America Bulletin, 113, settling and drowning of the carbonate platform, and 1142–1160. of the overthrusting of the ophiolites on the Maya DONELICK, R. A., RODEN, M. K., MOOERS, J. D., Block during the Mesozoic (Guatemala). Newsletter CARPENTER,B.S.&MILLER, D. S. 1990. Etchable Stratigraphy, 30, 33–43. length reduction of induced fission tracks in apatite at FOURCADE, E., PICCIONI, L., ESCRIBA,J.&ROSSELO, room temperature (¼23 8C): crystallographic orien- E. 1999. Cretaceous stratigraphy and palaeoenviron- tation effects and ‘initial’ mean lengths. Nuclear ments of the Southern Pete´n Basin, Guatemala. Cretac- Tracks and Radiation Measurements, 17, 261–265. eous Research, 20, 793–811. DONNELLY, T. W., HORNE, G. S., FINCH,R.C.& FUHRMAN,M.L.&LINDSLEY, D. H. 1988. Ternary LO´ PEZ-RAMOS, E. 1990. Northern Central America: feldspar modeling and thermometry. American Miner- the Maya and Chortis blocks. In:DENGO,G.& alogist, 73, 201–215. N. AMERICA–CARIBBEAN PLATE BOUNDARY 289

GANGULY, J., CHENG,W.&TIRONE, M. 1996. Thermo- Cuba: evidence for rapid uplift of the western Greater dynamics of aluminosilicate garnet solid solution: new Antilles. Journal of Geology, 112, 521–542. experimental data, an optimized model, and thermo- HARLOW, G. E. 1994. Jadeitites, albitites and related rocks metry applications. Contributions to Mineralogy and from the Motagua Fault Zone, Guatemala. Journal of Petrology, 126, 137–151. Metamorphic Geology, 12, 49–68. GARCI´A-CASCO, A., TORRES-ROLDA´ N, R. L., MILLA´ N, HARLOW, G. E., HEMMING, S. R., AVE´ LALLEMANT, G., MONIE´ ,P.&SCHNEIDER, J. 2002. Oscillatory H. G., SISSON,V.B.&SORENSEN, S. S. 2004. Two zoning of eclogitic garnet and amphibole, Northern high-pressure–low-temperature serpentinite-matrix Serpentinite melange, Cuba: a record of tectonic me´lange belts, Motagua fault zone, Guatemala: a instability during subduction. Journal of Metamorphic record of Aptian and Maastrichtian collisions. Geology, 20, 581–598. Geology, 32, 17–20. GARCI´A-CASCO, A., LAZARO, C., ROJAS AGRAMONTE, HERNA´ NDEZ-TREVIN˜ O, T., TORRES DE LEO´ N, R., Y., KRO¨ NER,A.&NEUBAUER, F. 2007. From Aptian SOLI´S-PICHARDO, G., SCHAAF, P., HERNA´ NDEZ- onset to Danian demise of subduction along the north- BERNAL,M.S.&MORALES-CONTRERAS,J.J. ern margin of the Caribbean Plate (Sierra del Convento 2004. Edad de la Formacio´n Chapolapa en la localidad Melange, Eastern Cuba). AGU Fall Meeting, Sup- del Rı´o Cochoapa al este del Ocotito, estado de plement, 88/52, Abstract T11D-06. Guerrero. IV Reunio´n Mexicana de Ciencias de la GHENT, E. D., STOUT, M. Z., BLACK,P.M.&BROTH- Tierra, Juriquilla, Quere´taro, 338. ERS, R. N. 1987. Chloritoid bearing rocks with blues- HERRMANN, U., NELSON,B.K.&RATSCHBACHER,L. chists and eclogites, northern New Caledonia. Journal 1994. The origin of a terrane: U–Pb systematics and of Metamorphic Petrology, 5, 239–254. tectonics of the Xolapa complex (southern Mexico). GOMBERG, D. N., BANKS,P.O.&MCBIRNEY,A.R. Tectonics, 13, 455–474. 1968. Guatemala: preliminary zircon ages from the HEY, M. 1954. A new review of the chlorites, Mineralogi- Central Cordillera. Science, 162, 121–122. cal Magazine, 30, 225–234. GORDON, M. B. 1993. Revised Jurassic and Early Cretac- HODGES,K.V.