spikings et al. 2014

7/24/2019 Spikings Et Al. 2014

1/45

GR Focus Review

The geological history of northwestern South America: from Pangaea to the earlycollision of the Caribbean Large Igneous Province (29075 Ma)

Richard Spikings a,, Ryan Cochrane b, Diego Villagomez c, Roelant Van der Lelij d, Cristian Vallejo e,Wilfried Winkler f, Bernado Beate g

a Department of Earth and Environmental Science, University of Geneva, Rue des Maraichers 13, Geneva 1205, Switzerlandb Thomson Reuters, London, UKc Tectonic Analysis Ltd., Geneva, Switzerlandd Norges Geologiske Underskelse, 7491 Trondheim, Norwaye Geostrat S.A., Quito, Ecuadorf ETH Zrich, Geological Institute, ETH Zentrum, NO E 59, Sonneggstrasse 5, CH-8092 Zrich, Switzerlandg

Facultad de Geologa, Minas y Petrleos, Escuela Politcnica Nacional, A.P. 17-01-2759, Quito, Ecuador

a b s t r a c ta r t i c l e i n f o

Article history:

Received 24 March 2014Received in revised form 4 June 2014Accepted 25 June 2014Available online xxxx

Handling Editor: M. Santosh

Keywords:

PangaeaSouth AmericaTectonic reconstruction

GeochronologyGeochemistryThermochronology

Northwestern South America preserves a record of the assembly of western Pangaea, its disassembly andinitiation of the far western Tethys Wilson Cycle, subsequent Pacic margin magmatism and ocean plateaucontinent interaction since the Late Cretaceous. Numerous models have been presented for various timeslices although they are based on either spatially restricted datasets, or dates that are inaccurate estimatesof the time of crystallisation. Here we review a very large quantity of geochronological, geochemical,thermochronological, sedimentological and palaeomagnetic data that collectively provide tight constraints forgeological models. These data have been collected over a trench (Pacic)-parallel distance of N1500 km(Colombia and Ecuador), and reveal important temporal trends in rifting and subduction. The temporal frame-workfor our model constraints are obtainedfrom robust, concordant zircon U-Pb ages of magmatic rocks during29075 Ma. The Late Cretaceous thermal history of the margin (b350 C) is described by 40Ar/39Ar andssiontrack data, and the higher temperature and thus older (pre-75 Ma) history are constrained by apatite U-Pbthermochronology. Variations in the isotopic compositions of Hf (zircon), Nd (whole) and O (quartz) withtime have been used to track the evolution of the source of magmatism, and are used as proxies for crustalthickness. Atomic chemical compositions, combined with isotopes and dense mineral assemblages are used todifferentiate between continental and oceanic environments. These data show that rifting within westernPangaea started at 240 Ma, leading to sea oor spreading between blocks of Central and South America by216 Ma.Pacic activemargin commenced at 209 Ma, andcontinued until 115 Ma above an east-dippingsubduc-tion zone that wasrolling back, attenuatingSouth Americaand formingnew continental crust. Theopeningof theSouth Atlantic drove South America westwards, compressed the Pacic margin of northwestern South Americaat 115 Ma and obducted an exhumed subduction zone. Passive margin conditions prevailed until the OceanicPlateau and its overlying intra-oceanic arc (The Rio Cala Arc) collided and accreted to South America at 75 Ma.

2014 Published by Elsevier B.V. on behalf of International Association for Gondwana Research.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 02. Geological framework of northwestern South America (Colombia and Ecuador) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 03. Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 04. Triassic: the disassembly of Pangaea and the formation of a passive margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

4.1. Historical perspective and occurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 04.2. Geochronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

4.2.1. Cordillera Real of Ecuador and Cordillera Central of Colombia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

Gondwana Research xxx (2014) xxxxxx

Corresponding author. Tel.: +41(0)223793176.E-mail addresses:[email protected](R. Spikings),[email protected](R. Cochrane),[email protected](D. Villagomez),[email protected]

(R. Van der Lelij),[email protected](C. Vallejo),[email protected](W. Winkler),[email protected](B. Beate).

GR-01278; No of Pages 45

http://dx.doi.org/10.1016/j.gr.2014.06.004

1342-937X/ 2014 Published by Elsevier B.V. on behalf of International Association for Gondwana Research.

Contents lists available atScienceDirect

Gondwana Research

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / g r

Please cite this article as: Spikings, R., et al., The geological history of northwestern South America: from Pangaea to the early collision of theCaribbean Large Igneous Province (29075 Ma), Gondwana Research (2014),http://dx.doi.org/10.1016/j.gr.2014.06.004
http://dx.doi.org/10.1016/j.gr.2014.06.004http://dx.doi.org/10.1016/j.gr.2014.06.004http://dx.doi.org/10.1016/j.gr.2014.06.004mailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]://dx.doi.org/10.1016/j.gr.2014.06.004http://www.sciencedirect.com/science/journal/1342937Xhttp://dx.doi.org/10.1016/j.gr.2014.06.004http://dx.doi.org/10.1016/j.gr.2014.06.004http://www.sciencedirect.com/science/journal/1342937Xhttp://localhost/var/www/apps/conversion/tmp/scratch_7/Unlabelled%20imagehttp://dx.doi.org/10.1016/j.gr.2014.06.004http://localhost/var/www/apps/conversion/tmp/scratch_7/Unlabelled%20imagemailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]://dx.doi.org/10.1016/j.gr.2014.06.004


2/45

4.2.2. Comparison with the ages of Permian and Triassic rocks in Venezuela and Peru . . . . . . . . . . . . . . . . . . . . . . . . . 04.3. Geochemistry of the granites and migmatites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

4.3.1. Cordillera Real of Ecuador and Cordillera Central of Colombia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 04.3.2. Comparison with Permian and Triassic rocks in Venezuela and Peru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

4.4. Geochemistry of the amphibolites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 04.5. Zircon Hf isotope geochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

4.5.1. Zircon Hf isotope geochemistry of the granites and migmatitic leucosomes . . . . . . . . . . . . . . . . . . . . . . . . . . . 04.5.2. Zircon Hf isotope geochemistry of the amphibolites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 04.5.3. Comparison with Zircon Hf isotope compositions in Peru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

4.6. Thermal histories during the Triassic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 04.7. Interpretation: Permian and Triassic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 04.7.1. Arc magmatism and metamorphism during 290240 Ma along western Pangaea . . . . . . . . . . . . . . . . . . . . . . . . 04.7.2. Initiating the disassembly of western Pangaea during 240200 Ma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

4.8. Conjugate margins to northwestern Gondwana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 04.9. Rifting between North and South America . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

5. Latest TriassicLower Cretaceous: arc magmatism and tectonic switching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05.1. Historical perspective and occurrence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

5.1.1. Latest TriassicJurassic granitoid intrusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05.1.2. Late JurassicEarly Cretaceous rocks to the west of the Jurassic intrusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

5.2. Geochronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05.2.1. Latest Triassic and Jurassic intrusions: Cordillera Real, Cordillera Central and the Santander Massif . . . . . . . . . . . . . . . . 05.2.2. Early Cretaceous magmatic and sedimentary rocks: Cordillera Real and Cordillera Central . . . . . . . . . . . . . . . . . . . . 05.2.3. Comparison with Peru and the Merida Andes of Venezuela . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

5.3. Geochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05.3.1. Latest Triassicearliest Cretaceous granitoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

5.3.2. Early Cretaceous igneous rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05.3.3. Comparison with magmatic rocks from Peru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

5.4. The tectonic setting during the latest TriassicJurassic (210145 Ma) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05.4.1. Why is there a gap in the Jurassic arc in Peru? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

5.5. The tectonic setting during the Early Cretaceous (145115 Ma) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05.6. Compression during the Early Cretaceous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05.7. The Chaucha Terrane and the Taham Terrane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05.8. Comparison with Peru (145115 Ma) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

6. The tectonic history of northwestern South America during 11575 Ma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 06.1. The formation of the Caribbean Large Igneous Province and its collision with South America. . . . . . . . . . . . . . . . . . . . . . . . 0

6.1.1. Geochemistry and geochronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 06.1.2. Time of initial accretion with South America . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 06.1.3. The nature of the CLIPSouth America suture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

1. Introduction

The northwestern South American plate hosts a Grenvillian base-ment, which was modied during the amalgamation and disassemblyof Pangaea, subsequent prolonged active margin magmatism and thecollision of the voluminous Caribbean Large Igneous Province, whichadded new crust to South America. This manuscript is mainly a reviewof a very large quantity of data, although some new U-Pb (apatite)and 40Ar/39Ar dates are presented. These data are used to generate

robust constraints for any model that describes the disassemblyand fragmentation of western Pangaea, the subsequent evolution ofthe Pacic margin offshore northwestern South America during the

JurassicEarly Cretaceous, and the early evolution of the Caribbeanregion and its interaction with South America. The review is organisedinto sections according to geological time, and compares the evolutionof northwestern South America (north of 5S) with the margin of Peruduring 29075 Ma.

