Para a realização dessa pesquisa foi desenvolvido um estudo da espessura crustal nos Andes setentrionais e na parte norte dos Andes centrais (entre 10ºN e 9ºS). Foram usadas reflexões da Moho de seis sismos de profundidade média para achar a espessura crustal na área de estudo. As ondas pmP e smS (reflexões na Moho) são encontradas como precursoras das fases profundas pP e sS de sismos registrados em estações localizadas a distâncias telessísmicas. A metodologia utilizada exigia sismos com profundidade maiores que 100 km e magnitude maiores que 6. Entretanto, devido à baixa amplitude apresentada pelas ondas pmP e smS, foi necessário realizar um empilhamento de vários sismogramas para realçar as chegadas das ondas analisadas. A área estudada foi dividida em três partes: área norte entre 6ºN e 0º(na Colômbia), área central entre 2ºS e 5ºS (no Equador) e área sul entre 6ºS e 9ºS (na borda entre Peru e Brasil). Na área norte foram encontradas espessuras crustais entre 26 e 56 km em quatro pontos de reflexão de dois sismos; na área central foram encontradas espessuras da crosta entre 40 e 63 km para três pontos de reflexão de um sismo; finalmente, na região sul, foram encontradas espessuras crustais entre 35 e 40 km para 7 pontos de reflexão de 3 sismos. Foi obtido a relação Vp/Vs = 1.79 ± 0.16 para a região norte dos Andes, usando as diferenças de tempo sS-smS e pP-pmP. Os resultados obtidos vão aumentar o banco de dados de espessura crustal nos Andes setentrionais e na parte norte dos Andes centrais, como também possibilitaram uma melhora do modelo de espessura crustal desenvolvido por Assumpção et al. (2013) para América do Sul, permitindo assim a analise e comparação entre os dados obtidos nesse estudo e o modelo crustal de Assumpção et al. (2013). / In this work we studied the crustal thickness in northern Andes and northern central Andes (between 10ºN and 9ºS). We analyzed the reflections from the underside of the Moho for six intermediate earthquakes occurred in the study area to estimate the crustal thickness at the bounce point. The pmP and smS phases (reflections at the Moho) are identied as precursors to the depth phase pP and sS, respectively, when recorded at teleseismic distances. This method require events of magnitude > 6 and depth > 100 km. In order to better identify those two reflections, it was necessary to stacking traces from different stations. The study area was divided in three sections: northern area between 6ºN - 0º (Colombia), in this area we analyzed two earthquakes and the crustal thicknesses in four bounce points were found; central area between 2ºS- 5ºS (Ecuador), in this area we studied one earthquake and the crustal thicknesses in three bounce points were found; southern area between 6ºS - 9ºS (Peru-Brazil border), where we analyzed three earthquakes and the crustal thicknesses in seven bounce points were found. Crustal thickness found in the southern area varied from 35 to 40 km, from 40 to 63 km in the central area and from 26 to 56 km in the northern area. We estimated a Vp/Vs ratio of 1.79±0.16 for the northern Andes using the time differences sS-smS and pP-pmP relation. Our analysis complements the data base of Moho depths in the Northern Andes and also, it allows a comparison with the model of crustal thickness estimated by Assumpção et al. (2013) confirming the Moho depth on both studies.

