sign in sign up
<<
Home , Atacama Desert, Gypsisol

Neogene climate change and uplift in the Atacama ,

Jason A. Rech ⎤ ⎥ Department of Geology, Miami University, Oxford, Ohio 45056, USA Brian S. Currie ⎦ Greg Michalski Department of Earth and Atmospheric Sciences, Purdue University, West Lafayette, Indiana 47907, USA Angela M. Cowan Department of Geology, Miami University, Oxford, Ohio 45056, USA

ABSTRACT elevations above ϳ2800 m, but do not cause The relationship between Andean uplift and extreme desiccation of the west coast of rainfall in the central Atacama. is important for understanding the interplay between climate and tectonics The Calama Basin is located on the eastern in the Central , yet it is poorly understood. Here we use morphological char- margin of the Atacama, ϳ150 km from the acteristics, salt chemistry, and mass independent fractionation anomalies (⌬17O values) in Pacific Coast at elevations between 2200 and dated to reconstruct a middle Miocene climatic transition from semiaridity to 3500 m (Fig. 1). Precipitation in the center of extreme hyperaridity in the Atacama Desert. Paleosols along the southeastern margin of the basin (2200 m) is ϳ4 mm/yr, whereas the Calama Basin change from calcic with root traces, slickensides, and gleyed along the eastern margin (3350 m) precipita- horizons to an extremely mature salic with pedogenic nitrate. We interpret this tion is ϳ50 mm/yr. transition, which occurred between 19 and 13 Ma, to represent a change in precipitation from Ͼ200 mm/yr to Ͻ20 mm/yr. This drastic reduction in precipitation likely resulted PALEOSOLS IN THE CALAMA BASIN from uplift of the Central Andes to elevations Ͼ2 km; the uplift blocked moisture from We examined Miocene strata and Quater- the South American summer monsoon from entering the Atacama. The mid-Miocene Gyp- nary landforms along the southeastern margin sisol with pedogenic nitrate is located at elevations between 2900 and 3400 m in the Cal- of the Calama Basin for evidence of pedogen- ama Basin, significantly higher than modern nitrate , which occur below ϳ2500 m. esis. Miocene gypcretes were first reported in Modern and soils in this elevation zone contain soil carbonate and lack ped- this by Hartley and May (1998). We ogenic gypsum and nitrate. We infer that Ͼ900 m of local surface uplift over the past 10 identified Miocene and Quaternary paleosols m.y. displaced these nitrate paleosols relative to modern nitrate soils and caused a return developed on substrates of alluvial fan and to wetter conditions in the Calama Basin by decreasing local air temperatures and creating flood-plain deposits, and basement bedrock. an orographic barrier to Pacific air masses.

Keywords: Atacama Desert, Andes, paleosols, Calama Basin, soil nitrate.