Wide disagreements exist over the tectonic origin of voluminousmagmatic units, including Triassicanatectites, JurassicEarlyCretaceousarc rocks, obducted M-HP/LT rocks and allochthonous units that com-prise the western cordilleras and the forearc. These contrasting inter-pretations result in signicantly different interpretations for platereconstructions during the TriassicLate Cretaceous (e.g. Litherland

et al., 1994; Spikings et al., 2001; Pratt et al., 2005; Pindell and

Kennan, 2009; Villagmez and Spikings, 2013; Cochrane et al., 2014a).Contrasting models partly exist because of the misinterpretation ofK/Ar and Rb/Sr dates that were published in the 1980's and 1990's asaccurateestimates of crystallisation age, ignoringthe effects of daughterisotope loss. We discard K/Ar and Rb/Sr dates in favour of recently pub-lished concordant zircon U-Pbdates,which are moreaccurateestimatesof crystallisation age. The U-Pb dates are combined with geochemicaland isotope data, sedimentological data and eld relationships toconstrain the magmatic source regions and tectonic environment

within which the rocks formed. The tectonic histories are subsequentlyinvestigated using thermochronological and palaeomagnetic data.We show that western Pangaea started to disassemble by rifting of

continental crust of Central America from South America at ~240 Ma,and thatthese hadcompletelyseparated by ~216 Ma.The northwesternmargin of South America remained passive until ~209 Ma withinPangaea, and arc magmatism occurred during 209114 Ma, accompa-nying the separation of North and South America at ~180 Ma. The

Jurassic magmas formed in a continental arc, which questions previousinterpretations that place theJurassic trench far from the location of the

Jurassic arcs, due to the presence of suspect continental terranes. Wedraw a single east-dipping subduction zone during 209114 Ma,which retreated oceanward and extended the South American margin,culminating in compression that drove rock uplift and exhumation.

Finally, we present evidence for an east-facing intra-oceanic arc, which

2 R. Spikings et al. / Gondwana Research xxx (2014) xxxxxx

http://dx.doi.org/10.1016/j.gr.2014.06.004http://dx.doi.org/10.1016/j.gr.2014.06.004


3/45

formed on an intra-oceanic plateau prior to its collision and accretionwith South America at ~75 Ma, resulting in growth of the continent.Our interpretation differs from other models that rely on large-scaleplate reconstructions to constrain the positions of continents and sub-duction zones.

2. Geological framework of northwestern South America (Colombia

and Ecuador)

The South American Plate forms a relict part of western Gondwana,and formed during the opening of the Central Atlantic, South Atlanticand the Inter-American Gap (Gulf of Mexico and the proto-Caribbean)during 180120 Ma (E.g.Eagles, 2007). The Atlantic margin remainspassive whereas the western margin became active at ~500480 Ma(e.g.Pankhurst et al., 2000; Van der Lelij, 2013), soon after the openingof the Iapetus Ocean during 570535 Ma (Cawood et al., 2001). TheNorthern plate margin formed during riftingfromYucatan in the Middle

Jurassic(Pindell et al., 2005), forming the region of the proto-Caribbean.At the present time, the Northern Andes are separated from the CentralAndes by the Huancabamba Deection (at ~6S), which marksa distinctchange in the strike orientation of the Andean chain (see inset inFig. 1).

The oldest rocks exposed in northwestern South America areGrenvillian aged gneisses, which are dispersed in inliers throughoutthe Eastern Cordillera of Colombia, Santander Massif and the SierraNevada de Santa Marta (Fig. 1; e.g.Restrepo-Pace et al., 1997), wherethey are considered to form part of the Chibcha Terrane (e.g.ToussaintandRestrepo,1994), although none crop outin Ecuador. ThePhanerozoicrocks, which form the focus of this review, can be separated into arelatively undifferentiated oceanic Late Cretaceous sequence, which isfaulted against older, differentiated continental crust.

The Late Cretaceous oceanicrocks formthe basementto the WesternCordillera and the forearcs of Colombia (Calima Terrane;Fig. 1; e.g. Kerret al., 1997) and Ecuador (Pallatanga-Pion Terrane; e.g.Vallejo et al.,2009). Geochemical, isotopic and geochronological data suggest thatthese ultramac and mac rocks formed in an oceanic hot-spot settingduring 9987 Ma (e.g.Kerr et al., 1997; Vallejo et al., 2006; Villagmez

et al., 2011), and that they are equivalent to the oceanic plateau rocksthat form the Caribbean Plate (e.g.Sinton et al., 1998). Field relation-ships andU-Pb zircon dates show that theoceanic plateauwas intrudedby an east-facing intra-oceanic arc prior to their collision with SouthAmerica in the Campanian (e.g.Feininger and Bristow, 1980; Vallejoet al., 2006), although this is not consistent with the plate reconstruc-tions ofLebratet al.(1987)and Pindelland Kennan (2009),whosuggestthat arcs at this time were west-facing and were also intruding thecontinental margin of South America. The accretion of allochthons ofthe Caribbean Large Igneous Province added at least 5x1061x107 km3

of new crust to the South American Plate (Cochrane, 2013).The Early Cretaceous continental margin hosts N-S trending linear

belts that are exposed within the Cordillera Central of Colombia,and the Cordillera Real of Ecuador (Fig. 1). Traversing eastwards from

the Campanian suture, these are M-HP/LT complexes of amphibolites,blueschists and eclogites, Early Cretaceous arc rocks of theQuebradagrande and Alao sequences, undifferentiated Palaeozoicrocks, which underwent anatexis in the Triassic, foliated EarlyCretaceous arc plutons, and large unfoliated Jurassic batholiths alongthe easternank of the Cordillera Central in Colombia, and CordilleraReal in Ecuador. Further east, the Eastern Cordillera of Colombia hasno equivalent topographic feature in Ecuador, and it includes the highplains of the Santander Massif in the north (Fig. 1). The SantanderMassif hosts the oldest segment of the latest TriassicEarly Cretaceouscontinental arc sequence, which has no equivalent assemblage inEcuador.Litherland et al. (1994)suggest that these belts are all intectonic contact, and these allochthonous units were juxtaposed duringcompression at 140120 Ma. Alternatively,Pratt et al. (2005)suggest

that the contacts are intrusive, and the rock units within Ecuador are

autochthonous, which is similar to the model proposed byVillagmezand Spikings (2013)andCochrane (2013)for Ecuador and Colombia.

The retro-arc foreland basins of the Middle and Lower MagdallenaValley basins in Colombia, and the Oriente Basin of Ecuador preserve arecord of the evolution of the margin during the JurassicRecent, andthese are utilised throughout the review to support or negate varioushypotheses.

3. Methodology

This review presents a very large quantity of data that was mainlypreviously peer reviewed and published, andthe details of the method-ologiesused by each study are provided in the respective publications. Asummary of the geochronological data and pertinent geochemical andisotopic data is presented in Tables 1 and 2, andthe complete geochem-ical dataset used to plot all of the geochemicalgures is provided as asupplementary le.

A majority of the geochronological, isotopic and geochemicaldata for the period spanning 290100 Ma (Mikovi et al., 2009;Villagmez et al., 2011; Cochrane, 2013; Cochrane et al., 2014a,b) wasacquired at laboratories at the Universities of Geneva and Lausanne(Switzerland), and at the Goethe Universitt Frankfurt. Accuracy andexternal reproducibility of the methods were determined by analysingi) Harvard 91500 (Wiedenbeck et al., 1995) and Pleovice zircon(Slma et al., 2008) during geochronology, ii) GJ-1 and Pleovice zircon(Slma et al., 2008) during Hf isotopic analyses, iii) standard JNdi-1 forNd isotopic analyses. Other geochronological analyses were performedat laboratories at Curtin University, the University of Grenoble (Rielet al., 2013), the Australian National University (Vinasco et al., 2006;Restrepo et al., 2011), the University of Arizona (Bustamante et al.,2010; Cardona et al., 2010; Weber et al., 2010) and WestflischeWilhelms-Universitt Munster (Lu-Hf;John et al., 2010). All zirconU-Pb dates that are used in the interpretations are concordant, anddene a single age population with respect to their mean squareweighted deviate statistic. The methodology used for each measure-ment is highlighted inTables 1 and 2.