Para a realização dessa pesquisa foi desenvolvido um estudo da espessura crustal nos Andes setentrionais e na parte norte dos Andes centrais (entre 10ºN e 9ºS). Foram usadas reflexões da Moho de seis sismos de profundidade média para achar a espessura crustal na área de estudo. As ondas pmP e smS (reflexões na Moho) são encontradas como precursoras das fases profundas pP e sS de sismos registrados em estações localizadas a distâncias telessísmicas. A metodologia utilizada exigia sismos com profundidade maiores que 100 km e magnitude maiores que 6. Entretanto, devido à baixa amplitude apresentada pelas ondas pmP e smS, foi necessário realizar um empilhamento de vários sismogramas para realçar as chegadas das ondas analisadas. A área estudada foi dividida em três partes: área norte entre 6ºN e 0º(na Colômbia), área central entre 2ºS e 5ºS (no Equador) e área sul entre 6ºS e 9ºS (na borda entre Peru e Brasil). Na área norte foram encontradas espessuras crustais entre 26 e 56 km em quatro pontos de reflexão de dois sismos; na área central foram encontradas espessuras da crosta entre 40 e 63 km para três pontos de reflexão de um sismo; finalmente, na região sul, foram encontradas espessuras crustais entre 35 e 40 km para 7 pontos de reflexão de 3 sismos. Foi obtido a relação Vp/Vs = 1.79 ± 0.16 para a região norte dos Andes, usando as diferenças de tempo sS-smS e pP-pmP. Os resultados obtidos vão aumentar o banco de dados de espessura crustal nos Andes setentrionais e na parte norte dos Andes centrais, como também possibilitaram uma melhora do modelo de espessura crustal desenvolvido por Assumpção et al. (2013) para América do Sul, permitindo assim a analise e comparação entre os dados obtidos nesse estudo e o modelo crustal de Assumpção et al. (2013). / In this work we studied the crustal thickness in northern Andes and northern central Andes (between 10ºN and 9ºS). We analyzed the reflections from the underside of the Moho for six intermediate earthquakes occurred in the study area to estimate the crustal thickness at the bounce point. The pmP and smS phases (reflections at the Moho) are identied as precursors to the depth phase pP and sS, respectively, when recorded at teleseismic distances. This method require events of magnitude > 6 and depth > 100 km. In order to better identify those two reflections, it was necessary to stacking traces from different stations. The study area was divided in three sections: northern area between 6ºN - 0º (Colombia), in this area we analyzed two earthquakes and the crustal thicknesses in four bounce points were found; central area between 2ºS- 5ºS (Ecuador), in this area we studied one earthquake and the crustal thicknesses in three bounce points were found; southern area between 6ºS - 9ºS (Peru-Brazil border), where we analyzed three earthquakes and the crustal thicknesses in seven bounce points were found. Crustal thickness found in the southern area varied from 35 to 40 km, from 40 to 63 km in the central area and from 26 to 56 km in the northern area. We estimated a Vp/Vs ratio of 1.79±0.16 for the northern Andes using the time differences sS-smS and pP-pmP relation. Our analysis complements the data base of Moho depths in the Northern Andes and also, it allows a comparison with the model of crustal thickness estimated by Assumpção et al. (2013) confirming the Moho depth on both studies.

The modern foreland basin straddling the eastern margin of the Andean orogen is the prime example of a retro-arc foreland basin system adjacent to a subduction orogen. While widely studied in the central and southern Andes, the spatial and temporal evolution of the Cenozoic foreland basin system in the northern Andes has received considerably less attention. This is in part due to the complex geodynamic boundary conditions, such as the oblique subduction and accretion of the Caribbean plates to the already complex interaction between the Nazca and the South American plates. In the Colombian Andes, for example, a foreland basin system has been forming since ~80 Ma over an area previously affected by rift tectonics during the Mesozoic. This setting of Cenozoic contractile deformation superposed on continental crust pre-strained by extensional processes thus represents a natural, yet poorly studied experimental set-up, where the role of tectonic inheritance