INTRODUCTION maintained through the existence of a strong The extreme aridity of the Atacama Desert orographic , the mid-Miocene ini- in northern Chile has been associated with the tiation of hyperaridity has implications for the uplift of the Central Andes and the subsequent uplift history of the Andes. blocking of moisture from the Amazon during the mid-Miocene (e.g., Alpers and Brimhall, STUDY AREA 1988; Sillitoe and McKee, 1996). Recent pa- The Atacama Desert is between the central leoclimatic studies from the Atacama, how- Andes and Pacific Ocean in northern Chile ever, have questioned this relationship. Hart- (Fig. 1). Several factors produce the extreme ley and Chong (2002) suggested a late hyperaridity of the Atacama today, including Pliocene age for the initiation of hyperaridity the Andean rain-shadow effect, the coastal and argued that desiccation of the west coast of South America was unrelated to Andean temperature inversion, and the latitudinal po- uplift. Dunai et al. (2005) proposed that hy- sition of this region (Houston and Hartley, peraridity commenced during the late Oligo- 2003). The hyperarid core of the Atacama re- Ͻ cene (ca. 25 Ma) and that this aridity could ceives 3 mm/yr precipitation, does not sup- have been a significant contributor to the re- port vascular plants, and contains soils with gional uplift in the Cordillera, as proposed by high concentrations of soluble salts, including Lamb and Davis (2003). Hoke et al. (2004), sulfates, chlorides, and nitrates. We use the however, identified a major shift toward more term extreme hyperaridity to refer to compa- arid conditions ca. 10 Ma and suggested that rable climatic conditions in the geologic desiccation at that time was related to the up- record. lift of the . In order to deter- Occasional precipitation events in the Ata- mine the relation and potential feedbacks be- cama generally result from Pacific air masses tween Central Andean uplift and climate that migrate northward from the westerly pre- change, we need records of Neogene precipi- cipitation belt. Along the eastern margin of the Figure 1. Location of Atacama Desert, Cal- tation and elevation. Here we present evidence Atacama (ϳ2500 m), precipitation is Ͼ20 ama Basin, and soil nitrate mines in north- from paleosols in the Calama Basin for ex- mm/yr and is associated with the South Amer- ern Chile. Red circles (1–5) identify loca- ican summer monsoon (SASM). SASM air tions of stratigraphic sections (Fig. 2). treme hyperaridity by ca. 12 Ma along the Shaded relief digital elevation model was eastern margin of the Atacama Desert. As ex- masses spill over the central Andes and gen- created by NASA/JPL/NIMA with data from treme hyperaridity in the Atacama today is erate precipitation on the eastern Atacama at Shuttle Radar Topography Mission.

᭧ 2006 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected]. Geology; September 2006; v. 34; no. 9; p. 761–764; doi: 10.1130/G22444.1; 3 figures; Data Repository item 2006159. 761 cm across and 2 m deep. These fractures are generally filled with eolian and as well as salts, and are similar to fractures in modern Atacama salic soils (Ericksen, 1981). Petrographic analysis of the Bym horizon shows a complex history of precipitation and dissolution of pedogenic salts, also analogous to modern soils in the Atacama. Between 1.5 and 3.5 m depth in the salic Gypsisol, and in some locations superimposed on top of older paleosols, is a salic horizon that contains as much as 2.5% NO3, 0.5% Cl, and trace amounts of (Fig. 3). Soils on top of Quaternary landforms along the southeastern margin of the Calama Basin are that contain pronounced calcic or petrocalcic horizons (stage II to stage IV de- velopment within ϳ25 cm of the surface), but lack argillic, gypsic, and salic horizons. These soils generally occur on the surface of alluvial Figure 2. Stratigraphic sections of Neogene paleosols from southeastern margin of Calama fans and fluvial terraces. Basin. Bym—petrogypsic horizon.