&SPEAR, F. S. 1982. Geothermometry, eous (pre-Yojoa Group) stratigraphy of the Chortis geobarometry and the Al2SiO5 triple point at block: paleogeographic and tectonic implications. Mt. Moosilauke, New Hampshire. American Mineral- GCSSEPM Foundation 13th Annual Research Confer- ogist, 67, 1118–1134. ence, Proceedings, 143–154. HOERNLE,K.&HAUFF, F. 2005. Oceanic igneous com- GORDON,M.B.&MUEHLBERGER, W. R. 1994. Rotation plexes. In:BUNDSCHUH,J.&ALVARADO, G. (eds) of the Chortis block causes dextral slip of the Guayape Central America: Geology, Resources, and Natural fault. Tectonics, 13, 858–872. Hazards. Taylor and Francis, Abingdom, 523–547. GOSE, W. A. 1985. Paleomagnetic results from Honduras HOLLAND, T. J. B. 1980. The reaction albite ¼ jadeite þ and their bearing on Caribbean tectonics. Tectonics, 4, quartz determined experimentally in the range of 565–585. 600 8C–1200 8C. American Mineralogist, 65, GRAFE, F., STANEK,K.P.ET AL. 2001. Rb–Sr and 129–134. 40 39 Ar– Ar mineral ages of granitoid intrusives in the HOLLAND, T. J. B. 1990. Activities of components in Mabujina unit, Central Cuba: thermal exhumation omphacitic solid solution. Contributions Mineralogy history of the Escambray Massif. Journal of Geology, Petrology, 105, 446–453. 109, 615–631. HOLLAND,T.J.B.&BLUNDY, J. D. 1994. Non-ideal GRAHAM,C.M.&POWELL, R. 1986. A garnet- interactions in calcic amphiboles and their bearing on hornblende geothermometer: calibration, testing, and amphibole-plagioclase thermometry. Contributions to application to the Pelona Schists, Southern California. Mineralogy and Petrology, 116, 208–224. Journal of Metamorphic Geology, 2, 13–31. HOLLAND,T.J.B.&POWELL, R. 1998. An internally GRAY, G. G., POTTORF, R. J., YUREWICZ, D. A., consistent thermodynamic data set for phases of petro- MAHON, K. I., PEVEAR,D.R.&CHUCHLA,R.J. logical interest. Journal of Metamorphic Geology, 16, 2001. Thermal and chronological record of syn- to 309–343. post-Laramide burial and exhumation, Sierra Madre HORNE, G., ATWOOD,M.&KING, A. 1974. Stratigraphy, Oriental, Mexico. In:BARTOLINI, C., BUFFLER, sedimentology, and paleoenvironment of Esquias R. T. & CANTU´ -CHAPA, A. (eds) The Western Gulf Formation of Honduras. American Association of of Mexico Basins: Tectonics, Sedimentary Basins, Petroleum Geologists Bulletin, 58, 176–188. and Petroleum Systems. American Association of HORNE, G., CLARK,G.&PUSHKAR, P. 1976a. Pre- Petroleum Geologists, Memoirs, 75, 159–181. Cretaceous rocks of northwestern Honduras: basement GREEN,T.H.&HELLMAN, P. L. 1982. Fe–Mg partition- terrane in Sierra de Omoa. American Association of ing between coexisting garnet and phengite at high Petroleum Geologists Bulletin, 60, 566–583. pressure, and comments on a garnet–phengite HORNE, G., PUSHKAR,A.&SHAFIQULLAH, M. 1976b. geothermometer. Lithos, 15, 253–266. Laramide plutons on the landward continuation of the GUZMA´ N-SPEZIALE, M. 2001. Active seismic defor- Bonacca ridge, northern Honduras. Publicaciones mation in the grabens of northern Central America Geologicas del ICAITI (Guatemala), 5, 84–90. and its relationship to the relative motion of the IGN. 1978. Instituto Geogra´fico Nacional, Mapa Geolo´- North America–Caribbean Plate boundary. Tectono- gico de Guatemala. Escala 1 : 50 000, Hoja Rı´o physics, 337, 39–51. Hondo, Guatemala. HALL, C. M., KESLER,S.E.ET AL. 2004. Age and tectonic INDARES,A.&MARTIGNOLE, J. 1985. Biotite garnet setting of the Camagu¨ey volcanic-intrusive arc, geothermometry in the granulite facies: the influence 290 L. RATSCHBACHER ET AL.