on the development of foreland basin systems can be evaluated. However, a detailed documentation of the early foreland basin evolution in this part of the Andes has thus far only been accomplished in the more internal sectors of the orogen. In this study, I integrate new structural, sedimentological and biostratigraphic data with low-temperature thermochronology from the eastern sector of the Colombian Andes, in order to provide the first comprehensive account of mountain building and related foreland basin sedimentation in this part of the orogen, and to assess as to what extent pre-existent basement anisotropies have conditioned the locus of foreland deformation in space and time. In the Medina Basin, along the eastern flank of the Eastern Cordillera, I integrated detailed structural mapping and new sedimentological data with a new chronostratigraphic framework based on detailed palynology that links an eastward-thinning early Oligocene to early Miocene syntectonic wedge containing rapid facies changes with an episode of fast tectonic subsidence starting at ~30 Ma. This record represents the first evidence of topographic loading generated by slip along the principal basement-bounding thrusts in the Eastern Cordillera to the west of the basin and thus constrains the onset of mountain building in this area. A comprehensive assessment of exhumation patterns based on zircon fission-track (ZFT), apatite fission-track (AFT) analysis and thermal modelling reveals the location of these thrust loads to have been located along the contractionally reactivated Soapaga Fault in the axial sector of the Eastern Cordillera. Farther to the east, AFT and ZFT data also document the onset of thrust-induced exhumation associated with contractional reactivation of the main range-bounding Servita Fault at ~20 Ma. Associated with this episode of orogenic growth, peak burial temperature estimates based on vitrinite reflectance data in the Cenozoic sedimentary record of the adjacent Medina Basin documents earlier incorporation of the western sector of the basin into the advancing fold and thrust belt. I combined these new thermochronological data with published AFT analyses and known chronologic indicators of brittle deformation in order to evaluate the patterns of orogenic-front migration in the Andes of central Colombia. This spatiotemporal analysis of deformation reveals an episodic pattern of eastward migration of the orogenic front at an average rate of 2.5-2.7 mm/yr during the Late Cretaceous-Cenozoic. I identified three major stages of orogen propagation. First, following initiation of mountain building in the Central Cordillera during the Late Cretaceous, the orogenic front propagate eastward at slow rates (0.5-3.1 mm/yr) until early Eocene times. Such slow orogenic advance would have resulted from limited accretionary flux related to slow and oblique (SW-NE-oriented) convergence of the Farallon and South American plates during that time. A second stage of rapid orogenic advance (4.0-18.0 mm/yr) during the middle-late Eocene, and locally of at least 100 mm/yr in the middle Eocene, resulted from initial tectonic inversion of the Eastern Cordillera. I correlate this episode of rapid orogen-front migration with an increase in the accretionary flux triggered by acceleration in convergence and a rotation of the convergence vector to a more orogen-perpendicular direction. Finally, stagnation of the Miocene deformation front along former rift-bounding reactivated faults in the eastern flank of the Eastern Cordillera led to a decrease in the rates of orogenic advance. Post-late Miocene-Pliocene thrusting along the actively deforming front of the Eastern Cordillera at this latitude suggests averaged Miocene-Holocene orogen propagation rates of 1.2-2.1 mm/yr. In addition, ZFT data suggest that exhumation along the eastern flank of the orogen occurred at moderate rates of ~0.3 mm/yr during the Miocene, prior to an acceleration of exhumation since the Pliocene, as suggested by recently published AFT data. In order to evaluate the relations between thrust loading and sedimentary facies evolution in the foreland, I analyzed gravel progradation in the foreland basin system. In particular, I compared one-dimensional Eocene to Pliocene sediment accumulation rates in the Medina basin with a three-dimensional