Late Quaternary Tectonics, Incision, and Landscape Evolution of the Calchaquí River Catchment, Eastern Cordillera, NW Argentina

Transcription

1 Late Quaternary Tectonics, Incision, and Landscape Evolution of the Calchaquí River Catchment, Eastern Cordillera, NW Argentina by James Anderson McCarthy A thesis submitted in conformity with the requirements for the degree of Master of Applied Science Graduate Department of Earth Sciences University of Toronto Copyright by James Anderson McCarthy 2014

2 ii Late Quaternary Tectonics, Incision, and Landscape Evolution of the Calchaquí River Catchment, Eastern Cordillera, NW Argentina James Anderson McCarthy Master of Applied Science Graduate Department of Earth Sciences University of Toronto 2014 Abstract In this study we use field investigations, analysis of longitudinal river profiles, and 10 Be-derived erosion rates and paleo-erosion rates to examine the Quaternary landscape evolution of the Calchaquí River Catchment of the southernmost Eastern Cordillera, in NW Argentina. The spatial distribution of erosion rates, normalized steepness indices, concavity indices, and knickpoints reflect active tectonics and resistant lithologies exposed along preexisting structural heterogeneities. Shortening is distributed across multiple structures and controls local baselevels. Field studies document active faults and ~100m of channel incision in <300 kyr. Catchment mean erosion rates and paleo-erosion rates are not markedly different, suggesting that Quaternary climate changes have not significantly influenced erosion rates at cosmogenic nuclide time scales. Collectively, our results demonstrate that the rate and style of landscape evolution in the southern Eastern Cordillera is primarily controlled by Quaternary tectonics and pre-orogenic structure, thus complicating regional investigations of tectonic and climatic feedbacks. ii

3 iii Acknowledgments I would first like to thank my supervisor, Dr. Lindsay M. Schoenbohm, for giving me opportunities to learn from her, to do field work in NW Argentina, to do research in the exciting and dynamic field of Tectonic Geomorphology, to work with her other amazing students (Mark, Neil, and Renjie), and to get to know her family. I especially appreciate Dr. Schoenbohm s ability to trust in me and to let me take this project in my own direction. I would also like to thank Dr. Paul Bierman and Dr. Dylan Rood, of the University of Vermont and the University of Glasgow, respectively, for their help with the attached manuscript. Thank you to my supervisory committee members, Dr. Pierre Robin and Dr. Nick Eyles, for their comments and support throughout my degree program. I would also like to thank Dr. Bodo Bookhagen (UCSB) and Dr. John Gosse (Dalhousie) for useful comments. Thanks to Santiago Uriburu Quintana for assistance in the field. Financial support for this work was provided by Dr. Schoenbohm through NSF and NSERC grants. I thank the Geological Society of America for awarding me a Graduate Student Research Grant (specifically the John Montagne Quaternary Geomorphology Award), which enabled additional analyses that are critical to my conclusions. Conference support was provided by the University of Toronto Mississauga Department of Chemical & Physical Sciences, as well as the University of Toronto School of Graduate Studies. Personal support was provided through a Connaught International Student Scholarship and by the Department of Earth Sciences. Lastly, thanks to all my friends in the Department of Earth Sciences for pulling me through. iii

4 iv Table of Contents Acknowledgments... iii Table of Contents... iv List of Tables... vii List of Figures... viii List of Appendices... xii Chapter 1: Introduction Evolution of the Eastern Puna Plateau & Retroarc Foreland Central Andean Plateau Southern Eastern Cordillera Tectonic-Climatic Interactions Longitudinal River Profile Analysis Background Theory Interpretation Terrestrial Cosmogenic Nuclide (TCN) Chronology Theory Application & Limitations Dating Stable Landforms Measuring Catchment Mean Erosion Rates Chapter 2: Late Quaternary Tectonics, Incision, and Landscape Evolution of the Calchaquí River Catchment, Eastern Cordillera, NW Argentina Abstract Introduction...22 iv

5 v 2.3 Landscape Analysis As A Tool For Evaluating Tectonics and Climate in Spatially Heterogeneous Regions Geologic Setting Structural Evolution Quaternary Climate & Geomorphology Methods Field Studies Longitudinal River Profile Analysis Terrestrial Cosmogenic Nuclide ( 10 Be) Chronology Paleo-Erosion Rates Results Field Studies River Profile Analysis Be Catchment Mean Erosion Rates Be Depth Profiles Paleo-Erosion Rates Discussion Controls on River Morphology Controls on Landscape Evolution of the Pucará Valley Tectonic Implications Conclusions Chapter 3: Concluding Remarks Methodological Considerations River Profile Analysis & Catchment-Mean Erosion Rates TCN Depth Profiles...55 v

6 vi 3.2 Future Research...56 References...58 Appendices...67 Appendix A: Supporting Information for Field Studies...67 Appendix B: Geologic Map of the Pucará Valley...76 Appendix C: 10 Be Analytical Results...77 Appendix D: Supporting Information for 10 Be Depth Profiles...78 Appendix E: Supporting Information for River Profile Analysis...86 Appendix F: Explanation of Attached Digital Material...91 vi

7 vii List of Tables Table Be Concentrations, Catchment-mean production rates, Catchment mean erosion rates, and corresponding topographic and climatic characteristics Table 2. Vertical incision rates and catchment mean paleo-erosion rates derived from 10 Be depth profile ages and inheritance, respectively. See section for methodology vii

8 viii List of Figures Figure 1. Characteristics of the Altiplano Plateau and Puna Plateau. Top panel shows (in km) topography, crustal thickness, mantle lithosphere thickness, and depth of the subducting Nazca Plate, along a coastline-parallel section (bottom panel). In the bottom panel, thick black line denotes political boundaries, dark grey denotes areas above 3km, light gray zones are thinskinned fold-and-thrust belts and cross-hatched zones are thick-skinned foreland provinces. Eastwest (vertical) rule shows area of high seismic-wave attenuation. From Allmendinger et al. (1997)... 2 Figure 2. Shaded relief maps of the Central Andes, showing (a) the names and extent of morphotectonic provinces (SBS = Santa Bárbara System), and (b) mean annual precipitation (m yr -1 ). Dashed vertical black bars along the y-axes represent the latitudinal zone in which the subducting plate transitions north to south from steep slab to shallow slab geometry (Cahill and Isacks, 1992). Solid lines represent the latitudinal zone of flat-slab subduction (Barazangi and Isacks, 1976). Red boxes indicate the approximate location of the study area. Base figure from Strecker at al. (2007)... 4 Figure 3. Structural and topographic comparison of A) the thin-skinned retroarc fold-and-thrust belt of Bolivia and B) the thick-skinned retroarc foreland of NW Argentina, which is the focus of this study. From Strecker et al. (2011)... 6 Figure 4. Schematic model of an aridity-driven tectonic-climatic feedback system, proposed for the eastern margin of the Puna-Altiplano Plateau and the northern border of the Tibetan Plateau. In Stage 1 (top), continued crustal shortening drives uplift of the frontal range, creating an orographic barrier to precipitation. In Stage 2 (middle), channels in Basin B are defeated due to aridity-driven reductions in stream power, and cessation of sediment export leads to crustal loading and forelandward propagation of deformation. In Stage 3 (bottom), Basins A and B coalesce and an orographic barrier to precipitation begins to isolate Basin C. Figure from Sobel et al. (2003)... 9 Figure 5. Oblique view of the frontal Himalaya and the abrupt physiographic transition which corresponds with the Main Central Thrust (MCT). Longitudinal profile steepness indices (ksn, viii

9 ix see section for description) correspond with this physiographic transition, and have been used to argue for out-of-sequence deformation along the MCT. MFT, Main Frontal Thrust; MBT, Main Boundary Thrust. From Kirby and Whipple (2012) Figure 6. Longitudinal river profiles and slope-area scaling (inset) under A) differing concavity indices (θ) and B) differing normalized steepness indices (ksn). From Kirby and Whipple (2012) Figure 7. Vertical-step and slope-break knickpoint morphology in (a & c) profile form and (b & d) slope-area space. From Kirby and Whipple (2012) Figure 8. Idealized attenuation curves for production of 10Be in A) rock ( ρ = 2.7 g cm-3) and B) soil (ρ = 1.2 g cm-3). Figure from Bierman and Nichols (2004) Figure 9. At isotopic steady-state, where the production of a given TCN in a catchment is equal to the export of that TCN, the mass flux (dm/dt) can be calculated with the measured TCN concentration (C, in atoms g-1) of river sediments at the catchment outlet. Catchment mean erosion rate (mm ky-1) is determined by dividing the mass flux by the catchment area and bedrock density. Erosion rates are spatially non-uniform across the catchment, but sediment is well mixed in the channel network such that a sample represents a catchment-integrated TCN concentration. Figure from von Blanckenburg (2005) Figure 10. Composite digital elevation model and shaded relief map of the south central Andes with major tectonomorphic provinces outlined in black. Thicker black line delineates internally drained Puna Plateau from the externally drained Eastern Cordillera and Sierras Pampeanas. Yellow line outlines the Calchaqui River catchment (CRC). Red box outlines the Pucará Valley, where field studies were focused. SBS = Santa Barbara System. CG = Cerro Galán Caldera Figure 11. Quaternary strath terraces and pediment surfaces in the Pucará Valley. Depositional ages derived from cosmogenic 10Be depth profiles. Numbered soil pits are described in TABLE SOILS. JVT = Jasimaná-Vallecito Thrust. SQT = Sierra de Quilmes Thrust. PT = Pucará Thrust. See Auxiliary Material for complete geologic map. Fault nomenclature and structure modified from (Carrera and Munoz, 2008). Area shown by red box in Figure Regional ix

10 x Figure 12. Photograph from the site of the Q5 depth profile in Figure 11, looking approximately northeast. Foreground shows Q5 strath terrace beveled into sedimentary rocks of the Tertiary Payogastilla Group, which rest in angular unconformity over Cretaceous Pirgua Group redbeds. In the background Q2, Q3, Q4, Q6 and Q7 surfaces are beveled into both Tertiary and Cretaceous sedimentary units. Rio Pucará flows from right (south) to left (north). Note monoclinal structure within Cretaceous units beneath the Q7 surface. High ranges are composed of the Neoproterozoic Puncoviscana Formation Figure 13 (opposite). (a) Shaded relief map of the CRC, 10Be Catchment mean erosion rate samples and corresponding subcatchments (labeled) from this study and Bookhagen and Strecker (2012). Stream network derived from ASTER 30m DEM and a minimum accumulation of 35,000 pixels (1.05 km2). (b) Lithologic divisions, major faults, and knickpoints in the CRC. Knickpoints according to legend in (d). Dashed lines are newly mapped faults. CNT = Cerro Negro Thrust (Carrapa et al., 2011). (c) 10Be catchment mean erosion rates, in mm kyr-1. Sample locations as per legend in (a). (d) Normalized channel steepness indices and knickpoints in the CRC. See text for description of knickpoint typology and channel regression parameters. Labeled streams are displayed in profile in Figure Streams Figure 14. Correlations between catchment mean erosion rates and catchment mean annual precipitation, catchment area, catchment mean slope, and catchment mean elevation. See Table 1 for data Figure 15. In situ 10Be depth profiles and monte carlo simulator results for age, inheritance, and surface erosion rates when run for 100,000 solutions at 1 sigma uncertainty, according to parameters described in the text and appendices. Black line is the best fit. Gray lines are 100,000 model solutions. Solid black dots are subsurface samples used in the model simulations. Hollow dots are surface sediment samples that were analyzed, but not used in model simulations due to evidence of bioturbation. Hollow square represents a quartz cobble amalgamation (n=85) sample that was simularly excluded from model simulations Figure 16. Selected longitudinal river profiles and corresponding local slope/drainage area regressions. Individual segments are bound by knickpoints or confluences with trunk streams and were regressed with a reference concavity of Resulting normalized steepness indices and x

11 xi raw concavity indices are displayed for each segment. Question marks identify faults with unknown dip. In slope-area space light and dark blue lines represent forced and unforced regressions, respectively. See Figure 13 for stream locations. CNT = Cerro Negro Thrust; PT = Pucara Thrust; JVT = Jasimana-Vallecito Thrust Figure 17. Concavity indices and mean annual rainfall in the CRC. See Figure 13 for knickpoint classification. TRMM precipitation data from Bookhagen and Strecker (2008). Labeled streams are displayed in profile in Figure Streams Figure 18. Vertical distribution of knickpoints in the CRC. See Figures 13 and 17 for plan view xi

12 xii List of Appendices Appendix A: Supporting Information for Field Studies Appendix B: Geologic Map of the Pucará Valley Appendix C: 10 Be Analytical Results Appendix D: Supporting Information for 10 Be Depth Profiles Appendix E: Supporting Information for River Profile Analysis Appendix F: Explanation of Attached Digital Material xii

13 1 Chapter 1: Introduction This study investigates the Quaternary landscape evolution of a large (12,860 km 2 ) catchment in the Eastern Cordillera of NW Argentina, immediately east of the Central Andean Puna Plateau. Due to the position of the study area with respect to large scale (100s of km) tectonic, climatic, and surface process transitions, investigation of the style and rates of both tectonic deformation and erosion speak to various outstanding geologic questions, including the kinematics of Andean deformation and plateau growth, the role of geodynamic processes in the topographic evolution of the Puna Plateau, and the dynamic interactions between climate and tectonics. At smaller scales (e.g. 10s of km), my results evaluate the capacity of arid landscapes to achieve erosional steady state, as well as the relative importance of lithology, climate, and relief in controlling local erosion rates. Many of these questions are addressed in a manuscript set for submission to the Journal of Geophysical Research Earth Surface, which comprises Chapter 2 of this thesis. Chapter 1 provides important background for that work and for additional conclusions in Chapter 3. Some of the content in Chapter 1 is reiterated in Chapter 2, although I limited repetition as much as possible while retaining clarity. 1.1 Evolution of the Eastern Puna Plateau & Retroarc Foreland Central Andean Plateau The Central Andean Puna-Altiplano Plateau is the highest and largest plateau in the world to in a Cordilleran arc setting, and investigations into the temporal and spatial evolution of this orogen have improved understanding of tectonic, geodynamic and climatic controls on topography at multiple scales (Allmendinger et al., 1997). At the plate-tectonic scale, the geometry of the subducting Nazca Plate exerts a first-order control on plateau extent. The Nazca plate dip angle shallows to ~10 at either end of the plateau (15 S & 28 S), whereas the plate dips ~30 beneath the plateau (Fig. 1) (Cahill and Isacks, 1992). These zones of flat subduction are spatially coincident with young, buoyant oceanic plateaus and the absence of arc volcanism, suggesting that the age and thickness of subducting oceanic crust provide strong controls on orogen morphology (Barazangi and Isacks, 1976; Gutscher et al., 2000). 1

14 2 Figure 1. Characteristics of the Altiplano Plateau and Puna Plateau. Top panel shows (in km) topography, crustal thickness, mantle lithosphere thickness, and depth of the subducting Nazca Plate, along a coastline-parallel section (bottom panel). In the bottom panel, thick black line denotes political boundaries, dark grey denotes areas above 3km, light gray zones are thin-skinned fold-and-thrust belts and cross-hatched zones are thick-skinned foreland provinces. East-west (vertical) rule shows area of high seismic-wave attenuation. From Allmendinger et al. (1997). 2

15 3 Within the plateau, differences in topography, magmatism, lithospheric structure and temporal evolution define two distinct regions; the Altiplano to the North and the Puna to the south (Figs. 1 and 2) (Allmendinger et al., 1997). The Altiplano Plateau is characterized by ~3.2 km mean elevation, extremely low relief, crustal shortening >500 km, crustal thicknesses >60 km and total lithospheric thicknesses >80 km; in contrast, the Puna Plateau is characterized by ~4 km mean elevations, rugged topography with multiple, isolated depocenters, km crustal shortening, crustal thicknesses 55 km and reduced lithospheric thickness, especially in the central Puna (~60 km), where mantle lithosphere may be absent (McQuarrie, 2002; Tassara et al., 2006; Zhou et al., 2013). Additionally, the Altiplano and Puna plateaus exhibit significant differences in the timing of uplift and deformation. Paleoaltimetry and sedimentology suggest that uplift of the Altiplano began ~25 Ma, but was punctuated by brief, rapid periods of uplift (>1 km) between ~10 and 6 Ma in the northern Altiplano and between ~16 and 9 Ma in the southern Altiplano (Allmendinger et al., 1997; Garzione et al., 2008; Garzione et al., 2014). The uplift and deformation history of the Puna is less thoroughly resolved. Volcanic glass paleoaltimetry, clumped isotope thermometry and detrital thermochronology of sedimentary rocks in the retroarc foreland (southern Eastern Cordillera, Fig. 2) suggest that the Puna reached near-modern elevations between 36 Ma and at least ~10 Ma, and deformation had reached the Eastern Cordillera by ~14 Ma (Carrapa et al., 2012; Canavan et al., 2014; Carrapa et al., 2014a) In the Puna Plateau, shortening is insufficient to explain the crustal thicknesses observed, suggesting that magmatic additions and/or geodynamic processes play important roles in the evolution of the orogen (McQuarrie, 2002). For example, lower crustal flow from the Altiplano and underplating of material from subduction erosion of the forearc have been proposed to explain the discrepancy between crustal shortening and crustal thickness in the Puna (Kley and Monaldi, 1998). A well-documented Late Miocene kinematic shift from contraction to extension suggests that gravitational spreading processes, likely driven by reduced Nazca-South America plate convergence rate, control the rate and style of deformation on the Puna Plateau (Allmendinger et al., 1997; Schoenbohm and Strecker, 2009; Lanza et al., 2013). Additionally, the combination of high mean elevations, volcanic rock geochemistry and partial or total absence of mantle lithosphere beneath the central Puna Plateau suggests that small-scale lithospheric 3