DISCUSSION All paleosols were classified using the system oxides and organic matter. At the Rı´o Seco Calcic Vertisols with gleyed horizons and of Mack et al. (1993). Four classes of paleo- locality, calcic Argillisols occur in a gravelly root traces form in poorly drained, vegetated sols are present. Middle-upper Miocene strata . These paleosols have diffuse flood plains with seasonal precipitation. These contain calcic Vertisols and calcic Argillisols, calcic horizons (stage I) and argillic horizons soils do not occur in the Atacama today, but which are overlain by a thick, well-developed composed of montmorillonite and illite. are present to the south in central Chile (ϳlat salic Gypsisol in all locations examined. Cal- A well-developed salic Gypsisol, hereafter 28Њ–34ЊS). The calcic Argillisols, with a cisols are present within Quaternary deposits defined as the Barros Arana geosol, is pre- coarse-grained parent material, form in basin and geomorphic surfaces. served across the southeastern margin of the margin environments (alluvial fan and baja- The oldest paleosols we investigated are in- Calama Basin (Figs. 1 and 2). This das) where the water table is lower. terbedded in the lower-middle Miocene El Loa formed on top of basement bedrock (Paleozoic The Barros Arana geosol formed across the Formation (ca. 20 Ma). These paleosols are volcanics) at the El Hotel and Agua de la Teca southeastern margin of the Calama Basin, on exposed at the Barros Arana and Rı´o Seco lo- localities, and in Miocene alluvial fan deposits top of pediments, alluvial fans, and bajadas. Ͼ calities (Fig. 2) and consist of calcic Vertisols at other localities (Fig. 2). The Gypsisol is 3 This paleosol was originally described by and calcic Argillisols, respectively. The calcic m thick and contains a 1.5–6-m-thick petro- Hartley and May (1998, p. 361), who inter- ϳ Vertisols are 1 m thick and contain red ar- gypsic horizon (Bym). The petrogypsic hori- preted it as ‘‘a subsurface crust formed by a Ͼ 2 gillic horizons with montmorillonite and illite zon has a high bulk density, 2.0 g/cm , and combination of hydromorphic and illuvial pro- ϳ . Argillic horizons have angular contains 10–45 wt% SO4,or 20%–90% cesses and subject to periodic exhumation and blocky structure with clay skins, slickensides, gypsum (Fig. 3). Sulfate concentrations are .’’ This interpretation was based and pseudoanticlinal fractures. These soils highest at the top of the paleosol profile, primarily on evidence of surficial exposure also have soil carbonate (stage II; 1–3 cm nod- where clasts commonly float in a gypsum ma- (salt fractures) and hydromorphic precipitation ules), gleyed mottling and gleyed horizons (5– trix. The paleosol has large v-shaped salt frac- of gypsum (poikilitic textures). Poikilitic tex- 15 cm thick), and root traces with manganese tures, or dikes, that are as much as 35 tures, however, are common in salic soils in the Atacama, most of which form in with low water tables and are not influenced by hydromorphic processes. Poikilitic textures in these soils likely result from dissolution and recrystallization of soil salts produced by me- teoric waters that infiltrate and evaporate in pore spaces and fractures. These pedogenic salt textures, although recognized in previous studies, have caused confusion and debate concerning the origin and genesis of Atacama nitrate soils (e.g., Searl and Rankin, 1993; Er- icksen, 1994). We interpret the formation of the Barros Arana geosol to be comparable to modern salic soils in the Atacama (Ericksen, 1981). These soils generally occur on stable Figure 3. Anion concentrations with depth of soluble salts from top of Barros Arana geosol landscape surfaces such as alluvial fans and from the five measured localities identified in Figure 2. pediments along basin margins, similar to the