sedimentary budget based on the interpretation of ~1800 km of industry-style seismic reflection profiles and borehole data tied to the new chronostratigraphic framework. The sedimentological data from the Medina Basin reveal rapid accumulation of fluvial and lacustrine sediments at rates of up to ~ 0.5 mm/yr during the Miocene. Provenance data based on gravel petrography and paleocurrents reveal that these Miocene fluvial systems were sourced by Upper Cretaceous and Paleocene sedimentary units exposed to the west, in the Eastern Cordillera. Peak sediment-accumulation rates in the upper Carbonera Formation and the Guayabo Group occur during episodes of gravel progradation in the proximal foredeep in the Early and Late Miocene. I interpreted this positive correlation between sediment accumulation and gravel deposition as the direct consequence of thrust activity in the Servita-Lengupá Fault. This contrasts with current models relating gravel progradation to episodes of tectonic quiescence in more distal portions of foreland basin systems and calls for a re-evaluation of tectonic history interpretations inferred from sedimentary units in other mountain belts. In summary, my results document a late Eocene-early Miocene eastward advance of the topographic loads associated with the leading edge of deformation in the northern Andes of Colombia. Crustal thickening of the Eastern Cordillera associated with initiation of thrusting along the Servitá Fault illustrates that this sector of the Andean orogen acquired ~90% of its present width already by the early Miocene (~20 Ma). My data thus demonstrate that inherited crustal anisotropies, such as the former rift-bounding faults of the Eastern Cordillera, favour a non-systematic progression of foreland basin deformation through time by preferentially concentrating accommodation of slip and thrust-loading. These new chronology of exhumation and deformation associated with specific structures in the Colombian Andes also constitutes an important advance towards the understanding of models for hydrocarbon maturation, migration and trap formation along the prolific petroleum province of the Llanos Basin in the modern foredeep area. / Das Vorlandbecken, das sich an der östlichen Flanke der Anden erstreckt, ist ein prototypisches Beispiel für ein Retro-Arc-Vorlandbecken eines Subduktionszonenorogens. Im Gegensatz zu den südlichen und zentralen Anden, wurde die zeitliche und räumliche Entwicklung dieses känozoischen Systems im nördlichen Teil des Orogens weit weniger untersucht. Dies liegt unter anderem an den komplexen geodynamischen Randbedingungen, wie der schrägen Subduktion und Anlagerung der karibischen Plattengrenzen an die südamerikanische und Nazca-Platte, deren Interaktion ebenfalls komplex ist und durch unterschiedliche Konvergenzrichtungen und –geschwindigkeiten gekennzeichnet ist. Aufgrund dieser Verhältnisse hat auch die Oberplatte eine sehr differenzierte tektonische Entwicklung erfahrens. In den kolumbianischen Anden hat sich zum Beispiel seit ca. 80 Milllionen Jahren ein Vorlandbeckensystem in einem Gebiet gebildet, das während des Mesozoikums durch Rifttektonik geprägt war. Dieses Gebiet, in dem kompressive Deformation die Strukturen vorheriger extensionaler Prozesse z.T. reaktiviert und überlagert, ist daher ein natürliches, wenn auch bisher wenig erforschtes Naturlabor, um zu untersuchen, wie sich tektonische bedingte Anisotropien auf die Entwicklung von Vorlandbeckensystemen auswirken können und Änderungen in den Ablagerungsräumen und in der Faziesverteilung von Sedimenten hervorrufen. In dieser Arbeit präsentiere ich neue strukturelle, sedimentologische und biostratigraphische Daten zusammen mit neuen Informationen zur Exhumationsgeschichte mit Hilfe von Niedrigtemperatur-Thermochronologie aus dem östlichen Teil der kolumbianischen Anden, um zum ersten Mal eine vollständige Darstellung der Gebirgsbildung und zugehöriger Vorlandbeckensedimentation in diesem Teil der Anden zu liefern. Zusätzlich wird untersucht, zu welchem Ausmaß bereits existierende krustale Anisotropien den Ort der Vorlanddeformation in Raum und Zeit bestimmt haben. Im Medina Becken, an der östlichen Flanke der östlichen Kordillere, habe ich detaillierte strukturelle Kartierungen und neue sedimentologische Daten mit einem neuen chronostratigraphischen Rahmen, der