全球变化知识资源中心

The occurrence of Holocene changes in the South American Summer Monsoon (SASM) in the Peruvian Andes has been well established in paleoclimatological records such as speleothem, lake and glacier records. How river systems responded to these events has, however, hardly been investigated. Here, we present evidence based on sedimentological, stratigraphical and geomorphological data as well as radiometric (optically stimulated luminescence and radiocarbon) dating that rivers from the Peruvian Andes are extremely sensitive to changes in SASM activity. The presence of lake sediments shows that from at least 12,000 years ago until 4153 +/- 988 years ago a lake was present in the Mantaro River valley, possibly due to damming by a glacier or glacial landforms. A reconstruction of fluvial terrace profiles shows that 4418 +/- 500 years ago the Mantaro and its tributary, the Cunas River, incised and laid down sediments simultaneously as a response to changes in regional base-level and increased SASM activity. The latter was largely subdued during a large part of the Holocene as evidenced by paleoclimatological records in the region. Between 2245 +/- 217 years ago and the present, the frequency of SASM events increased drastically and both rivers formed the majority of their fluvial terraces. In total, over the past 4000 years, the Mantaro River formed five terrace levels and the Cunas River formed seven main terraces. Locally, 11 terrace levels were recognised. Terrace formation occurred at intervals of approximately 250 to 300 years between 2245 +/- 217 and 1188 +/- 60 years ago and approximately every 150 years after 824 +/- 66 years ago until the present. A comparison with paleoclimatological data shows that sedimentation events correlate to periods of increased precipitation and glacier retreat in the Peruvian Andes, whereas phases of incision are attributed to continuous adjustments in base-level fall. Thirty-four metres of incision has occurred since 4418 +/- 500 years ago averaging 7.7 mm yr(-1). A comparison with data from other river systems in the Peruvian Andes shows that many rivers responded in a similar way to centennial-scale variations in SASM activity. Fluvial activity is thus not related to interannual variations in the El Nino Southern Oscillation (ENSO) contrary to previous proposals.