of Ti and Al in biotite. American Mineralogist, 70, of equilibria among garnet-clinopyroxene-kyanite- 272–278. phengite-coesite/quartz. Journal of Metamorphic JAMES, K. H. 2006. Arguments for and against the Pacific Geology, 22, 579–592. origin of the Caribbean Plate: discussion, finding LEAKE, B. E., WOOLEY,A.R.ET AL. 1997. Nomenclature for an inter-American origin. Geological Acta, 4, of amphiboles. Report of the subcommittee on amphi- 279–302. boles of the Intern. Mineral. Association Commission JONCKHEERE, R. 2003a. On the densities of etchable on new minerals and mineral names. European fission tracks in a mineral and co-irradiated external Journal of Mineralogy, 9, 623–665. detector with reference to fission-track dating of min- LUDWIG, K. R. 2003. Isoplot 3.01. A Geochronological erals. Chemical Geology, 200, 41–58. Toolkit for Microsoft Exel. Berkeley Geochronology JONCKHEERE, R. 2003b. On the methodical problems in Center Special Publications, 4, 1–70. estimating geological temperature and time from MACHORRO, R. 1993. Geology of the northwestern part measurements of fission tracks in apatite. Radiation of the Granados Quadrangle, Central Guatemala. Measurements, 36, 43–55. MSc thesis, University of Texas, El Paso. JONCKHEERE,R.&WAGNER, G. A. 2000. The KTB MALONE, J. R., NANCE, R. D., KEPPIE,J.D.&DOSTAL, apatite fission-track profile: the significance of bore- J. 2002. Deformational history of part of the Acatla´n hole data in fission-track analysis. Geological Society Complex: Late Ordovician Early Permian orogenesis of Australia Abstracts, 58, 193–194. in southern Mexico. Journal of South American JORDAN, B. R., SIGURDSSON,H.ET AL. 2007.Petrogenesis Earth Sciences, 15, 511–524. of Central American Cenozoic ignimbrites and MANN, P. 2007. Overview of the tectonic history of associated Caribbean Sea tephra. In:MANN, P. (ed.) northern Central America. In:MANN, P. (ed.) Geo- Geologic and Tectonic Development of the Caribbean logic and Tectonic Development of the Caribbean Plate Boundary in Northern Central America. Geo- Plate Boundary in Northern Central America. logical Society of America, Special Papers, 428, Geological Society of America, Special Papers, 428, 151–179. 1–19. KEPPIE, D. J. 2004. Terranes of Mexico revisited: a 1.3 MANTON, W. I. 1996. The Grenville of Honduras. Geo- billion year odyssey. International Geology Review, logical Society of America, Annual Meeting Abstracts 46, 765–794. with Program, A-493. KEPPIE, D. J., NANCE,R.D.ET AL. 2004. Mid-Jurassic MANTON,W.I.&MANTON, R. S. 1984. Geochronology tectonothermal event superposed on a Palaeozoic geo- and Late Cretaceous–Tertiary Tectonism of Hon- logical record in the Acatla´n complex of southern duras. Direccion General de Minas e Hidrocarburos, Mexico: hotspot activity during the breakup of Honduras, 1–55. Pangea. Gondwana Research, 7, 239–260. MANTON,W.I.&MANTON, R. S. 1999. The southern KEPPIE, J. D., NANCE, R. D., FERNA´ NDEZ-SUA´ REZ, J., flank of the Tela basin, Republic of Honduras. In: STOREY, C. D., JEFFRIES,T.E.&MURPHY,J.B. MANN, P. (ed.) Caribbean Basins. Sedimentary 2006. Detrital zircon data from the eastern Mixteca Basins of the World, 4. Elsevier, Amsterdam, 219–236. terrane, southern Mexico: evidence for an MARTENS, U., ORTEGA-OBREGO´ N, C., ESTRADA,J.