16 4 foundering events also play an important role in the evolution of the Puna Plateau (Schoenbohm and Strecker, 2009; Lanza et al., 2013; Zhou et al., 2013). Recent interpretations of seismic tomography indicate a descending and detached high-velocity block, supporting this conclusion (Bianchi et al., 2013). The degree to which gravitational spreading, extension, and lithospheric foundering of the Central Puna Plateau influence deformation within the adjacent retroarc foreland, which includes the southern Eastern Cordillera and northern Sierras Pampeanas (Fig. 2), is unresolved. Regional kinematic analyses indicate Pliocene shifts from subvertical to subhorizontal extension in many localities throughout the Eastern Cordillera and Sierras Pampeanas (Marrett et al., 1994; Schoenbohm and Strecker, 2009; Pearson et al., 2012; Lanza et al., 2013). However, Pliocene - Quaternary shortening has also been documented in the region (Strecker et al., 1989; Hilley and Strecker, 2005; Carrera and Munoz, 2008; Hain et al., 2011; Santimano and Riller, 2012). Figure 2. Shaded relief maps of the Central Andes, showing (a) the names and extent of morphotectonic provinces (SBS = Santa Bárbara System), and (b) mean annual precipitation (m yr -1 ). Dashed vertical black bars along the y- axes represent the latitudinal zone in which the subducting plate transitions north to south from steep slab to 4

17 5 shallow slab geometry (Cahill and Isacks, 1992). Solid lines represent the latitudinal zone of flat-slab subduction (Barazangi and Isacks, 1976). Red boxes indicate the approximate location of the study area. Base figure from Strecker at al. (2007) Southern Eastern Cordillera The southern Eastern Cordillera is a bi-vergent fold and thrust belt, characterized by basementinvolved reverse faults that preferentially occur along preexisting structural heterogeneities, including inverted Cretaceous rift structures and earlier metamorphic fabrics (Grier et al., 1991; Strecker et al., 2007; Carrera and Munoz, 2008; Santimano and Riller, 2012). Uplifted basement blocks are composed of Precambrian metasedimentary units, Paleozoic granitoids, and redbeds from the Cretaceous Salta Rift (Grier et al., 1991; Coutand et al., 2006). Cenozoic sedimentary rocks in intramontane basins within the region, which are derived from the Precambrian to Cretaceous units, reflect a spatiotemporal transition from proximal foredeep to wedge-top to intramontane basin as well as the transition to increasingly arid climate due to the uplift of orographic barriers to precipitation (Bywater-Reyes et al., 2010; Carrapa et al., 2012). U-Pb zircon ages and sedimentary provenance studies indicate that Andean shortening reached the western edge of the study area in the Eocene, and roughly propagated from west to east until the Pliocene, at which point deformation was primarily accommodated by the Santa Barbara System to the east (Carrapa et al., 2012 and references therein). However, the inherited structural heterogeneity of the southern Eastern Cordillera, northern Sierras Pampeanas, and Santa Barbara System has led to significant differences in the timing of deformation over distances as small as 50km, therefore presenting a fundamental difference with the more uniformly deforming thinskinned Bolivian retroarc foreland basin system (Fig. 3) (Hongn et al., 2007; DeCelles et al., 2011; Carrapa et al., 2012). A resulting debate exists as to whether the southern Cordillera Oriental region formed as part of a continuous or broken foreland, the answer to which has important implications for Andean fault kinematics and Andean paleoclimate (Strecker et al., 2011; Carrapa et al., 2012). 5

18 6 Figure 3. Structural and topographic comparison of A) the thin-skinned retroarc fold-and-thrust belt of Bolivia and B) the thick-skinned retroarc foreland of NW Argentina, which is the focus of this study. From Strecker et al. (2011) 6

19 7 For example, crustal heterogeneities in NW Argentina may have limited the crustal flexural response of the advancing orogenic wedge, such that creation of accommodation space and subsequent sedimentation are spatially limited relative to the wide foreland basin in Bolivia (Strecker et al., 2011). Shortening along reactivated high angle faults in NW Argentina creates relief that impedes penetration of moisture to wedge-top and intramontane basins, whereas moisture transport in the Bolivian system is more uniformly distributed across the wedge-top (Figure 2) (Strecker et al., 2007). The lack of internal moisture transport in NW Argentina, except along structurally controlled lows in the landscape, may explain the well-documented history of alternating sediment storage and excavation in intramontane basins in this region (Hilley and Strecker, 2005; González Bonorino and Abascal, 2012). Structural control on climate (through the uplift of orographic barriers to precipitation) may in turn influence the rate and style of deformation by decreasing erosional efficiency within intramontane basins (see section 1.2) (Sobel et al., 2003; Hilley et al., 2005; Hilley and Strecker, 2005; Bookhagen and Strecker, 2008). 1.2 Tectonic-Climatic Interactions The surface of Earth represents a dynamic balance between spatially and temporally variant forces that build (e.g. tectonics) and reduce (e.g. erosion) topography. Both tectonics and climate are capable of creating relief and modifying the landscape, and the effects of different tectonic or climatic regimes on sedimentary systems and topography can often be described through simple physical arguments and confirmed through observation of the sedimentary record (Miall, 2006). However, the dynamic nature of earth s sedimentary system necessitates an understanding of the interaction between these various controls, as negative and positive feedbacks in complex geologic environments may serve to diminish or amplify sedimentary processes (e.g., Sobel et al., 2003). Tectonic deformation influences global, regional, and local climate over various time scales. Across >10 Myr timescales, migration of tectonic plates controls the distribution of continental lithosphere across the Earth. Concentration of continental lithosphere in large landmasses at the poles, such as with Gondwana during the late Paleozoic, limits solar radiation to landmasses and thus promotes the formation of ice sheets and glacial climates (Caputo and Crowell, 1985). The 7

20 8 motion of continental landmasses also controls atmospheric and oceanic circulation patterns. The development of the Antarctic circumpolar current and the subsequent ice-growth on Antarctica are attributed to the drift of South America and Australia in the Cenozoic (Barker and Burrell, 1977; Kennett, 1977; Raymo and Ruddiman, 1992). On timescales relevant to this study, tectonics influences local climate by creating orographic barriers to precipitation (Sobel et al., 2003). Orographic barriers cause aridity leeward of high ranges, for example north of the Himalaya and west of the central Andes (Bookhagen and Burbank, 2006; Bookhagen and Strecker, 2008). The role of climate in driving tectonics is less thoroughly established. The capacity for climate to influence the rates, magnitude, and styles of tectonic deformation is primarily suggested due to spatial coincidence between intense precipitation and high crustal exhumation rates in the Himalaya and in Taiwan (Beaumont et al., 2001). However, high uplift rates may be spatially coincident with high precipitation rates and high erosion rates, but such coincidence may simply reflect the creation of an orographic barrier through tectonic shortening (thus enhancing local rainfall), while uplift occurs independently of surface erosion rates. Numerical modeling indicates that climate drives tectonics by focused erosion that rapidly removes mass from an advancing orogenic wedge (i.e. fold-and-thrust belt), lowering lithostatic pressure above faults and thus enhancing rates of deformation within the orogenic wedge (Davis et al., 1983; Whipple, 2009). This erosion-driven deformation may also be achieved through ductile flow of the middle crust (in contrast to brittle deformation by faulting), a condition that is proposed for the Himalaya (Beaumont et al., 2001). Conversely, the lack of both erosion and sediment export due to extreme aridity may affect tectonics; in active orogens, it is thought that the development of internal drainage due to aridity can topographically load the lithosphere, increasing frictional forces on faults within the orogenic wedge and thus drive deformation to the foreland (Sobel et al., 2003). Furthermore, positive feedbacks between sediment trapping, basin elevation and aridification allow for significant aggradation in basins that remain hyrdologically connected, further increasing topographic load within the orogenic wedge (Hain et al., 2011). Modeling of aridity-driven channel defeat, sediment trapping, and wedge-propagation suggests that these feedbacks contribute to the large widths of the Puna Plateau and Tibetan Plateau (Fig. 4) (Sobel et al., 2003; Hilley and Strecker, 2005). Attempts to explain tectonic (and associated 8

21 9 topographic) evolution of complex landscapes through climate change, beyond simple isostatic adjustment, are made more compelling if pronounced climate changes can be matched to sustained (>1 Myr) changes in uplift rate or pattern (Burbank et al., 2003; Whipple, 2009). Figure 4. Schematic model of an aridity-driven tectonic-climatic feedback system, proposed for the eastern margin of the Puna-Altiplano Plateau and the northern border of the Tibetan Plateau. In Stage 1 (top), continued crustal shortening drives uplift of the frontal range, creating an orographic barrier to precipitation. In Stage 2 (middle), channels in Basin B are defeated due to aridity-driven reductions in stream power, and cessation of sediment export leads to crustal loading and forelandward propagation of deformation. In Stage 3 (bottom), Basins A and B coalesce and an orographic barrier to precipitation begins to isolate Basin C. Figure from Sobel et al. (2003) 9

22 Longitudinal River Profile Analysis Background Given that the overall relief structure of a landscape is primarily set by the fluvial channel network, analysis of longitudinal river profile morphology is now the standard method for topographic interpretations of tectonics (Whipple and Tucker, 1999). Longitudinal river profiles are adjusted to balance erosion and uplift and have readily predictable forms under different climatic, lithologic, and transient conditions, making them particularly useful in tectonic geomorphology. For example, longitudinal river profile analysis has been used to demonstrate out-of-sequence deformation in the Himalaya (Fig. 5), perhaps influenced by climate (Seeber and Gornitz, 1983; Wobus et al., 2006c; Kirby and Whipple, 2012), and also pulsed uplift along the eastern margin of the Tibetan Plateau (Kirby et al., 2003; Schoenbohm et al., 2004). Longitudinal river profiles are more useful than hillslope gradients in tectonic interpretations due to higher erosion thresholds: beyond ~200 mm kyr -1, an erosion rate which is exceeded in many active orogens, hillslopes reach threshold gradients and fail, whereas longitudinal river profiles continue to steepen up to ~600 mm kyr -1 (Ouimet et al., 2009; Portenga and Bierman, 2011). Figure 5. Oblique view of the frontal Himalaya and the abrupt physiographic transition which corresponds with the Main Central Thrust (MCT). Longitudinal profile steepness indices (ksn, see section for description) 10

23 11 correspond with this physiographic transition, and have been used to argue for out-of-sequence deformation along the MCT. MFT, Main Frontal Thrust; MBT, Main Boundary Thrust. From Kirby and Whipple (2012) Theory Along a bedrock channel reach of uniform lithology, uplift rate, and climate, river profiles are expected to exhibit a power law relationship between slope and drainage area, such that S = k s A -θ (1) where S is the local channel gradient, k s is the channel steepness index, A is the contributing drainage area, and θ is the concavity index. At steady state, this relationship reflects a dynamic equilibrium between sediment supply and sediment export, such that uplift rate (U) and channel erosion (E) are equal. Channel erosion is typically modeled as a function of substrate erodibility (K) and bed shear stress, which is itself a function of contributing drainage area (A) and local gradient (S) and modified by m and n, positive constants which reflect differences in erosion process, channel geometry, and basin hydrology (Whipple and Tucker, 1999; Whipple, 2001): E = K A m S n (2) Solving for channel slope, and assuming that uplift and erosion are balanced (i.e. assuming steady-state) yields the following relationship, which is similar in form to Equation 1: S = (U/K) 1/n A -m/n (3) Comparing equations 1 and 3 reveals that steepness indices should vary due to differences in uplift rate, lithology, and climate, while concavity indices should be influenced by factors controlling the sediment transport ability of the profile such as drainage basin shape and downstream changes in channel width vs. discharge (Kirby and Whipple, 2012). Concavity indices typically fall into a narrow range ( ) when channels have uniform climate, uplift rate, and lithology along their length. In contrast, steepness indices vary widely with differing uplift rates, matching theoretical predictions of incision models (Whipple, 2004 and references therein). In order to accurately compare channels and channel segments of varying drainage area, steepness indices are normalized to a user-specified reference concavity (Figure XX) (Kirby and Whipple, 2012). 11

24 12 Figure 6. Longitudinal river profiles and slope-area scaling (inset) under A) differing concavity indices (θ) and B) differing normalized steepness indices (ksn). From Kirby and Whipple (2012). When along-channel changes in lithology, climate, or uplift rate occur, sharp breaks (knickpoints) in the profile separate segments with different steepness and concavity indices (Fig. 6). Similarly, if a temporal change in uplift rate or climate occurs, a transient knickpoint will develop at the basin outlet, and propagate upstream as an incisional wave, separating the newly equilibrated lower reaches from upper reaches equilibrated with previous conditions (Whipple and Tucker, 1999; Schoenbohm et al., 2004). Typically transient knickpoints exhibit slope-break morphology while spatially-bound knickpoints exhibit vertical-step morphology (Fig. 6), but distinguishing between spatially-bound or transient knickpoints is difficult based only on analysis of channel form. However, the distribution of knickpoints in the landscape can be diagnostic; a transient knickpoint will split at tributary junctions as it heads upstream, and its descendants will be found at similar elevations, while spatially-bound knickpoints will typically 12

25 13 occur along discrete shear zones or lithologic boundaries (Wobus et al., 2006a). The collective analysis of channel steepness and concavity indices, knickpoint distribution and form, and supporting landscape information (e.g. lithologic and climatic variability) provides the basis for using river profile morphology to interpret spatial and temporal changes in uplift rates (e.g. Schoenbohm et al., 2004; Harkins et al., 2007). Figure 7. Vertical-step and slope-break knickpoint morphology in (a & c) profile form and (b & d) slope-area space. From Kirby and Whipple (2012) Interpretation Tectonic interpretations of longitudinal river profile morphology require a priori knowledge of along-channel changes in lithology and climate, as changes in substrate erodibility (through climate or lithology) can be identical to changes in uplift rate (Cyr et al., 2014). For example, an increase in uplift rate (U) or a decrease in substrate erodibility (K, which can be achieved through drier climate or stronger lithology) will both result in the steepening of the channel profile (see equation 3). However, in complex landscapes, the covariance of normalized steepness indices (k sn ) and catchment mean erosion rates can distinguish lithologic and tectonic controls on profile form (Cyr et al., 2014). At steady state, channel gradients are adjusted to the 13

26 14 competing influence of uplift rates (U) and channel erodibility (K) such that the change in channel elevation over time is zero. It follows that steepness indices are high along channel segments with either high uplift rates or resistant lithologies. In the case of high uplift rate, the channel segment will erode at a high rate to keep pace with the uplifting block, while, in the case of resistant lithology, the channel segment will erode at a low rate. Therefore, high and low catchment mean erosion rates should indicate tectonic or lithologic control, respectively (Cyr et al., 2014). 1.4 Terrestrial Cosmogenic Nuclide (TCN) Chronology Theory Cosmogenic nuclides (e.g. 10 Be and 26 Al) are rare isotopes produced by cosmic ray bombardment of elements in the atmosphere and within rocks and soil in the near-surface environment. Cosmogenic nuclides produced in the atmosphere (known as meteoric ) are later deposited on the Earth s surface through precipitation and dustfall. Conversely, cosmogenic nuclides produced in rocks and soils are known as in situ or terrestrial cosmogenic nuclides (TCN). Due to increasing accuracy in estimates of cosmogenic nuclide production rates, measurement of cosmogenic nuclide activities in rocks and soils allows for the dating of stable surfaces or the calculation of erosion/deposition rates (Lal, 1991; Gosse and Stone, 2001). This study measures in situ production of 10 Be in sediments to date pediments and terraces, and constrain catchment-mean erosion rates. The production rate of cosmogenic nuclides in rock or sediments attenuates exponentially with increasing depth and density in rock or soil, thereby restricting the measurable production of cosmogenic nuclides to the near-surface environment (<5 m) (Fig. 8). The production rate (P x ) of a cosmogenic nuclide at a given depth (x) in soil or rock of known density (ρ) is moderated by a characteristic attenuation factor (Λ), so that: P x = P 0 e -xρ/ᴧ The production rate at the surface (P 0 ) varies as a function of latitude, altitude, and a shielding factor, which takes into account the geometry of both the surface and surrounding topography 14