Dense mineral assemblages andssion-track data for the period

after 100 Ma (e.g.Spikings et al., 2000, 2010) were obtained at laborato-riesat ETH-Zrich. A majorityof geochemicaland geochronological datafromthe Caribbean Large Igneous Provincewasobtainedfromlaborato-ries at the University of Geneva (40Ar/39Ar;Luzieux et al., 2006), theAustralian National University (U-Pb;Vallejo et al., 2006), and theUniversity of Lausanne (Mamberti et al., 2003).

4. Triassic: the disassembly of Pangaea and the formation of apassive margin

Triassic rocks within the Cordillera Real, AmotapeTerrane (Ecuador)and Cordillera Central (Colombia) are dominated by widely dispersedoutcrops of variably foliated granites, gneissic granites and migmatites,and less abundant amphibolites, ultramac rocks and meta-

sedimentary rocks (Fig. 2). Several studies have shown that themagmatic and metamorphic rocks that formed during the Triassic aregeochemically distinct from younger magmatic rocks, and formed in adifferent tectonic environment.

4.1. Historical perspective and occurrence

Triassic rocks within the Cordillera Real and Amotape Terrane ofEcuador include granitoids of the Tres Lagunas and Moromoro units,migmatites of the Sabanilla unit, geographically scattered amphiboliticdykes and sills (Piedras and Monte Olivo units), and sedimentaryrocks of the Piuntza unit (Fig. 2). The granites wererst described byColony and Sinclair (1932). Mapping by the British Geological Surveyduring 19861993 linked these occurrences into a semi-continuous

belt, and they were grouped into the Loja Terrane (Litherland et al.,

3R. Spikings et al. / Gondwana Research xxx (2014) xxxxxx



4/45
http://dx.doi.org/10.1016/j.gr.2014.06.004http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80


5/45

geochronological and geochemical data from the igneous and meta-morphic rocks in Ecuador (Table 1). The Piuntza unit consists of meta-morphosed and skarnied siliciclastic rocks, tuffs and limestones thathost Triassic bivalves (Litherland et al., 1994). The unit is exposedbeyond the structural limits of the Loja Terrane ofLitherland et al.(1994)along the easternank of the southern Cordillera Real (Fig. 2),where it is surrounded by the Jurassic Zamora Batholith although thenature of the contact is either unknown or unreported.

Widely dispersed and variably deformed Permian and Triassicmeta-granitoids, ultra-macmac rocks and metasedimentary rocksoccur within the northern Cordillera Central of Colombia (Fig. 2).These rocks were initially described by Hall et al. (1972),Feiningeret al. (1972)andGonzlez (1980), who considered them to be Permo-Triassic on the basis of K/Ar dates.Restrepo and Toussaint (1988)suggested that the Permo-Triassic rocks dened the basement of thefault-bounded Taham Terrane, and placed them within the CordilleraCentral Polymetamorphic Complex (Restrepo and Toussaint, 1982).The Taham Terrane ofRestrepo and Toussaint (1988)is bound by theOt-Pericos fault to the east, which separates it from Grenvillian agedmetamorphic basement of the Chibcha Terrane (e.g. Ordoez-Carmonaet al., 1999), and the San Jernimo Fault to the west, which separates itfrom the Quebradagrande Arc (Fig. 1).Maya and Gonzlez (1995)andVillagmez et al. (2011)group the Triassic metamorphosed igneousand sedimentary rocks into the Cajamarca Complex, which will beadopted in this manuscript. The geological map of Colombia (Gmezet al., 2007) reveals a paucity of Triassic lithologieswithin the CordilleraCentral southof the Ibagu Fault (Fig. 2). Vinasco et al. (2006), Martnez(2007), Cardona et al. (2010), Montes et al. (2010), Weber et al. (2010),Restrepo et al. (2011),Villagmez et al. (2011)andCochrane et al.(2014a)present a large quantity of geochemical data and concordantzircon U-Pb dates (Table 1,Fig. 2) from the Cajamarca Unit, conrmingthe Permo-Triassic crystallisation ages of the igneous rocks. Similar tothe Cordillera Realin Ecuador, the Permo-Triassic rocks of the CordilleraCentral intrude and are faulted against Palaeozoic metamorphic rockssuch as the La Miel Unit (e.g.Restrepo et al., 1991; Villagmez et al.,2011), although these will not be considered further in this review.Martnez (2007)report a series of metagabbros and amphibolites in

the northern Cordillera Central, which they attribute to a Triassicophiolitic sequence, referred to as the Aburr Ophiolite (Fig. 2).

Permo-Triassic igneous and metamorphic rocks have also beenrecognised in the Guajira Peninsula, Sierra Nevada de Santa Marta andat the base of boreholes drilled though thePlato-San Jorge Basin locatednorthof theCordillera Central (Cardona et al., 2010; Montes et al., 2010;Weber et al., 2010;Fig. 1). Granites from the basement of the Plato-San

Jorge Basin are mildly deformed, while the intrusions from the SierraNevada de Santa Marta are mylonitised.

4.2. Geochronology

4.2.1. Cordillera Real of Ecuador and Cordillera Central of ColombiaEarly attempts to date the Permo-Triassic crystalline rocks utilised

the K/Ar and Rb/Sr methods (e.g.Feininger et al., 1972; Hall et al.,1972McCourt et al., 1984; Restrepo et al., 1991; Litherland et al.,1994; Ordoez and Pimentel, 2002), resulting in a large scatter of agesspanning between the PermianTertiary due to variable degrees ofdaughter isotope loss. This review of geochronological work is restrictedto moreaccurate andpeer-reviewedmeasurementsof thecrystallisationages of granitoids and mac intrusions, which have been provided bynumerous concordant zircon and few monazite U-Pb dates (Table 1),obtained using TIMS, SHRIMP and LA-ICPMS. Unless otherwise stated,the LA-ICPMS and SHRIMP dates that are reported here were obtainedfrom the rims of zircons, and are considered to date either the mostrecent phase of magmatic crystallisation or the most recent metamor-phic event that crystallised zircon.

Fourteen metagranites and migmatites of the Tres Lagunas and

Sabanilla units in the Cordillera Real and Amotape Complex of

Ecuador yield concordant zircon and monazite U-Pb dates rangingbetween 207.6 9.2 Ma and 247.2 4.3 Ma (Figs. 2 and 3a;Table 1;Litherland et al., 1994; Aspden et al., 1995; Chew et al., 2008; Rielet al., 2013; Cochrane et al., 2014a). These dates overlap with concor-dant zircon U-Pb dates obtained from twenty six gneissic granites andpegmatites exposed in the Cordillera Central, Sierra Nevada de SantaMarta and the Guajira Peninsula in Colombia, which range between222 10 Ma and 288.1 4.5 Ma (Vinasco et al., 2006; Cardona et al.,

2010; Montes et al., 2010; Weber et al., 2010; Villagmez et al., 2011;Restrepo et al., 2011; Cochrane et al., 2014a;Figs. 2 and 3a). A majorityof these crystalline rocks are Triassic, although six rocks from northernColombia yield Permian ages. Ordonez-Carmona et al. (2001) reporta Sm-Nd whole rockgarnet isochron date of 226 17 Ma from agranulite in the Taham Terrane, which they interpret as a cooling agefollowing peak metamorphism. Multi-phase, plateau40Ar/39Ar dates(Table 1) from Triassic granites and migmatites in Colombia andEcuador (Spikings et al., 2001; Vinasco et al., 2006; Cochrane et al.,2014a) are younger than the U-Pb dates obtained from the same rocks,and Triassic dates span between 213.7 0.9 Ma and 243 4 Ma(Table 1). The 40Ar/39Ar dates reect the time of cooling of each rockthrough mineral-specic argon partial retention zones (i.e. 550300 C;hornblende, muscovite and biotite) subsequent to crystallisation andmetamorphic retrogression.

Concordant zircon U-Pb dates of amphibolites and a plagiogranitefrom the Cordillera Real and Cordillera Central range between216.6 0.4 Ma and 243 4 Ma (Fig. 3d;Noble et al., 1997; Vinascoet al., 2006; Martnez, 2007; Cochrane et al., 2014a). Within Ecuador,these are exposed as dykes and sills (e.g. the Piedras and Monte Olivounits;Fig. 2), whereas they are more massive in the Taham Terrane ofthe Cordillera Central. The youngest of these ages was obtained from aplagiogranite that formed by hydrothermal metamorphism of theAburr Ophiolite in northern Colombia, and thus is a minimum age forthe ophiolite (Martnez, 2007).