& Ordovician–Mississippian continental rise and a VALLE, M. 2007. Metamorphism and metamorphic Permo-Triassic clastic wedge adjacent to Oaxaquia. rocks. In:BUNDSCHUH,J.&ALVARADO, G. (eds) International Geology Review, 48, 97–111. Central America: Geology, Resources, and Natural KESLER, S. E., CAMPELL,I.H.&ALLEN,CH. M. 2005. Hazards. Taylor and Francis, Abingdom, 485–522. Age of the Los Ranchos Formation, Dominican Repub- MARTENS, U., WEBER,B.&VALENCIA, V. 2010. U–Pb lic: timing and tectonic setting of primitive island arc geochronology of lower Paleozoic beds in the south- volcanism in the Caribbean region. Geological western Maya block, Central America: its affinity Society of America Bulletin, 117, 987–995. with Avalonian-type Peri-Gondwanan terranes. Geo- KETCHAM, R. A., DONELICK,R.A.&DONELICK,M.B. logical Society of America Bulletin, in press. 2000. AFTSolve: a program for multikinetic modeling MASSONNE, H.-J. & SZPURKA, Z. 1997. Thermodynamic of apatite fission-track data. Geological Materials properties of white micas on the basis of high-pressure Research, 2, 1–32. experiments in the system K2O–MgO–Al2O3 –SiO2 – KOHN,M.&SPEAR, F. 1990. Two new geobarometers for H2O and K2O–FeO–Al2O3 –SiO2 –H2O. Lithos, 41, garnet-amphibolites, with application to southeastern 229–250. Vermont. American Mineralogist, 75, 89–96. MESCHEDE,M.&FRISCH, W. 1998. A plate-tectonic KREBS, M., STANEK,K.P.ET AL. 2007. Age of high model for the Mesozoic and Early Cenozoic pressure metamorphism from the Escambray Massif, history of the Caribbean Plate. Tectonophysics, 296, Cuba. Abstract, Goldschmidt Conference, Cologne, 269–291. 2007. Cambridge Publications, A522. MESCHEDE, M., FRISCH, W., HERRMANN,U.& KRETZ, R. 1983. Symbols for rock-forming minerals. RATSCHBACHER, L. 1996. Stress transmission across American Mineralogist, 68, 277–279. an active plate boundary: an example from southern KROGH, E. J. 2000. The garnet-clinopyroxene Fe/Mg Mexico. Tectonophysics, 266, 81–100. geothermometer: an updated calibration. Journal of MMAJ, METAL MINING AGENCY OF JAPAN. 1980. Metamorphic Geology, 18, 211–219. Report on Geology Survey of the Western Area, KROGH,E.J.&TERRY, M. P. 2004. Geothermobarometry Olancho. Japan International Cooperation Agency, of UHP and HP eclogites and schists: an evaluation Government of Japan, 1–138. N. AMERICA–CARIBBEAN PLATE BOUNDARY 291

MORA´ N-ZENTENO, D. 1992. Investigaciones isoto´picas Mexico and Caribbean evolution: eight controversies de Rb–Sr y Sm–Nd en rocas cristalinas de la resolved. Geologica Acta, 99, 303–341. eregio´n de Tierra Colorada–Acapulco–Cruz– PLAFKER, G. 1976. Tectonic aspects of the Guatemala Grande, Estado de Guerrero. PhD thesis, Universidad earthquake of 4 February 1976. Science, 193, Autonoma de Mexico, Mexico, DF. 1201–1208. MORA´ N-ZENTENO, D. J., CORONA-CHA´ VEZ,P.& PLYUSNINA, L. P. 1982. Geothermometry and geobaro- TOLSON, G. 1996. Uplift and subduction erosion in metry of plagioclase-hornblende bearing assemblages. southwestern Mexico since the Oligocene: Pluton geo- Contributions Mineralogy Petrology, 80, 140–146. barometry constraints. Earth and Planetary Science RATSCHBACHER, L., RILLER, U., MESCHEDE, M., Letters, 141, 51–65. HERRMANN,U.&FRISCH, W. 1991. Second look at MORA´ N-ZENTENO, D. J., TOLSON,G.ET AL. 1999. suspect terranes in southern Mexico. Geology, 19, Tertiary arc-magmatism of the Sierra Madre del Sur, 1233–1236. Mexico, and its transition to the volcanic activity of RILLER, U., RATSCHBACHER,L.