Earth and Planetary Science Letters 278 (2009) 112

Contents lists available at ScienceDirect

Earth and Planetary Science Letters

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

Long-term erosion and exhumation of the Altiplano Antioqueo, Northern Andes (Colombia) from apatite (UTh)/He thermochronology
Sergio A. Restrepo-Moreno a,, David A. Foster a, Daniel F. Stockli b, Luis N. Parra-Snchez c
a b c

Department of Geological Sciences, University of Florida, Gainesville, Florida 32611, USA Department of Geology, University of Kansas, Lawrence, Kansas 66045, USA ICNE, Universidad Nacional de Colombia, Medelln, Colombia

a r t i c l e

i n f o

a b s t r a c t
The Antioqueo Plateau (AP) in the northern Cordillera Central, Colombia, is the largest high elevation erosional surface in the Northern Andes. Apatite (UTh)/He thermochronometry (AHe) of samples collected from two elevation proles spanning 2 km of exhumed crustal sections reveal the long-term erosional exhumation of the AP. Sample proles exhibit AHe ages that increase with elevation from ca. 22 Ma ( 760 m) at the bottom of regional scarps to ca. 49 Ma ( 2350 m) on top of the AP. A marked inection point in age versus elevation data at ca. 25 Ma denes the bottom of the exhumed post-Oligocene He partial retention zone (He-PRZ). Elevation-invariant ages below ca. 25 Ma record the onset of rapid exhumation and surface uplift of the AP that led to river incision. A subtle change in slope within the He-PRZ, ca. 41 Ma, is interpreted as a less intense, exhumation-related cooling episode. These two exhumation pulses coincide with the ProtoAndina and Pre-Andina orogenic phases previously proposed for the Colombian Andes, and are synchronous with tectonically driven exhumation events reported for the Peruvian, Bolivian and Argentinean Andes, and for some orogenic systems in the Caribbean. The pulses are correlated with variations in the rates of convergence between Nazca (Farallon) and South America documented for the Middle Eocene and the Late Oligocene suggesting continental-scale controls on uplift and denudation throughout the Andean range. AHe data provide an average erosion rate of 0.04 mm/yr for the last 25 million years. Erosion rates during the exhumation pulses were in the order of 0.20.4 mm/yr. Similarity between AHe proles indicates the whole AP was uplifted and exhumed as a coherent structural block, corroborating previous structural evidence for the rigidity and coherence of this crustal block in the Northern Andes. Our results are in agreement with tectonostratigraphic data in the Magdalena and Cauca basins and with proposed scenarios for paleogeographic evolution in the Northern Andes. 2008 Elsevier B.V. All rights reserved.

Article history: Received 6 October 2007 Received in revised form 9 September 2008 Accepted 24 September 2008 Editor: R.W. Carlson Keywords: (UTh)/He dating relict landscape morphotectonics erosion/exhumation rates Northern Andes Antioqueo Plateau

1. Introduction Dening the morphotectonic evolution of the Andean Cordillera is critical for understanding the tectonic processes driving uplift, climatic inuences of the topography, the source of sediment to oceans and continental sedimentary basins, and changes in paleogeography (Hoorn et al., 1995; Gregory-Wodzicki, 2000; Harris and Mix, 2002; Gmez et al., 2005; Sobolev and Babeyko, 2005). Despite its importance in the geodynamic evolution of a geologically complex region, the Northern Andes in Colombia remain poorly studied compared to many segments of the Andes to the South. Little is known about the morphotectonic evolution of some major provinces, including the Antioqueo Eastern Massif (Botero, 1963), which is a polymetamorphic complex (Restrepo and Toussaint, 1982), intruded by large Mesozoic calc-alkaline plutons of the Antioqueo and Ovejas
Corresponding author. 241 Williamson Hall PO Box 112120, Gainesville, FL 326112120, USA. Tel.: +1 352 3922231; fax: +1 352 3929294. E-mail address: sergiorm@u.edu (S.A. Restrepo-Moreno). 0012-821X/$ see front matter 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.epsl.2008.09.037

Batholiths (Gonzlez, 2001). Geomorphologically, the Antioqueo Eastern Massif encompasses the Altiplano Antioqueo (Arias, 1995; Page and James, 1981), hereafter the Antioqueo Plateau (AP), which is the largest high elevation erosional surface in the Northern Andes. The AP is preserved as a wide, slightly dissected surface extending for tens of kilometers, locally and deeply incised by the MedellnPorce uvial system (Figs. 1 and 3). A better understanding of the long-term landscape development and erosional exhumation history of the AP is needed for dening the Cenozoic morphotectonic and uplift history of the Northern Andes, the paleogeographic evolution of the region, and the relationships between erosion and tectonics (Van der Hammen, 1960; Van der Hammen et al., 1973; Duque-Caro, 1990; Hoorn et al., 1995; Gregory-Wodzicki, 2000; Gmez et al., 2005; Hooghiemstra et al., 2006). The ndings of this study also have implications for rening morphotectonic response to variations in convergence rates between Nazca (Farallon) and South American plates during the Cenozoic (Pardo-Casas and Molnar, 1987). Elevated erosional surfaces affected by deep and localized uvial incision, such as the AP (Figs. 1 and 3), constitute an important source

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112

Fig. 1. Shaded relief map of the Northern Andes, Colombia. The Andean mountain chain in Colombia is divided into three separate ranges: Cordillera Occidental (Western Range), Cordillera Central (Central Range), and Cordillera Oriental (Eastern Range). The Antioqueo Plateau in the northern Cordillera Central is part of the Antioqueo Eastern Massif (Botero, 1963), which is bounded by major fault zones Cauca-Romeral fault system (R) to the west and Palestina fault system (P) to the east. The Guaicramo (G) and Bocon (B) faults dene the limit between the Andean Block and the autochthonous South American Plate (craton).

of morphotectonic information, retrievable from low-temperature thermochronologic data. Such surfaces, usually referred to as relict surfaces, have played an important role in the reconstruction of the morphotectonic evolution of active (Kennan et al., 1997; Coltorti and Ollier, 1999; Ollier and Pain, 2000) and passive continental margins (Peulvast and Sales, 2005). Elevated, relict erosional surfaces, indicate that portions of an orogen have not achieved equilibrium between erosion and tectonics (Coltorti and Ollier, 1999; Ollier and Pain, 2000). It is generally thought that these surfaces develop by planation close to base level (Davis, 1899) and are subsequently uplifted via tectonic processes to their present position (Arias, 1995; Coltorti and Ollier, 1999). The non-equilibrium character of these denudational features implies lagged response to tectonic forcing, which allows their use for reconstructing regional, long-term erosional exhumation as well as patterns of surface/rock uplift and topographic evolution (Clark et al., 2005; Schildgen et al., 2007). Cenozoic uplift and erosion in the northern Cordillera Central is recorded by sedimentary units (e.g., associated litho facies and uvial paleoows) in the Middle Magdalena Valley to the east (Van der Hammen, 1960; Gmez et al., 2005), the Lower Magdalena Valley to the north (Van der Hammen, 1960; Reyes et al., 2002), and the Cauca valley to the west (Grosse, 1926). The traditional interpretation suggests that since the latest Cretaceous the eastern ank of the Cordillera Central has operated as the source area for synorogenic detritus to the Middle and Lower Magdalena Valleys (Reyes et al., 2002; Gmez et al., 2005). The region was also considered as a source of terruginous material for the Terciario Carbonfero de Antioquia during the OligoMiocene (Grosse, 1926), a coal-bearing sedimentary sequence later renamed Amag Formation (Gonzlez, 2001), situated in the Cauca valley. Cenozoic uplift in the Northern Andes of Colombia has been also invoked as the cause of substantial paleogeographic changes that took place in the region, particularly, the reorientation reorganization of the Orinoco and Amazon uvial systems (Hoorn et al., 1995).