A majority of U-Pb dates of zircons extracted from meta-granitesand migmatites range between 240 and 230 Ma (Fig. 3a, d), and a com-parison withlatitude (Fig.3d) does notreveal anytrends for theTriassicperiod, with the exception of a possible increase in age in far northern

Colombia within the Guajira Peninsula. Permian ages are restricted toexposures in the Sierra Nevada de Santa Marta, and faulted blocks inthe region of the Ibagu Fault (Fig.2). This may reect exposure, or per-haps approximate the primary distribution of Permian magmaticintrusions.

Some concordant206Pb/238U dates determined by in-situ methodswere also obtained from the cores of zircon grains that were identiedusing cathodoluminescence. A frequency analysis of the distribution ofthese older dates from the granitoids and migmatitic leucosomes yieldsbroad peaks at 420580 Ma and 9501200 Ma (Fig. 3b), which are theages of protolith rocks and inherited grains. These age peaks are typicalof the distribution of dates obtained from detrital zircons from mostPalaeozoic terranes along western South America (e.g.Chew et al.,2007). The younger age group broadly corresponds with the age of the

Famatinian Arc, the Braziliano Orogeny and the timing of rifting duringthe fragmentation of Rodinia. The Famatinian arc (~510415 Ma;E.g. see zircon U-Pb ages presented inPankhurst et al., 2000; Cardonaet al., 2007; Chew et al., 2007; Bahlburg et al., 2009; Mikoviet al.,2009; Villagmez et al., 2011; Van der Lelij, 2013) formed during thesubduction of Pacic lithosphere beneath western South America sub-sequent to the fragmentation of Rodinia, and has been recorded inVenezuela, Colombia, Peru and Argentina (see previous citations).Inherited zircons with U-Pb dates spanning the 450650 Ma rangealso occur in Cretaceous and Tertiary sedimentary rocks of the AmazonForeland Basin in Ecuador (Martin-Gombojav and Winkler;, 2008),although intrusions of the Famatinian arc have not been recorded inEcuador, and within Colombia they are only recorded within the north-ern Cordillera Central (440470 Ma; La Miel orthogneiss;Villagmez

et al., 2011), Quetame, Floresta and Santander massifs (Horton et al.,




6/45

Table 1

Summary of data collected from Permo-Triassic rocks of Ecuador and Colombia.

Sample Unit Lithology Latitude N-S dm's'' Longitude W dm's'' 206Pb/238U age 2(Ma) MSWD 40Ar/39Ar age 2(Ma)

S-type granites and migmatitic leucosomes

Ecuador

Eastern Cordillera

09RC25 Tr. Lagunas metagranite S 1 23' 51" 78 21' 15" 233.7 0.8 1.109RC31 Tr. Lagunas metagranite S 0 22' 33" 78 08' 32" 234.4 0.9 0.809RC42 Sabanilla metagranite S 4 27' 43" 79 08' 52" 247.2 4.3 3.0

09RC53 Tr. Lagunas metagranite S 3 9' 24" 78 48' 45" 231.0 1.9 2.109RC44 Sabanilla paragneiss S 4 29' 02" 79 08' 55"09RC45 Sabanilla Paragneiss S 3 58' 41" 79 01' 15"09RC56 Tr. Lagunas metagranite S 1 23' 57" 78 22' 08" 235.0 1.5 3.011RC03 Agoyan fm. metagranite N 0 23' 24" 77 51' 44" 207.6 9.2 1.9Tres Lagunas Tr. Lagunas granite 227.3 2.2

Amotape Complex

09RC40 Moromoro migmatite S 3 42' 16" 79 51' 07" 237.7 5.2 4.6 214.6 0.9m

VI-08-12 La Bocana migmatite 226.0 1.3PU-08-10 La Bocana migmatite S 3 42' 58" 80 03' 18" 223.2 2.2AV-08-31 La Bocana migmatite 229.3 2.4AV-08-28d La Bocana migmatite S 3 40' 41" 79 54' 14" 225.7 6.5Moromoro Moromoro granite 227.5 0.8

Colombia

Central Cordillera

10RC04 Cajamarca metagranite N 4 19' 24" 75 12' 07" 277.6 1.6 1.210RC40 Cajamarca metagranite N 5 53' 13" 75 25' 28" 236.1 3.3 3.7 221.8 1.0m

10RC41 Cajamarca metagranite N 6 01' 08" 75 07' 28" 234.1 1.2 1.210RC42 Cajamarca metagranite N 5 59' 17" 74 55' 37" 244.6 2.4 2.310RC43 Cajamarca metagranite N 5 58' 34" 74 54' 02" 245.0 2.0 0.6 213.7 0.9m

10RC53 Cajamarca metagranite N 7 00' 56" 75 22' 28" 236.4 1.8 3.010RC66 Cajamarca qtz-Schist N 5 08' 20" 75 09' 47"10RC69 Cajamarca metagranite N 5 09' 27" 75 07' 57" 255.7 1.5 1.210RC71 Cajamarca pegmatite N 5 07' 34" 74 54' 38" 236.0 0.6 0.9DV65 Cajamarca metagranite N 5 59' 16" 74 55' 34" 240.9 1.5 0.6DV82 Cajamarca metagranite N 4 17' 16" 75 13' 59" 275.8 1.5 3.0DV02 Cajamarca paragneiss N 4 46' 42" 74 57' 54" 238582DV18 Cajamarca gneiss N 4 28' 19" 75 33' 18" 236.2 6.3 0.6DV19 Cajamarca quartzite N 4 28' 19" 75 33' 18" 2311163Abejorral Abejorral gneiss 250 10#

Palmitas Palmitas gneiss 240 4#

Amaga Amaga granite 227.6 4.5 1.4La Honda La Honda granite 218.7 0.3b

El Buey El Buey granite 219.3 0.3m

Manizales Manizales granite 229.7 0.5h

GSI1 Santa Isabel gneiss N 6 57' 34" 74 45' 13" 226.7 1.6 1.2GN1 Nechi gneiss N 8 10' 13" 74 46' 55" 236.4 6.6 2.1PALM-1 Palmas migmatite N 6 09' 14" 75 32' 36" 222 10#

Sierra Nevada de Santa Marta

A14 St. M. mylonite granite 288.1 4.5 1.0A48 St. M. mylonite granite 276.5 5.1 1.8EAM-12-05 St. M. mylonite granite 264.9 4.0 0.0

Plato-San Jorge Basin

Cicuco-2a unknown granite N 9 16' 25" 74 38' 53" 241.6 3.9 3.9Cicuco-3 unknown granite N 9 17' 39" 74 38' 52" 241.6 3.9 6.0Lobita 1 unknown granite N 9 18' 30" 74 41' 31" 239.6 2.9 0.6

Guajira Peninsular

AVO-03 Uray Gneiss gneiss 247.6 4.1 0.5AVO-06 Uray Gneiss gneiss 245.6 3.9 0.5

Amphibolites

Ecuador

10RC28 Chinchina amphibolite N 5 03' 05" 75 34' 25" 224.7 1.9 0.8

11RC04 Monte Olivo amphibolite N 0 23' 24" 77 51' 44"11RC10 Monte Olivo amphibolite S 1 23' 56" 78 22' 52" 231.9 3.2 1.611RC14 Piedras amphibolite S 3 39' 9" 79 50' 35" 222.7 6.3 1.9

JR148 Piedras amphibolite nr nr 221 17.0Colombia

10RC39 Santa Elena amphibolite N 5 54' 06" 75 24' 31"10RC39A Santa Elena amphibolite N 5 53' 52" 75 24' 37" 239.7 2.4 1.910RC50 Tr. Intrusive amphibolite N 6 09' 26" 75 44' 31"AC32B El Picacho plagiogranite 216.6 0.4 0.7CMK040A El Picacho meta-gabbroPadua Padua amphibolite 243 4h

Abbreviations: b (biotite), h (hornblende), m (muscovite), wr (whole rock); A/CNK (Molecular Al2O3/CaO + Na2O + K20); (La/Yb)n (normalized to N-MORB)87Sr/86Sr 2 s.d. (ext. reproducibility) = 0.0007%; 143Nd/144Nd =b0.0005%; 206Pb/204Pb = 0.12 %.Dates acquired by LA-ICPMS (Villagmez et al., 2011; Cochrane et al., 2014a), TIMS (Litherland et al., 1994; Aspden et al., 1995), SHRIMP (Vinasco et al., 2006; Restrepo et al., 2011).Monazite date.#Date obtained from the youngest zircon when a large spread of zircon ages were obtained due to xenocrystic contamination.