27 15 (Lal, 1991; Bierman and Steig, 1996; Balco et al., 2008). The latitudinal and altitudinal controls on cosmogenic production rate reflect variations in geomagnetic field strength and atmospheric thickness, respectively. With a known surface production rate, the production rate at a given depth can thus be calculated. A production rate at depth (x), coupled with a measured cosmogenic nuclide activity at depth (x) (surface samples can be used), provides the basis for using TCN as chronometers. Figure 8. Idealized attenuation curves for production of 10Be in A) rock ( ρ = 2.7 g cm-3) and B) soil (ρ = 1.2 g cm-3). Figure from Bierman and Nichols (2004). 15

28 Application & Limitations Research in the previous two decades demonstrates both the utility and the expanding application of cosmogenic nuclides in determining rates of landscape evolution and in dating landforms. Considering only studies that use in situ produced 10 Be, applications include dating terraces and alluvial fans, improving glacial chronologies, determining rates of bedrock erosion, and constraining rates of soil formation (Bierman et al., 2002 and references therein). However, correct interpretation of nuclide activities depends on the application of an appropriate geomorphic model to a given study (Bierman and Nichols, 2004). Determining the validity of assumptions inherent to a geomorphic model requires data obtained through traditional geomorphic and sedimentologic analyses. Thus, cosmogenic nuclide analyses do not replace older techniques, but rather work in conjunction to place more precise constraints on the evolution of a landscape. The two most basic applications of TCN model continuous surface exposure and steady-state erosion, allowing for the calculation of either minimum exposure ages or steady-state erosion rates. In the case of a sudden and continuous exposure (e.g. by a landslide or glacial retreat), and assuming no erosion since initial exposure, a model exposure age (t) is determined by the measured cosmogenic radionuclide activity (N) and the decay constant (λ) (Lal, 1991; Bierman et al., 2002). N x = (P x /λ)(1 e -λt ) In the case of steady-state erosion, assuming small, high-frequency erosion events, the erosion rate (ε) is determined by the following equation (Lal, 1991): N x = e -xρ/ᴧ [(P x )/(λ + ρᴧ -1 ε)] The validity of modeled exposure ages and erosion rates depend on the satisfaction of assumptions inherent to each model. However, surfaces commonly have complex exposure and burial histories, and experience variable erosion rates through time (Anderson et al., 1996; Bierman and Steig, 1996). As a result, most TCN studies employ refined versions of these models, specific to the geomorphology and research objectives in a given study area. 16

29 17 Furthermore, the history of a landform is not always made apparent through geomorphic and sedimentologic observation, thereby encouraging the development of statistical analyses of multi-run models to determine the most likely exposure age and/or erosion rate at a sample site (e.g., Hidy et al., 2010) Dating Stable Landforms To date incised fluvial terraces and pediments in our study, we use the depth-profile technique demonstrated by Anderson et al. (1996), and refined by Hancock et al. (1999), in which at least 2 sand samples, one from the surface and the other(s) at known depth(s) (extending greater than 1 m), are analyzed. This technique, by virtue of the difference between post-depositional nuclide production rates at the surface and at depth (ΔP), allows for the quantification of the surface age (T): T = (1/λ)ln[ΔP/(ΔP λδn)] This technique also permits the quantification of the average cosmogenic nuclide inheritance, which is the nuclide concentration attained prior to deposition in the target surface. Quantitatively, inheritance (N in ) is the difference between the nuclide concentration of a surface or subsurface sample (N s or N ss ) and the product of the production rate (at the surface or at depth) and the depositional age of the landform: N in = N ss P ss T = N s P s T The inheritance thus represents the sum of nuclide production during a previous exposure-burial cycle (if applicable), during the exhumation of the clast from the source area, and during transport to the final depositional site (Anderson et al., 1996). Not accounting for inheritance can lead to the overestimation of landform ages. Using sand or an amalgamation of clasts (>30), rather than single clasts, is an important step as it accounts for the variance in exposure histories among clasts; due to the stochastic nature of particle trajectories in a sedimentary system, different clasts moving from a similar source area to a depositional site can experience vastly different exposure histories (Anderson et al., 1996). Despite the concerns regarding variable inheritance, the dating of alluvial fan and terrace surfaces with TCN depth-profiles are powerful tools, and have been used to estimate fault slip 17

30 18 rates, paleoclimatic evolution, and the relative importance of tectonics and climate in a given landscape s evolution (e.g., Hancock et al., 1999; Vassallo et al., 2005; Hetzel et al., 2006). Indeed, the inheritance itself is useful, as it can indicate the rates at which sediments are exhumed and transported (Anderson et al., 1996). Varying inheritance between cobbles and sands has been used to indicate different sources; the variable inheritance reflects differing transport rates due to landscape position (Schmidt et al., 2011) Measuring Catchment Mean Erosion Rates Of particular interest to tectonic geomorphology is the application of TCN in the determination of catchment mean erosion rates. The ability to measure erosion rates of individual catchments at year timescales has enabled a greater understanding of the spatiotemporal controls on erosion, and provided a bridge between modern erosion rate estimates (e.g. sediment yield measurements, dam records) and long-term (>10 5 years) estimates of exhumation (e.g. thermochronology) (Bierman and Steig, 1996; Granger et al., 1996). For example, along the eastern margin of the Puna Plateau in NW Argentina, where rainfall varies from >2 m yr -1 in the foreland to 0.1 m yr -1 in the orogen interior (Fig. 2), Bookhagen and Strecker (2012) measured a 10 fold difference in 10 Be-derived catchment mean erosion rates, suggesting that precipitation was the primary control on erosion rates in the region. In the Appalachian Great Smoky Mountains, Matmon et al. (2003) found that erosion rates derived from modern sediment yields, 10 Be catchment mean erosion rates, and Mesozoic fission track exhumation rates were similar, suggesting that erosion rates have been relatively constant in the Mesozoic and Cenozoic. In contrast, Paleozoic exhumation rates were an order of magnitude higher. These results suggest that the Appalachians have remained high due to deep crustal roots and isostatic adjustment to erosion, rather than more recent tectonics (Matmon et al., 2003). To measure catchment mean erosion rates, a sample of sediment from a modern river is analyzed. This approach assumes that, although certain areas of a given catchment may erode at different rates, sediments are well mixed in transport such that a sample contains volumes of a target mineral from each subcatchment proportional to their relative erosion rates. In other words, the catchment is in isotopic steady state, where the rate of isotope production (in the target mineral) is equal to the rate of isotope export from the catchment (Fig. 9) (Bierman and 18

31 19 Steig, 1996). This approach also assumes availability of the target mineral in similar quantities and grain sizes throughout the catchment. When the target mineral is non-uniformly distributed (e.g. in catchments of mixed lithology), corrections need to be made (e.g., Safran et al., 2005). In practice, 10 Be is most commonly used to determine catchment mean erosion rates due to its favorably long half-life (1.4 Myr), its production in an abundant and chemically resistant target mineral (quartz), and its high analytical precision (von Blanckenburg, 2005). Figure 9. At isotopic steady-state, where the production of a given TCN in a catchment is equal to the export of that TCN, the mass flux (dm/dt) can be calculated with the measured TCN concentration (C, in atoms g-1) of river sediments at the catchment outlet. Catchment mean erosion rate (mm ky-1) is determined by dividing the mass flux by the catchment area and bedrock density. Erosion rates are spatially non-uniform across the catchment, but sediment is well mixed in the channel network such that a sample represents a catchment-integrated TCN concentration. Figure from von Blanckenburg (2005). 10 Be concentrations in river sediments are converted to erosion rates by estimating the mean 10 Be production rate of the contributing drainage area. Given the non-linear dependence of 10 Be 19

32 20 production rates on elevation, mean production rates are often calculated by quantifying basin hypsometry or by computer codes which calculate the production rate of each pixel in a digital elevation model of the catchment; using mean elevation will result in significant inaccuracies in high relief environments (Bierman and Nichols, 2004; Bookhagen and Strecker, 2012). Catchment mean erosion rate (ε) is related to catchment mean production rate (P) and a measured TCN concentration in sediment (C) by the following equation: ε = (P/C λ)(λ/ρ) Catchment mean erosion rates integrate over a period dependent on the erosion rate itself, because the erosion rate sets the residence time for a parcel of rock in the landscape. Averaging timescales range from 10 2 to 10 5 year timescales from tectonically active orogens to stable cratons (von Blanckenburg, 2005). Comprehensive review of published 10 Be catchment mean erosion rates (n = 1149) reveals that, on a global scale, erosion rate correlates most strongly with basin slope (R 2 = 0.33), and exhibits no significant relationship with mean annual temperature, mean annual precipitation, basin area and basin latitude (Portenga and Bierman, 2011). Correlations improve when considering basins from similar climatic, tectonic or lithologic settings. The large scale meta-analyses by Portenga and Bierman (2011) indicate that landscape complexity (due to mixed lithology, climatic gradients, and non-steady state conditions) must be considered in interpretations of 10 Be catchment-mean erosion rates (Safran et al., 2005; Hilley and Coutand, 2010; Bookhagen and Strecker, 2012). 20

33 21 2 Chapter 2: Late Quaternary Tectonics, Incision, and Landscape Evolution of the Calchaquí River Catchment, Eastern Cordillera, NW Argentina James A. McCarthy, Department of Earth Sciences, University of Toronto, Toronto, Ontario, Canada Lindsay M. Schoenbohm, Chemical & Physical Sciences, University of Toronto Mississauga, Mississauga, Ontario, Canada AND Department of Earth Sciences, University of Toronto, Toronto, Ontario, Canada Paul R. Bierman, Department of Geology and Rubenstein School of the Environment and Natural Resources, University of Vermont, Burlington, VT, USA Dylan Rood, Scottish Universities Environmental Research Centre, University of Glasgow, East Kilbride, Scotland, UK NOTE ON JOURNAL: JGR-Earth Surface focuses on the physical, chemical, and biological processes that affect the form and function of the surface of the solid Earth over all temporal and spatial scales, including fluvial, eolian, and coastal sediment transport; hillslope mass movements; glacial and periglacial activity; weathering and pedogenesis; and surface manifestations of volcanism and tectonics (from the website). 21

34 Abstract In this study we use field investigations, systematic analysis of longitudinal river profiles, 10 Bederived catchment mean erosion rates, and paleo-erosion rates and vertical incision rates from 10 Be depth profiles to examine the late Quaternary landscape evolution of the Calchaquí River Catchment (CRC) of the Eastern Cordillera, NW Argentina. We find that the spatial distribution of erosion rates, normalized steepness indices, and concavity indices reflect active tectonics and the exposure of resistant lithologies along preexisting structural heterogeneities in the study region. Abundant knickpoints are spatially coincident with tectonic and/or lithologic discontinuities, indicating local base-level control by thrust faulting that is distributed across multiple structures. Field studies document active faults, corroborating our interpretations of river profiles. Field studies also document the progressive abandonment of pediment and strath terraces, resulting in ~100 m of channel incision in <300 kyr. Catchment mean erosion rates and paleo-erosion rates are similar, suggesting Quaternary climate changes have not influenced erosion rates at ~10 ka time scales. Collectively, our data demonstrate that the rate and style of landscape evolution in the southern Eastern Cordillera is primarily driven by Quaternary tectonic deformation and inherited structural heterogeneity, complicating interpretations of tectonicclimatic feedbacks on the eastern margin of the Puna Plateau. We speculate that out-of-sequence shortening in the Calchaquí River Catchment, and perhaps localized extension on the plateau margin, reflect gravitational spreading processes on the Puna Plateau. 2.2 Introduction The increasing availability of high-resolution topographic data and improvements in our ability to measure basin-scale erosion permit the establishment of empirical relationships between topographic metrics and erosion rates in modern landscapes (Portenga and Bierman, 2011; Bookhagen and Strecker, 2012; Kirby and Whipple, 2012). These advances in our understanding of surface processes and surface process thresholds elucidate tectonic and climatic controls on erosion, setting the stage for field-based investigations of the coupling between climate and tectonics (Wobus et al., 2006a; Ouimet et al., 2009; Whipple, 2009). However, establishing a causal link between climatically-moderated erosion and tectonics is complicated by the preexisting geologic complexity commonly observed in natural landscapes, such that systematic 22

35 23 analyses must be carried out to isolate lithologic, climatic, and tectonic controls on topography and erosion (e.g., Hilley and Coutand, 2010). Furthermore, basin-scale erosion rates are typically integrated over millennia, so the degree to which modern climate data reflect measured erosion rates are dependent on the frequency and magnitude of past climate changes (Bierman and Steig, 1996). With these challenges in mind, we investigate the Late Quaternary landscape evolution of the southernmost Eastern Cordillera, a tectonomorphic province of the Central Andes (Figure Regional). Bordering on the Puna Plateau to the west, the Sierras Pampeanas to the south, and the Santa Bárbara System to the east, the study area lies within the west-east transition from high plateau to complex retroarc foreland (Allmendinger et al., 1997). Pronounced climatic gradients exist as well along the eastern margin of the central Andes due to orographic shielding of easterly moisture-bearing winds, resulting in stark differences in surface processes and erosional efficiency between the plateau and the foreland (Strecker et al., 2007; Bookhagen and Strecker, 2012). As a result, the southern Eastern Cordillera and nearby regions provide an ideal natural laboratory in which to examine the control exerted by preexisting structural fabrics, variable lithology, and geodynamic processes on topography, as well as the potential interactions between climatically-moderated surface processes and tectonics. We employ field investigations, longitudinal river profile analysis, 10 Be catchment-mean erosion rates, and estimates of paleoerosion rates derived from 10 Be depth-profiles to examine the various controls on erosion and topography in the Calchaquí River Catchment (CRC). We focus our field studies in the lower Pucará Valley, an intramontane basin within the CRC (Figure Regional). 23

36 24 Figure 10. Composite digital elevation model and shaded relief map of the south central Andes with major tectonomorphic provinces outlined in black. Thicker black line delineates internally drained Puna Plateau from the externally drained Eastern Cordillera and Sierras Pampeanas. Yellow line outlines the Calchaqui River catchment (CRC). Red box outlines the Pucará Valley, where field studies were focused. SBS = Santa Barbara System. CG = Cerro Galán Caldera. 2.3 Landscape Analysis As A Tool For Evaluating Tectonics and Climate in Spatially Heterogeneous Regions Fluvial channel network morphology and catchment mean denudatuion rate are sensitive indicators of both tectonic and climatic forcing (Whipple and Tucker, 1999). For example, in a region of uniform lithology and climate, the steepness of bedrock river channels should vary due to differences in uplift rate, while channel concavity should be influenced by factors controlling the sediment transport ability of the profile such as drainage basin shape and downstream 24

37 25 changes in channel width vs. discharge (Kirby and Whipple, 2012). When along-channel changes in lithology, climate, or uplift rate occur, sharp breaks (knickpoints) in the profile separate segments with different steepness and concavity. Similarly, if a temporal change in uplift rate or climate occurs, a transient knickpoint will develop at the basin outlet, and propagate upstream as an incisional wave, separating the newly equilibrated lower reaches from upper reaches equilibrated with previous conditions (Whipple and Tucker, 1999; Schoenbohm et al., 2004). In isolation, longitudinal river profile analysis cannot explicitly distinguish the relative effects of tectonics, lithology, climate and transient perturbations on profile form, but the incorporation of supporting information (e.g. lithologic and climatic data) can provide the keys to diagnosing these influences in complex landscapes (Kirby and Whipple, 2012). In particular, the covariance of normalized steepness indices and 10 Be catchment mean erosion rates can distinguish lithologic and tectonic controls on channel steepness (Cyr et al., 2014). High channel steepness and high erosion rates along a discrete channel segment indicate higher uplift rate whereas high channel steepness and (comparatively) low erosion rates indicate locally resistant substrate (Cyr et al., 2014). The covariance of 10 Be catchment mean erosion rates and channel steepness, together with a priori knowledge of the distribution of lithology and precipitation, allow us to evaluate the dominant controls on landscape evolution in the CRC. 2.4 Geologic Setting Structural Evolution The southern Eastern Cordillera is a bi-vergent fold and thrust belt, characterized by basementinvolved reverse faults that preferentially occur along preexisting structural heterogeneities, including inverted Cretaceous rift structures and earlier metamorphic fabrics (Grier et al., 1991; Strecker et al., 2007; Carrera and Munoz, 2008; Santimano and Riller, 2012). Basement uplifts are composed of Precambrian metasedimentary units, Paleozoic granitoids, and sedimentary rocks related to the Cretaceous Salta Rift (Grier et al., 1991; Coutand et al., 2006). Deposition of Cenozoic sedimentary rocks in intramontane basins within the CRC reflects eastward propagation of the orogenic front from late Eocene to Pliocene (Coutand et al., 2006; Carrapa et al., 2012). Pliocene to Quaternary deformation was primarily accommodated by the Santa 25