auf detaillierter Palynologie basiert, verknüpft. Dieser Bezugsrahmen verbindet einen nach Osten hin ausdünnenden, syntektonischen früholigozänen bis frühmiozänen Keil, welcher rasche Faziesänderungen enthält, mit einer Phase schneller tektonischer Subsidenz, die vor ca. 30 Millionen Jahren beginnt. Dieser hier erarbeitete Datensatz stellt den ersten Beweis einer tektonisch bedingten Subsidenz dar, die durch Bewegungen entlang der Haupüberschiebungen an der Westgrenze des Vorlandes stattfanden. Dadurch wird das Einsetzen der Gebirgsbildung in diesem Gebiet zeitlich eingegrenzt. Eine umfassende Auswertung von Exhumationsmustern, die auf Zirkon- (ZFT) und Apatit-Spaltspuraltern (AFT) sowie thermischen Modellierungen beruhen, zeigt, daß diese Überschiebungsbahnen und die bedeutende Aufschiebungstätigkeit und tektonische Auflast entlang der reakivierten, vormals extensionalen Servita-Störung, im zentralen Bereich der östlichen Kordillere liegen. Weiter östlich dokumentieren AFT und ZFT Daten den Einsatz einer durch Überschiebungen hervorgerufenen Exhumation, die mit einer kompressiven Reaktivierung der großen Servita-Störung vor ca. 20 Millionen Jahren zusammenhängt. Vitrinitreflexionsdaten aus dem känozoischen Sedimentationsdatensatz des benachbarten Medina Beckens zeigen eine bedeutende Absenkung in dieser Region, bei der der westliche Sektor des Beckens schon im Anfangsstadium der orogenen Entwicklung in den nach Osten wandernden Falten- und Überschiebungsgürtel einbezogen wurde. Ich verbinde diese neuen thermochronologischen Daten mit veröffentlichten AFT Analysen und bekannten chronologischen Indikatoren für Spröddeformation, um die räumlich-zeitlichen Muster in der Entwicklung der Gebirgsfront in den Anden Zentralkolumbiens zu charakterisieren. Diese Analyse der Deformation zeigt ein episodisches Muster in der östlich gerichteten Migration der Gebirgsfront, mit einer durchschnittlichen Rate von 1.8-3.4 mm/a am Übergang von der späten Kreide zum frühen Känozoikum. Ich habe dabei drei Hauptabschnitte des lateralen Orogenwachstums identifiziert. Zuerst wandert die Gebirsfront, nach dem Beginn der Gebirgsbildung in den Zentralkordilleren während der späten Kreidezeit, ostwärts mit niedrigen Raten (0.3-3.3 mm/a) bis ins frühe Eozän. Ein solches langsames laterales Wachstum des Orogens resultiert aus Akkretionsprozessen im Zuge einer langsamen und schrägen (SW-NO orientiert) Konvergenz der Farallon- mit der südamerikanischen Platte. Eine zweite Phase schnellen Fortschreitens der Gebirgsfront mit Raten von 5.3-13.3 mm/a, lokal sogar bis zu 100 mm/a, fand während des mittleren/späten Eozäns statt und resultierte aus einer beginnenden tektonischen Inversion der östlichen Kordillere. Ich verbinde diese Phase rascher Gebirgsfrontmigration mit einem erhöhten Akkretionsfluß, der durch eine Beschleunigung der Konvergenz sowie einer Rotation des Konvergenzvektors in eine mehr rechtwinklige Richtung ausgelöst wurde. Letztlich führte eine Stagnation der Deformationsfront im Miozän entlang von ehemals riftbegrenzenden, reaktivierten Störungen an der östlichen Flanke der östlichen Kordillere zu einer Abschwächung der Raten der Gebirgsfrontmigration. Aus Überschiebungen des späten Miozän/Pliozän entlang der aktiv deformierten Front der östlichen Kordillere kann man auf durchschnittliche Bewegungsraten der Gebirgsfront von etwa 1.5-2.1 mm/a im Zeitraum Miozän bis Holozän schließen. Außerdem deuten ZFT Daten darauf hin, daß Exhumation entlang der östlichen Flanke des Orogens mit mittleren Raten von ungefähr 0.3 mm/a während des Miozäns stattfand. Im Pliozän erfolgte daraufhin eine Beschleunigung der Exhumation, wie kürzlich veröffentlichte AFT Daten nahelegen. Um die Beziehung zwischen tektonischer Auflast aufgrund der Verkürzung im Orogen und Evolution der sedimentären Fazies im Vorland zu untersuchen, habe ich die Progradation von Konglomeraten im Vorlandbeckensystem detailliert analysiert. Insbesondere habe ich eindimensionale Raten von Sedimentakkumulation vom Eozän bis zum Pliozän im Medina Becken mit einem dreidimensionalen Sedimenthaushalt verglichen. Dieser wurde aus der Interpretationen mit einer Gesamtlänge von ~2500 km seismischer Reflexionsprofile sowie Bohrlochdaten, verbunden mit dem neuen chronostratigraphischen Bezugssystem der sedimentären Ablagerungen, gewonnen. Die