&FRISCH, W. 1992. the Trans-Mexican Volcanic Belt. Journal of South Left-lateral transtension along the Tierra Colorada American Earth Sciences, 12, 513–535. deformation zone, northern margin of the Xolapa NIETO-SAMANIEGO, A. F., ALANIZ-A´ LVAREZ, S. A., magmatic arc of southern Mexico. Journal of South SILVA-ROMO, G., EGUIZA-CASTRO,M.H.& American Earth Sciences, 5, 237–249. MENDOZA-ROSALES, C. C. 2006. Latest Cretaceous ROBINSON, K. L., GASTIL, R. G., CAMPA,M.F.& to Miocene deformation events in the eastern Sierra RAMIREZ-ESPINOSA, J. 1989. Geochronology of Madre del Sur, Mexico, inferred from the geometry basement and metasedimentary rocks in southern and age of major structures. Geological Society of Mexico and their relation to metasedimentary rocks America Bulletin, 118, 238–252. in Peninsular California. Geological Society of ORTEGA-GUTIE´ RREZ, F., ELI´AS-HERRERA, M., America Abstracts with Programs, 21, 135. MACI´AS-ROMERO,C.&LO´ PEZ, R. 1999. Late Ordo- ROGERS, R. D. 2003. Jurassic–Recent tectonic and strati- vician–Early Silurian continental collisional orogeny graphic history of the Chortis block of Honduras and in southern Mexico and its bearing on Gondwana– Nicaragua (northern Central America). PhD thesis, Laurentia connections. Geology, 27, 719–722. University of Texas, Austin, TX. ORTEGA-GUTIE´ RREZ, F., SOLARI, L. A., SOLE´ , J., ROGERS, R. D., KA´ RASON,H.&VAN DER HILST,R.D. MARTENS, U., GO´ MEZ-TUENA, A., MORA´ N-ICAL, 2002. Epeirogenic uplift above a detached slab in S., REYES-SALA,M.&ORTEGA-OBREGO´ N,C. northern Central America. Geology, 30, 1031–1034. 2004. Polyphase, high-temperature eclogite-facies ROGERS, R. D., MANN, P., SCOTT,R.W.&PATINO,L. metamorphism in the Chuacu´s complex, Central 2007a. Cretaceous intra-arc rifting, sedimentation, Guatemala: petrology, geochronology, and tectonic and basin inversion in east-central Honduras. In: implications. International Geology Review, 46, MANN, P. (ed.) Geologic and Tectonic Development 445–470. of the Caribbean Plate Boundary in Northern ORTEGA-GUTIE´ RREZ, F., SOLARI,L.A.ET AL. 2007. The Central America. Geological Society of America, Maya–Chortı´s boundary: a tectonostratigraphic Special Papers, 428, 89–128. approach. International Geology Review, 49, ROGERS, R. D., MANN, P., EMMET,P.A.&VENABLE, 996–1024. M. E. 2007b. Colon fold belt of Honduras: evidence ORTEGA-OBREGO´ N, C., SOLARI, L., ORTEGA- for Late Cretaceous collision between the continental GUTIE´ RREZ, F., SOLE´ -VIN˜ AS,J.&GO´ MEZ-TUENA, Chortis block and intra-oceanic Caribbean Arc. In: A. 2004. Caracterizacio´n estructural, petrolo´gica y MANN, P. (ed.) Geologic and Tectonic Development geocronolo´gica de la zona de cizalla “Baja Verapaz”, of the Caribbean Plate Boundary in Northern Guatemala. Libro de Resu´menes, IV Reunio´n Nacional Central America. Geological Society of America, de Ciencias de la Tierra, Juriquilla, 2004, 204. Special Papers, 428, 129–149. ORTEGA-OBREGO´ N, C., SOLARI, L., KEPPIE, J. D., ROGERS, R. D., MANN,P.&EMMET, P. A. 2007c. Tec- ORTEGA-GUTIE´ RREZ, F., SOLE´ ,J.