38 26 Barbara System to the east (Hilley and Strecker, 2005; Coutand et al., 2006; González Bonorino and Abascal, 2012). Aridity similarly propagated eastward due to uplift of orographic barriers to precipitation (Coutand et al., 2006; Bywater-Reyes et al., 2010; Carrapa et al., 2012). The Pucará Valley, like other intramontane basins in the CRC, is defined by N-S trending contractional structures (Fig. 11). On the west, the Jasimaná Vallecito Thrust, an inverted Cretaceous normal fault, carries Cretaceous sedimentary rocks over Holocene sediments (Coutand et al., 2006). On the east, the Sierra de Quilmes Thrust carries Precambrian basement over Cretaceous rift strata (Carrera and Munoz, 2008). Cenozoic strata of the Pucará Valley record the evolution from a distal to proximal foredeep from Late Eocene to Middle Miocene (Carrapa et al., 2012). Eastward propagation of deformation led to the development of a wedgetop basin from approximately Ma, and further shortening of the wedge-top after 10 Ma led to the development of the modern intramontane physiography (Coutand et al., 2006; Carrera and Munoz, 2008; Carrapa et al., 2012). 26

39 27 Figure 11. Quaternary strath terraces and pediment surfaces in the Pucará Valley. Depositional ages derived from cosmogenic 10Be depth profiles. Numbered soil pits are described in TABLE SOILS. JVT = Jasimaná-Vallecito Thrust. SQT = Sierra de Quilmes Thrust. PT = Pucará Thrust. See Auxiliary Material for complete geologic map. Fault nomenclature and structure modified from (Carrera and Munoz, 2008). Area shown by red box in Figure Regional. 27

40 Quaternary Climate & Geomorphology The CRC is characterized by an arid, intramontane climate, reflecting the effects of significant orographic barriers to precipitation and highly seasonal rainfall. Mean annual precipitation in the CRC is <250 mm yr -1, but most rainfall occurs in the summer months, when a seasonal lowpressure system brings humid northeasterly and easterly winds to the region (Trauth et al., 2003b; Bookhagen and Strecker, 2008). Interannual variability in precipitation is significant (±75%), and driven primarily by ENSO and the Tropical Atlantic Sea-surface Temperature Variability (TAV) (Trauth et al., 2003b). Cooler and more humid periods occurred throughout the Quaternary, increasing landslide-frequency, expanding glacial and periglacial zones, and apparently increasing overall catchment erosional efficiency (Bobst et al., 2001; Haselton et al., 2002; Trauth et al., 2003a; Fritz et al., 2004; May and Soler, 2010; Bookhagen and Strecker, 2012). The geomorphology of the CRC and nearby regions reflect arid, highly seasonal climate and relief >1000 m in intramontane basins. Aerial photography reveals abundant pediment surfaces and alluvial fans throughout the CRC, most of which are incised by modern channels. For example, in the Pucará Valley, incision and base-level lowering of ~100 m have abandoned a sequence of pediments and strath terraces (this study). Similar evidence for Quaternary incision is well documented in the Sierras Pampeanas and Santa Barbara System (Strecker et al., 1989; Hilley and Strecker, 2005; González Bonorino and Abascal, 2012). Pedogenesis is weak, and soils are dominated by soil carbonate (May and Soler, 2010, and this study). Periglacial processes are restricted to areas over 4500 m elevation, but this limit may have been depressed as much as 900 m during the Pleistocene, as evidenced by broad convex range crests and moraines in the Sierra de Quilmes and northwestern CRC (Haselton et al., 2002). 2.5 Methods Field Studies Field studies were focused in the lower Pucará Valley with the goal of characterizing neotectonic structures and Quaternary landscape evolution. We conducted structural and geomorphic mapping of the valley on aerial photography and ASTER 30 m topography. Geology was compiled from existing maps by Carrera and Muñoz (2008) and Coutand et al. (2006). 28

41 29 Additionally, we described soils at 13 sites within the valley. We selected sites to ensure investigation of soils at various pediment levels, and described soils according to USDA soil taxonomy guidelines (Staff, 2010). Descriptions are solely morphological and geochemical classification metrics (e.g. weight percent CaCO 3 ) are inferred. Reported stages (Appendix A) of pedogenic carbonate and gypsum accumulation follow the morphological classification scheme of Gile et al. (1966). We determine desert pavement indices (PDI) according to methods developed by Al-Farraj and Harvey (2000). See Appendix A for more thorough description of PDI methodology Longitudinal River Profile Analysis We rely on digital topographic data and coupled ArcGIS and Matlab scripts to derive normalized channel steepness indices (k sn ) and concavity indices (θ) ( Following methods outlined by Wobus et al. (2006a), we extracted channel topographic data from 30 m ASTER topography (NASA), removed data irregularities, smoothed channel data along a 450 m moving average window, determined local slopes over a 10 m vertical interval, and set a minimum drainage area of 3000 pixels. The above parameters balance our desires to preserve channel topographic complexity, remove artifacts in digital topographic data associated with high relief landscapes, and exclude channel headwaters that are dominated by debris-flow processes (Wobus et al., 2006a). Channel steepness and concavity indices are determined by linear regression of local channel slope and drainage area after log transformation. We normalize steepness indices to a reference concavity of 0.45, following empirical and theoretical predictions for detachment limited systems (Whipple and Tucker, 2002). We identify individual segments along a profile by the occurrence of major knickpoints or downstream confluences with larger trunk streams, and regress the data from each segment to derive k sn and θ. Considering the large scale of our analysis, we selected knickpoints that are conspicuous in log-slope/log-area plots and in topographic profiles. We classify knickpoints according to morphology: slope-break knickpoints, vertical step knickpoints, the base of a convex reach, and the top of a convex reach (see Kirby and Whipple, 2012). We also classify knickpoints genetically, based on their spatial coincidence with significant tectonic (e.g faults) and/or lithologic boundaries (e.g. transition from crystalline basement to Tertiary sedimentary rock), giving rise to four knickpoint types: lithologic, tectonic, lithotectonic, and undefined. We 29

42 30 specifically focus on slope-break and undefined knickpoints, because they may represent transient channel responses to an external forcing (e.g. Harkins et al., 2007) Terrestrial Cosmogenic Nuclide ( 10 Be) Chronology Analytical Procedures In this study we isolate and analyze in situ produced cosmogenic 10 Be in quartz to determine catchment mean erosion rates and date stable geomorphic surfaces. Samples were processed at the University of Vermont Cosmogenic Nuclide Laboratory using standard analytical methods (see auxiliary materials and for detailed methodology). First, quartz was purified for 10 Be analysis using mineral separation procedures modified from Kohl and Nishiizumi (1992). For Beryllium isolation, samples were prepared in batches that contained a full-process blank and 11 unknowns. We used between 11.6 and 23.0 g of purified quartz for analysis. We added ~250 µg of 9 Be carrier made from beryl at the University of Vermont to each sample. After isolation, Be was precipitated at ph 8 as hydroxide gel, dried, ignited to produce BeO, ground, and packed into copper cathodes with Nb powder at 1:1 molar ratio for accelerator mass spectrometry (AMS) measurements. 10 Be/ 9 Be ratios were measured at the Scottish University Environmental Research Center and were normalized to NIST standard with an assumed ratio of based on a half life of 1.36 My. The average measured sample ratio ( 10 Be/ 9 Be) was 947 x and AMS measurement precisions, including blank corrections propagated quadratically, averaged 1.9 %. The blank correction is an inconsequential part of most measured isotopic ratios (<0.7% on average, maximum 2.0%). The CRONUS N standard was run with these samples and returned a concentration of 2.31±0.06 x 10 5 atoms g -1, consistent with values reported by other labs Be Catchment Mean Erosion Rates We contribute five new 10 Be-derived catchment mean erosion rates from the Pucará River catchment and its subcatchments (see Figure 13 for locations of samples BW1,2,3,5, and 6). Samples were collected from bars within active streams. For each sample, we determined the contributing drainage area using GIS software, and 10 Be production rates were calculated for each pixel of a DEM at 250 m resolution. Our Matlab code incorporates elevation, shielding, and 30

43 31 muonogenic production for each pixel, but relies on mean latitude for each catchment. The code follows the scaling scheme of Lal (1991) and a sea-level high-altitude surface production rate of 4.76 atoms g -1 yr -1 (Nishiizumi et al., 2007; Hidy et al., 2010). We calculate erosion rates using a sample density of 2.6 g cm -3 and an attenuation length of 160 g cm -2 (von Blanckenburg, 2005). The uncertainties which accompany our reported erosion rates reflect the uncertainties in both AMS measurements and production rates. Major lithologies in the CRC are quartz rich, so we make no corrections for variably distributed quartz (Sparks et al., 1985; Francis et al., 1989; Do Campo and Guevara, 2005; Marquillas et al., 2005; Coutand et al., 2006). In addition to our own samples, we re-analyze seven previously published catchment-mean erosion rates in the CRC (Bookhagen and Strecker, 2012). Using reported sample locations and nuclide concentrations, we recalculate production rates and erosion rates using the same methods as for our own samples, and find that recalculated and reported values differ by <8%. Similarly, inputting mean latitudes and elevations for each catchment into the CRONUS calculator (rather than using a pixel-by-pixel code) produces erosion rates that differ from our results by <11% (Balco et al., 2008). For sampled catchments which contain sampled subcatchments (BW5 and M2), we calculate the differential erosion rate by area-weighting erosion rates from the contributing subcatchments (Granger et al., 1996) Be Depth Profiles To date pediment surfaces, we hand-excavated 2 m deep pits for the sampling of cosmogenic nuclide ( 10 Be) depth profiles at three locations (Fig. 11). Soil pits were dug at geomorphically stable sites, with minimal evidence for erosion, bioturbation, and complex shielding histories. However, the absence of bar & swale topography, the presence of Av horizons, and the heavily dissected nature of the pediment surfaces throughout the valley collectively suggest some degree of surface degradation at all sites. We sampled ~1 kg of sand-sized grains in ~2 cm thick horizons at 0, 50, 100, 150, and 200 cm depths, across the width of the pit. All samples were field-sieved to remove the < 250 µm fractions, which made up minor portions (<25%) of the total soil mass. To determine surface exposure age, inheritance and erosion rate for each depth profile, we employ the Monte Carlo simulator developed by Hidy et al. (2010). Results reported are from 31

44 32 100,000 model simulations at 1σ uncertainties, based on model parameters described in the auxiliary materials. Although average AMS uncertainty for the data set was <2%, we assigned nuclide concentration uncertainties of 5% for all depth-profile samples, to reflect errors in sampling (e.g. sample depth and thickness), laboratory analysis (e.g. balance errors), geomorphic variability (e.g. bioturbation/cryoturbation, shielding variations), and also systematic errors (e.g. temporal variation in cosmic ray flux, and scaling uncertainty) (Gosse and Phillips, 2001). Model inputs of density and associated uncertainties are based off of previous field determinations in similar soil types with similar ranges of carbonate and gypsum development (Reheis, 1987; Reheis et al., 1995; Hidy et al., 2010) Paleo-Erosion Rates We use inheritance values for each depth-profile to calculate catchment mean paleo-erosion rates. Comparing paleo-erosion rates to modern erosion rates should evaluate whether Quaternary climate changes significantly affected sediment transfer rates in the CRC. We calculate catchment mean 10 Be production rates via the methods described above, deriving paleodrainage basins from modern topography. For each depth profile, we calculate 10 Be concentrations of a paleo-sample with the maximum, minimum and modal solutions for inheritance, corrected for radioactive decay of 10 Be (using the appropriate minimum, maximum, and modal depositional ages, respectively). We find no evidence for stream captures or major drainage reorganization in the Pucará River catchment, suggesting that the use of modern topography is valid, especially given much larger uncertainties in inheritance. We report maximum, minimum, and modal erosion rates for each profile. We also derive vertical incision rates for the Pucará River, using minimum, maximum and modal ages for each depth-profile. We estimate vertical incision as the difference in elevation between the modern Pucará River floodplain and each dated surface, projecting the surface to the modern floodplain, using a reference slope of 3.5. The reference slope reflects the results of differential GPS surveys we conducted across pediments and strath terraces in the Pucará Valley. We use Trimble differential GPS equipment with <10 cm vertical and ~1 m horizontal precision. 32

45 Results Field Studies The semi-arid Pucará Valley contains seven abandoned and incised geomorphic surfaces (Q1 Q7, youngest to oldest) from 5 m to ~100 m above modern base-level, that record multiple pulses of incision in the late Quaternary (Fig. 12). Abandoned pediments and strath terraces dominate the landscape, we find no evidence for significant depositional intervals, and valley incision continues currently. Structural mapping reveals a series of blind and emergent thrusts east of the valley (syncline) axis, active in the Quaternary (Fig. 11). The Pucará Thrust visibly offsets Q3 surfaces, although differential GPS transects across the fault measure vertical displacement <1 m (see Appendix XX). In the southern end of the field area, we observe heavily dissected surfaces, steep rivers, and deeply incised canyons spatially coincident with a N-S striking monocline, suggesting Quaternary activity along a blind thrust. However, additional differential GPS transects of pediment surfaces between these two areas do not reveal any clear signal of deformation (e.g. tilting, oversteepening), suggesting that late Quaternary deformation within the lower Pucará Valley is of relatively low magnitude. 33

46 34 Figure 12. Photograph from the site of the Q5 depth profile in Figure 11, looking approximately northeast. Foreground shows Q5 strath terrace beveled into sedimentary rocks of the Tertiary Payogastilla Group, which rest in angular unconformity over Cretaceous Pirgua Group redbeds. In the background Q2, Q3, Q4, Q6 and Q7 surfaces are beveled into both Tertiary and Cretaceous sedimentary units. Rio Pucará flows from right (south) to left (north). Note monoclinal structure within Cretaceous units beneath the Q7 surface. High ranges are composed of the Neoproterozoic Puncoviscana Formation. Soils in the study area classify broadly as aridisols, and range from Ustic Haplocambids on modern surfaces to Ustic Haplocalcids, Ustic Petrocalcids, Leptic Haplogypsids, and Ustic Petrogypsids on the abandoned surfaces (see Appendix A). The differences between these soil taxons reflect differing degrees of pedogenic accumulation of either carbonate or gypsum. Carbonate and Gypsum reach stage III and incipient stage IV morphology on the highest (Q6 Q3) surfaces, do not exceed Stage II on lower (Q2 Q1) surfaces, and exhibit minimal carbonate accumulation on modern surfaces. Similarly, desert pavements exhibit greater development on 34

47 35 the oldest surfaces, although the differences are minimal, likely because of destructive forces acting on pavements such as vegetation, surface erosion, and human modification. Importantly, the correlations between relative landform age and degrees of pedogenic salt and pavement development indicate that arid or semi-arid conditions in the study area are long-lived and that past humid phases, at least locally, were not significant enough (>750 mm MAP) to cause major dissolution of soil carbonate or gypsum (Gile et al., 1966; Royer, 1999; Buck and Van Hoesen, 2002) River Profile Analysis We analyzed 77 streams in the Calchaquí River catchment, giving rise to 147 separately regressed segments, and 75 knickpoints (Fig. 13). Normalized steepness indices range from 28 to >1000, with a mean k sn of 175. Mean concavity index is 0.9, with a maximum of 28 and minimum values <0 (convex) (Fig. 17). The highest steepness indices occur in a narrow band within and between the high crystalline ranges in the western half of the study area. These steep segments vary greatly in morphology; some are small tributaries to the Calchaqui River, running perpendicular to the structural grain within crystalline bedrock (e.g. STR13 on Figure 13), some are parallel to the structural grain, running within sedimentary rocks in valleys bound by thrust faults (e.g. STR11 on Figure 13), and others represent a combination of those morphologies (e.g. STR16 on Figure 13). A common feature to all steep segments (and the corresponding catchments) is that they cross one or more N to NW striking thrust faults within the high Eastern Cordillera. The lowest steepness indices are generally observed in the eastern part of the catchment along small tributaries to the Calchaquí River (Fig. 13). Many of these tributaries are segmented, with knickpoints and convexities coincident with the Cerro Negro Thrust and other west-vergent thrust faults. We also observe low normalized steepness indices in the southwestern CRC. These segments are typically bound by prominent lithotectonic or lithologic kickpoints (Figures 13D and Streams S67, S65, and S1) coincident with two prominent NW striking lineaments. Morphologically, we classified 25 knickpoints as vertical-step knickpoints, 10 as slope-break knickpoints, and the remaining 40 as high and low bounds on convex channel reaches (Kirby and Whipple, 2012). From a genetic standpoint we classified 10 knickpoints as lithologic, 15 as 35