We present thermochronological evidence from apatite (UTh)/He analysis from two vertical proles in the AP (Fig. 3) that reveal the erosional exhumation history of the plateau with implications for the timing of tectonic pulses of the Andes and the topographic evolution of the range during and after uplift. 2. Study site: the Antioqueo Plateau 2.1. Physiographic and Geologic overview From west to east, the Colombian Andes are composed of three main mountain belts known as Cordillera Occidental, Cordillera Central, and Cordillera Oriental. These Cordillera are separated by two inter-Andean depressions occupied by the Cauca and Magdalena rivers (Figs. 1 and 2). The Magdalena River yields 1000 t/km2 yr of sediment to the delta which is the highest sediment yield of any large river in South America along the Caribbean and Atlantic coasts (Restrepo and Syvitski, 2006). Such sediment yield is roughly equivalent to denudation rates on the order of 1 mm/yr. Similar gures have been reported for the AP based on a temporal series of bathymetric surveys in ve reservoirs in the AP (Giraldo, 2005). Augmented erosion in the Colombian Andes has been attributed to anthropogenic activities, e.g., deforestation, agriculture, and development of infrastructure (Restrepo and Syvitski, 2006). However, quantitative data regarding long-term, background erosion rates for the area are inexistent. The three major cordilleras belong to two distinct geologic domains that are separated by the Romeral Fault system (Figs. 1 and 2). The western province consists of accreted oceanic crust and is exposed in the Cordillera Occidental and the lower western ank of the Cordillera Central. The eastern province is underlain by continental basement and is exposed throughout most of Cordillera Central and Cordillera Oriental (Taboada et al., 2000; Tapias et al., 2006). The Cordillera Oriental is a fold-and-thrust belt that consists of

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112

polydeformed, continental Precambrian and Paleozoic metamorphic and igneous rocks overlain by thick Paleozoic to Mesozoic sedimentary sequences (Gonzlez, 2001; Taboada et al., 2000). The Cordillera Central is characterized by modern volcanic activity related to the subduction of the Nazca plate and a basement consisting of a MesozoicCenozoic plutonic arc. This study focuses on a cordilleran segment located from 545 to 7 latitude North without modern volcanism. Within the study area (Fig. 2), the Cordillera Central encompasses a pre-Mesozoic to Mesozoic polymetamorphic complex (Restrepo and Toussaint, 1982) composed of low- to mediumgrade metamorphic rocks of the Cajamarca Complex (Maya and Gonzlez, 1995), and high-grade metamorphic rocks of the El Retiro Group and Las Palmas Gneiss (Gonzlez, 2001). Lower Paleozoic or Precambrian age has also been proposed for this metamorphic core (Hall et al., 1972). The metamorphic basement is covered by the unmetamorphosed shallow marine to epicontinental sedimentary sequences San Luis (Feininger et al., 1972), San Pablo, La Soledad (Hall et al., 1972), and Abejorral (Etayo-Serna, 1985), that are Early Cretaceous in age (132127 Ma). Metamorphic and sedimentary units were intruded at a shallow crustal level ( 4 km) by Meso Cenozoic batholiths, including the Batolito Antioqueo and Batolito de Ovejas plutons, and overlain by Late Eocene continental-arc volcanic rocks (Gonzlez, 2001). Finally, Late Neogene volcaniclastic sequences

associated with modern volcanic activity in the Cordillera Central mantle the region (Arias, 1995). The northern portion of the Cordillera Central is considered to be part of an exotic terrane, which was affected by DevonianCarboniferous, PermoTriassic (Hercynian), and Cretaceous tectono-metamorphic events (Restrepo and Toussaint, 1982, 1988; Toussaint, 1993). The Antioqueo and Ovejas batholiths are texturally and compositionally homogenous. Approximately 95% of the samples collected from this units have been classied as quartzdioritegranodiorite (Feininger and Botero, 1982; Saenz, 2003). Previously published thermochronologic data from the Antioqueo Batholith and other associated plutons include biotite K/Ar ages, which range from 68 3 Ma (Feininger and Botero, 1982) to 90 6 Ma (Restrepo, 1975), three biotite Rb/Sr ages from 56 7.4 to 66 9.2 Ma (Fujiyoshi et al., 1976); two whole-rock Rb/Sr ages of 82 8 and 98 27 Ma (Ordez and Pimentel, 2001); zircon ssiontrack dates from 49.1 2.5 to 67.1 2.1 Ma and fourteen apatite ssiontrack ages varying between 49.1 1.2 and 28.7 1.5 Ma (Saenz, 2003). The large spectrum of ages can be explained by the wide range of closure temperatures of the thermochronometric systems employed and should be interpreted as cooling rather than formation ages. Although the Rb Sr age of 98 27 Ma has been proposed as the best estimate of crystallization age for the Antioqueo batholith (Ordez and Pimentel, 2001), recent zircon UPb spot analyses (LA-ICP-MS) give ages between

Fig. 2. Geologic units of Colombia and geology of the study site, northern Cordillera Central. The area is dominated by the polymetamorphic Cajamarca Complex (Pz), small patches of marine sedimentary sequences Early Cretaceous in age (Kisp, Ksls, Kissl) and the Late Cretaceous calk-alkaline intrusion of the Antioqueo Batholith and associated stocks (Ksta, dotted grey). Other units include: Q = Quaternary uvial and alluvial deposits; Ngc + gas = (Neogene sediments from Combia Formation (volcaniclastic continental) and Amag Formation (uvial); Ngro = Mesa Formation (detrital, continental). Ksvb = Volcanic Barroso Formation (volcanic, marine) and Ksu = Urrao Formation (clastic-chemical, marine) belonging to the western oceanic province. Jsde = Segovia Batholith, Jts = Sonsn Batholith,P = metamorphic rocks of Neoproterozoic age. The Cauca Almaguer Fault is a paleosuture zone which represents the boundary between the Eastern Continental Province (approximately coinciding with the Cordillera Central) and the Western Oceanic Province (approximately coinciding with the Cordillera Occidental). (Geologic information was adapted from Gonzlez, 2001 and Tapias et al., 2006.)

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112

topographic margins expose up to 3200 m of crustal section with slopes in excess of 50. Most of the planar topography of the AP coincides with the Antioqueo Batholith, one of the largest (7600 km2) and most continuous lithological units in the Cordillera Central. The prominent topographic features within the AP, like the Las Baldas Range and Pramo de Belmira in the western margin of the plateau, are held up by the more resistant metamorphic units (Figs.13). About 150 km south of the study area, the plateaus low-relief rapidly transitions into a more rugged portion of the Cordillera Central dominated by active volcanoes. This part of the orogen, in clear contrast to the AP, has a geometry that bears resemblance to a symmetric pyramid terminated in higher summits reaching up to 5000 m. The AP has been interpreted to have developed as an erosion surface near sea level in Paleocene time and that was uplifted in discrete episodes during the Cenozoic, and modied by uvial activity combined with the unique weathering properties of granitic rocks in a tropical setting (Scheibe, 1919; Arias, 1995). While this is the prevailing view, the AP's morphotectonic history is still poorly constrained, and the age of the surface is strongly debated. Despite the fact that exhumation and paleoelevational history of the AP remain virtually undocumented, remnants of shallow, marine sedimentary sequences formed during an Early Cretaceous marine transgression point to the origin of the region as a low-elevation, low-relief domain. No other elevated low-relief erosional surfaces as extensive as the AP have been reported in other parts of the Northern Andes. Nevertheless, similar geomorphic features of lesser extent, referred to in the literature as elevated planation surfaces (Coltorti and Ollier, 1999), high altitude paleosurfaces (Kennan et al., 1997), and erosion plateaus (Ollier and Pain, 2000), are found in Ecuador, Peru, and Bolivia. Quantitative data that constrain erosion rates, landscape evolution, and uplift history of these important geomorphic features is still scarce.
Fig. 3. Three dimensional representation (top) and simplied topographic/geologic cross section AA (bottom) across the Antioqueo Plateau. The at portion of the plateau has an average elevation of 2,500 m. MP = MedellnPorce uvial valley, CG = Cauca Gorge (CG). The CG coincides with the CaucaRomeral (CaucaAlmaguer) fault system (shear zone). Sampled proles indicated with black and open small circles. Pramo de Belmira within Las Baldas Range is at, 3600 m of elevation. (Shaded relief DEM was generated from GTOPO-30 elevation data. (Geologic information was obtained from Gonzlez, 2001; and Tapias et al., 2006).

