In the Stallard model of erosional regimes, "a complex interplay of chemical, physical, and biological processes" generate weathering products (1995a, 1995b). More precisely, these researchers claimed that maximum chemical denudation rates are achieved in the weathering-limited regime, with vegetation maintains a relatively thin soil on steep slopes, trapping water and generating bioacids in contact with fresh rock. With erosion triggered by earthquakes, treefalls, and so forth, fresh rock is reexposed for the regeneration of soil (Stallard 1992). In contrast, in the transport-limited regime, thick highly weathered soils ultimately slow down chemical denudation by shielding fresh rock buried deep below the surface from percolating water. Stallard (1995b) pointed out that in flat terrains vegetation has two contradictory effects on chemical denudation: its stabilization of thick leached soils tends toward its reduction, whereas the production of bioacids and penetrating root systems tends toward its increase. In any case the estimated chemical denudation rates even in flat terrains with thick soils developed on granitic rocks are still some two orders of magnitude higher than a computed two-dimensional bare rock case. It is interesting to speculate that this mechanism could explain at least part ofthe cooling apparently tied to global periods of orogenesis (e.g., possibly at 2.5 Ga; see Knauth and Clemens 1995). More mountains could magnify biotic influence on weathering if the Stallard model is correct.
Estimated chemical denudation rates of siliceous rocks from the humid tropics of South America support the Stallard model (figure 4-8). More recent data from the Guayana shield (Edmond et al. 1995) and Congo basin (Gaillardet et al. 1995) provide additional support for this scenario. In both cases, significantly lower chemical denudation rates were inferred from rivers draining relatively flat terrain. In contrast, White and Blum (1995) found no correlation between chemical denudation rates of granitic rocks and relief in their study of 68 watersheds. They concede that their data set contains only one tropical watershed (rain forest in Puerto Rico). Drever and Clow (1995) suggest that White and Blum's negative finding with respect to the influence of relief may have resulted from the absence of really low relief areas (e.g., Amazon basin), or that "noise" in data prevented the emergence of a correlation. The basic validity of the Stallard model will be assumed in the discussion of the evolutionary history of the biotic enhancement of weathering in chapter 8.
An even more radical view of the factors influencing weathering has emerged, namely that climate has very little to do with determining the rates of chemical denudation and carbon dioxide consumption by silicate weathering, but rather the predominant controls are relief and lithology (Edmond and Huh 1997; Huh and Edmond 1997; Allegre et al. 1997; Gaillardet et al. 1997). This contention is based on inferred carbon dioxide consumption rates from a global data base of river and watershed rock chemistry. Silicate and carbonate weathering contributions to rivers are distinguished by using the different strontium isotopic signatures of each lithology.
However, supporters of a strong climatic effect on weathering rates have noted that the carbonate weathering contribution is often underestimated in watersheds with predominant silicate lithology (Berner and Berner 1997). For example, significant outcrops of marble occur in the cold Aldan River watershed in Siberia (Berner and Berner 1997), which Huh and Edmond
Was this article helpful?