48 36 tectonic, 26 as lithotectonic, and 24 as undefined. We find no clear correlation between knickpoint morphology and knickpoint genesis, and note that only 1 undefined knickpoint also has slope-break morphology. See auxiliary materials for stream profile figures, stream profile regression data, and knickpoint data. Figure 13 (opposite). (a) Shaded relief map of the CRC, 10Be Catchment mean erosion rate samples and corresponding subcatchments (labeled) from this study and Bookhagen and Strecker (2012). Stream network derived from ASTER 30m DEM and a minimum accumulation of 35,000 pixels (1.05 km2). (b) Lithologic divisions, major faults, and knickpoints in the CRC. Knickpoints according to legend in (d). Dashed lines are newly mapped faults. CNT = Cerro Negro Thrust (Carrapa et al., 2011). (c) 10Be catchment mean erosion rates, in mm kyr-1. Sample locations as per legend in (a). (d) Normalized channel steepness indices and knickpoints in the CRC. See text for description of knickpoint typology and channel regression parameters. Labeled streams are displayed in profile in Figure Streams. 36

49 37 37

50 Be Catchment Mean Erosion Rates Catchment mean erosion rates range from 145 ± 19 mm kyr -1 to 27 ± 3 mm kyr -1 (Table 1), indicating that the averaging time scales of the sampled catchments range from 4 23 kyr (von Blanckenburg, 2005). Erosion rates do not correlate significantly with catchment mean annual precipitation, catchment area, or catchment mean elevation, but show a modest correlation with catchment mean slope (Fig. 14). Comparing catchment mean erosion rates with lithology (Fig. 13) reveals that catchments dominated by resistant lithologies (e.g. crystalline bedrock) exhibit some of the highest (e.g. STR13) and lowest (e.g. BW3) erosion rates in the study, suggesting that lithologic resistance alone cannot explain the spatial variation in erosion rates in the CRC. Table Be Concentrations, Catchment-mean production rates, Catchment mean erosion rates, and corresponding topographic and climatic characteristics. Sample Name Sample Latitude Sample Longitude Sample Elevation, m 10Be Concnetration, atoms g -1 10Be Concentration 1σ, atoms g -1 Mean Production Rate, atoms g -1 yr -1 Mean Production rate 1σ, atoms g -1 yr -1 Erosion Rate, mm kyr -1 Erosion Rate 1σ, mm kyr -1 BW E E BW E E BW E E BW E E BW5 * E E+04 N/A N/A BW E E M E E M2* E E+03 N/A N/A STR E E STR E E STR E E STR E E STR E E STR E E Sample Name Apparent Age, kyr Centroid Latitude Centroid Longitude Mean elevation (m) Mean Precipitation (mm yr -1 ) Drainage Area (km -1 ) Mean Slope (degrees) Mean 1km radius relief, m Mean 5km radius relief, m BW N/A BW BW BW BW5 *

51 39 BW N/A M M2* STR N/A STR STR STR STR STR Figure 14. Correlations between catchment mean erosion rates and catchment mean annual precipitation, catchment area, catchment mean slope, and catchment mean elevation. See Table 1 for data. 39

52 Be Depth Profiles At all three sites, surface samples exhibit low 10 Be concentrations compared to depth-profile attenuation curves (Fig. 15) (Anderson et al., 1996). Previous work suggests that low surface concentrations reflect bioturbation, so we exclude the surface samples from our depth-profile simulations (Hidy et al., 2010). Although such exclusion significantly increases the range of solutions (the uncertainty), the modal ages produced (hereafter referred to as best-fit ages) more accurately reflect depositional ages. The depth-profile simulator yields best fit ages of 42.9 ka, 96.3 ka, and ka for our Q2, Q5, and Q6 surfaces, respectively, therefore agreeing with geomorphic relative-age constraints. Additionally, simple calculations using the formulations of Anderson et al. (1996), which assume no surface erosion, yield ages of 41.8 ka, 85.7 ka, and 141 ka (see auxiliary materials for methods and inputs), suggesting that we have robust age signals. Here we report uncertainty as the maximum and minimum solutions to acknowledge the full range of model solutions, but refer the reader to appendix C to view frequency histograms for age, inheritance, and surface erosion rate. Figure 15. In situ 10Be depth profiles and monte carlo simulator results for age, inheritance, and surface erosion rates when run for 100,000 solutions at 1 sigma uncertainty, according to parameters described in the text and 40

53 41 appendices. Black line is the best fit. Gray lines are 100,000 model solutions. Solid black dots are subsurface samples used in the model simulations. Hollow dots are surface sediment samples that were analyzed, but not used in model simulations due to evidence of bioturbation. Hollow square represents a quartz cobble amalgamation (n=85) sample that was simularly excluded from model simulations. AR13-01 is located on a Q2 strath terrace, consisting of ~4 m of channel sands and lag deposits which lie in angular unconformity over Miocene age sedimentary rock (Angastaco Fm.). The soil consists of a coarse desert pavement, underlain by a vesicular A horizon (Av), which is underlain by a Bk horizon that diffusely transitions to a C horizon. We find only minor field evidence for bioturbation at this site, but the weak stratification and uniformity of the soil framework grains makes identification of vertical mixing difficult. Model simulations yield a modal age of 42.9 ka and a modal surface erosion rate of 2.4 mm kyr -1, thereby indicating that ~10 cm of erosion has occurred at this site. AR13-02 is located on a Q5 fluvial strath terrace sourced dominantly from Paleozoic granitoids and Tertiary volcanics southwest of the Pucará valley. This deposit consists of couplets of fine and coarse grained layers, similar to AR13-01, but the sedimentology is partially obscured by significant carbonate accumulation. The soil consists of a pavement layer over a shallow, weakly developed Av horizon over a massive and root-limiting Bkk horizon over a Bk horizon which diffusely transitions to a C horizon. The shallow depth to the Bkk horizon (10 cm) suggests that significant erosion of the surface has occurred, prompting us to input a wide range (10 90 cm) for the total erosion threshold parameter in the depth profile simulator (Royer, 1999). We report a modal age of 96.3 ka and a modal surface erosion rate of 3.7 mm kyr -1, which yield an erosion estimate of ~36 cm, confirming our suspicion of surface degradation at this site. AR13-03 is located on a Q6 pediment surface and is notable for its coarse sedimentology and its pedogenic gypsum content. The lower portion of the alluvial deposit is a clast-supported pebble to cobble conglomerate, with moderate internal stratification and moderate sorting within individual strata, indicating that it was deposited by sheetflood processes (Blair and McPherson, 1994). The upper part of the pit appears to be a storm deposit, likely of similar age to the underlying material, given similar degrees of soil development. This event scoured a channel into the existing alluvial surface, and deposited a poorly sorted, matrix-rich conglomerate with no noticeable stratification. The soil consists of an Av over a Byy horizon over a Byk horizon. 41

54 42 Similar to AR13-02, we place large bounds on the total erosion threshold parameter for AR Depth profile simulations yield a modal age of ka and a modal erosion rate of 1.2 mm kyr -1, suggesting that ~20 cm of erosion has occurred on this surface. See auxiliary materials for annotated pit photos and model input parameters for each depth profile Paleo-Erosion Rates Vertical incision estimates for our Q2, Q5, and Q6 depth profiles are 11, 70, and 76 m, respectively, yielding vertical incision rates of 270, 730, and 480 mm kyr -1 when modal ages are used (Table 2). Catchment mean paleo-erosion rates derived from modal inheritances of the three profiles are 98, 52, and 50 mm kyr -1 respecticely. The depth profile AR13-01 (Q2) paleodrainage reaches the edge of the Puna Plateau. At this site modal values suggest that the incision rate is ~2.5 times higher than the catchment mean paleo-erosion rate, although the rates overlap given their large uncertainty. Depth profile AR13-02 (Q5) has a paleo-drainage nearly identical in extent to catchment BW3, and yields a vertical incision rate at minimum 3.5 times the catchment mean paleo-erosion rate at this site. The paleo-drainage for depth profile AR13-03 (Q6) is a local tributary for the Pucará River, similar to the modern BW1 catchment. The modal vertical incision rate (480 mm kyr -1 ) is nearly an order of magnitude higher than the catchment mean paleo-erosion rate (50 mm kyr -1 ) at this site. Table 2. Vertical incision rates and catchment mean paleo-erosion rates derived from 10 Be depth profile ages and inheritance, respectively. See section for methodology. Depth Profile Total Incision (m) Vertical Incision Rates Mode (mm kyr -1 ) Maximum (mm kyr -1 ) Minimum (mm kyr -1 ) Inherited Catchment Mean Erosion Rates Mode (mm kyr -1 ) Maximum (mm kyr -1 ) Minimum (mm kyr -1 ) JM-AR JM-AR JM-AR N/A Discussion Here we interpret our results with respect to the distribution of lithology, faults, and precipitation in the CRC. We find that the spatial distribution of steepness indices and catchment mean 42

55 43 erosion rates indicate strong lithologic control on erosion and topography throughout the catchment. We also find evidence for active tectonics, particularly in the western CRC. The distribution of anomalously high concavity indices also supports active faulting, as does our analysis of knickpoints. Paleo-erosion rates indicate different uplift rates across major faults in the CRC, and are similar to modern rates, while vertical incision rates may reflect transient landscape adjustment or variable lithologic resistance. Lastly we consider the tectonic implications of out-of-sequence deformation in the retroarc foreland Controls on River Morphology Normalized Channel Steepness Indices In the eastern part of the catchment, normalized steepness indices are controlled, at least in part, by lithology (Figure 13). For example, in catchment M2, where less resistant sedimentary rocks and Quaternary alluvium dominate, we observe low steepness indices (<200), but a high catchment erosion rate (118 ± 16 mm kyr -1 ), indicating that low steepness indices reflect weaker lithologies rather than low uplift rates in the eastern CRC. STR3, a small eastern catchment that has its headwaters in more resistant crystalline rock, has similarly low steepness indices but a significantly lower erosion rate than M2 (41 ± 5 mm kyr -1 ), further supporting the notion that erosion rates in the eastern CRC are predominantly controlled by lithologic resistance. In the western CRC, high k sn values are measured in crystalline rocks within the high ranges bordering the plateau (Figure 13). To some extent this may reflect greater lithologic resistance to erosion, but we note that erosion rates vary widely across the western catchments, suggesting that spatially variable uplift rates may also control steepness indices. For example, in catchment STR13, which is principally composed of crystalline bedrock, tributaries to the Calchaquí River exhibit high (>200) k sn values, and the catchment erodes at a high rate (142 ± 18 mm kyr -1 ), suggesting that uplift rates are higher here than in other parts of the CRC (e.g. catchment STR3). Catchment STR11 (Luracatao River) similarly exhibits high erosion rate and steepness indices. In contrast, we measure low erosion rates (<40 mm kyr -1 ) in catchments BW2 and BW3. In these two catchments low steepness indices occur above prominent knickpoints that coincide with a major NW striking lineament, upstream of which we also observe small Quaternary basins. Below these lithotectonic knickpoints we observe steep segments with anomalously high 43

56 44 concavity indices (see below for discussion of channel concavities). This suggests that BW2 and BW3 exhibit such low erosion rates because the majority of their drainage areas lie above a major fault which separates areas of low and high uplift rates (Willenbring et al., 2013). In the SW catchments (BW2, BW3, BW5, and STR16) we observe evidence for a major NW-SE striking fault. It is mapped in its southern end (within BW2, BW3, and part of BW5) but we use our data to infer its extension to the north (Figure 13 B). For example, we suspect that the upper reaches of catchment BW5 (e.g. S67; Figure 16), would exhibit similarly low catchment mean erosion rates if sampled at or above the prominent lithotectonic knickpoints along the fault. Catchment STR16 also records a low catchment mean erosion rate, which suggests tectonic isolation similar to catchments BW2 and BW3. Although direct evidence for a fault is obscured by Tertiary ignimbrites, knickpoint distribution and form suggest that a previously unidentified fault (parallel to the dominant structural grain) divides two zones of differing uplift rate (Figure 13). In addition to the dominant tectonic controls, climate also exerts some control on channel steepness in the CRC. Bookhagen and Strecker (2012) demonstrated that correcting for the effect of spatially variable precipitation on discharge significantly influences the distribution of normalized steepness indices in this region. However, we find that the steepest channel reaches in our analysis closely match those identified in the precipitation-corrected analysis by Bookhagen and Strecker (2012), indicating that climatic corrections would not significantly affect our interpretations (see Figure DR8 in Bookhagen and Strecker, 2012). This is expected given the relatively uniform precipitation of the CRC compared to the steep precipitation gradient that was the subject of the previous work. We find that our systematic investigation of river profiles, which uses a significantly shorter channel-smoothing window (450 m vs. 5 km used by Bookhagen and Strecker, 2012), allows for analysis of more spatially discrete (e.g. lithologic and tectonic) controls on channel morphology. 44

57 45 Figure 16. Selected longitudinal river profiles and corresponding local slope/drainage area regressions. Individual segments are bound by knickpoints or confluences with trunk streams and were regressed with a reference concavity of Resulting normalized steepness indices and raw concavity indices are displayed for each segment. Question marks identify faults with unknown dip. In slope-area space light and dark blue lines represent forced and unforced regressions, respectively. See Figure 13 for stream locations. CNT = Cerro Negro Thrust; PT = Pucara Thrust; JVT = Jasimana-Vallecito Thrust Concavity Indices and Non-Uniform River Profile Morphology A key assumption in tectonic interpretations of normalized steepness indices in bedrock channels is that lithology, climate, and uplift rate are uniform along a given channel reach, and that abrupt 45

58 46 changes are marked by knickpoints. When this is true, concavity indices typically fall into a relatively narrow range ( ) (Kirby and Whipple, 2012). However, when uplift or climate gradients exist, or when transitions from detachment-limited to transport-limited conditions occur, concavity indices can vary widely (Whipple, 2004). Our results for all streams (n = 147) yield an anomalously high mean concavity of 0.9, with respect to theoretical expectations, which rises further to >2 when we exclude convex segments (e.g. θ < 0) (Figure 17). Here we address the spatial distribution of concavity indices in the CRC, and the factors promoting such high channel concavities. The Calchaquí River itself is a well-graded profile (Figure 16). The lower segment exhibits a concavity index within the expected range for river profiles in tectonically active orogens (0.53), while the upper segment has a slightly low concavity index (0.34), likely reflecting the influence of debris-flows and/or high sediment flux in the upper most part of the catchment (Whipple, 2004). The Calchaquí River flows through (and actively incises) sedimentary rock and Quaternary alluvium, and also crosses the Cerro Negro Thrust, but we note no major breaks across lithologic or tectonic boundaries. The narrow range of concavity and the well graded profile suggests that the Calchaquí River is equilibrated to the prevailing climatic and tectonic conditions and thus is in steady-state (Whipple et al., 2013). Small tributaries to the Calchaquí River typically have concavities between 0.3 and 1, within the normal range of incising rivers (Figure 17). Low concavities (<0.4) likely reflect the effects of debris-flow processes and incision thresholds, especially for smaller catchments which undergo periglacial processes in their headwaters. Higher concavities (0.7<θ<2) likely reflect downstream reductions in both lithologic resistance (thus, an increase in K) and uplift rate, as well as transitions to alluvial conditions at the range front (Whipple, 2004 and references therein). In the CRC all three of these conditions are common, as rivers typically originate in fault-bounded crystalline ranges that are bordered by Tertiary-Quaternary intramontane basins. 46