2.3. Tectonic setting Orogeny in the Northern Andes is primarily driven by the collision of allochthonous terrains and interactions between several tectonic plates and micro plates (Taboada et al., 2000; Cortes and Angelier, 2005). This has resulted in heterogeneous deformation compared to the Central Andes (Taboada et al., 2000; Trenkamp et al., 2002; Coates et al., 2004; Cortes and Angelier, 2005). Most authors agree that the regional tectonic history has been dominated by the subduction of two oceanic plates, Nazca and Caribbean, beneath continental South America and by interactions of two lithospheric provinces, namely, the North Andean and the Panama-Choc blocks (Duque-Caro, 1990; Taboada et al., 2000; Trenkamp et al., 2002; Coates et al., 2004; Cortes and Angelier, 2005) (Fig. 4). The Nazca plate is undergoing rapid steep subduction along the length of the ColombiaEcuador trench at a rate of 54 mm/a, and is responsible for the volcanism in the Cordillera Central (Taboada et al., 2000; Trenkamp et al., 2002). The Caribbean plate is being subducted slowly from the northwest at 20 mm/a, at a shallow angle, and without an associated volcanic arc (Trenkamp et al., 2002; Cortes and Angelier, 2005). During Paleocene until Miocene, the collision of the Caribbean plate with South America resulted in uplift in the Cordillera Central (Toussaint, 1999). Starting in the Neogene, subduction was blocked in the collision area between the Panama Choc block and the Cordillera Occidental of Colombia and this ongoing collision is producing deformation in the three cordilleras (Taboada et al., 2000; Cortes and Angelier, 2005). Modern stress regimes are accommodated by major structures within the North Andean Block, which is part of the South American plate (Taboada et al., 2000), and within the PanamaChoc block (Duque-Caro, 1990; Coates et al., 2004; Cortes and Angelier, 2005). The Bocon and Guaicramo faults separate the North Andean Block proper from the South American Plate (Figs. 1 and 4).The North

71 1.2 and 77 1.7 Ma suggesting that formation occurred close to the transition CampanianMaastrichtian (Restrepo et al., 2007). 2.2. The Antioqueo Plateau The AP is an extensive (N5000 km2) low-relief geomorphic domain, which appears to be a relict surface located in the northernmost portion of the Cordillera Central (Arias, 1995). It is anked in the east and west by the Magdalena and Cauca river valleys, respectively (Fig. 1). Mean elevation in the AP is 2500 m and it possesses subdued internal topography characterized by rolling hills with local relief b40 m and slopes b10. Its overall geometry resembles an asymmetric, truncated pyramid with its steepest margin corresponding to the Cauca River canyon (Figs. 1 and 3). The AP is dominated by a dendritic drainage network and developed over gently rolling terrain where most hill slopes are mantled by saprolitic horizons up to 100 m thick. Inselbergs (e.g., bornhardts and tors) are conspicuous, sporadic geomorphic features throughout the AP and in some cases create local relief N300 m. To the north and northeast the AP descends in a series of step-like, less extensive erosional surfaces into the lowlands of the Lower Cauca and Magdalena Rivers. Internal low-relief of the AP is interrupted by steep regional escarpments and deeply incised gorges (Figs. 1 and 3). In some localities, steep

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112

Fig. 4. Modern tectonic setting. Generalized map of tectonic plates and crustal blocks around the Northern Andes of Colombia. Arrows and numbers indicate, respectively, the direction and velocity of plate's motion. (Adapted from Cortes and Angelier, 2005; Coates et al., 2004; Taboada et al., 2000; Duque-Caro, 1990; Trenkamp et al., 2002.).

Andean Block is escaping rigidly at 6 2 mm/a to the northeast with the South American craton acting as a rigid buttress while the PanamaChoc block is in active collision at a rate of 25 mm/yr (Trenkamp et al., 2002). Finally, three lithospheric blocks with different relative strengths are invoked to describe the kinematics of the Northern Andes of Colombia (Montes et al., 2005). The area that encompasses the AP is characterized as a lithospheric unit of high rigidity (Montes et al., 2005). Two major variations in the rate of convergence between the Nazca (Farallon) and South American plates have been reported for the Eocene and Oligocene (Hey, 1977; Wortel, 1984; Pardo-Casas and Molnar, 1987; Somoza, 1998). These changes may have determined morphotectonic, deformational and magmatic trends in the whole Andes (Mgard, 1984; Pilger, 1984). Collision of the PanamaChoc block is considered as the trigger of the modern topographic conguration of the Northern Andes and has been invoked as the cause of the most recent phase of uplift and exhumation in the region (EuAndina Phase of Van der Hammen, 1960), as well as the driver of Late MiocenePliocene uplift of the Cordilleras Occidental and Oriental that has produced most of the present day structural relief (Duque-Caro, 1990; Kroonenberg et al., 1990; Taboada et al., 2000; Gmez et al., 2005; Hooghiemstra et al., 2006). The Cordillera Central may represent a crustal-scale, positive ower structure (Taboada et al., 2000; Gmez et al., 2005). The age of major deformation of the Cordillera Central at the latitude of this study is pre-middle Eocene, as evidenced by the widespread Middle Magdalena Valley unconformity (Gmez et al., 2005). Morphotectonic evolution of the Cordillera Central has exerted a major control on the tectonostratigraphic development of the Magdalena and protoCordillera Oriental basins (Gmez et al., 2005). Elevated areas in the Cordillera Central were a source of detrital material for basins on both

sides of this range. Good examples are the Late Cretaceous to present sedimentary sequences in the Magdalena Basin to the east (Reyes et al., 2002; Gmez et al., 2005) and the continental, coal-rich Amag Formation with N1500 m of sediments deposited on the west side of the Cordillera Central (Grosse, 1926; Gonzlez, 2001). 3. Apatite (UTh)/He thermochronology AHe dating is a well-established thermochronometer with temperature sensitivity between ~40 C and 80 C (House et al., 1998; Farley, 2002; Ehlers and Farley, 2003; Stockli, 2005), a range that is lower than that of any other routinely used thermochronometer. Assuming geothermal gradients of 2030 C/km and mean annual surface temperature of 15 C, this temperature range is equivalent to depths of 1.23 km. AHe analysis has been used in diverse tectonic settings to investigate several geologic processes in the upper crust, including landscape and topographic development (House et al., 1998; Clark et al., 2005), orogenic exhumation in collisional orogens (Reiners et al., 2003), uplift/erosion in transpressive environments (Spotila, 2005), development of rifted margins and escarpments (Persano et al., 2002), and footwall exhumation in extensional settings (Stockli et al., 2000; Stockli, 2005). AHe thermochronometry of samples collected from near-vertical elevation proles enables study of the timing, magnitude, and rate of cooling of rocks as they traverse the upper 14 km of the earth's crust, the realm strongly inuenced by upper-crustal tectonic and erosional processes (Ring et al., 1999; Ehlers and Farley, 2003; Spotila, 2005; Stockli, 2005; Reiners and Brandon, 2006). Improved reconstructions of erosional exhumation by AHe are achieved where (1) geothermal gradients have remained relatively stable through time (Moore and England, 2001); (2) denudation rates are not excessively high

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112

Table 1 Summary of (UTh)/He results for Matasanos and La Garca vertical transects Sample number SR-9 SR-11 SR-15 SR-19 SR-6 SR-2 SR-40 SR-41 SR-CC3 SR-CC2 SR-CC1 SR-26 SR-31 SR-32 SR-45 SR-46 SR-48 SR-44 Age (Ma) 43.6 43.4 48.9 40.7 33.7 36.6 31.4 25.1 24.2 23.9 22.8 46.7 42.9 41.3 45.7 40.8 32.2 26.6 (Ma) 2.2 2.2 2.4 2.0 1.7 1.8 1.6 1.3 1.2 1.2 1.1 2.3 2.1 2.1 2.3 2.0 1.6 1.3 Std. dev. 1.6 2.5 3.9 2.8 2.5 0.8 2.5 3.5 2.6 2.8 1.5 3.4 1.1 0.5 1.5 2.7 1.8 3.8 A/G 3/3-2 3/2-1 5/2 5/3-2 3/3 3/2 4/2 3/2 3/2 3/2 3/2 3/3-2 3/3 3/2 4/2 3/3 4/3-1 3/2 Mass (mg) 2.8 2.8 4.6 4.0 9.4 4.0 3.6 3.7 8.0 3.1 3.6 5.7 5.6 4.3 4.1 9.2 4.6 3.5 U (ppm) 17.5 13.5 24.8 17.7 20.3 8.4 44.4 42.5 18.7 29.0 19.0 14.1 14.9 13.1 8.3 10.7 14.7 16.7 Th (ppm) 9.5 7.7 3.4 8.0 29.5 10.8 55.4 56.7 39.7 18.2 15.0 15.3 15.1 13.9 9.7 11.3 15.3 15.9 He (ncc/mg) 76.5 54.0 112.3 54.0 82.3 34.5 156.7 115.6 63.4 67.0 39.7 76.1 68.2 57.6 42.0 49.2 49.1 45.3 Ft 0.70 0.70 0.73 0.73 0.72 0.70 0.71 0.70 0.80 0.70 0.70 0.75 0.71 0.70 0.71 0.74 0.70 0.69 Elev. (m) 2370 2230 2100 1990 1710 1520 1500 1380 1070 1000 760 2350 2170 2110 2015 1900 1710 1640 Lat N/Lon Wtransect 6.46/75.37 M 6.46/75.37 M 6.45/75.37 M 6.45/75.36 M 6.43/75.37 M 6.47/75.38 M 6.41/74.44 M 6.41/75.41 M 6.76/75.12 M 6.80/75.14 M 6.86/75.18 M 6.38/75.59 G 6.37/75.59 G 6.38/75.60 G 6.36/75.59 G 6.36/75.59 G 6.35/75.58 G 6.34/75.58 G

The column marked as A/G indicates the number of aliquots (A) and the number of grains (G) per aliquot analyzed for each sample, for instance, in sample SR-9 a total of 3 aliquots were analyzed and each aliquot was prepared using two to three grains. Average mass of aliquots run for each sample is indicated in column Mass (mg). 147Sm concentrations (ppm) in all samples analyzed were virtually 0 or below the detection limits.