&MORA´ N,S. tonic terranes of the Chortis block based on integration 2008. Middle–Late Ordovician magmatism and of regional aeromagnetic and geologic data. In:MANN, Late Cretaceous collision in the southern Maya P. (ed.) Geologic and Tectonic Development of the block, Rabinal-Salama´ area, central Guatemala: Caribbean Plate Boundary in Northern Central implications for North America–Caribbean Plate America. Geological Society of America, Special tectonics. Geological Society of America Bulletin, Papers, 428, 65–88. 120, 556–570. ROJAS-AGRAMONTE, Y., KRO¨ NER, A., GARCI´A- PERCHUK,L.L.&LAVRENT’EVA, J. V. 1983. Exper- CASCO, A., ITURRALDE-VINENT, M. A., WINGATE, imental investigations of exchange equilibria in the M. T. D. & LIU, D. Y. 2006. Geodynamic implications system cordierite–garnet–biotite. In:SAXENA,S. of zircon ages from Cuba. Geophysical Research (ed.) Kinematics and Equilibria in Mineral Reactions. Abstracts, 8, 04943; SRef-ID: 1607-7962/gra/ Springer, Berlin, 199–239. EGU06-A-04943. PINDELL, J. K., CANDE,S.C.ET AL. 1988. A plate- ROSENCRANTZ, E., ROSS,M.I.&SLATER, J. G. 1988. kinematic framework for models of Caribbean evol- Age and spreading history of the Cayman Trough as ution. Tectonophysics, 155, 121–138. determined from depth, heat flow, and magnetic PINDELL, J. L., KENNAN, L., STANEK, K. P., MARESCH, anomalies. Journal Geophysical Research, 93, W. & DRAPER, G. 2006. Foundations of Gulf of 2141–2157. 292 L. RATSCHBACHER ET AL.

RYBURN, R. J., RAHEIM,A.&GREEN, D. H. 1976. Deter- STANEK, K. P., RIßE, A., RENNO, A., ROMER,R.& mination of the PT paths of natural eclogites during GRAFE, F. 2005. The history of the Great Antillean metamorphism – record of a subduction. Lithos, 9, Island Arc: example from Central Cuba. Abstract, 161–164. Colloquium on Latin American Geosciences 2005, SA´ NCHEZ-ZAVALA, J. L., ORTEGA-GUTIE´ RREZ, F., Potsdam, Terra Nostra, 1. KEPPIE, J. D., JENNER,G.A.&BELOUSOVA,E. STEINER,M.B.&WALKER, J. D. 1996. Late Silurian 2004. Ordovician and Mesoproterozoic zircon from plutons in Yucatan. Journal Geophysical Research, the Tecomate Formation and Esperanza granitoid, 101, 17727–17735. Acatla´n complex, southern Mexico: local provenance STIPP, M., STU¨ NITZ, H., HEILBRONNER,R.&SCHMID, in the Acatla´n and Oaxacan complexes. International S. M. 2002. The eastern Tonale fault zone: a natural Geology Review, 46, 1005–1021. laboratory for crystal plastic deformation of quartz SCHAAF, P., MORA´ N-ZENTENO, D., HERNA´ NDEZ- over a temperature range from 250 to 7008C. Journal BERNAL, M. S., SOLIS-PICHARDO, G., TOLSON,G. of Structural Geology, 24, 1861–1884. &KO¨ HLER, H. 1995. Paleogene continental margin SUTTER, J. F. 1979. Late Cretaceous collisional tectonics truncation in southwestern Mexico: geochronological along the Motagua fault zone, Guatemala. Geological evidence. Tectonics, 14, 1339–1350. Society of America, Abstracts with Programs, 11, SCHAAF, P., WEBER, B., WEIS, P., GROß, A., ORTEGA- 525–526. GUTIE´ RREZ,F.&KO¨ HLER, H. 2002. The Chiapas TALAVERA-MENDOZA, O., RUIZ, J., GEHRELS, G. E., Massif (Mexico) revised: new geologic and isotopic MEZA-FIGUEROA, D., VEGA-GRANILLO,R.& data for basement characteristics. Neues Jahrbuch CAMPA-URANGA, M. F. 2005. U–Pb geochronology Geologie Pala¨ontologie, Abhandlungen, 225, 1–23. of the Acatla´n complex and implications for the SCHMID, R., FRANZ, L., OBERHA¨ NSLI,R.