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://dx.doi.org/10.1016/j.gr.2014.06.004http://dx.doi.org/10.1016/j.gr.2014.06.004http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


7/45

Hf zircon 2 Nd w.r 2 (87Sr/86Sr)i wr 2 (206Pb/204Pb)i wr 2 18O () 2 Th/U zircon 2 A/CNK wr (La/Yb)n wr Publication

10.5 to 3.2 15.3 0.2 0.26 0.1 1.99 13.24 Cochrane et al. (2014a)11.0 to +3.2 15.1 0.2 0.04 0.1 1.40 13.50 Cochrane et al. (2014a)5.3 to 0.5 16.8 0.2 0.69 0.5 1.23 10.58 Cochrane et al. (2014a)

2.63 0.43 0.24 0.1 1.19 12.68 Cochrane et al. (2014a)1.37 4.65 Cochrane et al. (2014a)

Cochrane et al. (2014a)6.0 to +1.7 12.1 0.2 0.14 0.1 2.24 6.92 Cochrane et al. (2014a)16.3 to 9.0 15.1 0.2 0.01 0.0 Cochrane et al. (2014a)

Litherland et al. (1994)

7.5 to +0.8 0.42 0.5 2.38 11.36 Cochrane (2013)Riel et al. (2013)

1.85 Riel et al. (2013)0.13 Riel et al. (2013)0.10 1.50 Riel et al. (2013)

Aspden et al. (1995)

1.96 0.31 13.6 0.2 1.27 0.6 1.18 16.23 Cochrane et al. (2014a)6.57 0.66 17.4 0.2 0.08 0.1 1.73 8.19 Cochrane et al. (2014a)

9.5 to

0.2 13.1 0.2 0.23 0.1 1.27 11.49 Cochrane et al. (2014a)8.2 to +1.4 13.1 0.2 0.35 0.1 1.33 12.00 Cochrane et al. (2014a)11.7 to 3.1 0.42 0.4 1.36 15.70 Cochrane et al. (2014a)5.9 to +3.1 15.9 0.2 0.30 0.2 1.56 14.27 Cochrane et al. (2014a)

1.84 12.63 Cochrane et al. (2014a)3.16 0.7 15.6 0.2 1.10 0.2 1.70 12.81 Cochrane et al. (2014a)6.0 to +0.4 0.31 0.1 Cochrane et al. (2014a)5.9 to +0.7 0.26 0.2 Cochrane et al. (2014a)3.7 to +0.3 0.66 0.1 Cochrane et al. (2014a)

Villagmez et al. (2011)Villagmez et al. (2011)Villagmez et al. (2011)

0.82 Vinasco et al. (2006)0.25 Vinasco et al. (2006)0.30 Vinasco et al. (2006)

Vinasco et al. (2006)Vinasco et al. (2006)Vinasco et al. (2006)

0.19 Restrepo et al. (2011)0.23 Restrepo et al. (2011)0.24 Restrepo et al. (2011)

0.73 Cardona et al. (2010)0.57 Cardona et al. (2010)

Cardona et al. (2010)

nr Montes et al. (2010)nr Montes et al. (2010)nr Montes et al. (2010)

0.20 Weber et al. (2010)0.59 Weber et al. (2010)

13.31 0.25 9.83 0.70354 17.520938 0.20 0.1 0.66 1.41 Cochrane et al. (2014a)

0.61 2.59 Cochrane et al. (2014a)6.3 to +11.2 5.03 0.71470 18.707878 0.19 0.1 0.63 1.71 Cochrane et al. (2014a)15.00 0.29 9.79 0.70271 17.754038 0.32 0.2 0.61 0.81 Cochrane et al. (2014a)

0.52 Noble et al. (1997)

8.98 0.70430 18.119529 0.82 2.34 Cochrane et al. (2014a)4.8 to +10.0 4.13 0.70535 18.298843 0.62 2.02 Cochrane et al. (2014a)

10.18 0.70243 16.607997 0.50 0.49 Cochrane et al. (2014a)3.4 0.70448 0.97 8.00 Martnez (2007)8.4 0.61 0.64 Martnez (2007)

Vinasco et al. (2006)




8/45


9/45


10/45

2010; Van der Lelij, 2013). Field relationships (Litherland et al., 1994)suggest that the mesosomal rocks of the Triassic migmatites andS-type granites within Ecuador are considered to be sedimentary rocksof the Palaeozoic, fossil bearing Chiguinda and Isimanchi units of theCordillera Real (Fig. 2). Thesesparsely studied sequences yield a detritalzircon U-Pb age spectrum that has the same age peaks (Fig. 3c;Chewet al., 2008), although the tectonic setting within which these sequenceswere deposited is undetermined. The Brasiliano metamorphic belts

(Cordani et al., 2003) formed during the late Neoproterozoic amalgam-ation of Gondwana, and may have supplied some detritus to westernSouth America. However, these belts are located in eastern SouthAmerica, and a lack of evidence for detritus being sourced from theintervening Amazonia Craton suggests that the Brasiliano Orogenicbelts were not a major source region (Chew et al., 2008). Finally,all magmatism associated with Neoproterozoic extension is mac(e.g. the Puncoviscana fold belt in northwestern Argentina;Omariniet al., 1999), which ledChew et al. (2008)to suggest that it is unlikelythat these rocks were a major contributor of zircons to Palaeozoicsequences along western South America.

4.2.2. Comparison with the ages of Permian and Triassic rocks in Venezuela

and PeruVan der Lelij (2013)report concordant zircon U-Pb dates (LA-

ICPMS) fromfour granitoid intrusions anda dacitic lava fromthe MeridaAndes of Venezuela (Fig. 1). These dates range between 202.0 1.8 Ma(La Quinta Fm.) and 243.5 3.4 Ma, and overlap with dates obtainedfrom Colombia and Ecuador (Fig. 3d). No Permian concordant zirconU-Pb dates have been reported from the Merida Andes. Rhyolites andgranites of the El Baul massif in Venezuela yield zircon U-Pb dates thatspan between 283.3 2.5 Ma and 291.1 3.1 Ma (Viscarret et al.,2009),andazirconU-Pbageof272.22.6wasobtainedfromagraniticintrusion in the Paragauana Peninsula (Van der Lelij, 2013).

Voluminous, partly migmatised Late PermianTriassic magmaticintrusions are exposed throughout the southern and central EasternCordillera of Peru (Mikovi et al., 2009). Zircon U-Pb dates rangebetween 223 and 285 Ma (Fig. 3d;Mikoviet al., 2009; Reitsma, 2012),with a peak at 240260 Ma. The crystallisation ages show a southward

younging trend, and the oldest plutons south of 11.5S are younger than245 Ma. The Mitu Group of the central and southern Eastern Cordilleraof Peru hosts abundant Triassic sedimentary and volcanic sequences.Volcanic tuffs in the south yield concordant zircon U-Pb (LA-ICPMS)dates ranging between 234.3 0.3 Ma and 238.7 1.8 Ma (Mikoviet al., 2009; Reitsma, 2012), and Chew et al. (2005) report a zirconU-Pb age of 219.7 1.8 Ma from a rhyolite of the MituGroup in centralPeru. Detrital zircons extractedfrom oxidisedterrigeneoussedimentaryrocks of the Mitu Group yield minimum crystallisation dates rangingbetween 217.2 4.1 Ma and 250.7 4.9 Ma, which constrain theirmaximum stratigraphic ages (Reitsma, 2012). Several authors proposethat the Mitu Group was deposited within a rift (Mgard, 1978;Laubacher et al., 1988;Reitsma, 2012), andReitsma (2012)suggeststhat the rift formed in a back-arc basin setting.

Romero et al. (2013)recently published a concordant zircon U-Pbage of 243 0.1 Ma from a basalt exposed in Macab Island offshorenorthern Peru (~8S).