sedimentologischen Daten aus dem Medina Becken zeugen von rascher Akkumulation von fluviatiler und lakustriner Sedimente mit Raten von bis zu 0.5 mm/a während des Miozäns. Provenienzanalysen mittels Konglomerat-Petrographie und Paläoströmungsmessungen belegen, daß diese miozänen fluviatilen Systeme des Miozäns durch die Erosion sedimentärer Einheiten aus der oberen Kreide und dem Paläozän generiert wurden, die im Westen der östlichen Kordillere aufgeschlossen sind. Die höchsten Sedimentationsraten in der oberen Carbonera Formation und der Guayabo Gruppe finden sich während Episoden von Konglomeratprogradation der proximalen Vortiefe im frühen und späten Miozän. Ich interpretiere diese positive Korrelation zwischen Sedimentakkumulation und Konglomeratablagerung als direkte Konsequenz von Überschiebungstektonik an der Servita-Lengupá-Störung. Diese Interpretation ist allerdings im Gegensatz zu gängigen Sedimentationsmodellen Modellen, die eher eine tektonische Ruhephase mit der Progradation grober Schüttungen in den distalen Bereichen der Vorlandbecken in Verbindung bringen. Dies bedeutet, daß Interpretationen der aus Faziesverteilungen gewonnenen tektonischen Entwiklungsschritte eines Orogens auch in andeen Regionen neu bewertet werden müssen. Zusammengefaßt dokumentieren meine Ergebnisse, daß die Überschiebungsfront sowie die durch Einengung generierte Topographie und Auflast der Überschiebungsblöcke in den nördlichen kolumbianischen Anden während des späten Miozäns bis zum frühen Miozän ostwärts gewandert ist. Einengung und Krustenverdickung der östlichen Kordillere, verbunden mit beginnender Aktivität entlang der Servitá Störung, deutet an, daß dieser Bereich der Anden schon nahezu 90% seiner derzeitigen Breite bereits im Miozän (20 Ma) erreicht hattte. Die hier vorgestellten Daten zeigen also, daß ererbte krustale Anisotropien ein diachrones Voranschreiten der Vorlandbeckendeformation begünstigen. Dies geschieht durch Konzentration der Bewegungsverteilung an ererbten Störungen sowie lokalen Spannungsänderung im Vorland durch tektonische induzierte Auflasten. Diese neue Charakterisierung der Deformationsabfolge im Vorland der Anden bedeutet auch einen großen Schritt vorwärts in Richtung des Verstehens von Modellen, die das Reifen und die Wanderung von Kohlenwasserstoffen sowie die Entstehung von Ölfallen entlang der produktiven Petroleumprovinz im Llanos Becken der rezenten Vortiefe beschreiben.

Water security requires that sufficient quantities of water be available at critical times. This is particularly challenging for high-intensity urban and agricultural settings. In underdeveloped nations, streamflow is commonly the preferred water source, as it is readily available and delivered cost-free to users. Yet, the sources of these critical streamflows are often unknown. This issue is salient in the Northern Andes, where basic knowledge of controlling factors for the quantity, quality, and timing of runoff is lacking. High-elevation headwaters are the primary catchment areas in the Northern Andes, but the extent of water providing to municipalities in the Northern Andes is unknown. In this study, the contribution of water derived from the upper watershed to the streamflow in the Tulúa River which supplies water to 200,000 people in the city of Tulúa was quantified. The river runs 72 km through urban, agricultural, and industrial land use in the Central Cordillera of the Colombian Andes. We collected 32 and 34 water samples in August and November, respectively. The water samples were representative of high-elevation headwaters runoff, shallow groundwater discharge, and streamflow throughout the watershed. Samples were analyzed for dissolved constituents and stable isotopes. The dissolved constituents were used in mass-balance mixing models to identify the source of streamflow in the lower watershed of the Tulúa River, where it the river supports a large municipality. Results indicate that approximately 50% surface runoff largely originates as high-elevation headwater runoff, including high-elevation settings where páramos dominate the land cover. These findings underscore the need for source-water protection efforts in the upper watershed, including the páramos. This project serves as a model for other páramo derived watersheds, where source-water protection is a critical challenge.