762 GEOLOGY, September 2006 interpreted paleolandscape setting of the Bar- ilar to values observed at depth in the Barros for relict geomorphic landforms in the Ata- ros Arana geosol. The majority of regions Arana geosol (0.4%–2.4%; Fig. 3) are restrict- cama (Hoke et al., 2004; Nishiizumi et al., with well-developed salic horizons do not ed mostly to localities where precipitation is 2005). have high water tables and are not influenced 5–10 mm/yr. Nitrate concentrations of soils in Our results do not support the late Oligo- by upward capillary migration of soluble salts. the hyperarid core of the Atacama are gener- cene age for the cessation of erosion proposed Rather, salts originate from eolian dust that is ally greater than this. Mass independent frac- by Dunai et al. (2005) on the basis of 21Ne translocated down into the soil profile during tionation anomalies of oxygen (⌬17O) in soil concentrations in fluvial gravels, nor with the rare precipitation events (Rech et al., 2003). nitrate also allow us to constrain paleoprecip- late Pliocene age postulated by Hartley and In addition, the Barros Arana geosol has many itation values (Michalski et al., 2004). The Chong (2002) based on the sedimentological of the characteristics of extremely mature ni- ⌬17O values of soil nitrate measures the rela- record. Hartley and Chong (2002) concluded trate soils in the modern Atacama, including tive proportion of nitrate derived from pho- that Miocene fluvio-lacustrine and alluvial fan large vertical salt fractures, angular salt- tochemical reactions in the atmosphere deposits in the Atacama are indicative of a shattered gravels floating in a gypsum matrix, (ϩ23‰; Michalski et al., 2003) to nitrate semiarid climate, the formation of thick evap- and a comparable distribution of soluble salts formed or modified by moisture-dependent orites after 6 Ma identifies a phase of in- (i.e., high concentrations of near-surface sul- microbial processes (0‰; Michalski et al., creased aridity, and the cessation of fluvio- fates and greater concentrations of chlorides 2004). Soil nitrate in the hyperarid core of the lacustrine and alluvial fan sedimentation and nitrates at depth; Ericksen, 1981; Rech, Atacama, where precipitation is Ͻ3 mm/yr, between 4 and 3 Ma indicates the beginning unpublished data). has ⌬17O values that range from 13.7‰ to of hyperaridity. Unfortunately, there is not a Although the duration of soil development 21.6‰ (mean 17.6‰) (Michalski et al., 2004). good correlation between these depositional for modern nitrate soils in the Atacama is not Samples (n ϭ 10) of nitrate from the Barros environments and precipitation in the Ataca- precisely known, a maximum duration of ϳ5– Arana geosol have values that range from ma today. Fluvio-lacustrine sediments are be- 10 m.y. has been postulated on the basis of 4.6‰ to 14.4‰ (mean 9.9‰), suggesting ing deposited across the modern Atacama in Ar/Ar ages of underlying tuffs (Alpers and slightly wetter conditions than occur today in deposystems supported by groundwater re- Brimhall, 1988; Sillitoe and McKee, 1996). the hyperarid core of the Atacama (see GSA charged in the high Andes (Rech et al., 2002). The Barros Arana geosol is a paleolandscape Data Repository Table DR11). Combined, soil Hence, these wet depositional environments surface in the Calama Basin, and therefore nitrate concentrations and ⌬17O anomalies of bear no direct climatic implications for the represents varying amounts of time at differ- the Barros Arana geosol suggest paleoprecip- valley settings in which they are being depos- ent localities depending on local deposition itation values of 5–10 mm/yr. ited. salars are found in many lo- histories. The geosol is directly overlain by We estimate precipitation during the for- cations in the hyperarid Atacama, but also oc- the Sifon and Artola Ignimbrites (Fig. 2), mation of Quaternary Calcisols in the Calama cur throughout the Altiplano in regions that which have been dated as 8.3 and 9.4 Ma, Basin by using modern precipitation values