4.3. Geochemistry of the granites and migmatites

4.3.1. Cordillera Real of Ecuador and Cordillera Central of ColombiaMajor oxide, trace element and Rare Earth Element (REE) abun-

dances and oxygen isotope compositions (Table 1) have been obtainedfrom Permian and Triassic granites and migmatitic leucosomes fromColombia (Vinasco et al., 2006; Martnez, 2007; Cardona et al., 2010)and Ecuador (Litherland et al., 1994; Cochrane et al., 2014a). Theserocks span the boundaries of calcic and alkali-calcic differentiationtrends on the modied alkali-lime index of Peacock (1931;Fig. 4a),

with a compositional range of 6278 wt% SiO2. The same rocks plot

within the high-K calc-alkaline and calc-alkalineelds when comparingSiO2with K2O (Fig. 5a). The Triassic anatectites have strongly pera-luminous Aluminium Saturation Indices (ASI 0.972.38; calculatedusingManiar and Piccoli, 1989;Fig. 4b), while the Permian granitoidstendto cluster at slightly lower peraluminousand mildly metaluminousvalues (ASI 0.921.73, with a majority b1.1). The Triassic anatectitesyield elevated 18O quartz (Fig. 4c), which along with their high ASI in-dices places these granites and leucosomes within theS-type granite

eld ofChappell and White (1974)andHarris et al. (1997). N-MORBnormalized trace element abundances of Triassic granitoids fromEcuador and Colombia are identical (Fig. 4d), suggesting that there areno signicant along-strike changes in the fractionation and assimilationhistory of these high-SiO2melts. The trace elements are enriched inLight Ion Lithophile Elements (LILE), and negative Nb and Ta anomaliesare present in both the Permian and Triassic granitoids (Fig. 4e, f), sug-gesting that a subduction-derived component was incorporated intothese rocks. The Triassic granitoids yield slight negative Ba, Eu, Sr andTi anomalies, which suggest that plagioclase and Fe-Ti oxides have frac-tionated, and a positive Pb anomaly that may be derived from aprotolith within the continental crust. In contrast, the Permian granitesdo not yield Ba, Eu and Sr anomalies, although they do have negative Tianomalies, suggesting that they evolved via a different fractionationscheme. Trace element concentrations normalized to the compositionof average upper continental crust (Taylor and McLennan, 1995) plotclose to unity, corroborating the S-type character of these rocks(Fig. 4f). REE abundances in the Triassic granites and leucosomes nor-malized to N-MORB reveal light-REE enrichment with (La/Yb)n rangingbetween 2.3and 19.8, with a mildly positive correlation with206Pb-238Ucrystallisation age (Fig. 4h). (La/Yb)n ratios from Permian granites yielda larger range of 8.855.1, and the REE concentrations have a largerrange relative to N-MORB, compared to the Triassic rocks (Fig. 4g).

4.3.2. Comparison with Permian and Triassic rocks in Venezuela and PeruMigmatised granitoids within the Eastern Cordillera of Peru that

crystallised during 285223 Ma are high-SiO2granites, mildly meta-luminous to peraluminous (ASI 0.91.1;Fig. 4b), and yield K2O/Na2O

ratios that mainly range between 0.8 and 1.2 (Mikoviet al., 2009).Mikoviet al. (2009)report U-Pb zircon ages from monzogranites

in southern Peru (Cordillera de Carabaya) which range between 190and 216 Ma. These intrusions are geochemically distinct from theolder Permo-Triassic group (223285 Ma) because they are stronglyperaluminous (ASI 0.981.42;Fig. 4b), and yield anomalously highwhole rock K2O/Na2O ratios (0.552.22, with a majority N 1.20). Geo-chemically, the Permian and Triassic migmatites and granites of theCordillera Real of Ecuador and the Cordillera Central of Colombia(K2O/Na2O 0.772.93; ASI0.922.38) resemblethe Late Triassicgranitesof the Eastern Cordillera of Peru. Mikoviet al. (2009)combined thesemajor element characteristics with iron oxide number, SiO2(e.g.Frostet al., 2001) and trace element abundances, to classify the Permo-Triassic (285223 Ma) and Late Triassic plutons as late- to post-

orogenic.The shift froma well characterised, calc-alkaline Carboniferous arc in

Peru to Permian post-tectonic alkali feldspar granites (Mikoviet al.,2009) and ultimately alkaline bimodal volcanic rocks of the MituGroup is characteristic of lithospheric thinning (e.g.Xu et al., 2007).Mikovi et al. (2009) suggest that the Permo-Triassic granitoids(223285 Ma)formed by dehydration melting of the lowercrustduringbasaltic underplating (e.g. Sisson et al., 2005), which was driven by de-compression subsequent to break-off of the Carboniferous slab. Granit-oid intrusions and rift-related magmatism (Dalmayrac et al., 1980)initiated in the MiddleTriassic, forming the Mitu Riftand bimodal volca-nic rocks of the Mitu Group (Reitsma, 2012). Finally, the highlyperaluminous Late Triassic granites formed by melting of the fertile up-permost crust, consisting of an igneous protolith and a substantial sed-

imentary component.




11/45

4.4. Geochemistry of the amphibolites

A comparison of theabundance of K2O and SiO2 (Fig. 5a)inallofthemagmatic rocks that yield Triassic ages reveals the bimodal nature ofmagmatism within northwestern South America as part of Pangaea be-tween216.60.4Maand2434Ma( Fig.3d).Amphibolitic dykes andmassivemetagabbros fromthe Cordillera Realand Amotape Complex ofEcuador (Chinchina, Monte Olivo and Piedras units) and the Cordillera

Central of Colombia (Santa Elena, Padua and Aburr) yield low K2O(b0.5 wt%) relative to SiO2(4655 wt%), placing them within the tho-leiitic eld ofPeccerillo and Taylor (1976;Fig. 5a). However, a compar-ison of the immobile elements Th and Co (Fig. 5b) suggests thatamphibolitic dykes straddle the tholeiite and calc-alkalineelds, whilethe massive metagabbros of the Aburr Ophiolite plot in the tholeiiteeld. These discrepancies suggest that the amphibolitic dykes may bepartially altered, although this was not visible in hand specimen or inthin sections. The amphibolites and metagabbros have values of Zr/TiO2that are lower than 0.01 (Fig. 5c), placing them within the sub-alkaline basalteld, implying that the tholeiitic nature is primary. Theamphibolitic dykes are enriched in Ti relative to V (Fig. 5d), and plotin the MORB or back arc basin basalt (BABB)eld ofShervais (1982).However, the massive metagabbros of the Aburr Ophiolite (El Picachometagabbros) that yield very low K2O abundances of 0.020.12 wt%(Fig. 5a), plot closer to the arc eld. LILE abundances within theamphibolitic dykes from Colombia and Ecuador (Fig. 5e) are enriched(up to ~100 times) relative to N-MORB and lack signicant Nb and Taanomalies. The HFSE elements plot close to parity with N-MORB,which is consistent with the tectonic discrimination plots. Similarly,N-MORB normalised REE plots (Fig. 5f) for the amphibolitic dykes plotclose to MORB compositions, although the LREE are slightly enrichedwith (La/Yb)n ratios varying between 0.59 and 3.16, while the massivemetagabbros of the Aburr Ophiolite yield approximatelyat REE pat-terns that are slightly depleted relative to N-MORB. N-MORB normal-ised La/Yb ratios from all of these rock sequences show a progressivereduction with crystallisation age from 243 4 Ma to 216.6 0.4 Ma(Fig. 5g). Finally, whole rock Ndivalues for the amphibolitic dykesand the metagabbros range between 3.40 and 10.18, and become

more juvenile with younger crystallisation ages (Fig. 5h). The most ju-venile rocks are characteristic of MORBand BABBisotopic compositions.

A single amphibolite from theMonte Olivo unit in theCordillera Realof Ecuador yields a whole-rock87Sr/86Sri of 0.7147 (Fig.5i), which isex-tremely high relative to its143Nd/144Ndi of 0.5126and low La/Yb ratio of1.71 (Cochrane et al., 2014a). This is consistent with low temperaturealteration, which has preferentially mobilized the LILE but had a mini-mal effect on the REE.

4.5. Zircon Hf isotope geochemistry

Cochrane et al. (2014a) report Hf isotopic compositions from zirconsextracted from eighteen migmatitic leucosomes, more massive granit-oids and amphibolitic dykes throughout the Cordillera Real of Ecuador

and the Cordillera Central of Colombia (Table 1). The zircons, whichhave been dated by LA-ICP-MS (U-Pb), yield a large range of weightedmeanHfivalues of +15 and 20 (Fig. 6), which are consistent withcrustal recycling and the addition of new continental crust (Collinset al., 2011; Cochrane et al., 2014a).

4.5.1. Zircon Hf isotope geochemistry of the granites and

migmatitic leucosomes

Single leucosomes of migmatites and peraluminous granites gener-ally yield high, intra-sample variations (e.g.Hfi+3 to 11;Fig. 6a)within coeval magmatic rims that surround variably aged xenocrysticcores (Cochrane et al., 2014a), and within samples that lack oldercores. These variations are too large for magmatic zircons thatcrystallised from a single, well-mixed source (e.g.Gerdes et al., 2002),

andCochrane et al. (2014a)conclude that they are mainly derived

from multiple crustal sources, while some variation found in thexenocrystic bearing zircons may be due to the fractionation of Hf iso-topes between cores and overgrowths during melting (e.g.Gerdes andZeh, 2009). In contrast, a small proportion of granites yieldHfi (zircon)values that statistically dene a single population, suggesting that theywere derived from a distinct, homogeneous source. No correlation isfound between theHfivalues obtained from the rims of xenocryst-bearing and xenocryst- free zircons, and crystallisation age, which is

notsurprising giventhe heterogeneous natureof the source rocks with-in the crust.Hfi values obtained from xenocrystic zircon cores (Fig. 6a)span a large range and are representative of the sedimentary protolithsthat melted to form the anatectites. The large heterogeneity inHficor-roborates the large range in238U/206Pb dates of the protoliths.