(N5 mm/yr) as to promote signicant heat advection in the upper crust (Brown and Summereld, 1997; Mancktelow and Grasemann, 1997); (3) pre-existing topography was not rough, i.e., topographic amplitude N3 km, topographic wavelength b20 km at denudation rates N0.5 mm/yr (Stwe et al., 1994); and 4) surface relief amplitude has not changed signicantly since the rocks cooled through the closure temperature, as this has a strong effect on the slope of the age-elevation relationship and can lead to overestimation/underestimation of exhumation rates (Braun, 2002). Although AHe can be used to reconstruct exhumation histories even when these criteria are not met, uncertainties associated with these four factors are minimized if a vertical prole for sample collection is available and if samples are collected over a large range of elevations. Extended vertical proles allow a more thorough analysis of features displayed in age-elevation proles such as exhumation events, helium partial retention zone (He-PRZ), and paleogeothermal gradients, which permit better constraints on denudation (Stockli et al., 2000; Braun, 2002; Farley, 2002). The AP possesses geologic, morphologic and tectonic characteristics that make it a perfect scenario to undertake AHe low temperature thermochronology studies along vertical proles. 4. Methods The AHe technique is based on the accumulation of 4He ( particles) resulting from the radioactive decay of the parent isotopes 235U, 238U, 232 Th, and 147Sm. Thermokinetics of helium diffusion and -particle stopping distances in apatite (20 m) are well constrained so that measurements of 4He and its parent isotopes by mass spectrometry allows calculation of the system's closing time (House et al., 1999; Farley, 2000, 2002). Measurements were made by degassing multigrain aliquots through laser heating and evaluating 4He by isotope-dilution gas source mass spectrometry, followed by determination of U, Th, and Sm on the same crystals by isotope-dilution ICP-MS. Mean (UTh)/He ages were calculated on the basis of 35 apatite replicate analyses and reported age uncertainties (2) reect the reproducibility of replicate apatite analyses (6%) of laboratory standard samples (Stockli et al., 2000). AHe analyses were conducted on samples from two separate elevation proles located 40 km apart in the central portion of the AP. The La Garca and MatasanosPorce proles are situated on a regional scarp developed on the northern ank of the Medelln/Porce River canyon. Samples were collected at elevation intervals of 70 m,

spanning the largest possible range of paleocrustal depths (2 km) in the exposed structural relief along this margin of the AP (Fig. 3). Apatite concentrates were obtained through the conventional method of heavy liquid and magnetic susceptibility separation. Nineteen samples yielded good quality inclusion-free, euhedral, non-fractured apatite with N65 m in the prism width, suitable for AHe analysis (Farley, 2002; Ehlers and Farley, 2003). AHe apparent ages were measured at the (UTh)/He Laboratory of the University of Kansas following the methods described in Farley (2002), Reiners et al. (2003) and Stockli et al. (2000), and available as a Supplemental le. 5. Results U, Th and He concentrations, Ft correction for grain dimensions and alpha ejection corrected He ages are reported in Table 1 for the La Garca and MatasanosPorce proles. All individual AHe sample

Fig. 5. Relation between apparent apatite (UTh)/He ages and elevation integrated for La Garca (black squares) and Matasanos (gray triangles) vertical proles. Error bars are 2. Bottom of the exhumed He-PRZ marked by inection point at 1500 m. Upper boundary of the He-PRZ appears to be beyond the maximum elevation sampled. Slopes and R2 values of the regression lines by segments appear in front of the segment's symbol for the two exhumation pulses discussed in the text. Block arrows in gray indicate the duration of the Pre-Andina and Proto-Andina orogenic phases of Van der Hammen (1960).

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112

Fig. 6. Relation between apparent apatite (UTh)/He ages (this study; black squares) + AFT ages (Saenz, 2003; grey circles) and elevation. Error bars are 2. Gray dotted lines and block arrows indicate the extent of the Pre-Andina Phase (Van der Hammen, 1960).

aliquot analyses displayed excellent reproducibility. Similarity in the age-elevation relationships for both proles indicates consistency of the data sets and allowed us to combine AHe data into one composite vertical transect (Fig. 5). Most aliquots yielded ages with standard deviations b2.0 Ma. Only one sample, SR-11, yielded an aliquot with irreproducible 4He concentrations (see Data supplement) and AHe age of 1088.8 Ma. This was likely caused by parentless 4He hosted by mineral or uid U- and/or Th-bearing micro-inclusions, which is the most common reason for irreproducible ages (Stockli et al., 2000; Farley, 2002; Ehlers and Farley, 2003). Small, acicular mineral inclusions were found in some resin-mounted transparent crystals from sample SR-11 when analyzed under high magnication. Spontaneous ssion-track distributions in other apatite grains from the same separates show negligible zoning, therefore, -emission

corrections, which assume homogenous U and Th distributions (Farley, 2002), can be considered adequate for the population of grains analyzed. AHe ages in the La Garca vertical transect increase systematically with elevation from 26.61.3 Ma at the bottom of the scarp to 46.7 2.4 Ma at the level of the plateau. Virtually identical results were obtained in the Matasanos transect, with samples ranging in age from 22.8 1.1 Ma to 48.9 2.4 Ma (Fig. 5 and Table 1). Variations in AHe ages with elevation in this study are positively correlated and resemble theoretical He retention vs. depth curves (Wolf et al., 1998) and age-elevation relationships obtained for other He vertical-prole studies (Stockli et al., 2000; Reiners et al., 2003). Positive correlation in our age-elevation proles, similarity in slope between previous apatite ssion-track data (Saenz, 2003), and modern topography of the AP suggest that relief changes have been negligible since Eocene time, therefore, topographic corrections by the admittance ratio (Braun, 2002; Reiners et al., 2003) to our data are not required. Further, the combination of U and Th concentrations in our apatite grains (average 19 ppm) and previous apatite ssion-track ages reported for the same geologic province (2645 Ma, Saenz, 2003) fall within the gures recently proposed as the range over which excessive He retention due to radiation damage, and concomitant increases in AHe ages should be negligible (Green et al., 2006). Even though these are general guidelines, they enhance our condence that the He kinetics we used in interpreting our data are a reasonable assumption, so that none of our AHe apparent ages are potential overestimates of the time of cooling through the He partial retention zone. The age-elevation relationship for all data is characterized by three segments (Fig. 5). Segment 1 between 760 and 1600 m displays AHe ages from 22 to 26 Ma, exhibits very low data dispersion (R2 = 0.98), and has a slope of 0.24 mm/yr. Segment 2 at sample elevations between 1600 and 2000 m shows a broader range of AHe ages from 2641 Ma (R2 b 0.2) and a lower slope than the other two segments (0.02 mm/yr). Segment 3 at elevations N2000 m is characterized by apparent ages from 41 to 49 Ma, and data scatter higher than in segment 1 but lower than in segment 2 (R2 b 0.5) and a slope of 0.08 mm/yr.