are much colder and receive as much as 300 respectively (de Silva, 1989). These ignim- and precipitation values inferred from Quater- mm/yr precipitation (Stoertz and Ericksen, brites provide a minimum age (9.4 Ma) and nary rodent middens. Modern precipitation 1974). ϳ duration (1.1 m.y.) for soil development. We ranges from 30 to 50 mm/yr at the eleva- tions of the Miocene paleosols (2900–3400 estimate, based on the degree of ANDEAN UPLIFT AND CLIMATE m). Pleistocene rodent middens, however, in- relative to modern nitrate soils, that the Barros CHANGE dicate precipitation between 50 and 100 mm Arana geosol likely represents several million We suggest that the initiation of hyperar- ca. 10 ka (Betancourt et al., 2000). We esti- years (2–5 m.y.) of soil development. idity in the Calama Basin, which occurred be- mate average precipitation values associated tween ca. 19 and 13 Ma, was the result of with Pleistocene Calcisols to be ϳ50 mm/yr. PALEOPRECIPITATION IN THE uplift of the Andes to elevations that were CALAMA BASIN ANTIQUITY OF THE ATACAMA high enough to block moisture entrained in the Comparison of paleosols in the Calama Ba- The transition from ca. 20 Ma calcic Ver- SASM from entering the Atacama. We esti- sin with modern soils in the Atacama Desert tisols with root traces, soil carbonate, and mate that a minimum Andean paleoelevation ϳ and surrounding region allow us to reconstruct gleyed horizons to the ca. 13–8 Ma Barros of 2 km was necessary to cause extreme hy- time-averaged precipitation values during pe- Arana geosol with pedogenic nitrate clearly peraridity along the eastern margin of the Cal- riods of soil formation in the Neogene. Ver- marks the onset of extreme hyperaridity, with ama Basin. This estimation is based on ex- Њ tisols are not found in the Atacama Desert to- precipitation Ͻ20 mm/yr, in the Calama Ba- amination of modern rainfall between lat 22 ϳ Њ Њ day, but do occur in central Chile ( lat 29 – sin. This middle Miocene age for the onset of and 25 S across the Central Andes, which Њ ϳ 32 S), where precipitation is 250 mm/yr. We extreme desiccation in the Atacama Desert shows a significant decrease in precipitation at ϳ therefore estimate minimum precipitation val- broadly supports ages for the initiation of hy- an elevation of 2 km on the eastern slope of ues during the formation of these calcic Ver- peraridity inferred from the cessation of su- the Andes (Fig. DR1; see footnote 1). Atmo- Ͼ tisols to be 200 mm/yr. pergene mineralization and erosion of ca. 15 spheric modeling of precipitation along the We estimate precipitation during the for- Ma in northern Chile (Alpers and Brimhall, eastern flank of the Andes also indicates that mation of the Barros Arana geosol to have 1988; Sillitoe and McKee, 1996) and is slight- most moisture is precipitated out of air masses Ͻ Ͻ been 20 mm/yr. This estimate is based on ly older than proposed ages of 10 and 9 Ma at elevations 2 km (Masek et al., 1994). This the high concentrations of nitrate and the middle Miocene paleoelevation estimate of ⌬17O anomalies of nitrate. Nitrate within the 1GSA Data Repository item 2006159, Table ϳ2 km prior to 12 Ma for the Central Andes upper few meters of soils is only known to DR1, ⌬17O values of soil nitrate, and Figure DR1, suggests a slightly earlier uplift history than occur in the ’s driest today, such precipitation data for the Central Andes, is available those based on the isotopic composition of online at www.geosociety.org/pubs/ft2006.htm, or Њ as the Atacama, the dry valleys of , on request from [email protected] or Docu- soil carbonate near lat 18 S (Garzione et al., and parts of the in . Modern At- ments Secretary, GSA, P.O. Box 9140, Boulder, CO 2006; Ghosh et al., 2006). acama soils with concentrations of nitrate sim- 80301, USA. We suggest that the return to wetter condi-

