Aridity indices from oxygen isotopes

Africa never experienced the large temperature swings associated with glacial to interglacial shifts at higher latitudes, and the climate variable of most importance and influence seems to have been rainfall. An overall trend toward aridifi-cation has been inferred from the visible trends toward more open, grassier African landscapes over the last ~5-6 Ma, as discussed earlier. However, although aridity and forest diminishment may well be linked in certain respects (for example, maintenance of equatorial forests requires high rainfall), the two are not necessarily tightly connected in savanna habitats. In their report on the isotopic data from pedogenic carbonates from Turkana from Ma, Cerling et al. (1988) note that the presence of these carbonates through the sequence is indicative that arid conditions were present from the lower sections. Since the 813C data of pedogenic carbonates and fauna in the region shows that the proportions of C4 grasses remained fairly modest for millions of years after their first appearance, these data suggest strongly that tree density/open landscapes and aridity should be delinked. Studies of modern savanna ecosystems show that lower rainfall often favors more thicket and bushland rather than grass (Scholes and Walker 1993; Owen-Smith 1999). It remains important to establish independent indicators for aridity. A number of isotopic approaches have been, or are, in the process of development.

As illustrated in O Figure 9.2, the Turkana Basin pedogenic carbonate d18O sequence suggests that a change in East African rainfall patterning occurred around 1.8 Ma. Since 818O values prior to this time were more negative, and those thereafter more positive, the suggestion is that the dominant moisture source changed from a distant (perhaps the Atlantic Ocean) to a closer source, likely the Indian Ocean. The d18O values in the latter period are sufficiently enriched to suggest that processes of evaporation also exerted an influence (O Figure 9.2). This kind of information, therefore, provides some hints about development of aridity, but little detail.

Isotopic ratios from ostrich or ratite eggshells show great promise as aridity indicators, where the potential is to extract subtle and quantifiable indications of moisture shifts in the past. Johnson et al. (1997) used dual 815N and d18O data from ostrich eggshell in the Equus Cave hyena accumulation to document moisture fluctuations associated with Late Glacial to Holocene climate shifts. However, no similar studies have been undertaken for earlier periods of interest for hominin evolution, possibly because of the difficulties associated with the organic fraction in older material. Fortunately, data from modern and Late Pleistocene-Holocene archeological sites show that 818O alone provides a reasonable aridity index (Lee-Thorp and Talma 2000; Segalen 2003). d18O from the eggshells of earlier ratites has been used to show continuously high, but fluctuating aridity in the Namib from the Miocene to the Present (Segalen 2003; Segalen et al. 2006). The main constraint is the chronology, which of necessity is based on biostratigraphy and thus provides only rather crude time intervals. A related tool based on fossil tooth enamel is under development (Schoeninger et al. 2003; Levin et al. 2006), which employs 818O differences between animals that use environmental water, and those that rely on plant water and in turn reflect the influences of relative humidity (and aridity). Both of these tools hold potential but have not yet yielded information of direct consequence for addressing questions about climate trends and hominin evolution.

0 0

Post a comment