4.5.2. Zircon Hf isotope geochemistry of the amphibolitesCochrane et al. (2014a)reportHfi(zircon) values from four am-

phibolites which show a negative correlation with crystallisation age(Fig. 6b). The two older amphibolitic dykes (240232 Ma) yield alarge range in Hfi, with juvenile values (7.4 to 10) obtained frompatchy or unzoned (cathodoluminescence) zircons, and crustal compo-sitions (3.6to4.8) from zirconsthatexhibitoscillatory zoning, sim-ilar to the zircons extracted from the anatectites (Fig. 6c). The two

youngest amphibolites (225223 Ma) de

ne single populations ofHfi(13.315.0) fromunzonedzircons, which approach the depletedmantlearray. Crustal contamination of the mac melts during emplacementwas an important process in the petrogenesis of the older amphibolitesprior to ~ 225 Ma. However, there is no evidence for the assimilation ofsignicant continental crust after ~225 Ma. This interpretation is sup-ported by a negative correlation betweenHfi(zircon) and (La/Yb)n,(Fig. 6b) suggesting that the mac dykes trend towards MORB compo-sitions with a depleted mantle source, during 240223 Ma.

4.5.3. Comparison with Zircon Hf isotope compositions in PeruHf isotopic compositions obtained from Permian and Triassic

peraluminous granitoids of the Eastern Cordillera of Peru yield noclear trends with time, and Hft values range between 6 and 8

(Fig. 6d;Mikoviand Schaltegger, 2009). This range overlaps withthat obtained from Permian and Triassic granitoids of Ecuador andColombia, and individual, intra-sample spot analyses show a largespread, reecting the heterogeneity of the source rocks. Mean zirconHfi values from the Early and Late Triassic of 0.02 1.56 and1.96 1.56 suggest that this period was characterised by the addi-tion of isotopically juvenile, mantle derived magmas, which were un-derplating previously attenuated continental crust (Mikovi andSchaltegger, 2009). Calculated crustal Hf model ages for the granitoidswithin the Eastern Cordillera of Peru range between 1.4 and 1.0 Ga(Mikoviand Schaltegger, 2009), suggesting that the source rockswere basement that formed within the Sunsas Orogeny during theamalgamation of Rodinia.

4.6. Thermal histories during the Triassic

Numerous K/Ar dates have been obtained from Palaeozoic andTriassic magmatic rocks in the Cordillera Real of Ecuador (Litherlandet al., 1994) and the Cordillera Central of Colombia (Feininger et al.,1972; Hall et al., 1972; McCourt et al., 1984; Aspden et al., 1987).However, these dates cannot be unambiguously interpreted ascrystallisation ages given the potential for isotopic disturbance bythermally activated diffusive loss anduid assisted loss of the daughterisotopes. Within Ecuador, muscovite, biotite and whole rock K/Ar datesof the Tres Lagunas Granite and Sabanilla unit (migmatites) rangebetween 100 and 50 Ma (Litherland et al., 1994), revealing a cleardisturbance to the isotopic system. Unfortunately, we cannot extractuseful timeTemperature (tT) information from these data because

the degree of daughter isotope loss cannot be quantied. Nevertheless,




12/45

Ibag

uerela

ted

0

1N

2N

4N

8N

78W

2S

3S

4S

5S

80W

79W 227.32.2

PF

SabanillaMigmatite

Moromoro Migmatite& Piedras Amphibolite

Monte OlivoAmphibolites

Chinchina Stock

233.70.8

235.01.5

239.72.4

236.13.3221.81.0

236.00.6

275.81.5

231.93.2

247.24.3*

222.76.3

237.75.2214.60.9

234.40.9

234.11.2

238 - 2800

2345 - 2600

227.50.8

223-229

227.64.5

224.71.9

Tres LagunasGranite

Cajamarca Amphibolites

236.41.8

245.02.0*213.70.9

240.91.5

207.69.2

PF

CAF

244.62.4Santa ElenaAmphibolites

255.71.5

231.01.9

7476 75

ZumbaOphiolite

Aburr Ophiolite

0 100 kmColombia

Ecuador

Peraluminous granites and migmatites

(e.g. Tres Lagunas Granite, Sabanilla

Migmatite)

Amphibolites and ultramafic rocks

(e.g. Piedras unit, Monte Olivo unit)

Metalluminous granitoids

(mainly granodiorite).

Continental arc intrusions

Jurassic

Triassic

Undifferentiated para- and

ortho-, schists and gneisses

(Ecuador: Agoyn, Chiguinda, Piuntza

units.

Colombia: Cajamara Unit)

Palaeozoic - Triassic

244.62.4U-Pb zircon, monazite

LA-ICPMS, intrusions

233.74.8 - 2.6 GaU-Pb zircon, LA-ICPMS,

detrital zircons,

metasedimentary rocks

213.70.940Ar/39Ar plateau date

hornblende, muscovite,

biotite, intrusions

236.26.3

236.46.6

277.33.0U-Pb zircon

SHRIMP, intrusions220 - 600

226.71.9

25010*

229.70.5

2404*

216.60.4

2434

277.61.6

227.64.5

218.70.3219.30.3

220 - 1200

22118

OPF

M

I

P

Q

L

City

22210*

SJF

IF

247.24.3*Also dated by apatite

U-Pb for t-T analysis

LF

BF

PiuntzaUnit


http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://dx.doi.org/10.1016/j.gr.2014.06.004http://dx.doi.org/10.1016/j.gr.2014.06.004http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80


13/45

McCourtet al. (1984)and Litherland et al. (1994)interpret these datatoreect continental collision at ~ 120 and 6555 Ma.

Spikings et al. (2000, 2001)andVillagmez and Spikings (2013)present40Ar/39Ar (white mica, biotite, alkali feldspar) and ssion track(zircon, apatite) data from Triassic magmatic rocks in Ecuador andColombia, which were used to construct tT paths. However, the colli-sion of the Caribbean Large Igneous Province with South America at~75 Ma drove more than 350 C of cooling, and Triassic thermal histo-

ries were not preserved in the isotopic systems.Recently, Cochraneet al. (2014b) published apatite U-Pb data from aTriassic leucosome of the Sabanilla Unit (migmatite 09RC42; Table 1) ofsouthern Ecuador. The authors demonstrate that Pb was lost from theapatite grains by thermally activated diffusion, and thus the dates canbe combined with grain sizes and the diffusion properties of Pb in apa-tite to generate a series of plausible tTpaths(Fig.7a) usinga computedMonte Carlo algorithm, at temperatures N350 C. Those paths revealrapid cooling subsequent to anatexis at ~250 Ma. The leucosome subse-quently remained at temperatures lower than the Pb Partial RetentionZone throughout the Triassic. The same method has been applied to aperaluminous Triassic granite (10RC43;Table 1;Fig. 7b) from the Caja-marca Complex of central Colombia. Similarly, the computed tT pathsalso reveal rapid cooling subsequent to Triassic anatexis, after whichthe rock was colder than the apatite Pb Partial Retention Zone. Datafrom the latter sample are new, although the methodology is identicalto that presented inCochrane et al. (2014b), and the dataare presentedin a supplementaryle. Rapid cooling during the Triassic corroboratesthe indistinguishable U-Pb dates from apatites with a large range ingrain size (Fig. 7b).

Rapid cooling from anatectic temperatures to less than ~380 C isprobably mainly a consequence of thermal relaxation subsequent tothe removal of the heat source at a local geographic scale.Some compo-nent of cooling may also be a consequence of exhumation, and giventhat the samples remained colder than ~380 C throughout the remain-der of the Triassic and the Jurassic, it is likely that these particular sam-ples were at depths of15 km within the crust after ~220 Ma. Rapidcoolingduring ~250220 Ma temporally coincideswith a signicant in-crease in zirconHfiobtained from the amphibolitic dykes (Fig. 7a, b).