GEOLOGY, September 2006 763 tions in the Calama Basin, evidenced by Qua- sions, and Steven G. Driese, Carmala Garzione, and ocline: Journal of Geophysical Research, ternary Calcisols, rodent middens, and mod- Teresa Jordan for thoughtful reviews. v. 93, p. 13,841–13,854. Lamb, S., and Davis, P., 2003, Cenozoic climate ern precipitation values that are greater than REFERENCES CITED change as a possible cause for the rise of the middle Miocene paleoprecipitation estimates, Alpers, C.N., and Brimhall, G.H., 1988, Middle Andes: Nature, v. 425, p. 792–797, doi: was the result of local surface uplift. The up- Miocene climatic change in the Atacama De- 10.1038/nature02049. per elevation limit of modern nitrate soils sert, northern Chile: Evidence from supergene Lamb, S., Hoke, L., Kennan, L., and Dewey, J., mineralization at La Escondita: Geological So- 1997, Cenozoic evolution of the central Andes along the eastern margin of the Atacama is in Bolivia and northern Chile, in Burg, J.P., ϳ ciety of America Bulletin, v. 100, 2500 m. At the Purilactis locality, the Barros p. 1640–1656, doi: 10.1130/0016-7606(1988) and Ford, M., eds., Orogeny through time: Arana geosol crops out at an elevation of 3410 100Ͻ1640:MMCCITϾ2.3.CO;2. Geological Society [London] Special Publi- m. This suggests a minimum upward displace- Betancourt, J.L., Latorre, C., Rech, J.A., Quade, J., cation 121, p. 237–264. ment of 900 m over the past 9.4 m.y. This and Rylander, K., 2000, A 22,000-year record Mack, G.H., James, W.C., and Monger, H.C., 1993, of monsoonal precipitation form northern Classification of paleosols: Geological Society minimum uplift estimate is in accord with Chile’s Atacama Desert: Science, v. 289, of America Bulletin, v. 105, p. 129–136, doi: river-profile data from the western flank of the p. 1542–1546, doi: 10.1126/science.289. 10.1130/0016-7606(1993)105Ͻ0129: central Andes (18.5Њ–22ЊS) that indicate ϳ1 5484.1542. COPϾ2.3.CO;2. km of post–10 Ma uplift (Hoke et al., 2005). de Silva, S.I., 1989, Geochronology and stratigra- Masek, J.G., Isacks, B.L., Gubbels, T.L., and Field- Њ Ј ing, E J., 1994, Erosion and tectonics at the As late Neogene uplift developed in the ab- phy of ignimbrites from the 21 30 Sto 23Њ30ЈS portion of the Central Andes of north- margins of continental : Journal of sence of significant local upper-crustal defor- ern Chile: Journal of Volcanology and Geo- Geophysical Research, v. 99, 13,941–13,956. mation along the western flank of the Andes, thermal Research, v. 37, p. 93–131, doi: Michalski, G., Scott, Z., Kabiling, M., and Thie- it is likely driven by ductile thickening of the 10.1016/0377-0273(89)90065-6. mens, M.H., 2003, First measurements and ⌬17 lower crust and convective removal of the un- Dunai, T.J., Gonza´lez Lo´pez, G.A., and Juez-Larre, modeling of O in atmospheric nitrate: Geo- J., 2005, Oligocene–Miocene age of aridity in physical Research Letters, v. 30, no. 16, doi: derlying lithospheric mantle (Isacks, 1988; the Atacama Desert revealed by exposure dat- 10.1029/2003GL017015,2003. Lamb et al., 1997; Garzione et al., 2006). ing of erosion-sensitive landforms: Geology, Michalski, G., Bo¨hlke, J.K., and Thiemens, M., Several factors associated with uplift likely v. 33, p. 321–324. 2004, Long term atmospheric deposition as contributed to the wetter conditions observed Ericksen, G.E., 1981, Geology and origin of the the source of nitrate and other salts in the At- acama Desert, Chile: New evidence from in the Calama Basin during the Quaternary. Chilean nitrate deposits: U.S. Geological Sur- vey Professional Paper 1188, p. 1–37. mass-independent oxygen isotopic composi- Decreased air temperatures would have in- Ericksen, G.E., 1994, Discussion of a petrographic tions: Geochimica et Cosmochimica Acta, creased effective moisture. Continued uplift study of Chilean nitrates: Geological Maga- v. 68, p. 4023–4038, doi: 10.1016/j.gca. also would have created an orographic barrier zine, v. 131, p. 849–852. 