Fig. 7. Simplied sketch illustrating the extent of the crustal section removed since ca. 25 Ma (dotted area). Left panel represents crustal conditions at the onset of major exhumation 25 million years ago. Right panel is the actual condition with remnants of a Miocene surface that has begun to be obliterated by uplift related denudation during the most recent Andean phase (Eu-Andina IIIV, 12 MA to present). Deep and localized incision of the Medelln and Porce rivers is shown. Rock uplift discussed in the text is represented by the displacement of point A to A. Bottom of He-PRZ represents the paleodepth at which AHe ages were zero before onset of rapid exhumation at 25 Ma. PB = Pramo de Belmira, AP = Antioqueo Plateau.

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112

6. Discussion and interpretation Our AHe results complement apatite and zircon ssion track thermochronology from the Antioqueo Batholith and associated plutons providing additional detail to constrain the exhumation history of the region during Cenozoic time. The AHe data reveal a

marked cooling event at ca. 23 Ma (Fig. 5). We interpret the almost invariant He ages of segment 1 as representing a rapid exhumation pulse starting at 25 Ma. Segment 2 of the age-elevation prole is interpreted to be part of the lower segment of the OligoMiocene HePRZ and may represent a period of tectonic quiescence lasting 17 million years. The change in slope at 41 Ma ( 2000 m) is not

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112

Fig. 8. Simplied cartoon of cross-sections depicting the paleogeographic and geologic evolution of the northern Andes at the latitude of the study site. (a) HauterivianConiacian: the ancestral Cordillera Central was covered by a shallow sea and marine sedimentary units were being deposited. The Proto-Cordillera Occidental was part of the subducting plate separated from the Cordillera Central by the ancestral Romeral Fault System. (b) SantonianMaastrichtian: Increased convergence rates lead to the magmatism of the Antioqueo Batholith (Ksta), a shallow (epizonal) intrusion emplaced into PaleozoicPrecambrian crystalline basement and lower Cretaceous marine sequences (contact metamorphism). Initial phase of uplift in the region. Cordillera Central emerges leading to a regional unconformity and becomes source of sediment to eastern and western marginal basins (Proto Magdalena and Proto Cauca basins respectively). Progradation to the east begins in the eastern margin of the Cordillera Central. Subsidence and sedimentation are prevalent over the region of the Proto-Cordillera Oriental. (c) Paleocene, Laramic Phase, 60 Ma: Generalized coastal deposition took place, inversion of the eastern margin of the Cordillera Oriental basin is now pronounced. Cordillera Central continued to be gradually uplifted. Initial phases of uplift of the Cordillera Occidental begin. Reactivation of vertical movement along the CaucaRomeral and Palestina fault systems. (d) Lower to Middle Eocene, Pre-Andina Phases III, 5445 Ma: climax of Pre-Andean Orogeny in Colombia. No deposition taking place in most of Colombia. Exposure of the future Middle Magdalena Valley as depocenter. (e) Oligo-Miocene transition, Proto-Andina Phase, 2421 Ma: Major phase of uplift across the Colombian Andes, uplift on the three cordilleras. Amag formation begins to develop in the Cauca depression. Cordilleras Occidental and Oriental still low standing. Localized subsidence in the Cauca and Magdalena depocenters. Nazca plate subduction begins after splitting of the Farallon plate. (f) Miocene to Present, Eu-Andina Phases IIV, 180 Ma: Major uplift and initial inversion of the Cordillera Oriental by reactivation of inherited JurassicCretaceous steep and deeply rooted normal faults. Amag formation is folded, Colombian Andes attain the actual topographic conguration. The majority of the overburden of the Antioqueo Batholith has been removed to expose the bulk of this massive igneous unit. (Major events in this gure are compiled from: this study; Gmez, 2005; Gonzlez, 2001; Grosse, 1926; Pardo-Casas and Molnar, 1987; Restrepo and Toussaint, 1988; Restrepo et al., 2007; Tapias et al., 2006; Van der Hammen, 1961).

as well dened as the one at 25 Ma (1500 m). It is probably indicative of an exhumation event that was less signicant than the Oligocene episode. Apatite ssion-track data (Saenz, 2003), although not derived from vertical proles, provide additional insight to evaluate the AHe change in slope between segments 2 and 3 at ca. 41 Ma. All of the apatite ssion-track apparent ages are equal or older than AHe ages with a mean age of ca. 46 Ma (Fig. 6). The samples show unimodal ssion-track length populations with a mean value of 14.1 m indicating relatively rapid cooling in Eocene time (Saenz, 2003). This nding corroborates our interpretation that the segment 3 of the age-elevation plot records an Eocene exhumation event. The average

AHe age for segment 3 (41 Ma) is a minimum age for exhumation as these samples resided in the lower temperature part of the He PRZ when exhumation slowed. Nearly concordant AHe and apatite ssion-track ages over a crustal section of 1500 m (Fig. 6) indicate that rocks were exhumed from temperatures above the apatite ssion-track partial annealing zone (N110 C) to below most of the AHe PRZ. This suggests that rocks within this zone cooled ca. 50 C in 5 million years, which, at a gradient of 25 C/km, implies unroong on the order of 2 km, i.e., 0.4 mm/yr. Basal conglomerates in the Magdalena and Cauca basins deposited at this time also suggest signicant erosion of the AP during the Eocene (e.g., Gmez et al., 2005; Van der Hammen, 1960).

10

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112

The AHe results are consistent with interpretations that kilometerscale uplift and exhumation in the Northern Andes, as well as in other portions of the Andes, took place in discrete pulses in the Cenozoic (Steinman, 1929; Van der Hammen, 1960; Mgard et al., 1984). These exhumation events occurred during or just after rapid rates of convergence between the Farallon (Nazca) and South America (150 mm/yr) documented for the Middle Eocene and the Late Oligocene (Pardo-Casas and Molnar, 1987), suggesting a relationship between orogeny and subduction dynamics. Such pulses of exhumation are separated by interludes of tectonic quiescence of durations in excess of 15 million years, spans during which low-standing erosional surfaces (peneplain in the Davisian sense) can develop. The inection point between segments 1 and 2 of the age-elevation plot (Fig. 5) denes the bottom of He-PRZ established towards the end of the quiescence period between ca. 45 and 25 Ma when erosional processes lead to the development of the proto-AP. This pre-early Miocene He-PRZ was exhumed at ca. 25 23 Ma. The base of this fossil He-PRZ (80 C paleo-isotherm) is at a modern depth relative to the present AP surface of 1000 m. Before exhumation rates increased at ca. 25 Ma, the break in slope was at a depth of about 2.6 km, assuming a geothermal gradient of 25 C/km and a mean surface temperature of 15 C. This implies rock uplift of approximately 1.5 km (Fig. 7). Based on the assumed paleogeothermal gradient and the temperature difference across the He-PRZ (40 C), the top of the pre-Miocene He-PRZ is probably at 3100 m, which is the elevation average in the Las Baldas Range on top of the AP (Fig. 3). If our assumptions for the geothermal gradient and pre-existing topography are correct, only about 900 m of plateau have been removed by erosion in the last 25 Ma, yielding a low erosion rate for an active orogen (0.04 mm/yr). Erosion in the summits of the Las Baldas Range (e.g., Pramo de Belmira) must have been very limited, i.e., less than 200 m in the last 25 million years (0.008 mm/yr), and such surfaces probably represent a position closer to a true relict of the original plateau. Our interpretation implies very contrasting efciencies in erosion on the granitic units of the Antioqueo Batholith (0.04 mm/yr) relative to the metamorphic units in Las Baldas Range (0.008 mm/yr) where less than 200 m of crustal material (and in some cases b100 m) have been removed since Late Oligocene. The spatially restricted remnants of Early Cretaceous marine sedimentary strata present in the area are portions of the overburden removed by erosion during the development of the present AP since 25 Ma, as well as during previous exhumation pulses in the Eocene and Paleocene. The low values for background erosion rates derived from our study are in agreement with denudation rates for subdued relief and soil mantled regions suggested by other authors (Ahnert, 1970; Summereld and Hulton, 1994) supporting the idea that the AP is a relict surface that has not adjusted to modern erosional conditions of the orogen. Such rates are in clear contrast relative to proposed gures of modern erosion in the region, i.e., ~ 1 mm/yr (Giraldo, 2005; Restrepo and Syvitski, 2006). The discrepancy may be explained by the intensication of erosional process due to anthropogenic inuence in the Magdalena basin, a problem not addressed in detail in this investigation, but that certainly warrants further consideration. Chronology and intensity of exhumation pulses of the AP documented in this study coincide remarkably well with the timing and relative magnitude of the orogenic phases Proto-Andina (24 Ma) and Pre-Andina II ( 43 Ma) proposed for the Northern Andes of Colombia by Van der Hammen (1960) based on a stratigraphic and paleobotanical study. Further, these phases of erosion occurred at the same time as the discrete orogenic events in Peru known as the Incaic in the Eocene, and The Quechua 1 in the Early Miocene (Mgard, 1984). The Late OligoceneEarly Miocene exhumation event recorded in our study is consistent with documented erosional phases for the Peruvian Andes and the Bolivian Eastern Cordillera, where GregoryWodzicki (2000) reported an erosive phase at ca. 22 Ma that removed 2 km of overburden. For the Eastern Cordillera of the Andes of

