Erosion of an active fault scarp leads to drainage capture in the Amazon region, Brazil Pedro Val, 1 * Clauzionor Silva, 2 David Harbor, 3 Norberto Morales, 4 Felipe Amaral 2 and Tiago Maia 2 1 Department of Earth Sciences, Syracuse University, Syracuse, NY 13244, USA 2 Federal University of Amazonas (UFAM), Geosciences Department, Manaus, AM, Brazil 3 Washington and Lee University, Geology Department, Lexington, VA, USA 4 São Paulo State University, UNESP, Rio Claro, SP, Brazil Received 21 October 2012; Revised 11 November 2013; Accepted 12 November 2013 *Correspondence to: Pedro Val, Department of Earth Sciences, Syracuse University, Syracuse, NY 13244, USA. E-mail: pfval@syr.edu ABSTRACT: Far from the continental margin, drainage basins in Central Amazonia should be in topographic steady state; but they are not. Abandoned remnant fluvial valleys up to hundreds of square kilometers in size are observed throughout Amazonia, and are evi- dence of significant landscape reorganization. While major Late Miocene drainage shifts occurred due to initiation of the transcontinen- tal Amazon River, local landscape change has remained active until today. Driven either by dynamic topography, tectonism, and/or climatic fluctuations, drainage captures in Amazonia provide a natural experiment for assessing the geomorphic response of low- slope basins to sudden, capture related base-level falls. This paper evaluates the timing of geomorphic change by examining a drainage capture event across the Baependi fault scarp involving the Cuieiras and Tarumã-Mirim River basins northwest of the city of Manaus in Brazil. A system of capture-related knickpoints was generated by base-level fall following drainage capture; through numerical modeling of their initiation and propagation, the capture event is inferred to have occurred between the middle and late Pleistocene, consistent with other studies of landscape change in surrounding areas. In low-slope settings like the Amazon River basin, base-level fall can increase erosion rates by more than an order of magnitude, and moderate to large river basins can respond to episodes of base-level fall over timescales of tens to hundreds of thousands of years. Copyright © 2013 John Wiley & Sons, Ltd. KEYWORDS: Amazonia; response time; drainage capture; Pleistocene; Baependi fault scarp Introduction An increasing number of studies in the Amazon River basin re- veal remarkable evidence of landscape changes over scales of hundreds of square kilometers (Rossetti and Góes, 2008; Mantelli et al., 2009; Hayakawa et al., 2010). In Central Amazonia, several studies suggest that such changes are driven by Quaternary tectonism (Franzinelli and Igreja, 2002; Rossetti et al., 2005; Almeida-Filho and Miranda, 2007; Silva et al., 2007), but stop short in constraining timing of deformation and its connection to landscape change. Shephard et al. (2010) suggested that intra-continental deformation occurs through the reactivation of old structures by mantle-convection driven dynamic topography. This caused a Late Miocene rever- sal of the Amazon River drainage system, and resulted in the development of the modern eastward-flowing transcontinental Amazon River system we see today (Shephard et al., 2010). These studies show that parts of the Amazon’s landscape might not be in steady state, despite being in a low relief cratonic interior. Furthermore, these studies show that tectonism and dynamic topography in the last 6 Ma drove landscape change across Amazonia. Constraining the nature of geomorphic pro- cesses, and the age and response time of landscape change in Amazonia is central to the recognition of topographic equilib- rium state, and to mechanisms of its perturbation. Given dense vegetation and difficult access, acquisition of new field data in Amazonia is challenging. Consequently, most studies of landscape change rely solely on remotely sensed data, including digital elevation models (DEMs) to map linea- ments as indicators of structural trends. Thus, resolving which processes drive landscape change are unconvincing in the ab- sence of measured amounts and timing of tectonic events. For example, interpretation of lineaments and abandoned channel valleys seen in satellite images led Silva et al. (2007) and Almeida-Filho and Miranda (2007) to suggest that, during the Quaternary, the lower Rio Negro once occupied at least two different positions in what is now the Negro-Solimões River interfluve (Figure 1). While these interpretations provide an important understanding of the modern landscape, such an ap- proach does not allow for a determination of timing nor quan- titative assessment of the geomorphic processes that are acting to shape it. As our main objective is to better constrain the mechanisms of drainage reorganization in this area, it is neces- sary to investigate other landscape elements. Analysis of river profiles is an ideal approach to quantifying geomorphic processes, especially where access to the field is limited (Snyder et al., 2000; Bishop et al., 2005; Harbor et al., 2005; Hu et al., 2010). Rate of relief generation governs the pace at which rivers incise, and thus governs how landscapes evolve through time. Therefore, changes in relief, whether by EARTH SURFACE PROCESSES AND LANDFORMS Earth Surf. Process. Landforms (2013) Copyright © 2013 John Wiley & Sons, Ltd. Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/esp.3507