59 47 Figure 17. Concavity indices and mean annual rainfall in the CRC. See Figure 13 for knickpoint classification. TRMM precipitation data from Bookhagen and Strecker (2008). Labeled streams are displayed in profile in Figure Streams. Extreme concavities (>2) occur along segments that are in the hanging walls of major thrust faults, just downstream of lithotectonic knickpoints. Downstream lithologic changes commonly occur along these segments, but in some cases we observe no such change (e.g. S65; Figure 16), suggesting that the faults which bound these segments are active, and gradual downstream reductions in uplift rates drive the high concavities (Whipple, 2004). We also observe downstream increases in precipitation (increasing K) along many high-concavity reaches, but we find the magnitude of increase to be too small (<500 mm yr -1 ) to have a significant effect (Figure 17) (Schlunegger et al., 2011; Bookhagen and Strecker, 2012). Channel convexities (θ<0) are 47

60 48 also closely associated with tectonic features in the CRC. In most cases channel convexities are short (<10 km) and occur across faults. However, some convex reaches can be as long as 50 km, and tend to run sub-parallel, or at low angles, to faults in the study area (e.g. S1; Fig. 16). Steepness indices are usually low above convex reaches and higher below, providing evidence that these convexities represent transitions from zones of low to high uplift rate (Whipple et al., 2013). Overall we find that deviations from the expected range of concavity indices in erosive landscapes (0.4<θ<0.7) can be reasonably well explained by the structurally controlled distribution of lithology in the CRC; resistant crystalline ranges bound by faults are the headwaters for streams, and lower reaches flow through less resistant (and potentially more slowly uplifting) sedimentary rocks and alluvium, leading to high concavities. In some cases, increasing downstream precipitation may contribute to this affect. Channel convexities are associated with discrete tectonic features, and may separate regions of low and high uplift rate. In particular, we argue that deformation is most active in the western half of the CRC, along a narrow band of NNW-SSE striking reverse faults Knickpoint Genesis, Form, and Distribution The majority of knickpoints (51 of 75) in the study area are spatially coincident with tectonic and/or lithologic discontinuities along channels, providing further evidence that channel morphology in the CRC primarily reflects structurally-controlled, lithologic heterogeneity. However, we identify 24 knickpoints of undefined genesis, which can be employed to evaluate the passage of transient channel responses in a landscape (e.g. Schoenbohm et al., 2004; Crosby and Whipple, 2006; Harkins et al., 2007). Such an approach relies on the prediction that a drop in base-level or change in uplift-rate/climate will create a slope-break knickpoint that migrates up the channel network at a horizontal rate dependent on contributing drainage area (that is, stream power), and at a fixed vertical rate. Transient knickpoints can therefore be identified in the landscape by uniform elevations and map-view positions in the channel network (Wobus et al., 2006b). Importantly, this approach assumes that concavity indices do not respond to rock uplift rates. Our analysis suggests that concavity indices in the CRC do indeed reflect changes in uplift 48

61 49 rates, so the migration of transient knickpoints in our study area may not produce uniform elevations. This approach also assumes detachment-limited conditions throughout channel reaches; in transport-limited erosional systems, transient responses are characterized by a gradual change in channel gradient along the entire reach, making transient and steady-state morphologies indistinguishable (Whipple and Tucker, 2002). Although our field observations support detachment-limited conditions along steep channel reaches in the CRC, we note that transportlimited conditions are dominant in the intramontane Pucará Valley (evidenced by mixed bedrock-alluvial channel morphology and >3 m thick sedimentary cover on abandoned strath terraces) (Whipple and Tucker, 2002). Similar transport-limited segments likely occur in other intramontane basins in the CRC. Figure 18. Vertical distribution of knickpoints in the CRC. See Figures 13 and 17 for plan view. Given these complexities, we find little evidence for transience in our analysis of knickpoint distribution and form. Undefined knickpoints (those which are not associated with discrete lithologic and/or tectonic discontinuities) are observed across a wide range of elevations (Fig. 18), and do not exhibit physical relationships that predict transient knickpoint behavior, such as 49

62 50 the power law relationship between knickpoint contributing drainage area and knickpoint distance upstream of tributary mouths (i.e horizontal celerity) (Harkins et al., 2007). We find nine knickpoints clustered at approximately 4000 m elevation, but they are linearly aligned and are parallel to previously mapped thrusts, suggesting structural rather than temporal control (Wobus et al., 2006a) Controls on Landscape Evolution of the Pucará Valley Spatial Controls In the Pucará Valley, vertical incision rates are local measurements, integrated over long timescales (as much as 300 kyr in the case of the Q6 pit), while catchment mean paleo-erosion rates integrate over much larger areas, but short (<20 kyr) periods. As a result, discrepancies between vertical incision rates and paleo-erosion rates may reflect variations in tectonic, climatic, and lithologic controls on erosion in different areas of the catchment. Our analyses of the Q2 and Q5 depth profiles reveal that the lower Pucará Valley has best-fit vertical incision rates 2.5 to 13 times higher than catchment mean paleo-erosion rates, suggesting that the lower Pucará Valley has eroded at a higher rate than its headwaters for the last ~100 ka (the approximate age of the Q5 surface). We acknowledge the difficulty in comparing vertical incision and catchment mean denudation, but this explanation is supported by our analysis of modern denudation rates and channel steepness indices, which are lower in the headwaters and higher in the Pucará Valley (Harkins et al., 2007). Differential rates of denudation can most easily be explained by differing uplift rates across major faults (Figure 13). The Q6 depth profile, excavated on a pediment surface derived from a smaller catchment area, provides a more local estimate of catchment-mean paleo-erosion rate than do the Q2 and Q5 depth profiles. Vertical incision and paleo-erosion rates may therefore be expected to agree, assuming long-term topographic steady-state (Dortch et al., 2011). However, at this site, as at our other sites, best-fit vertical incision rate is nearly an order of magnitude higher than catchment mean paleo-erosion rates. This discrepancy may reflect a low erosion rate at the time of surface deposition (~160 ka), and a subsequent sustained increase in erosion rate. Alternatively, this discrepancy in erosion rate may reflect substantially lower erosion rate in the sediment source area (Sierra de Quilmes) than in the Cretaceous and Tertiary sedimentary rocks where pediment 50

63 51 gravels are deposited (Figure 11). Previous reconstructions of a landslide-dammed paleo-lake suggest that, during cyclic short-term (annual to decadal) humid phases (e.g. La Niña), erosion rates increase significantly (10-15%) in Cretaceous Quaternary sedimentary units while erosion rates in the high crystalline ranges remain relatively constant (Trauth et al., 2003b). Over year timescales, this effect may intensify the overall relief of the Eastern Cordillera and explain the discrepancy between vertical incision rate and catchment mean paleo-erosion rate Temporal Controls Given the lack of transient signals in modern streams, we assume that rates of tectonic uplift have been relatively constant across the <300 ka timescale of our paleo-erosion rate analyses. Differences between modern and paleo-erosion rates should therefore reflect climatic changes, which have occurred frequently throughout the Quaternary in this region (Bobst et al., 2001; Trauth et al., 2003a; Fritz et al., 2004). Our catchment mean paleo-erosion rates ( mm kyr -1 ) are not markedly different from modern catchment mean erosion rates ( mm kyr - 1 ), suggesting that climate has not significantly influenced erosion rates over this period (Table 2), or that short term climate changes in the Quaternary were not significant enough to affect erosion rates in the CRC on cosmogenic nuclide timescales. This finding contrasts with recent reconstructions of a ~6700 year sedimentary record from a landslide-dammed paleo-lake that existed during the humid Minchin Phase (25 to 40 ka), which yields catchment mean erosion rates an order of magnitude higher than modern rates in the CRC (Bookhagen and Strecker, 2012). Given the large uncertainties associated with our depth-profile surface ages, we cannot associate paleo-erosion rates with discrete climate intervals. Further, 10 Be catchment mean erosion rates and paleo-erosion rates are averaged over 4 23 kyr timescales in this study, and thus may integrate across multiple climate phases (Bierman and Steig, 1996) Tectonic Implications Our analysis of longitudinal river profiles, catchment mean erosion rates, and paleo-erosion rates provide strong evidence that Quaternary tectonic deformation influences the rate and style of landscape evolution in the Eastern Cordillera. Coupled steepness indices and catchment mean 51

64 52 erosion rates, high concavity indices, linearly-aligned knickpoints, and ponded Quaternary sediment above knickpoints all point to differential uplift across a band of N-S to NW-SE trending reverse faults in the western CRC. The orientation of faults within this band reflects preexisting structural anisotropies within the crystalline bedrock, and reactivated Cretaceous rift structures (Grier et al., 1991; Hongn et al., 2007; Santimano and Riller, 2012; Carrapa et al., 2014b). Further, field investigations in the lower Pucará valley reveal an active reverse fault, an active blind thrust, and locally deeply incised (~100 m) pediment surfaces, supporting our interpretations of active shortening in the western CRC. In the southeastern CRC, lithotectonic knickpoints, high channel concavities, and channel convexities suggest that the Cerro Negro Thrust and other west-vergent thrusts are also active (Figure 13). Therefore, we argue that Quaternary shortening is active throughout the CRC, along most major faults in the study area. This assertion is supported by field evidence for shortening in subcatchments within the CRC and in adjacent areas (Strecker et al., 1989; Hilley and Strecker, 2005; Carrera and Munoz, 2008; Hain et al., 2011; Santimano and Riller, 2012). Quaternary shortening in the CRC has implications for tectonic and kinematic models of the Eastern Cordillera. Active shortening in the interior of the thick-skinned orogenic wedge across the southern Puna (DeCelles et al., 2011) could reflect reduced surface slopes, possibly due to increased erosional efficiency (Davis et al., 1983; Whipple, 2009). However, erosion rates in the Eastern Cordillera have likely decreased or not changed in the Late Quaternary due to increased aridity since the Minchin Phase (Bookhagen et al., 2001), which would favor eastward propagation of deformation rather than internal deformation (Bookhagen and Strecker, 2012). This suggests that localized shortening in the Eastern Cordillera is driven by kinematic (e.g. changing slab geometry) or geodynamic (e.g. gravitational spreading) processes (Schoenbohm and Strecker, 2009) rather than climatic changes. Irrespective of the causes for shortening in the CRC, the style of deformation we observe suggests a localized coupling between tectonics and climatically-moderated erosion. Numerical modeling of mountain belts bound by preexisting high-angle faults (as is the case for the western CRC) shows that, as shortening occurs, dry (less erosive) conditions favor the build-up of surface slopes, which in turn drives deformation to adjacent structures, while humid conditions focus erosion and deformation along a discrete orographic front (Willett, 1999; Hilley et al., 2005). 52

65 53 Morphologically, dry conditions are reflected by wide mountain ranges with intramontane basins and distributed deformation along many inherited structures (Hilley and Coutand, 2010), such as can be observed in the CRC (this study) and in the northernmost Sierras Pampeanas, which border the study area to the south (Fig. 10) (Sobel and Strecker, 2003; Hilley et al., 2005). We argue that, in these regions, preexisting structural heterogeneities are the dominant control on mountain range structure, and the current dry climate favors accommodation of shortening across multiple structures (Hilley and Coutand, 2010). We also present speculative evidence for active extension in the southwestern CRC. In the upper reaches of the Pucará River catchment, we identify a previously unmapped NNW-striking fault (Figs. 13 and 17). We did not observe this fault in the field, and so cannot constrain its dip. However, this fault is parallel to strike-slip and extensional faults on the Puna Plateau mapped by Schoenbohm and Strecker (2009), to minor Quaternary strike-slip faults (not shown) in the Cachi Range that are coincident with knickpoints in our analysis (Pearson et al., 2012), and to a major fault zone immediately north of the study area, which records Quaternary strike-slip faulting and extension on the Puna Plateau (Lanza et al., 2013). Plio-Quaternary strike-slip and extensional tectonics in NW Argentina have been attributed to gravitational spreading on the Puna Plateau, potentially in response to lithospheric foundering (Schoenbohm and Strecker, 2009; Zhou et al., 2013). Regardless of the morphology of this newly mapped fault, continued displacement in the current dry climate could lead to upstream channel defeat and basin isolation, and ultimately morphologic incorporation into the Puna Plateau (Humphrey and Konrad, 2000; Sobel et al., 2003) 2.8 Conclusions In this study we use field investigations, systematic analysis of longitudinal river profiles, 10 Bederived catchment mean erosion rates, and paleo-erosion rates and vertical incision rates from 10 Be depth profiles to examine the late Quaternary landscape evolution of the Calchaquí River Catchment. The distribution of high normalized steepness indices, abrupt lithotectonic knickpoints, and variable catchment mean erosion rates demonstrate that incision and sediment routing in this landscape are largely controlled by active tectonics and the structural juxtaposition 53

66 54 of variably resistant lithologies. Anomalously high channel concavities, typically observed in the hanging walls of thrust faults, reflect some combination of downstream decreases in uplift rate, decreases in bedrock resistance (through lithologic and/or climatic changes), and transitions from bedrock to alluvial channel reaches. Lithologic and tectonic controls, including preexisting structural heterogeneity, obscure the effects of spatially variable climate on erosion and river profile morphology, but aridity in the CRC may contribute to the distributed pattern of deformation (Hilley et al., 2005). We find no clear evidence for transience in the landscape, but along-channel fluctuations between detachment-limited and transport-limited conditions restrict our ability to evaluate whether erosion and uplift are balanced. Knickpoints reveal that previously unidentified faults subparallel to the dominant structural grain provide important base-level controls on the uppermost reaches of the western CRC. Aggradation behind these uplifting blocks occurs to keep pace with deformation, but continued tectonic isolation of baselevel and low precipitation rates could lead to channel defeat, internal drainage, and incorporation into the Puna Plateau. We speculate that pervasive shortening in the CRC and perhaps localized extension reflect gravitational spreading on the Puna Plateau. Future kinematic analyses may elucidate the controls on active shortening in the CRC, and Quaternary paleoclimatic analyses may better evaluate the coupling of climate and tectonics in the Central Andean retroarc foreland, but our findings suggest that a catchment scale understanding of the controls on erosion is a prerequisite to regional analyses of tectonic and climatic interactions. 54

67 55 3 Chapter 3: Concluding Remarks 3.1 Methodological Considerations River Profile Analysis & Catchment-Mean Erosion Rates The results of this study demonstrate that successful application of longitudinal river profile analysis in complex landscapes can be achieved through systematic analysis of river profiles in multiple orientations with regards to structural trends (Kirby and Whipple, 2012). Given the variable lithology, climate, and surface process domains (e.g. transport-limited and detachmentlimited channels) that we observed in the study area, the incorporation of catchment-mean erosion rates further improved our ability to evaluate the spatial distribution of deformation in the study area (Cyr et al., 2014). That being said, our arguments may have been more strongly supported had we sampled river sediment directly at knickpoints, thereby measuring erosion rates along discrete channel segments identified in our river profile analysis. This speaks to the need for carrying out longitudinal river profile analysis before field work. We selected catchment mean erosion rate sample sites based on qualitative assessment of landscape relief (e.g. aerial photography, Google Earth), but longitudinal river profile analysis provides a quantitative measure of relief and thus a higher likelihood of separating regions of different erosion rates (and by extension, uplift rates). For future investigations we suggest that researchers analyze longitudinal river profiles first, draft testable hypotheses (see Cyr et al., 2014), and then sample for catchment-mean erosion rates in locations that will allow for the evaluation of such hypotheses TCN Depth Profiles The large uncertainties associated with our TCN depth-profile ages limit our ability to investigate discrete climate intervals in the past, so we cannot definitively evaluate the effect of Quaternary climate changes on catchment erosion rates. To improve depth-profile precision we suggest that the surface sample (0 cm) be replaced with a subsurface sample that is definitively deeper than the zone of bioturbation. In this study a sample at 20 cm would be of sufficient depth for all three profiles. Communication with J. Gosse suggests that this sampling methodology is 55

68 56 of increasing favor among researchers who frequently work with TCN depth profiles (Hidy et al., 2010). 3.2 Future Research This work advances the application of longitudinal river profile analysis by using catchmentmean erosion rates to explore the relative effects of variable lithology and uplift rate on profile form. This technique was recently proved by Cyr et al. (2014), where uplift rates and lithologic resistance were previously known, thereby setting the stage for the application of longitudinal river profile analysis into environments of increasing complexity. This technique should advance our understanding of the ways that landscape relief, tectonics, and climate control erosion rates. When paired with hillslope erosion rates, thermochronologic estimates of erosion, and paleoerosion rates from 10 Be depth profiles, our understanding of the controls on erosion and sediment routing in active orogens should further improve (Portenga and Bierman, 2011; Kirby and Whipple, 2012). I believe that continued work in the Calchaquí River Catchment could advance this technique, and also improve our knowledge of the Quaternary landscape evolution of the southern Eastern Cordillera. For example, erosion rate sampling of additional perched low-relief subcatchments in the western CRC (similar to BW2 and BW3) would further test our hypothesis that the low relief landscapes reflect markedly lower uplift rates. Field investigations at the transitions to those landscapes (i.e. at lithotectonic knickpoints) would allow one to assess the dip of faults and therefore assess whether extension or shortening is occurring, which has implications for gravitational spreading processes on the Puna (Schoenbohm and Strecker, 2009). Dating more abandoned surfaces with TCN depth profiles (in the Pucará and beyond), using the revised sampling methodology from section 3.1.2, should help evaluate the controls on base-level in the CRC: if terraces in other subcatchments within the CRC have similar ages that suggests a synchronous base-level control, most likely climate, whereas a diachronous response would most likely reflect tectonic control (Hilley and Strecker, 2005). Numerical modeling of channel incision processes, constrained by field analyses, would permit the derivation of bedrock erodibility (K) and therefore allow us to estimate uplift rate (U) (Sobel et al., 2003; Hilley and Strecker, 2005). Numerical modeling would also be able to evaluate whether the climatically- 56