central Peru apatite ssion track data record two exhumation pulses during the Neogene at ca. 21 Ma and between 12 Ma and the Present, which removed about 4 to 6 km of overburden in the past 30 million years (Laubacher and Naeser, 1994). Similarly, an exhumation pulse at 40 Ma was reported by Benjamin et al. (1987) in his study of ssiontrack thermochronology for the Bolivian Andes, whereas analogous timing for Eocene cooling events associated with erosion have been indicated for the Sierras Pampeanas in Argentina (Coughlin et al., 1998) and northern Bolivia (Barnes et al., 2006). Rapid exhumation events in mountain ranges in the Caribbean region (Rojas-Agramonte et al., 2006) have also been recognized for the Late Eocene and Mid to Late Oligocene. This evidence suggests that Cenozoic exhumation pulses in the Northern Andes and the Caribbean are controlled by continentalscale tectonics. The trigger of these accelerated pulses of erosion may be related to well documented changes in convergence rates between the Nazca (Farallon) and the South American plates (Hey, 1977; PardoCasas and Molnar, 1987; Somoza, 1998), which are in turn related to major reorganization of the plates such as the one caused by the break up of the Farallon Plate into Nazca and Cocos plates. The period of relative tectonic quiescence between the two exhumation pulses provides enough time for the development of an erosional surface on which the Terciario Carbonifero de Antioquia of OligoceneMiocene age was deposited (Grosse, 1926). Our results are also in agreement with the tectonostratigraphic development of sedimentary sequences in the Magdalena Inter Andean depressions (Reyes et al., 2002; Gmez et al., 2005) corroborating that the Cordillera Central was an elevated massif that functioned as an important source of detrital material to major sedimentary basins during the Paleogene and Neogene. Late PaleogeneNeogene sedimentary sequences found in the AP (Parra, personal communication) imply that uvial activity and concomitant obliteration of the original erosional surface in the AP have been in operation throughout the Neogene. AHe data in this study do not record postMiddle Miocene exhumation events, i.e., the EuAndina orogenic phase of Van der Hammen (1960), which has been invoked as the phase responsible for the modern conguration of the Andean range and which has also been recorded in many massifs throughout the Andes. The fact that the Eu-Andina phase is not veried by AHe data, in conjunction with low spatially-averaged, long-term erosion rates derived in this study, constitutes compelling evidence that the AP is in geomorphic disequilibrium. The modern expression of the AP developed through a long history (Fig. 8) of multiple phases of erosional surface formation at low elevation (peneplanation) beginning with its emergence in the SantonianEarly Masstrichtian (Van der Hammen, 1960). The Eocene and OligoceneMiocene exhumation events here reported were initiated after kilometer-scale orogenic uplift accompanied by uvial incision and erosion that removed part of the cover of the Antioqueo Batholith including the majority of the Cretaceous marine formations. Although there seems to be an agreement in regards to the timing of the major phases of surface uplift-denudation for the Andes, paleoelevation of the region remains for the most part undocumented. Gregory-Wodzicki (2000) suggested that the Colombian Andes was at 40% of its modern elevation by 4 Ma. Similar results, e.g., uplift of the Colombian Andes concentrated in the last 10 Ma and intensied in the Pliocene, were obtained by Van der Hammen et al. (1973) and by Hooghiemstra et al. (2006) from paleobotany studies in the Cordillera Oriental. We suggest that the age of the most recent peneplain in the AP is Miocene. During the Pre-Andina (Eocene) and Proto-Andina (OligoMiocene) phases the area experienced kilometer-scale uplift followed by erosion surface development (relief obliteration) during intervals of tectonic quiescence. The actual elevation of the AP was attained with the onset of the Eu-Andina phase, particularly the sub-phases II through IV (12 Ma to present) when the erosion surface raised from close to sea level (e.g., 500 m) to ca. 3600 m. However, reliable paleoelevation data for the Cordillera

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112

11

Central is scant so the question about the progression of elevation within the context of orogenic pulses and uplift phases in the region has only been addressed indirectly, particularly in relation to the evolution of depocenters in the Magdalena Basin and its development from a foreland basin to an Interandean basin during the Cenozoic (Gmez et al., 2005; Reyes et al., 2002). Deep incision in the AP along the MedellnPorce river is probably a recent feature developed during surface uplift in the Pliocene. Modern geomorphic expression of the AP is the result of recent uvial processes controlled by nick point propagation. The closest point to a relict Mesozoic surface may be found at the summits of Pramo de Belmira, although glacial erosion during PaleoceneHolocene glacial periods may have removed some of it (100 m). The saprolitic cover of the AP is absent above 3300 m elevation where the landscape is dominated by remnants of glacial landforms and truncated summits. Similarity in the age versus elevation plots for both proles (Fig. 5) is interpreted as indicating exhumation of the entire AP as a discrete unit. Our results support the hypothesis that the Northern Cordillera Central is part of a structurally coherent crustal block within the orogen (Montes et al., 2005). The rigidity of the Northern Cordillera Central block, where the AP is located, may have controlled the nearly circular outcrop pattern of the Antioqueo batholith. Cross-sections presented by Cortes and Angelier (2005) show a major structural zone at 7N/7577W with high-angle, reverse faults anking the AP (Palestina and Romeral fault systems in the east and west respectively). These structures are inherited from the basement and bound the AP block. Rigidity of the block and bounding faults may have controlled the structural continuity of the area as well as the mode of emplacement and overall shape of the Antioqueo batholith, thus imposing litho-structural control in the geomorphic evolution of the AP. Reactivation of these major crustal structures during periods of enhanced convergence may have facilitated uplift of the AP as a coherent block. The Cenozoic uplift and exhumation history of the northern Cordillera Central of Colombia resulted from plate-tectonic reorganization and directly affected the morphotectonic evolution of the AP as well as the development of the sub-Andean fold-and-thrust belt and the ll and tectonostratigraphic evolution of the intramontane and foreland basins to the east (Magdalena, Llanos Basin) and west (Cauca). This is particularly clear at the end of the Oligocene when the Farallon plate broke up into the Nazca and Cocos plates (Wortel,1984; Pardo-Casas and Molnar, 1987). Augmented convergence rates induced synchronous tectonic and geomorphic activity on the overriding plate (i.e., uplift, exhumation, deformation, shortening, etc.) from Argentina to Colombia. A similar effect was also observed during an earlier phase of rapid convergence in the middle Eocene. Additional control to the more recent morphotectonic history in the area is related to the west-to-east motion of the Caribbean plate and the collision of the PanamaChoc block at 10 Ma, which closed the Panama isthmus between 3.7 and 3.4 Ma (Duque-Caro, 1990) triggering the most recent orogenic phase in the region and giving the Colombian Andes its present topographic conguration (Kroonenberg et al., 1990). These main phases of uplift of the Northern Andes led to important paleogeographic events (Hoorn et al., 1995) and to erosion of a large volume of sediments resulting in thick basal sandstones and molasse successions in the sub-Andean basins (Van der Hammen, 1960; Gmez et al., 2005). Age control for two of these major periods is accurately provided by our AHe data tting well the spatiotemporal framework for the evolution of the Colombian Andes in the context of continental tectonics. Acknowledgements This work is part of Sergio Restrepo's Ph.D. investigation. Sergio wants to thank his family for continued support. The study was funded by the Compton Foundation, American Geological Institute, National Science Foundation (SEAGEP Program), Corporacin para la Investigacin y el Ecodesarrollo Regional (NGO Colombia), University of Florida

