Vignettes > Stream response to Climate Change, Atacama Desert, Chile

Stream response to Climate Change, Atacama Desert, Chile

Jason A Rech, Craig Tully, Claudio Latorre
Miami University, Ohio; Pontificia Universidad Católica de Chile
Author Profile

Shortcut URL:


Continent: South America
Country: Chile
UTM coordinates and datum: none


Climate Setting: Arid
Tectonic setting: Continental Arc
Type: Process, Stratigraphy, Chronology

Figure 1 Location of the Atacama Desert, Chile Details

Figure 2 An example of a perennial stream in the Atacama Desert with in-stream wetlands. Details

Figure 3 Rodent midden with inset photographs of (a)plant macrofossils and (b) the rodent Phyllotis (leaf-eared mouse) who makes middens in the Atacama. Details

Figure 4 Fluvial terraces along Rio San Salvador used to changes in channel height over time. Inset diplays organic units with plant macrofossils used for radiocarbon dating. Details

Figure 5 Comparison between changes in paleo-precipitation and stream channel morphology (figure adapted from Rech et al., 2003). Precipitation is reconstructed from rodent middens (Latorre et al., 2002; 2006). Periods of stream aggradation and incision are derived from radiocarbon dating of fluvial terraces and mapping of fluvial stratigraphy. Details


Climate on earth is constantly changing. Earth's climate can change gradually over millions of years (tectonic-scale) due to changes in greenhouse gases or the slow movement of tectonic plates, or climate can change periodically over tens of thousands of years as a result of slight changes in earth's orbit (orbital-scale climate change). Clear evidence for dramatic orbital-scale climate change includes the large ice sheets that covered the northern portions of North America and Europe just 20,000 years ago. Climate, however, can also change on much shorter time scales, over decades, centuries, or millennia, for a wide variety of reasons.

As you would expect, the rate and nature of geomorphic processes also respond to climate change. For example, if an arid region becomes wetter, the rates of chemical weathering may increase or flash floods may occur more frequently. Although most geomorphic systems will respond to changes in climate, it can be difficult sometimes to predict how a specific system, or parts of a system, will respond. This is important to better understand and predict how geomorphic processes and associated systems will respond to future climate change.

In this study, we examined how perennial streams in the Atacama Desert have responded to short term climate changes (centuries to millennia) during the last 10,000 years (the Holocene Epoch). The Atacama Desert, situated between the Andes Mountains and Pacific Ocean in northern Chile, is probably the driest and oldest desert on earth (Figure 1). Large areas devoid of any vegetation recieve <3 cm of rainfall per year, and many areas only receive a rainfall event one or two times a decade! Some streams, however, have perennial stream flow as a result of high levels of ground water that is sourced from snow melt and precipitation in the Andes. These streams, however, are different than most in that although perennial, they only experience a few stream discharge events per year from precipitation falling on the western flank of the Andes. As a result, the beds of these streams are covered by dense wetland vegetation. These are called in-stream wetlands (Figure 2).

In particular, we were interested in whether these streams cut down into their stream bed (incised), or accumulated sediment (aggrade) during periods of wetter climate. Therefore, we needed an independent measure of precipitation in this region over the last 10,000 years. We decided to use fossil vegetation preserved in rodent middens to reconstruct past precipitation (Latorre et al., 2002). In arid lands, small rodents build nests using the surrounding local vegetation. These nests become cemented with the rodent's urine over time. These urine-encased nests become time capsules of past plant communities, the age of which can be determined by radiocarbon dating (Figure 3). When collected along the lower elevation limit of vegetation in the Atacama (which is controlled primarily by precipitation with temperature only having a minor influence), these vegetation assemblages can be used to reconstruct past precipitation amounts (Latorre et al. 2002).

We also needed to identify when streams in the Atacama incised into their stream beds and when they aggraded. We can determine when streams incised into their streambeds and when they aggraded by mapping and dating the fluvial terraces within these stream valleys (Figure 4). Fluvial terraces represent old, inactive floodplains of a stream. However, these fluvial landforms are notoriously difficult to date. We are fortunate that the terrace deposits of in-stream wetlands contain an abundance of terrestrial plant fossils (Figure 4 inset). By radiocarbon dating these plant fossils, and mapping the stratigraphy of the fluvial terrace deposits, we can reconstruct the history of incision and aggradation of several stream systems in the Atacama over the last 10,000 years (Rech et al. 2002; 2003). If climate change (paleoprecipitation) is really the cause of aggradation and incision in these stream systems, then we should be able to reproduce a similar history of stream incision in several different stream systems.

When we combined our records of precipitation and stream incision, we see that streams generally accumulate sediments during wetter periods, and cut into their beds during dry periods. Is this what you expected? Why?

Our current understanding of the climatic influence on these streams is that the dominant control on aggradation/incision is the resistance of the stream bed to erosion and not stream power (flow). When the water table is high perennial flow is present in the stream, the stream bed is armored by dense vegetation. However, during periods of drier climate, the water table drops due to reduced groundwater recharge in the Andes. Consequently, plants die and the in-stream wetland sediments, which are mostly silt, become extremely vulnerable to erosion and incision. After the stream has adjusted to the new (lower) water table, it will slowly accumulate sediments and aggrade until the next major drought and lowering of the water table. We now believe that we can use these stream deposits to reconstruct the history of major droughts in the area, and possibly identify the possible changes in ocean and atmospheric circulation that are responsible for causing major droughts in the Atacama.

It is important to note that not all streams will respond directly to changes in climate, or that they may respond very differently to climate than perennial streams with in-stream wetlands in the Atacama. In other regions, the response of a stream may be influenced primarily by land use changes, or tectonics. Or, streams may continuously be aggrading and then incising to maintain equilibrium and adjust to fluctuations in sediment load.

Associated References

  • Betancourt, Julio L., Van Devender, Thomas R., Martin, Paul S., 1990, Packrat middens: the last 40,000 years of biotic change, The University of Arizona Press, p. 467.
  • Latorre, Claudio, Betancourt, Julio L., Rylander, Kate A., Quade, Jay, Vegetation invasions into absolute desert: A 45,000 yr rodent midden record from the Calama-Salar de Atacama Basins, northern Chile (lat 22°-24°S), Geological Society of America Bulletin, 114(3):349-366
  • Latorre, Claudio, Betancourt, Julio L., Rylander, Arroyo, Mary T.K., 2006, Late Quaternary Vegetation and climate history of a perennial stream canyonin the Rio Salado Basin (22S) of northern Chile, Quaternary Research 65:450-466.
  • Rech, Jason A, Quade, Jay, and Betancourt, Julio L., 2002, Late Quaternary Paleohydrology of the Central Atacama Desert, Chile (22°-24°S), Geological Society of America Bulletin, 114(3):334-348.
  • Rech, Jason A., Pigati, Jeffrey S., Quade, Jay, and Betancourt, Julio L., 2003, Re-evaluation of Holocene deposits at Quebrada Puripica, northern Chile, Palaeogeography, Palaeoclimatology, Palaeoecology, 194:207-222.