&DONG,S. Paleozoic paleogeography and tectonic evolution of 2000. High Si-phengite, mineral chemistry and P-T southern Mexico. Earth and Planetary Science evolution of ultra-high-pressure eclogites and calc- Letters, 235, 682–699. silicates from the Dabie Shan, eastern PR China. TALAVERA-MENDOZA, O., RUIZ, J., GEHRELS, G. E., Geological Journal, 35, 185–207. VALENCIA,V.A.&CENTENO-GARCI´A, E. 2007. SCHNEIDER, J., BOSCH,D.ET AL. 2004. Origin and evol- Detrital zircon U–Pb geochronology of southern ution of the Escambray Massif (Central Cuba): an Guerrero and western Mixteca arc successions example of HP/LT rocks exhumed during intraoceanic (southern Mexico): new insights for the tectonic evol- subduction. Journal of Metamorphic Geology, 22, ution of southwestern North America during the Late 227–247. Mesozoic. Geological Society of America Bulletin, SCHWARTZ, D. P. 1976. Geology of the Zacapa quadran- 119, 1052–1065. gle and vicinity, Guatemala, Central America. PhD TOLSON, G. 2005. La falla Chacalapa en el sur de Oaxaca. thesis, State University of New York, Binghamton. Boletı´n de la Sociedad Geolo´gica Mexicana, 57, SOLARI, L. A., DOSTAL, J., ORTEGA-GUTIE´ RREZ,F.& 111–122. KEPPIE, J. D. 2001. The 275 Ma arc-related La TORRES, R., RUIZ, J., PATCHETT,P.J.&GRAJALES,J. Carbonara stock in the northern Oaxacan Complex of M. 1999. Permo-Triassic continental arc in eastern southern Mexico: U–Pb geochronology and geochem- Mexico: tectonic implications for reconstructions of istry. Revista Mexicana de Ciencias Geolo´gicas, 18, southern North America. In:BARTOLINI, C. W., 149–161. WILSON,J.L.&LAWTON, T. F. (eds) Mesozoic Sedi- SOLARI, L. A., KEPPIE, J. D., ORTEGA-GUTIE´ REZ, F., mentary and Tectonic History of North-Central ORTEGA-RIVERA, A., HAMES,W.E.&LEE, Mexico. Geological Society of America, Special J. K. W. 2004. Phanerozoic structures in the Grenville Papers, 340, 191–196. complex, southern Mexico: result of thick-skinned tec- TRIBOULET, C. 1992. The (Na–Ca)amphibole–albite– tonics. International Geology Review, 46, 614–628. chlorite–epidote–quartz geothermobarometer in the ´ ´ SOLARI, L. A., TORRES DE LEON, R., HERNANDEZ system S–A–F–M–C–N–H2O. 1. An empirical PINEDA, G., SOLE´ , J., SOLI´S-PICHARDO,G.& calibration. Journal of Metamorphic Geology, 10, HERNA´ NDEZ-TREVIN˜ O, G. 2007. Tectonic signifi- 545–556. cance of Cretaceous–Tertiary magmatic and structural TSUJIMORI, T., LIOU,J.G.&COLEMAN, R. G. 2004. evolution of the northern margin of the Xolapa Comparison of two contrasting eclogites from the Complex, Tierra Colorada area, southern Mexico. Motagua fault zone, Guatemala: southern lawsonite Geological Society of America Bulletin, 119, eclogite versus northern zoisite eclogite. Geological 1265–1279. Society of America Abstracts with Programs, 36, 136. SOLARI, L. A., ORTEGA-GUTIE´ RREZ,F.ET AL. 2009. TSUJIMORI, T., LIOU,J.G.&COLEMAN, R. G. 2005. U–Pb zircon geochronology of Palaeozoic units in Coexisting retrograde jadeite and omphacite in a Western and Central Guatemala: insights into the tec- jadeite-bearing lawsonite eclogite from the Motagua tonic evolution of Middle America. In:JAMES, K. H., Fault Zone, Guatemala. American Mineralogist, 90, LORENTE,M.A.&PINDELL, J. L. (eds) The 836–842. Origins and Evolution of the Caribbean Plate. Geo- TSUJIMORI, T., SISSON, V. B., LIOU, J. G., HARLOW, logical Society, London, Special Publications, 328, G. E. & SORENSEN, S. S. 2006. Petrologic characteriz- 295–313. ation of Guatemalan lawsonite eclogite: Eclogitization SPEAR, F. S. 1981. NaSi–CaAl exchange equilibrium of subducted oceanic crust in a cold subduction zone. between plagioclase and amphibole. Contributions to In:HACKER, B. R., MCCLELLAND,W.C.&LIOU, Mineralogy and Petrology, 72, 33–41. J. G. (eds) Ultrahigh-pressure Metamorphism: Deep N. AMERICA–CARIBBEAN PLATE BOUNDARY 293