Department of Geological Sciences, Geological Society of America (to Restrepo), and NSF EAR-0414817 (to Stockli). We thank Carlos Suescn and Camilo Polanco for their help in the eld, and Stephanie Brichau and others at the University of Kansas (UTh)/He Laboratory for laboratory assistance. The Geomorphology Group at ICNE-Universidad Nacional de Colombia provided unpublished information and valuable discussions. Juan D. Moreno and Patricia Monsalve were of great help in generating the graphics that accompany the text. Thanks are also due to the reviewers of this manuscript, B. Kohn and J. Hourigan, as their input and criticism helped improve the scientic quality of the article. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.epsl.2008.09.037. References
Ahnert, F., 1970. Functional relationships between denudation, relief, and uplift in large mid-latitude basins. Am. J. Sci. 268, 243263. Arias, L.A., 1995. El relieve de la zona central de Antioquia: un palimpsesto de eventos tectnicos y climticos. Revista Faculta de Ingeniera Universidad de Antioquia, pp. 924. Barnes, J.B., Ehlers, T.A., McQuarrie, N., O'Sullivan, P.B., Pelletier, J.D., 2006. Eocene to recent variations in erosion across the central Andean foldthrust belt, northern Bolivia: Implications for plateau evolution. Earth Planet. Sci. Lett. 248, 118133. Benjamin, M.T., Johnson, N.M., Naeser, C.W., 1987. Recent rapid uplift in the Bolivian Andes: evidence from ssion track dating. Geology 15, 680683. Botero, G., 1963. Contribucin al conocimiento de la geologa de la zona central de Antioquia. Anales Facultad de Minas Medelln. 57, 7101. Braun, J., 2002. Quantifying the effect of recent relief changes on ageelevation relationships. Earth Planet. Sci. Lett. 200, 331343. Brown, R.W., Summereld, M.A., 1997. Some uncertainties in the derivation of rates of denudation from thermochronologic data. Earth Surf. Process. Landf. 22, 239248. Clark, M.K., Maheo, G., Saleeby, J., A, F.K., 2005. The non-equilibrium landscape of the southern Sierra Nevada, California. GSA Today 15, 410. Coates, A.G., Collins, L.S., Aury, M.-P., Berggren, W.A., 2004. The geology of the Darien, Panama, and the late MiocenePliocene collision of the Panama arc with northwestern South America. Geol. Soc. Am. Bull. 116, 13271344. Coltorti, M., Ollier, C.F., 1999. The signicance of high planation surfaces in the Andes of Ecuador. In: Smith, B.J., Whalley, W.B., Warke, P.A. (Eds.), Uplift, Erosion and Stability: Perspectives on Long-term Landscape Development. Special Publications, vol. 162. Geological Society, London, pp. 239253. Cortes, M., Angelier, J., 2005. Current states of stress in the Northern Andes as indicated by focal mechanisms of earthquakes. Tectonophysics 403, 2958. Coughlin, T.J., O'Sullivan, P., Kohn, B.P., Holcombe, R.J., 1998. Apatite ssion-track thermochronology of the Sierras Pampeanas, central western Argentina: implications for the mechanism of plateau uplift in the Andes. Geology 26, 9991002. Davis, W.M., 1899. The geographical cycle. Geog. J. 14, 481504. Duque-Caro, H., 1990. The Choco Block in the northwestern corner of South America: structural, tectonostratigraphic and paleogeographic implications. J. South Am. Earth Sci. 3, 7184. Ehlers, T.A., Farley, K.A., 2003. Apatite (UTh)/He thermochronometry: methods and applications to problems in tectonic and surface processes. Earth Planet. Sci. Lett. 206, 114. Etayo-Serna, F., 1985. El lmite Jursico Cretcico en Colombia. Proyecto Cretcico. Publicaciones Geolgicas Especiales, vol. 16. INGEOMINAS, Bogot, pp. 23/2123/24. Farley, K.A., 2000. Helium diffusion from apatite: general behavior as illustrated by Durango Fuorapatite. J. Geophys. Res. 105, 29032914. Farley, K.A., 2002. (UTh)/He dating: techniques, calibrations, and applications. Rev. Mineral. Geochem. 47, 819844. Feininger, T., Botero, G., 1982. The Antioquian Batholith, Colombia. Publicacin Geolgica Especial INGEOMINAS, Bogot. 150 pp. Feininger, T., Barrero, D., Castro, N., 1972. Geologa de parte de los departamentos de Antioquia y Caldas (Sub-zona II-B). Bol. Geol. - INGEOMINAS 20, 1173. Fujiyoshi, A., Ishishaka, K., Heyese, Y., Tokoyuma, A., 1976. Metamorphic and igneous rocks from the Medelln-Yarumal and Santa Marta areas, Colombia, and their Rb/Sr ages. J. Geol. Soc. Jpn. 82, 559563. Giraldo, M.C., 2005. Sondeo automatizado de embalses. Revista Hidrometeorolgica Empresas Pblicas de Medelln 1, 121137. Gmez, E., Jordan, T.E., Allmendinger, R.W., Cardozo, N., 2005. Development of the Colombian foreland-basin system as a consequence of the diachronous exhumation of the Northern Andes. Geol. Soc. Am. Bull. 117, 12721292. Gonzlez, H., 2001. Mapa Geolgico del Departamento de Antioquia Escala 1:400.000, Memoria Explicativa, INGEOMINAS, Medelln, 240 pp. Green, P.F., Crowhurst, P.V., Duddy, I.R., Japsen, P., Holford, S.P., 2006. Conicting (UTh)/ He and ssion track ages in apatite: enhanced He retention, not anomalous annealing behavior. Earth Planet. Sci. Lett. 250, 407427. Gregory-Wodzicki, K.M., 2000. Uplift history of the central and Northern Andes: a review. Geol. Soc. Am. Bull. 112, 10911105.

12

S.A. Restrepo-Moreno et al. / Earth and Planetary Science Letters 278 (2009) 112 Restrepo, J.J., Toussaint, J.F., 1988. Terranes and continental accretion in the Colombian Andes. Episodes 11, 189193. Restrepo, S.A., Foster, D.A., Kamenov, G.A., P, O.S., 2007. Formation age and magma sources for the Antioqueo and Ovejas batholiths derived from UPb dating and Hfisotope analysis of zircon grains by LA-ICP-MS. GSA Abstr. Progr. 39 (6) GSA. Reyes, J.P., Rueda, M.C., Gmez, P.D., Mantilla, M., 2002. Distribucin y caractersticas estratigrcas de las unidades cronoestratigrcas Cretceas y Terciarias, Cuenca del Valle Inferior del Magdalena y Cinturones Plegados de San Jacinto y Sin., Memorias de la Segunda Convencion Tcnica de la Asociacin Colombiana de Gelogos y Geofsicos del Petrleo, 5 pp. Ring, U., Brandon, M.T., Lister, G.S., D, W.S. (Eds.), 1999. Exhumation Processes: Normal Faulting, Ductile Flow and Erosion. Geological Society Special Publication, London. 378 pp. Rojas-Agramonte, Y., Neubauer, F., Bojar, A.V., Hejl, E., Handler, R., Garca-Delgado, D.E., 2006. Geology, age and tectonic evolution of the Sierra Maestra mountains, southeastern Cuba. Geol. Acta 4, 123150. Saenz, E.A., 2003. Fission track thermochronology and denudational response to tectonics in the north of The Colombian Central Cordillera [Masters Thesis]: Matsue-shi, Shimane University Japan. Scheibe, R., 1919. Informe sobre resultados de la comisin cientca nacional en Antioquia. Compilacin Estudios Geolgicos Ociales en Colombia 1, 95165. Schildgen, T.F., Hodges, K.V., Whipple, K.X., Reiners, P.W., Pringle, M.S., 2007. Uplift of the western margin of the Andean plateau revealed from canyon incision history, southern Peru. Geology 35, 523526. Sobolev, S.V., Babeyko, A.Y., 2005. What drives orogeny in the Andes? Geology 33, 617620. Somoza, R., 1998. Updated Nazca (Farallon)South America relative motions during the last 40 My: implications for mountain building in the central Andean region. J. South Am. Earth Sci. 11, 211215. Spotila, J.A., 2005. Applications of low-temperature thermochronometry to quantication of recent exhumation in mountain belts. In: Reiners, P.W., Ehlers, T.A. (Eds.), Low-temperature Thermochronology: Techniques, Interpretations, and Applications. Rev. Mineral. Geochem., vol. 58. MSA, Chantilly, VA, pp. 449466. Steinman, G., 1929. Geologie von Peru. Karl Winter, Heidelberg. 448 pp. Stockli, D.F., 2005. Application of low-temperature thermochronometry to extensional tectonic settings. In: Reiners, P.W., Ehlers, T.A. (Eds.), Low-temperature Thermochronology: Techniques, Interpretations, and Applications. Rev. Mineral. Geochem., vol. 58. MSA, Chantilly, VA, pp. 411448. Stockli, D.F., Farley, K.A., Dimitru, T.A., 2000. Calibration of the apatite (UTh)/He thermochronometer on an exhumed fault block, White Mountains, California. Geology 28, 983986. Stwe, K., White, L., Brown, R., 1994. The inuence of eroding topography on steady-state isotherms. Application to ssion track analysis. Earth Planet. Sci. Lett. 124, 6374. Summereld, M.A., Hulton, N.J., 1994. Natural controls of uvial denudation rates in major world drainage basins. J. Geophys. Res.-Sol. Ea. 99, 1387113883. Taboada, A., Rivera, L.A., Fuenzalida, A., Cisternas, A., Philip, H., Bijwaard, H., Olaya, J., Rivera, C., 2000. Geodynamics of the Northern Andes: subductions and intracontinental deformation (Colombia). Tectonics 19, 787813. Tapias, J.G., Nivia, A., Jimnez, D.M., Montes, N.E., Seplveda, J., Osorio, J.A., Tejada, M.L., Mora, M., Gaona, T., Diederix, H., Uribe, H., 2006. Mapa Geolgico de Colombia 1:2'800.000, Instituto Colombiano de Geologa y Minera, Bogot, Colombia, pp. Toussaint, J.F., 1993. Evolucin geolgica de Colombia, Tomo I: Precmbrico y Paleozoico. Editorial Universidad Nacional de Colombia, Medelln. 229 pp. Toussaint, J.F., 1999. Importancia de los movimientos de rumbo dextrales desde el Neoproterozoico en Colombia. Bol. Cienc. Tierra 13, 110. Trenkamp, R., Kellogg, J.N., Freymueller, J.T., Mora, H.P., 2002. Wide plate margin deformation, southern Central America and northwestern South America, CASA GPS observations. J. South Am. Earth Sci. 15, 157171. Van der Hammen, T., 1960. Estratigrafa del Terciario y Maestrichtiano continentales y tectognesis de los Andes Colombianos, Informe No. 1279. Servicio Geolgico Nacional, Bogot. 128 pp. Van der Hammen, T., Werner, J.H., van Dommelen, H., 1973. Palynological record of the upheaval of the Northern Andes: a study of the Pliocene and Lower Quaternary of the Colombian Eastern Cordillera and the early evolution of its High-Andean biota. Rev. Paleobot. Palynol. 16, 1122. Wolf, R.A., Farley, K.A., Kass, D.M., 1998. A sensitivity analysis of the apatite (UTh)/He thermochronometer. Chem. Geol. 148, 105114. Wortel, M.J.R., 1984. Temporal and spatial variations in the Andean subduction zone. J. Geol. Soc. Lond. 141, 783791.