A ocorrência de uma grande reorganização de placas durante o Neógeno reflete a comlpexa interação entre as placas do Caribe, Farallon, e América do Sul, e o arco intra-oceânico de Panamá. Isto, tem influenciado o registro tectono-estratigráfico dos Andes nor-ocidentais da Colômbia, o qual tem sido escassamente estudado perto da sutura entre os domínios de Panamá o da América do Sul. Novos dados de petrografia de arenitos, análises de minerais densos, geocronologia detrítica em zircão, geoquímica de rocha total, e modelamento termal inverso de múltiplos termocronômetros, realizados em um transepto de 50 km com direção SE-NW, permitem reconstruir a história de deformação e exumação/erosão das rochas do Cretáceo Superior ao Plioceno do flanco ocidental da Cordilheira Ocidental. O início da exumação das rochas do Cretáceo Superior associadas com a margem continental Sul-americana, ocorreu entre 45 e 20 Ma, o qual é sugerido pelo resfriamento inicial desde temperaturas dentro da zona de retenção parcial do hélio em zircão (~180-200°C). Essas rochas, também sofreram um pulso de exumação considerável durante o Mioceno Médio, o qual caracteriza-se por taxas moderadamente rápidas entre 0.3 km/my e 0.7 km/my. Adicionalmente, um período de resfriamento rápido também é documentado com taxas máximas de exumação ~1.3 km/my. As rochas do Eoceno Médio do arco oceânico de Panamá, as quais estão separadas das rochas do Cretáceo Superior da Cordilheira Ocidental por uma faixa de deformação, registram um resfriamento inicial em ~15 Ma desde a zona de retenção parcial de hélio em apatita (APRZ, ~80-60°C, com taxas perto de 0.6 km/my. Termocronologia detrítica em secuências pós-colisionais e sedimentos modernos, também registram estes períodos de exumação. Os novos resultados podem ser relacionados com processos tectônicos de escala regional, que incluem: Convergência rápida e frontal no Oligoceno Superior, entre a recentemente formada placa de Nazca e a placa continental Sul-americana, depois da fragmentação da antiga placa de Farallon, a colisão do Arco de Panamá do Cretáceo Superior-Eoceneo durante o Mioceno Médio, o qual resultou exumação geral ao longo do segmento NW da placa Sul-americana, é o início da subducção da placa de Nazca é a instalação de um segmento plano ao norte durante o Mioceno Superior-Plioceno. Este é provavelmente o mecanismo responsável pelo soerguimento da superfície recente na Colômbia nor-oriental e a contínua deformação-soerguimento das serras costeiras ocidentais e das bacias orientais de foreland ao norte de 5°N. / The occurrence of major plate reorganization during the Neogene, reflects the complex interactions between the Caribbean, Farallon and South-American plates, and the Panama Arc. These, have influenceed the tectono-stratigraphic record of the northwestern Colombian Andes, which has been scarcely studied neart the suture between the Panama and South-American domains. New sandstone petrography, heavy minerals analyses, U-Pb detrital zircon geochronology, whole-rock geochemical data, and thermal inverse modeling of multiple thermochronometers, carried along a 50 km southeast-northwest transect, allow to reconstruct the deformation and exhumation/erosional history of the Late Cretaceous to Pliocene rocks from the western flank of the Western Cordillera. The onset of exhumation of Late Cretaceous sedimentary rocks, associated with the South-American margin, occurred between 45-20 Ma, as suggested by the initial cooling from temperatures within the zircon helium partial retention zone (ZPRZ, ~180-200°C). These rocks also experienced a substantial pulse of exhumation during the Middle Miocene, which is characterized by moderately rapid rates between ~0.3 km/my and 0.7 km/my. Moreover, a Pliocene period of rapid cooling is also documented with maximum exhumation rates of ~1.3 km/my. The Middle Eocene rocks from the intra-oceanic Panama Arc, which are separated by a highly-deformed zone from the Late Cretaceous rocks of the Western Cordillera, record an initial cooling at ca. 15 Ma from the apatite partial retention zone (APRZ, ~80-60°C) at rates around 0.6 km/my. Detrital thermochronology of post-collisional sequences and modern river sediments also record these exhumation events. The we results can be related to more regional-scale tectonic processes including: the Late Oligocene rapid and frontal convergence between the newly formed Nazca plate and the continental South-American plate following the fragmentation of the former Farallon plate, the Middle Miocene collision of the Cretaceous-Eocene Panama Arc that triggered generalized exhumation-deformation widespread in the upper NW South-American plate, and the Middle to Late Miocene initiation of the Nazca subduction and installation of a Late Miocene-Pliocene flat-slab to the north. The latter is likely to represent the mechanism that triggered surface uplift in eastern Colombia and ongoing uplift-deformation of western coastal ranges and eastern foreland basins north of 5°N