Continental Subduction. Geological Society of and paleomagnetic evolution of the western Maya America, Special Papers, 403, 147–168. block, SW Mexico. Geological Society of America VAN DEN BOOM, G. 1972. Petrofazielle Gliederung des Abstract with Programs, 37, 68. metamorphen Grundgebirges in der Sierra de WEBER, B., CAMERON, K. L., OSORIO,M.&SCHAAF,P. Chuacu´s, Guatemala. Beihefte Geologisches Jahrbuch, 2005. A Late Permian tectonothermal event in Gren- 122, 5–49. ville crust of the Southern Maya terrane; U–Pb VEGA-GRANILLO, R., TALAVERA-MENDOZA,O.ET AL. zircon ages from the Chiapas Massif, Southeastern 2007. Pressure–temperature–time evolution of Me´xico. International Geology Review, 47, 509–529. Paleozoic high-pressure rocks of the Acatla´n WEBER, B., IRIONDO, A., PREMO, W. R., HECHT,L.& Complex (southern Mexico): implications for the evol- SCHAAF, P. 2006. New insights into the history and ution of the Iapetus and Rheic Oceans. Geological origin of the southern Maya Block, SE Me´xico: Society of America Bulletin, 119, 1249–1264. U-Pb-SHRIMP zircon geochronology from meta- VENABLE, M. 1994. A geological, tectonic, and metallo- morphic rocks of the Chiapas Massif. International genetic evaluation of the Siuna terrane (Nicaragua). Journal of Earth Sciences; doi: 10.1007/ PhD thesis, University of Arizona, Tucson. s00531-006-0093-7. VILAND, J., HENRY, B., CALIX,R.&DIAZ, C. 1996. Late WEBER, B., VALENCIA, V. A., SCHAAF, P., POMPA- Jurassic deformation in Honduras: proposals for a MERA,V.&RUIZ, J. 2008. Significance of prove- revised regional stratigraphy. Journal of South Ameri- nance ages from the Chiapas Massif complex (SE can Earth Sciences, 9, 153–160. Mexico): redefining the Paleozoic basement of the VON BLANCKENBURG, F. V., VILLA, I., BAUR, H., Maya block and its evolution in a peri-Gondwanan MORTEANI,G.&STEIGER, R. H. 1989. Time cali- entourage. Journal of Geology, 6, 619–639. bration of a PT-path in the Western Tauern Window, WEILAND, T., SUAYAH,W.&FINCH, R. 1992. Petrologic Eastern Alps: the problem of closure temperatures. and tectonic significance of Mesozoic volcanic rocks in Contributions to Mineralogy and Petrology, 101, the Rı´o Wampu´ area, eastern Honduras. Journal of 1–11. South American Earth Sciences, 6, 309–325. WATERS,D.J.&MARTIN, H. N. 1993. Geobarometry in WILSON, H. H. 1974. Cretaceous sedimentation and orogeny phengite-bearing eclogites. Terra Abstracts, 5, in nuclear Central America. American Association of 410–411; updated calibration of 1996 at World Wide Petroleum Geologists Bulletin, 58, 1348–1396. Web Address: http://www.earth.ox.ac.uk/c.davewa/ YAN˜ EZ, P., RUIZ, J., PATCHETT, P. J., ORTEGA- ecbar.html. GUTIERREZ,F.&GEHRELS, G. E. 1991. Isotopic WAWRZYNIEC, T., MOLINA-GARZA, R. S., GEISSMAN, studies of the Acatla´n complex, southern Mexico: J. & IRIONDO, A. 2005. A newly discovered relic, implications for Paleozoic North American tectonics. transcurrent plate boundary – the Tonala shear zone Geological Society of America Bulletin, 103, 817–828.