Grosse, E., 1926. El Terciario Carbonifero de Antioquia. D. Reimer-E. Vohsen, Berlin. 361 pp. Hall, R., lvarez, J., Rico, H., 1972. Geologa de los departamentos de Antioquia y Caldas (Subzona II-A). Bol. Geol. - INGEOMINAS 20, 185. Harris, S.E., Mix, A.C., 2002. Climate and tectonic inuences on continental erosion of tropical South America, 013 Ma. Geology 30, 447450. Hey, R., 1977. Tectonic evolution of the CocosNazca spreading center. Geol. Soc. Am. Bull. 88. Hooghiemstra, H., Wijninga, V.M., Cleef, A.M., 2006. The paleobotanical record of Colombia: implications for biogeography and biodiversity. Ann. Mo. Bot. Gard. 93, 297. Hoorn, C., Guerrero, J., Sarmiento, G.A., Lorente, M.A., 1995. Andean tectonics as a cause for changing drainage patterns in Miocene Northern South America. Geology 23, 237240. House, M.A., Wernicke, B.P., Farley, K.A., 1998. Dating topography of the Sierra Nevada, California, using apatite (UTh)/He ages. Nature 396, 6669. House, M.A., Farley, K.A., Kohn, B.P., 1999. An empirical test of helium diffusion in apatite: borehole data from the Otway Basin, Australia. Earth Planet. Sci. Lett. 170, 463474. Kennan, L., Lamb, S., Hoke, L., 1997. High-altitude palaeosurfaces in the Bolivian Andes: evidence for late Cenozoic surface uplift. In: Widdowson, M. (Ed.), Palaeosurfaces: Recognition, Reconstruction and Paleoenvironmental Interpretation, vol. 120. Geological Society of London Special Publication, London, pp. 307324. Kroonenberg, S.B., Bakker, J.G.M., Van der Wiel, A.M., 1990. Late Cenozoic uplift and paleogeography of the Colombian Andes constraints on the development of highAndean biota. Geol. Mijnb. 69, 279290. Laubacher, G., Naeser, C.W., 1994. Fission-track dating of granitic rocks from the Eastern Cordillera of Peru: evidence for Late Jurassic and Cenozoic cooling. J. Geol. Soc. Lond. 151, 473483. Mancktelow, N., Grasemann, B., 1997. Time-dependent effects of heat advection and topography on cooling histories during erosion. Tectonophysics 270, 167195. Maya, M., Gonzlez, H., 1995. Unidades Litodmicas en la Cordillera Central de ColombiaInforme Unidad Operativa INGEOMINAS. INGEOMINAS, Medelln, pp. 4457. Mgard, F., 1984. The Andean orogenic period and its major structures in central and northern Peru. J. Geol. Soc. Lond. 141, 893900. Mgard, F., Noble, D.C., McKee, E.H., 1984. Multiple pulses of Neogene compressive deformation in the Ayacucho intermontane basin, Andes of Central Peru. Geol. Soc. Am. Bull. 95, 11081117. Montes, C., Robert, D., Hatcher, R.D., Restrepo-Pace, P.A., 2005. Tectonic reconstruction of the Northern Andean blocks: oblique convergence and rotations derived from the kinematics of the PiedrasGirardot area, Colombia. Tectonophysics 399, 221250. Moore, M.A., England, P.C., 2001. On the inference of denudation rates from cooling ages of minerals. Earth Planet. Sci. Lett. 158, 265284. Ollier, C.D., Pain, C.F., 2000. The Origin of Mountains. Routledge, London. 368 pp. Ordez, O., Pimentel, M., 2001. Consideraciones geocronolgicas e isotpicas del Batolito Antioqueo. Rev. Acad. Colomb. Cienc. Exactas Fs. Nat. 25, 2735. Page, W.D., James, M.E., 1981. The antiquity of the erosion surfaces and late Cenozoic deposits near Medelln, Colombia: Implications to tectonics and erosion rates. Rev. Ciaf 6, 421454. Pardo-Casas, F., Molnar, P., 1987. Relative motion of the Nazca (Farallon) and South American plates since Late Cretaceous time. Tectonics 6, 233248. Persano, C., Stuart, F.M., Bishop, P., Barfod, D.N., 2002. Apatite (U-Th)/He age constraints on the development of the Great Escarpment of the southeastern Australia passive margin. Earth Planet. Sci. Lett. 200, 7990. Peulvast, J.-P., Sales, V.C., 2005. Surfaces d'aplanissement et godynamique. Gomorphologie 4, 249273. Pilger, R.H., 1984. Cenozoic plate kinematics, subduction and magmatism: South American Andes. J. Geol. Soc. Lond. 141, 793802. Reiners, P.W., Brandon, M.T., 2006. Using thermochronology to understand orogenic erosion. Annu. Rev. Earth Planet. Sci. 34, 419466. Reiners, P.W., Zhou, Z., Ehlers, T.A., C, X., Brandon, M.T., Donelick, R.A., Nicolescu, S., 2003. Post-orogenic evolution of the Dabie Shan, Eastern China, from (UTh)/He and ssion-track thermochronology. Am. J. Sci. 303, 489518. Restrepo, J.J., 1975. Edades radiomtricas de algunas rocas de Antioquia, Colombia. Publicacin Especial Universidad Nacional, Medelln. 25 pp. Restrepo, J.D., Syvitski, J.P.M., 2006. Assessing the effect of natural controls and land use change on sediment yield in a major Andean River: the Magdalena Drainage Basin, Colombia. Ambio 35, 6574. Restrepo, J.J., Toussaint, J.F., 1982. Metamorsmos superpuestos en la Cordillera Central de Colombia., Memorias V Congreso Latinoamericano de Geologa, Buenos Aires, Argentina, 505512 pp.