A small lexicon of traditional weaving (telar de cintura), collected in Agallpampa (Otuzco, La Libertad) offers evidence against the idea of any linguistic identity being shared between the area of the now extinct Culle language, and the Cajamarca Valley. Yet, such a link is supported by the isolation of a grammatical element, traced to the Culle language in both areas: diminutive suffix –ash–, as in cholasho ‘young little man’ and chinasha ‘young little woman’. These contrasting data shed new light on a discussion begun by Torero (1989) about the existence of particular languages in the central Cajamarca area (languages Den and Cat); and continued by Adelaar with Muysken (2004), who suggest that cases of lexical community between the Culle geographical nucleus and indigenous words of Cajamarca Quechua, which cannot be traced back to Quechua idiomatic sources, suggest that a Culle substratum holds for the Cajamarca Valley. This paper argues that, in order to solve this apparent paradox, it is necessary to focus on this issue in terms of linguistic strata: i.e., different stages of idiomatic hegemony before Quechua and Spanish were established in the region. The oldest stratum would be associated with Den, and the more recent, albeit still prior to the Quechua and Spanish periods, would be Culle. Based upon archaeological research in the area and on the recent association of Cajamarca Quechua with the Huari expansion (Adelaar 2012), I suggest that the chronological distance between both strata must be deep, since Culle would have been established in the region long before the Northern Huari expansion took place. Nonetheless, the existence of Quechua-Den mixed toponyms precludes this hypothesis being applied to the whole Cajamarca territory, especially its southwest area (Contumazá). / Un breve repertorio léxico del telar de cintura, recogido en Agallpampa (Otuzco, La Libertad), aporta evidencia contraria a la idea de que existió identidad idiomática entre la zona de emplazamiento de la extinta lengua culle y el valle de Cajamarca, en el departamento del mismo nombre. En cambio, abona a favor de esta propuesta la identificación de un elemento gramatical atribuible al culle en ambas zonas: el sufijo diminutivo –ash–, como en cholasho ‘muchachito’ y chinasha ‘muchachita’. Este contraste constituye una ilustración del debate abierto por Torero (1989) sobre la existencia de idiomas indígenas particulares en las provincias centrales cajamarquinas (las lenguas den y cat) y continuado por Adelaar, con la col. de Muysken (2004), quien ha planteado que los ejemplos de comunidad léxica entre el núcleo de la zona culle y las palabras indígenas del quechua cajamarquino que no pueden ser atribuidas al fondo idiomático quechua sugieren la existencia de un sustrato culle en el valle de Cajamarca. Este artículo argumenta que para resolver esta aparente paradoja, es necesario pensar en términos de estratos lingüísticos, es decir, en diferentes etapas de hegemonía idiomática previas a la presencia del quechua y del castellano en dicho territorio. El estrato más antiguo correspondería al fondo idiomático den, mientras que el posterior, previo al advenimiento del quechua y el castellano, correspondería al culle. Partiendo de la investigación arqueológica realizada en la zona y de la reciente atribución del quechua cajamarquino a la avanzada huari (Adelaar 2012), se sostiene que la separación temporal entre ambos estratos debió de ser prolongada, ya que el culle tendría que haberse asentado en la zona mencionada antes de la expansión norteña de Huari. Sin embargo, la existencia de toponimia mixta quechua-den previene contra la posibilidad de generalizar esta hipótesis al territorio cajamarquino en su conjunto, especialmente al sector sureño occidental (Contumazá).