2004.04.009. capable of intercepting air masses that migrate Garzione, C.N., Molnar, P., Libarkin, J.C., and Nishiizumi, K., Caffee, M.W., Finkel, R.C., Brim- MacFadden, B.J., 2006, Rapid late Miocene hall, G., and Mote, T., 2005, Remnants of a northward from the southern Westerlies. rise of the Bolivian Altiplano: Evidence for fossil alluvial fan landscape of Miocene age removal of mantle lithosphere: Earth and Plan- in the Atacama Desert of northern Chile using CONCLUSIONS etary Science Letters, v. 24, p. 543–556. cosmogenic nuclide exposure dating: Earth The initiation of extreme hyperaridity along Ghosh, P., Garzione, C.N., and Eiler, J.M., 2006, and Planetary Science Letters, v. 237, the eastern margin of the Atacama Desert is Rapid uplift of the Altiplano revealed through p. 499–507, doi: 10.1016/j.epsl.2005.05.032. 13C-18O bonds in paleosol carbonate: Science, Rech, J.A., Quade, J., and Betancourt, J.L., 2002, clearly recorded by the transition from calcic v. 311, p. 511–515. Late Quaternary paleohydrology of the central soils with root traces to salic (nitrate) Hartley, A.J., and Chong, G., 2002, Late Pliocene Atacama Desert, Chile (22–24ЊS): Geological in Miocene strata of the Calama Ba- age for the Atacama Desert: Implications for Society of America Bulletin, v. 114, sin. This transition represents a change in pre- the desertification of western South America: p. 334–348, doi: 10.1130/0016-7606(2002) 114Ͻ0334:LQPOTCϾ2.0.CO;2. cipitation from Ͼ200 mm/yr to Ͻ20 mm/yr Geology, v. 30, p. 43–46, doi: 10.1130/0091- 7613(2002) 0302.0. Rech, J.A., Quade, J., and Hart, W.S., 2003, Isotopic and occurred well before 9.4 Ma. We interpret Hartley, A.J., and May, G., 1998, Miocene Gypcre- evidence for the origin of Ca and S in soil this dramatic climate change to have taken tes from the Calama Basin, northern Chile: gypsum, anhydrite, and calcite in the Atacama place between 19 and 13 Ma and suggest that Sedimentology, v. 45, p. 351–364, doi: Desert, Chile: Geochimica et Cosmochimica it was the result of Andean uplift to a paleo- 10.1046/j.1365-3091.1998.0166e.x. Acta, v. 67, p. 575–586, doi: 10.1016/S0016- Ͼ Hoke, G.D., Isacks, B.L., Jordan, T.E., and Yu, J.S., 7037(02)01175-4. elevation ( 2 km). Although the initiation of 2004, Groundwater-sapping origin for the gi- Searl, A., and Rankin, S., 1993, A preliminary pet- hyperaridity could also be the result of global ant quebradas of northern Chile: Geology, rographic study of the Chilean nitrates: Geo- climate cooling and the strengthening of the v. 32, p. 605–608. logical Magazine, v. 130, p. 319–333. at that time, hyperaridity Hoke, G.D., Isacks, B.L., Jordan, T.E., and Blanco, Sillitoe, R.H., and McKee, E.H., 1996, Age of su- pergene oxidation and enrichment in the Chi- could not have developed along the eastern N., 2005, Geomorphic evidence for post-10 Ma uplift of the western flank of the central lean porphyry province: Economic Ge- margin of the Atacama without a strong An- Andes18Њ30Ј–22ЊS: Geological Society of ology and the Bulletin of the Society of dean rain-shadow effect in place. Therefore, America Abstracts with Programs, v. 37, Economic Geologists, v. 91, p. 164–179. the Barros Arana geosol places a minimum no. 7, p. 272. Stoertz, G.E., and Ericksen, G.E., 1974, Geology of age for the development of a strong Andean Houston, J., and Hartley, A.J., 2003, The Central salars in northern Chile: U.S. Geological Sur- Andean west-slope rainshadow and its poten- vey Professional Paper 811, 65 p. rain shadow in the Central Andes, which we tial contribution to the origin of hyperaridity estimate to be ϳ2 km. in the Atacama Desert: International Journal Manuscript received 18 November 2005 of Climatology, v. 23, p. 1453–1464, doi: Revised manuscript received 7 April 2006 ACKNOWLEDGMENTS 10.1002/ joc.938. Manuscript accepted 25 April 2006 We thank the Hampton Fund for support, Nicolas Isacks, B.L., 1988, Uplift of the Central Andean Blanco and Andrew Tomlinson for helpful discus- plateau and the bending of the Bolivian or- Printed in USA

764 GEOLOGY, September 2006