4.7. Interpretation: Permian and Triassic

4.7.1. Arc magmatism and metamorphism during 290240 Ma along

western PangaeaThe exposure of Permian andearly Triassic (290240 Ma) granitoids

within the Northern Andes is restricted to the Sierra Nevada de SantaMarta, Guajira Peninsula and the central Cordillera Central withinColombia, while almost all intrusions within the Cordillera Real andAmotape Terrane of Ecuador are240 Ma. The Aluminium SaturationIndex of the alkali-calcic to calcic granites that crystallised during290240 Ma straddles the peraluminous and metaluminous elds,while the d18O values of 14%17suggest that these rocks formed bypartialmelting of sedimentaryrocks. Zirconsyield magmatic Th/Uratios

of 0.261.27, suggesting that they have not recrystallized during subse-quent metamorphic events (Table 1;Fig. 8). The whole rock trace ele-ment abundances are characteristic of subduction related magmatism.Magmatism during thisperiod was not accompaniedby macdykeem-placement, and all rocks yield N58 wt% SiO2. The 290240 Ma graniteswithin the Sierra Nevada de Santa Marta (Cardona et al., 2010) andthe Cordillera Central of Colombia (Villagmez et al., 2011) areinterpreted to have formed above an east dipping Pacic subductionzone beneath Pangaea (Cochrane et al., 2014a;Fig. 8).

Plate reconstructions of western Pangaea during the Permian toEarly Triassic (Elas-Herrera and Ortega-Gutirrez, 2002; Weber et al.,2007) juxtapose Acatln, Oaxaquia and the Chortis Block againstnorth-western South America (Fig. 9). At the present time, Oaxaquia isconsidered to underlie central and southern Mexico, including theMaya Block. The Maya Block hosts extensive, undeformed Permiangranites, and deformed and foliated Permian granitic gneisses andmigmatites (Chiapas Massif; e.g. Solari et al., 2008), which underwent

rock uplift and erosion during the early Triassic (Schaaf et al., 2002).The undeformed and deformed Permian granites yield concordant U-Pb (zircon) dates ranging between 289 and 255 Ma (Yanez et al.,1991; Solari et al., 2001; Elas-Herrera and Ortega-Gutirrez, 2002;Ducea et al., 2004; Weber et al., 2007; Kirsch et al., 2012;Ortega-Obregon et al., 2013; Kirsch et al., 2014) and are considered toform part of a continental arc (e.g.Torres et al., 1999). We suggestthat the Permian, peraluminous granitoids exposed within north-western South America formed within the same tectonic regime, andare a continuation of the Permian belt that is exposed in southernMexico (e.g.Centeno-Garcia and Keppie, 1999; Dickinson and Lawton,2001; Kirsch et al., 2014). Remnants of Permian magmatism have alsobeen found within the Sierra de Perij (Dasch, 1982), Paraguana Penin-sula (Van der Lelij, 2013) and the El Baul Massif in Venezuela (Viscarretet al., 2009). The Permian granites and migmatites exposed in Colombiaformed in a different tectonic regime to similar lithologies that formedduring the Triassic, which was dominated by extension (see nextsection).

Weber et al. (2007) report concordant zircon U-Pb dates frommigmatites of the Maya Block (Chiapas Massif) of 251.8 3.8254.0 2.3 Ma, which they interpret to be a result of MP-HT metamorphismduring compression, and stacking within an orogenic wedge. The colli-sion event post-dates the amalgamation of Pangaea, which is recordedalong the diachronous Ouachita-Marathon suture (Fig. 9) that hadformedbytheEarlyPermian. Weberetal.(2007) attribute the compres-sional event to closure of a marginalbasin,which had previously formedduring extension that accompanied Permian arc magmatism. Similarly,Cardonaet al. (2010)attribute anatexis at ~250 Ma in the Sierra Nevadade Santa Marta (Fig. 1) to a compressional event. The cause of compres-

sion has been attributed to either terrane accretion, subduction of thick-ened, topographically prominent and buoyant oceanic lithosphere (e.g.Weber et al., 2007), or increased plate coupling during the waningstages of the amalgamation of Pangaea (Cardona et al., 2010).

The high-SiO2, Permo-Early Triassic granites and alkali feldspargranites of the Eastern Cordillera of Peru are interpreted to have formedwithin a continental arc setting during extensive lithospheric thinning(Sempere et al., 2002; Mikoviet al., 2009). Roll-back is considered tohavedriven anti-clockwise rotation of the forearc, dextral displacementand southward younging of the onset of extension along the Peruvianarc, all of which commenced at ~280 Ma (Mikoviet al., 2009). Perm-ian arc magmatism in Peru is consistent with arc magmatism that issporadically preserved within the Northern Andes, and moreabundant-ly within theconjugate margin that is nowexposed in southern Mexico.

Weber et al. (2007)propose that the Permian arc intrusions within theMaya Block were also emplaced during lithospheric thinning. However,their interpretation is based on the tectonic-switching model ofCollins(2002), who state that a thinned, hot lithosphere is a prerequisite forsubsequent crustal thickening and anatexis within a period of ~ 20 My.

Geochemical and isotopic evidence from Permo-Triassic intrusions inPeru provides no evidence for compression at ~250 Ma (Mikoviet al.,2009). Coeval compression and extension along different regions of thewestern margin of Gondwana, within Pangaea, spatially correlates with

Fig. 2.Geology of the Cordillera Real and Amotape Complex of Ecuador, and the Cordillera Central of Colombia, showing the distribution of Palaeozoic and Triassic rocks. The Jurassiccontinental arc is also shownfor reference. Concordant Permian andTriassic zircon U-Pband plateau 40Ar/39Ardatesand their uncertainties(2) obtained by various analyticalmethods(see Table 1) areshown(see referencesin Table1). ThePalaeozoic andTriassicrockswithin theCordillera Realof Ecuador aregrouped together withinthe Loja Terraneby Litherland et al.(1994). Cities, I: Ibagu, L: Loja, M: Medellin, P: Pasto,Q: Quito.Faults. Faults:BF: Baos Fault, CAF: Cauca-AlmaguerFaults, LF:Llanganates Fault, OPF: Ot-Pericos Fault, PF:Peltetec Fault.

Map compiled fromLitherland et al. (1994)andGmez et al. (2007).




14/45

300

290

280

270

260

250

240

230

220

210

200

1906 4 2 0 2 4 6 8 10 12

Ecuador Colombia

Triassic

Permian

Latitude (degrees)

Zircon U-Pb

age (Ma)2

Metagranite,migmatite

Amphibolite,

plagiogranite

A

SNSM

GPMA

EBM

PP

PeruVenez

NorthSouth

206Pb/238U age (Ma)

Frequency

10

0

100

220

280

340

400

460

520

580

640

700

760

820

880

940

1000

1060

1120

1180

Colombia (n = 360)

Ecuador (n = 288)

Colombia (n = 104)

Permian anatectite

Triassic anatectite

0

1

2

3

4

5

6

7

8

9

200 225 250 275 300

Colombia (n = 27)

Ecuador (n = 14)

PermianTriassic

Anatectites

206Pb/238U age (Ma)

Frequency

0

1

2

3

4

5

6

7

8

9

360

420

480

540

600

660

720

780

840

900

960

1020

1080

1140

1200

1260

1320

1380

1440

480 - 630 Ma 960 - 1200 Ma

N = 99

Palaeozoic metasedimentaryunits (melanosome)

206Pb/238U age (Ma)

Frequency

A

B C

D

Fig.3. A) 206Pb/238U agehistogramfor thetime ofanatexis ormetamorphiczircon growthfor granites, metagranites andmigmatites (leucosomes) inEcuador andColombia. B) 206Pb/238Uage histogram for Permian and Triassic S-type granitoids and migmatites (leucosomes) from the Cordillera Real (Ecuador) and the Cordillera Central (Colombia). Ages are single spotzircon ages determined using LA-ICPMS (Villagmez et al., 2011; Cochrane et al., 2014a), SHRIMP and SIMS (Vinasco et al., 2006; Chew et al., 2008; Restrepo et al., 2011).C) 206Pb/238U age histogram for detrital zircons from the Palaeozoic Chiguinda and Isimanchi metasedimentary units of the Cordillera Real of Ecuador (Chew et al., 2008).D)A comparison ofPermian and Triassic concordant zircon and monazite U-Pb dates with latitude along the Cordillera Real of Ecuador, Cordillera Central, Guajira Peninsula and the Sierra Nevadade Santa Marta of Colombia. The ranges of concordant zircon U-Pb dates obtained from granitoid intrusions and volcano-sedimentary rocks from Venezuela ( Van der Lelij, 2013) andthe Eastern Cordillera of Peru (Mikoviet al., 2009; Reitsma, 2012) are shown for comparison. Data and citations are presented inTable 1. EBM: El Baul Massif, GP: Guajira Peninsula,

MA: Merida Andes, PP: Paraguana Peninsula, SNSM: Sierra Nevada de Santa Marta.


http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/

spikings et al. 2014

Documents