69 57 moderated changes in erosional efficiency that have occurred in the Quaternary are significant enough to influence tectonics (Beaumont et al., 2001; Sobel et al., 2003; Hilley et al., 2005). 57

70 58 References Al-Farraj, A., and Harvey, A. M., 2000, Desert pavement characteristics on wadi terrace and alluvial fan surfaces: Wadi Al-Bih, UAE and Oman: Geomorphology, v. 35, no. 3-4, p Allmendinger, R. W., Jordan, T. E., Kay, S. M., and Isacks, B. L., 1997, The evolution of the Altiplano-Puna plateau of the Central Andes: Annual review of Earth and Planetary Sciences, v. 25, p Anderson, R. S., Repka, J. L., and Dick, G. S., 1996, Explicit treatment of inheritance in dating depositional surfaces using in situ 10 Be and 26 Al: Geology, v. 24, p Balco, G., Stone, J. O., Lifton, N. A., and Dunai, T. J., 2008, A complete and easily accessible means of calculating surface exposure ages or erosion rates from 10Be and 26Al measurements: Quaternary Geochronology, v. 3, p Barazangi, M., and Isacks, B. L., 1976, SPATIAL-DISTRIBUTION OF EARTHQUAKES AND SUBDUCTION OF NAZCA PLATE BENEATH SOUTH-AMERICA: Geology, v. 4, no. 11, p Barker, P. F., and Burrell, J., 1977, OPENING OF DRAKE PASSAGE: Marine Geology, v. 25, no. 1-3, p Beaumont, C., Jamieson, R. A., Nguyen, M. H., and Lee, B., 2001, Himalayan tectonics explained by extrusion of a low-viscosity crustal channel coupled to focused surface denudation: Nature, v. 414, no. 6865, p Bianchi, M., Heit, B., Jakovlev, A., Yuan, X., Kay, S. M., Sandvol, E., Alonso, R. N., Coira, B., Brown, L., Kind, R., and Comte, D., 2013, Teleseismic tomography of the southern Puna plateau in Argentina and adjacent regions: Tectonophysics, v. 586, no. 0, p Bierman, P., and Nichols, K., 2004, Rock to sediment - Slope to sea with 10-Be - Rates of landscape change: Annual review of Earth and Planetary Sciences, v. 32, p Bierman, P., and Steig, E. J., 1996, Estimating rates of denudation using cosmogenic isotope abundances in sediment: Earth Surface Processes and Landforms, v. 21, no. 2, p Bierman, P. R., Caffee, M. W., Davis, P. T., Marsella, K., Pavich, M., Colgan, P., Mickelson, D., and Larsen, J., 2002, Rates and timing of Earth surface processes from in situ-produced cosmogenic Be-10, in Grew, E. S., ed., Reviews in Mineralogy and Geochemistry: Beryllium; mineralogy, petrology, and geochemistry, Volume 50: Washington, DC, United States, Mineralogical Society of America and Geochemical Society, Washington, DC, p

71 59 Bobst, A. L., Lowenstein, T. K., Jordan, T. E., Godfrey, L. V., Ku, T. L., and Luo, S. D., 2001, A 106 ka paleoclimate record from drill core of the Salar de Atacama, northern Chile: Palaeogeography Palaeoclimatology Palaeoecology, v. 173, no. 1-2, p Bookhagen, B., and Burbank, D. W., 2006, Topography, relief, and TRMM-derived rainfall variations along the Himalaya: Geophysical Research Letters, v. 33, no. 8. Bookhagen, B., Haselton, K., and Trauth, M. H., 2001, Hydrological modelling of a Pleistocene landslide-dammed lake in the Santa Maria Basin, NW Argentina: Palaeogeography Palaeoclimatology Palaeoecology, v. 169, no. 1-2, p Bookhagen, B., and Strecker, M. R., 2008, Orographic barriers, high-resolution TRMM rainfall, and relief variations along the eastern Andes: Geophysical Research Letters, v. 35, no. 6. -, 2012, Spatiotemporal trends in erosion rates across a pronounced rainfall gradient: Examples from the southern Central Andes: Earth and Planetary Science Letters, v. 327, p Buck, B. J., and Van Hoesen, J. G., 2002, Snowball morphology and SEM analysis of pedogenic gypsum, southern New Mexico, USA: Journal of Arid Environments, v. 51, no. 4, p Burbank, D. W., Blythe, A. E., Putkonen, J., Pratt-Sitaula, B., Gabet, E., Oskin, M., Barros, A., and Ojha, T. P., 2003, Decoupling of erosion and precipitation in the Himalayas: Nature, v. 426, no. 6967, p Bywater-Reyes, S., Carrapa, B., Clementz, M., and Schoenbohm, L., 2010, Effect of late Cenozoic aridification on sedimentation in the Eastern Cordillera of northwest Argentina (Angastaco basin): Geology, v. 38, no. 3, p Cahill, T., and Isacks, B. L., 1992, SEISMICITY AND SHAPE OF THE SUBDUCTED NAZCA PLATE: Journal of Geophysical Research-Solid Earth, v. 97, no. B12, p Canavan, R. R., Carrapa, B., Clementz, M. T., Quade, J., DeCelles, P. G., and Schoenbohm, L. M., 2014, Early Cenozoic uplift of the Puna Plateau, Central Andes, based on stable isotope paleoaltimetry of hydrated volcanic glass: Geology. Caputo, M. V., and Crowell, J. C., 1985, MIGRATION OF GLACIAL CENTERS ACROSS GONDWANA DURING PALEOZOIC ERA: Geological Society of America Bulletin, v. 96, no. 8, p Carrapa, B., Bywater-Reyes, S., DeCelles, P. G., Mortimer, E., and Gehrels, G. E., 2012, Late Eocene-Pliocene basin evolution in the Eastern Cordillera of northwestern Argentina (25 degrees-26 degrees S): regional implications for Andean orogenic wedge development: Basin Research, v. 24, no. 3, p Carrapa, B., Huntington, K. W., Clementz, M., Quade, J., Schoenbohm, S. B. R. L. M., and Canavan, R. R., 2014a, Uplift of the Central Andes of NW Argentina associated with 59

72 60 upper crustal shortening, revealed by multi-proxy isotopic analyses: Tectonics, p. 2013TC Carrapa, B., Reyes-Bywater, S., Safipour, R., Sobel, E. R., Schoenbohm, L. M., DeCelles, P. G., Reiners, P. W., and Stockli, D., 2014b, The effect of inherited paleotopography on exhumation of the Central Andes of NW Argentina: Geological Society of America Bulletin, v. 126, no. 1-2, p Carrera, N., and Munoz, J. A., 2008, Thrusting evolution in the southern Cordillera Oriental (northern Argentine Andes): Constraints from growth strata: Tectonophysics, v. 459, no. 1-4, p Coutand, I., Carrapa, B., Deeken, A., Schmitt, A. K., Sobel, E. R., and Strecker, M. R., 2006, Propagation of orographic barriers along an active range front: insights from sandstone petrography and detrital apatite fission-track thermochronology in the intramontane Angastaco basin, NW Argentina: Basin Research, v. 18, no. 1, p Crosby, B. T., and Whipple, K. X., 2006, Knickpoint initiation and distribution within fluvial networks: 236 waterfalls in the Waipaoa River, North Island, New Zealand: Geomorphology, v. 82, no. 1-2, p Cyr, A. J., Granger, D. E., Olivetti, V., and Molin, P., 2014, Distinguishing between tectonic and lithologic controls on bedrock channel longitudinal profleiles using cosmogenic Be-10 erosion rates and channel steepness index: Geomorphology, v. 209, p Davis, D., Suppe, J., and Dahlen, F. A., 1983, MECHANICS OF FOLD-AND-THRUST BELTS AND ACCRETIONARY WEDGES: Journal of Geophysical Research, v. 88, no. NB2, p DeCelles, P. G., Carrapa, B., Horton, B. K., and Gehrels, G. E., 2011, Cenozoic foreland basin system in the central Andes of northwestern Argentina: Implications for Andean geodynamics and modes of deformation: Tectonics, v. 30. Do Campo, M., and Guevara, S. R., 2005, Provenance analysis and tectonic setting of late Neoproterozoic metasedimentary successions in NW Argentina: Journal of South American Earth Sciences, v. 19, no. 2, p Dortch, J. M., Dietsch, C., Owen, L. A., Caffee, M. W., and Ruppert, K., 2011, Episodic fluvial incision of rivers and rock uplift in the Himalaya and Transhimalaya: Journal of the Geological Society, v. 168, no. 3, p Francis, P. W., Sparks, R. S. J., Hawkesworth, C. J., Thorpe, R. S., Pyle, D. M., Tait, S. R., Mantovani, M. S., and McDermott, F., 1989, PETROLOGY AND GEOCHEMISTRY OF VOLCANIC-ROCKS OF THE CERRO GALAN CALDERA, NORTHWEST ARGENTINA: Geological Magazine, v. 126, no. 5, p Fritz, S. C., Baker, P. A., Lowenstein, T. K., Seltzer, G. O., Rigsby, C. A., Dwyer, G. S., Tapia, P. M., Arnold, K. K., Ku, T. L., and Luo, S. D., 2004, Hydrologic variation during the 60

73 61 last 170,000 years in the southern hemisphere tropics of South America: Quaternary Research, v. 61, no. 1, p Garzione, C. N., Auerbach, D. J., Jin-Sook Smith, J., Rosario, J. J., Passey, B. H., Jordan, T. E., and Eiler, J. M., 2014, Clumped isotope evidence for diachronous surface cooling of the Altiplano and pulsed surface uplift of the Central Andes: Earth and Planetary Science Letters, v. 393, no. 0, p Garzione, C. N., Hoke, G. D., Libarkin, J. C., Withers, S., MacFadden, B., Eiler, J., Ghosh, P., and Mulch, A., 2008, Rise of the Andes: Science, v. 320, no. 5881, p Gile, L. H., Peterson, F. F., and Grossman, R. B., 1966, MORPHOLOGICAL AND GENETIC SEQUENCES OF CARBONATE ACCUMULATION IN DESERT SOILS: Soil Science, v. 101, no. 5, p. 347-&. González Bonorino, G., and Abascal, L. D., 2012, Drainage and base-level adjustments during evolution of a late Pleistocene piggyback basin, Eastern Cordillera, Central Andes of northwestern Argentina: Geological Society of America Bulletin, v. 124, no , p Gosse, J. C., and Phillips, F. M., 2001, Terrestrial in situ cosmogenic nuclides: theory and application: Quaternary Science Reviews, v. 20, no. 14, p Gosse, J. C., and Stone, O., 2001, Terrestrial cosmogenic nuclide methods passing milestones toward paleo-altimetry: Eos, Transactions, American Geophysical Union, v. 82, no. 7, p. 82, 86, 89. Granger, D. E., Kirchner, J. W., and Finkel, R., 1996, Spatially averaged long-term erosion rates measured from in situ-produced cosmogenic nuclides in alluvial sediments: Journal of Geology, v. 104, no. 3, p Grier, M. E., Salfity, J. A., and Allmendinger, R. W., 1991, ANDEAN REACTIVATION OF THE CRETACEOUS SALTA RIFT, NORTHWESTERN ARGENTINA: Journal of South American Earth Sciences, v. 4, no. 4, p Gutscher, M. A., Spakman, W., Bijwaard, H., and Engdahl, E. R., 2000, Geodynamics of flat subduction: Seismicity and tomographic constraints from the Andean margin: Tectonics, v. 19, no. 5, p Hain, M. P., Strecker, M. R., Bookhagen, B., Alonso, R. N., Pingel, H., and Schmitt, A. K., 2011, Neogene to Quaternary broken foreland formation and sedimentation dynamics in the Andes of NW Argentina (25 degrees S): Tectonics, v. 30. Hancock, G. S., Anderson, R. S., Chadwick, O. A., and Finkel, R. C., 1999, Dating fluvial terraces with 10 Be and 26 Al profiles; application to the Wind River, Wyoming: Geomorphology, v. 27, no. 1-2, p

74 62 Harkins, N., Kirby, E., Heimsath, A., Robinson, R., and Reiser, U., 2007, Transient fluvial incision in the headwaters of the Yellow River, northeastern Tibet, China: Journal of Geophysical Research-Earth Surface, v. 112, no. F3. Haselton, K., Hilley, G., and Strecker, M. R., 2002, Average Pleistocene climatic patterns in the southern central Andes: Controls on mountain glaciation and paleoclimate implications: Journal of Geology, v. 110, no. 2, p Hetzel, R., Niedermann, S., Tao, M. X., Kubik, P. W., and Strecker, M. R., 2006, Climatic versus tectonic control on river incision at the margin of NE Tibet: Be-10 exposure dating of river terraces at the mountain front of the Qilian Shan: Journal of Geophysical Research-Earth Surface, v. 111, no. F3. Hidy, A. J., Gosse, J. C., Pederson, J. L., Mattern, J. P., and Finkel, R. C., 2010, A geologically constrained Monte Carlo approach to modeling exposure ages from profiles of cosmogenic nuclides: An example from Lees Ferry, Arizona: Geochemistry Geophysics Geosystems, v. 11, p. 18. Hilley, G. E., Blisniuk, P. M., and Strecker, M. R., 2005, Mechanics and erosion of basementcored uplift provinces: Journal of Geophysical Research-Solid Earth, v. 110, no. B12. Hilley, G. E., and Coutand, I., 2010, Links between topography, erosion, rheological heterogeneity, and deformation in contractional settings: Insights from the central Andes: Tectonophysics, v. 495, no. 1-2, p Hilley, G. E., and Strecker, M. R., 2005, Processes of oscillatory basin filling and excavation in a tectonically active orogen: Quebrada del Toro Basin, NW Argentina: Geological Society of America Bulletin, v. 117, no. 7-8, p Hongn, F., Papa, C. d., Powell, J., Petrinovic, I., Mon, R., and Deraco, V., 2007, Middle Eocene deformation and sedimentation in the Puna-Eastern Cordillera transition (23-26 S): Control by preexisting heterogeneities on the pattern of initial Andean shortening: Geology, v. 35, no. 3, p Humphrey, N. F., and Konrad, S. K., 2000, River incision or diversion in response to bedrock uplift: Geology, v. 28, no. 1, p Kennett, J. P., 1977, CENOZOIC EVOLUTION OF ANTARCTIC GLACIATION, CIRCUM- ANTARCTIC OCEAN, AND THEIR IMPACT ON GLOBAL PALEOCEANOGRAPHY: Journal of Geophysical Research-Oceans and Atmospheres, v. 82, no. 27, p Kirby, E., and Whipple, K. X., 2012, Expression of active tectonics in erosional landscapes: Journal of Structural Geology, v. 44, p Kirby, E., Whipple, K. X., Tang, W. Q., and Chen, Z. L., 2003, Distribution of active rock uplift along the eastern margin of the Tibetan Plateau: Inferences from bedrock channel longitudinal profiles: Journal of Geophysical Research-Solid Earth, v. 108, no. B4. 62

