Paleosols as proxies of paleoprecipitation

Climatic zonation of soils was a key element in the Russian origins of soil science (Jenny 1941), and a variety of relationships between particular soil features and climatic variables can be applied to East African paleosols in order to reconstruct

O Figure 13.5

Carbon isotopic (S13Corg) depth profiles of Kenyan Miocene pedotypes, showing strong surface humification in grassland paleosols (Chogo and Onuria pedotypes), subsurface humification in Alfisols (Tut) and mixing in vertic Inceptisols (Chido). Carbon isotopic data is from Bestland and Krull (1999) and Cerling et al. (1997a), and paleosols described by Retallack (1991a), Retallack et al (1995), and Bestland and Krull (1999)

paleoclimate. For example, depth to carbonate horizon (D in cm) is related to mean annual precipitation (P in mm) by formula (1) below (from Retallack 2005). This depth can be corrected for compaction due to overlying sediment using geological estimates of overburden and standard formulae (Sheldon and Retallack 2001). Also related to mean annual precipitation (P) is nutrient base content (C = Al2O3 / (Al2O3 + CaO + MgO + Na2O) in mol) of soil Bt horizons by formula (2) below (Sheldon et al. 2002).

P = 137.24 + 6.45D - 0.013D2 R2 = 0.52, S.E. = ±147mm (1)

P = 221.12 • e0197C R2 = 0.72, S.E. = ±182mm (2)

Chemical weathering also alters the mineral content of soils, especially their clay minerals, which begin as smectites and then lose cationic bases with further chemical weathering to become kaolinite (Retallack 2001). This indication of paleoprecipitation works best with noncalcareous soils, which are found in climates receiving more than 1,000-mm mean annual precipitation (Retallack 2004). In East Africa today, smectite is dominant in soils receiving less than 1,200-mm mean annual precipitation, and kaolinite dominant in wetter climates

(Mizota et al. 1988). Thus noncalcareous, smectitic soils define a limited paleo-climatic window of 1,000-1,200-mm mean annual precipitation.

My new compilation of Kenyan paleoprecipitation over the past 20 million years (O Figure 13.6b) includes previously published data on African depth to Bk (Wynn 2001, 2004a, b; Wynn and Retallack 2002; Retallack 2001b; Retallack et al. 2002), and paleosol chemical (Retallack et al. 1995,2002; Thackray 1989; Bestland 1990; Retallack 1991a; Wynn and Retallack 2002) and clay mineral composition (Retallack 1991a, Behrensmeyer et al. 2002), as well as published inferences from size and shape of fossil leaves (Jacobs 2002). This compilation is limited to data from around Lake Victoria for the early-middle Miocene, the Tugen Hills for the mid-late Miocene and the Turkana Basin for the Miocene to Quaternary. The geological time scale is from radiometric dating of these various fossil primate sites (Deino et al. 1990; Retallack 1991a, Jacobs and Deino 1996; Behrensmeyer et al. 2002; Hill et al. 2002).

These new data reveal not just one Neogene aridification event at about 7 Ma, as has long been implied by the "Tertiary pluvial hypothesis'' (Leakey 1952), the "Miocene lake hypothesis'' (Kent 1944), the "Miocene rain forest hypothesis'' (Andrews and Van Couvering 1975; Andrews 1996), and the "Late Miocene grassland hypothesis'' (Cerling 1992; Cerling et al. 1997a, b). These theories had already been discredited by discovery of Miocene desert dunes, shrubland snails, alkaline lakes, open-country grasses, grazing mammals, and grassland paleosols in East Africa (Pickford 1986a, 2002a; Retallack et al. 1990, 2002).

Instead the data (O Figure 13.6) reveal a Neogene paleoclimatic roller coaster of at least nine dry spells with intervening wet periods, of which humidity spikes at 16 and 13 Ma were the wettest of the last 20 million years. This new paleo-precipitation curve is similar to paleotemperature variations for Africa inferred from north-south oscillation through time of Ethiopian and Palearctic biogeo-graphic realms (Pickford 2002a). These new data are also similar to foraminiferal oxygen isotope curves from the deep sea (Zachos et al. 2001), commonly used as a basis for evaluating human evolution in Africa (de Menocal 2004), but the match is not precise (O Figure 13.6b and c). A general trend of extreme and volatile middle Miocene values, but subdued late Miocene to Quaternary values, is evident from both isotopic and paleosol data. The paleosol record reveals much greater variation in rainfall than would be inferred from carbon isotopic values of marine foraminifera, which are damped by global oceanic mixing with time lags of several thousand years. More profound damping is seen in oxygen isotopic values of marine foraminifera, which show a long-term increase unlike local rainfall and foraminiferal carbon records. This increase is plotted on reversed axes in O Figure 13.6d because it has been interpreted as a long-term temperature

O Figure 13.6

A 20 million year record of vegetation (a) and paleoprecipitation (b) from Kenya, compared with carbon (c) and oxygen (d) isotopic composition of marine foraminifera. Paleoprecipitation data from paleosols (b) is from depth to carbonate (open ellipses), clay mineral (diamonds), and chemical composition (squares) after Retallack (1991a), Retallack et al. (1995, 2002), Wynn (2001, 2004a, b), Wynn and Retallack (2001). Paleobotanical estimates from Jacobs (2002) and Jacobs and Deino (1996). Modern vegetation precipitation limits are from Anhuf et al. (1999)

decline (Zachos et al. 2001), but part of this long-term trend is due not just to temperature but to water recycling with plate tectonics (Veizer et al. 2001). The global oxygen isotope record also shows an increase after 3 Ma due to continental icecap sequestration of isotopically light oxygen, in addition to temperature effects (Zachos et al. 2001). Despite these problems, the East African paleosol record and global isotopic records present a very different concept of climatic variation experienced by our distant ancestors than the past idea of a seminal Late Miocene climatic event. Instead of a single origin of humanity at a turning point of environmental change, the new record implies rather that our lineage responded to a gauntlet of changing conditions with a variety of adaptations (O Table 13.1), as discussed later.

O Table 13.1

Geological age of African climatic events, selected adaptations, and hominoid diversity (D), origination (O), and extinction (E)

Age (Ma) Hominoid adaptations and extinctions D O E

20.2 dry Robust mandible for hard food (Rangwapithecus) 10 9 0

19.1 wet Low cusp molars for folivory (Nyanzapithecus) 5 0 5

17.7 dry Thick enamel for hard food (Afropithecus) 6 1 2

16.1 very wet Short back for suspension (Morotopithecus) 4 3 6

14.9 very dry Adducted hallux for ground walking (Kenyapithecus) 7 4 2

19.1 wet Low cusp molars for folivory (Nyanzapithecus) 5 0 5

17.7 dry Thick enamel for hard food (Afropithecus) 6 1 2

16.1 very wet Short back for suspension (Morotopithecus) 4 3 6

14.9 very dry Adducted hallux for ground walking (Kenyapithecus) 7 4 2

12.6 very wet

Thin enamel molars for soft food (Otavipithecus)

Was this article helpful?

0 0

Post a comment