75 63 Kley, J., and Monaldi, C. R., 1998, Tectonic shortening and crustal thickness in the Central Andes: How good is the correlation?: Geology, v. 26, no. 8, p Kohl, C., and Nishiizumi, K., 1992, Chemical isolation of quartz for measurement of in-situproduced cosmogenic nuclides: Geochimica et Cosmochimica Acta, v. 56, no. 9, p Lal, D., 1991, Cosmic ray labeling of erosion surfaces; in situ nuclide production rates and erosion models: Earth and Planetary Science Letters, v. 104, no. 2-4, p Lanza, F., Tibaldi, A., Bonali, F. L., and Corazzato, C., 2013, Space-time variations of stresses in the Miocene-Quaternary along the Calama-Olacapato-El Toro Fault Zone, Central Andes: Tectonophysics, v. 593, p Marquillas, R. A., del Papa, C., and Sabino, I. F., 2005, Sedimentary aspects and paleoenvironmental evolution of a rift basin: Salta Group (Cretaceous-Paleogene), northwestern Argentina: International Journal of Earth Sciences, v. 94, no. 1, p Marrett, R. A., Allmendinger, R. W., Alonso, R. N., and Drake, R. E., 1994, LATE CENOZOIC TECTONIC EVOLUTION OF THE PUNA PLATEAU AND ADJACENT FORELAND, NORTHWESTERN ARGENTINE ANDES: Journal of South American Earth Sciences, v. 7, no. 2, p Matmon, A., Bierman, P. R., Larsen, J., Southworth, S., Pavich, M., and Caffee, M., 2003, Temporally and spatially uniform rates of erosion in the southern Appalachian Great Smoky Mountains: Geology, v. 31, no. 2, p May, J. H., and Soler, R. D., 2010, Late Quaternary morphodynamics in the Quebrada de Purmamarca, NW Argentina: Quaternary Science Journal, v. 59, no. 1-2, p McQuarrie, N., 2002, Initial plate geometry, shortening variations, and evolution of the Bolivian orocline: Geology, v. 30, no. 10, p Miall, A. D., 2006, Sediments and Basins, Toronto, McGraw-Hill Ryerson Limited, 384 p.: Nishiizumi, K., Imamura, M., Caffee, M. W., Southon, J. R., Finkel, R. C., and McAninch, J., 2007, Absolute calibration of Be-10 AMS standards: Nuclear Instruments & Methods in Physics Research Section B-Beam Interactions with Materials and Atoms, v. 258, no. 2, p Ouimet, W. B., Whipple, K. X., and Granger, D. E., 2009, Beyond threshold hillslopes: Channel adjustment to base-level fall in tectonically active mountain ranges: Geology, v. 37, no. 7, p Pearson, D. M., Kapp, P., Reiners, P. W., Gehrels, G. E., Ducea, M. N., Pullen, A., Otamendi, J. E., and Alonso, R. N., 2012, Major Miocene exhumation by fault-propagation folding within a metamorphosed, early Paleozoic thrust belt: Northwestern Argentina: Tectonics, v

76 64 Portenga, E. W., and Bierman, P. R., 2011, Understanding Earth's eroding surface with 10 Be: GSA Today, v. 21, no. 8, p Raymo, M. E., and Ruddiman, W. F., 1992, TECTONIC FORCING OF LATE CENOZOIC CLIMATE: Nature, v. 359, no. 6391, p Reheis, M. C., 1987, Gypsic Soils on the Kane Alluvial Fans, Big Horn County, Wyoming: United States Geological Survey Bulletin, v. Report 1590-C, p. C1-C39. Reheis, M. C., Goodmacher, J. C., Harden, J. W., McFadden, L. D., Rockwell, T. K., Shroba, R. R., Sowers, J. M., and Taylor, E. M., 1995, QUATERNARY SOILS AND DUST DEPOSITION IN SOUTHERN NEVADA AND CALIFORNIA: Geological Society of America Bulletin, v. 107, no. 9, p Royer, D. L., 1999, Depth to pedogenic carbonate horizon as a paleoprecipitation indicator?: Geology, v. 27, no. 12, p Safran, E. B., Bierman, P. R., Aalto, R., Dunne, T., Whipple, K. X., and Caffee, M., 2005, Erosion rates driven by channel network incision in the Bolivian Andes: Earth Surface Processes and Landforms, v. 30, no. 8, p Santimano, T., and Riller, U., 2012, Kinematics of Tertiary to Quaternary intracontinental deformation of upper crust in the Eastern Cordillera, southern Central Andes, NW Argentina: Tectonics, v. 31. Schlunegger, F., Norton, K. P., and Zeilinger, G., 2011, Climatic Forcing on Channel Profiles in the Eastern Cordillera of the Coroico Region, Bolivia: Journal of Geology, v. 119, no. 1, p Schmidt, S., Hetzel, R., Kuhlmann, J., Mingorance, F., and Ramos, V. A., 2011, A note of caution on the use of boulders for exposure dating of depositional surfaces: Earth and Planetary Science Letters, v. 302, no. 1-2, p Schoenbohm, L. M., and Strecker, M. R., 2009, Normal faulting along the southern margin of the Puna Plateau, northwest Argentina: Tectonics, v. 28. Schoenbohm, L. M., Whipple, K. X., Burchfiel, B. C., and Chen, L., 2004, Geomorphic constraints on surface uplift, exhumation, and plateau growth in the Red River region, Yunnan Province, China: Geological Society of America Bulletin, v. 116, no. 7-8, p Seeber, L., and Gornitz, V., 1983, RIVER PROFILES ALONG THE HIMALAYAN ARC AS INDICATORS OF ACTIVE TECTONICS: Tectonophysics, v. 92, no. 4, p Sobel, E. R., Hilley, G. E., and Strecker, M. R., 2003, Formation of internally drained contractional basins by aridity-limited bedrock incision: Journal of Geophysical Research-Solid Earth, v. 108, no. B7. 64

77 65 Sobel, E. R., and Strecker, M. R., 2003, Uplift, exhumation and precipitation: tectonic and climatic control of Late Cenozoic landscape evolution in the northern Sierras Pampeanas, Argentina: Basin Research, v. 15, no. 4, p Sparks, R. S. J., Francis, P. W., Hamer, R. D., Pankhurst, R. J., Ocallaghan, L. O., Thorpe, R. S., and Page, R., 1985, IGNIMBRITES OF THE CERRO GALAN CALDERA, NW ARGENTINA: Journal of Volcanology and Geothermal Research, v. 24, no. 3-4, p Staff, S. S., 2010, Keys to Soil Taxonomy, 11th ed., in USDA-NRCS, ed.: Washington, DC. Strecker, M. R., Alonso, R. N., Bookhagen, B., Carrapa, B., Hilley, G. E., Sobel, E. R., and Trauth, M. H., 2007, Tectonics and climate of the southern central Andes, Annual review of Earth and Planetary Sciences, Volume 35: Palo Alto, Annual Reviews, p Strecker, M. R., Cerveny, P., Bloom, A. L., and Malizia, D., 1989, LATE CENOZOIC TECTONISM AND LANDSCAPE DEVELOPMENT IN THE FORELAND OF THE ANDES - NORTHERN SIERRAS PAMPEANAS (26-DEGREES-28-DEGREES-S), ARGENTINA: Tectonics, v. 8, no. 3, p Strecker, M. R., Hilley, G. E., Bookhagen, B., and Sobel, E. R., 2011, Structural, geomorphic and depositional characteristics of contiguous and broken foreland basins: Examples from the eastern flanks of the central Andes in Bolivia and NW Argentina, in Busby, C., and Azor, A., eds., Tectonics of Sedimentary Basins: Recent Advances: Cambridge, MA, Blackwell, p Tassara, A., Götze, H.-J., Schmidt, S., and Hackney, R., 2006, Three-dimensional density model of the Nazca plate and the Andean continental margin: Journal of Geophysical Research: Solid Earth, v. 111, no. B9, p. B Trauth, M. H., Bookhagen, B., Marwan, N., and Strecker, M. R., 2003a, Multiple landslide clusters record Quaternary climate changes in the northwestern Argentine Andes: Palaeogeography Palaeoclimatology Palaeoecology, v. 194, no. 1-3, p Trauth, M. H., Bookhagen, B., Muller, A. B., and Strecker, M. R., 2003b, Late pleistocene climate change and erosion in the Santa Maria Basin, NW Argentina: Journal of Sedimentary Research, v. 73, no. 1, p Vassallo, R., Ritz, J. F., Braucher, R., and Carretier, S., 2005, Dating faulted alluvial fans with cosmogenic Be-10 in the Gurvan Bogd mountain range (Gobi-Altay, Mongolia): climatic and tectonic implications: Terra Nova, v. 17, no. 3, p von Blanckenburg, F., 2005, The control mechanisms of erosion and weathering at basin scale from cosmogenic nuclides in river sediment: Earth and Planetary Science Letters, v. 237, no. 3-4, p Whipple, K. X., 2001, Fluvial landscape response time: How plausible is steady-state denudation?: American Journal of Science, v. 301, no. 4-5, p

78 66 -, 2004, Bedrock rivers and the geomorphology of active orogens: Annual review of Earth and Planetary Sciences, v. 32, p , 2009, The influence of climate on the tectonic evolution of mountain belts: Nature Geoscience, v. 2, no. 2, p Whipple, K. X., DiBiase, R. A., and Crosby, B. T., 2013, 9.28 Bedrock Rivers, in Shroder, J. F., ed., Treatise on Geomorphology: San Diego, Academic Press, p Whipple, K. X., and Tucker, G. E., 1999, Dynamics of the stream-power river incision model: Implications for height limits of mountain ranges, landscape response timescales, and research needs: Journal of Geophysical Research-Solid Earth, v. 104, no. B8, p , 2002, Implications of sediment-flux-dependent river incision models for landscape evolution: Journal of Geophysical Research-Solid Earth, v. 107, no. B2. Willenbring, J. K., Gasparini, N. M., Crosby, B. T., and Brocard, G., 2013, What does a mean mean? The temporal evolution of detrital cosmogenic denudation rates in a transient landscape: Geology, v. 41, no. 12, p Willett, S. D., 1999, Orogeny and orography: The effects of erosion on the structure of mountain belts: Journal of Geophysical Research: Solid Earth, v. 104, no. B12, p Wobus, C., Whipple, K. X., Kirby, E., Snyder, N., Johnson, J., Spyropolou, K., Crosby, B., and Sheehan, D., 2006a, Tectonics from topography: Procedures, promise, and pitfalls: Tectonics, Climate, and Landscape Evolution, v. 398, p Wobus, C. W., Crosby, B. T., and Whipple, K. X., 2006b, Hanging valleys in fluvial systems: Controls on occurrence and implications for landscape evolution: Journal of Geophysical Research-Earth Surface, v. 111, no. F2. Wobus, C. W., Whipple, K. X., and Hodges, K. V., 2006c, Neotectonics of the central Nepalese Himalaya: Constraints from geomorphology, detrital 40Ar/39Ar thermochronology, and thermal modeling: Tectonics, v. 25, no. 4, p. TC4011. Zhou, R. J., Schoenbohm, L. M., and Cosca, M., 2013, Recent, slow normal and strike-slip faulting in the Pasto Ventura region of the southern Puna Plateau, NW Argentina: Tectonics, v. 32, no. 1, p

79 67 Appendices Appendix A: Supporting Information for Field Studies Appendix A includes a description of desert pavement development index (PDI) methodology, PDI data, annotated photographs of depth profiles, and photographs of differing degrees of soil carbonate development. 1. Desert Pavement Methodology and Results 1.1 Methods We determined desert pavement indices according to the method developed by Al-Farraj and Harvey (2000). At each site of desert pavement analysis, we took nine photographs in gridfashion looking vertically down on the surface from a height of approximately 1m. Later these photographs were digitally merged using Adobe Photoshop and a 3X3 grid of arbitrary dimensions was placed atop the merged photo. At each grid intersection, we measured the multiple criteria of Al-Farraj and Harvey (2000). Thus each individual criterion is a mean of nine values, with exception of those criteria that describe the overall surface, in which case only one value is assigned. The mean values of the seven individual criteria are then averaged to yield a pavement development index (PDI). 1.2 PDI Results. See Figure 11 for locations. Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief Final Score 67

80 68 Averages Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief Final Score Averages Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief Final Score Averages Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief

81 Final Score Averages Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief Final Score Averages Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief Final Score Averages

82 70 Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief Final Score Averages Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief Final Score Averages Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief

83 Final Score Averages Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief Final Score Averages Pit Number Grid Intersection Relative clast size Sorting Angularity Clast Fracturing Depositional Fabrics Interlocking Clast relief Final Score Averages

84 72 2. Annotated Depth Profile Photographs Here we include annotated photographs of each depth profile referred to in the text. For AR13-02 and AR13-03, we merged two photographs using Adobe Photoshop software, so images are distorted and the scale increases towards the bottom of each pit. We sampled every 50 cm, marked by the red lines in each photo. We also include two photographs which detail the differing degree of soil carbonate development on Q2 (AR13-01) and Q5 (AR13-02) surfaces. 72

85 73 73

86 74 74

87 75 Soil pit descriptions from the Pucará Valley. Locations shown in Figure 11. Pit # Pit Type Pit Depth, cm Surface level Soil Type Pedogenic Salt Stage PDI 1 Soil 100 Q0 Ustic Haplocambid 0 n/a 2 Soil 50 Q1 Ustic Haplocambid Cosmo 200 Q2 Ustic Haplocalcid Soil 20 Q3 Ustic Haplocalcid Soil 50 Q3 Ustic Petrocalcid Soil 50 Q3 Ustic Haplocalcid Pavement 0 Q4 n/a n/a Soil 25 Q4 Ustic Haplocambid Cosmo 200 Q5 Ustic Petrocalcid Cosmo 195 Q6 Leptic Haplogypsid Pavement 0 Q6 n/a n/a Soil 25 Q6 Ustic Petrocalcid Soil 80 Q6 Ustic Petrogypsid 3.5 n/a 75

88 76 Appendix B: Geologic Map of the Pucará Valley Lithology and structure compiled from (2008) and Coutand et al. (2006). Topographic base map derived from ASTER 30 m digital elevation model. Strike and dips and updated structure are all from this study. 76

89 77 Appendix C: 10 Be Analytical Results 77

90 78 Appendix D: Supporting Information for 10 Be Depth Profiles Be Analytical Procedures Samples were processed at the University of Vermont Cosmogenic Nuclide Laboratory using standard analytical methods. Quartz was purified for 10 Be analysis using mineral separation procedures modified from Kohl and Nishiizumi (1992). Material was sieved to isolate the 250 to 850 µm grainsize and magnetically separated. Samples were ultrasonically etched in hot 6N HCl, washed, and then etched at least three more times in hot dilute (1%) HF-HNO 3 to preferentially dissolve all grains except quartz. Samples were then etched in (0.5%) HF-HNO 3 for 7 to 10 days. Quartz was tested for purity by inductively coupled plasma - optical emission spectroscopy, and additional HF-HNO 3 etches were performed until desired purity levels were reached. For Beryllium isolation, samples were prepared in batches that contained a full-process blank and 11 unknowns. See for detailed methods. We used between 11.6 and 23.0 g of purified quartz for analysis. We added ~250 µg of 9 Be carrier made from beryl at the University of Vermont to each sample and dissolved samples in 120 g of hot, concentrated HF. After dissolution and HF evaporation, samples were treated with HClO 4, and then HCl. We removed Fe in anion exchange columns and removed Ti, Be, Al, and B in cation exchange columns. Be was precipitated at ph 8 as hydroxide gel, dried, ignited to produce BeO, ground, and packed into copper cathodes with Nb powder at 1:1 molar ratio for accelerator mass spectrometry (AMS) measurements. 10 Be/ 9 Be ratios were measured at the Scottish University Environmental Research Center and were normalized to NIST standard with an assumed ratio of based on a half life of 1.36 My. The average measured sample ratio ( 10 Be/ 9 Be) was 947 x and AMS measurement precisions, including blank corrections propagated quadratically, averaged 1.9 %. The full process blank associated with the three batches in which these samples were run was 4.24±1.54 x Because all samples received similar amounts of carrier, the blank ratio was subtracted from the measured ratio for the sample. The blank correction is an inconsequential part of most measured isotopic ratios (<0.7% on average, maximum 2.0%). The CRONUS N standard was run with these samples and returned a concentration of 2.31±0.06 x 10 5 atoms g -1, consistent with values reported by other labs. 78

91 79 2. Depth-profile Input Parameters 2.1. JM-AR13-01 Sample Depth (cm) Sample Thickness (cm) 10Be Concentration (atoms/g) Uncertainty (%) Density (g/cm^3) JM-AR13-02 Sample Depth (cm) Sample Thickness (cm) 10Be Concentration (atoms/g) Density Depth (cm) Density Uncertainty (g/cm^3) Uncertainty (%) Density (g/cm^3) JM-AR13-03 Sample Depth (cm) Sample Thickness (cm) 10Be Concentration (atoms/g) Uncertainty (%) Density (g/cm^3) Density Uncertainty (g/cm^3) Density Uncertainty (g/cm^3) 79

92 80 3. Depth Profile User Interfaces 3.1. JM-AR

93 JM-AR

94 JM-AR

95 83 4. Depth Profile Solutions: Frequency Histograms 4.1. JM-AR

96 JM-AR

97 JM-AR