Technical Comments

Comment on “Intensifying Weathering and Land Use in Iron Age Central Africa”

+ See all authors and affiliations

Science  31 Aug 2012:
Vol. 337, Issue 6098, pp. 1040
DOI: 10.1126/science.1221820


Bayon et al. (Reports, 9 March 2012, p. 1219) claim that the “rainforest crisis” in Central Africa centered around 2500 years before the present “was not triggered by natural climatic factors” and that it was caused by widespread deforestation resulting from the arrival of the Bantu colonists. However, there is a consensus among palaeoecologists that this landscape change and the related physical erosion it caused was due mainly to a shift to more seasonal rainfall regime.

Bayon et al. (1) analyzed the fluctuations in climate and vegetation in the Congo Basin during the last 20 millennia using aluminum-potassium ratios (Al/K) and hafnium isotopic composition of a marine sediment core retrieved off the mouth of the Congo River. They claim that the highest values of these ratios are evidence of intense chemical weathering during intervals that experienced the greatest amounts of annual precipitation, stressing that the degree of chemical weathering at about 3000 years before the present (yr B.P.) exceeded that of all other previous episodes. Bayon et al. recognize “a major vegetation change occurred in Central Africa, when rainforest trees were abruptly replaced by savannas.” However, a consensus exists among palaeoecologists, palynologists, and archaeologists that this event resulted mainly from a change in the seasonal regime of rains that did not involve a change in the annual amount of precipitation (2, 3). In our opinion, the increase in physical erosion that Bayon et al. observe was related to a large climatic change for two main reasons: first, because of the abrupt character of the event and, second, due to its very large geographical extension, from the Gulf of Guinea to the eastern end of the Congo basin and farther to the western part of East Africa (2), and throughout all northern tropical Africa from Cameroon (4) to Chad and West Africa (5, 6). Bayon et al. suggest that this event was linked to the arrival of the Bantu colonists in Central Africa, which would have immediately produced large-scale deforestation due to agriculture and iron smelting. Moreover, Bayon et al. also confuse the issue by using the term “weathering” to qualify both the physical and chemical processes. Physical weathering must be specified every time as well as chemical weathering and, in the first case, terms such as “mechanical erosion” or “physical denudation” would be preferable. Furthermore, chemical weathering occurs first and enables the process of the physical weathering, but the processes generally are not linked immediately. A study of the sediment fluxes of big rivers, including the Congo, concluded that “an inverse correlation between weathering intensities and suspended sediment concentrations is observed showing that the regions having the highest rates of physical denudation produce the least weathered sediments” (7). Had there been increasing deforestation and physical erosion from the Bantu expansion, less chemical weathering, not more, should have occurred. Therefore, the Bantu expansion alone cannot explain the increases in this index seen by Bayon et al.

Assessing the chemical weathering ratio, the basis of Al/K, is an effective method in soil studies. This ratio rises when potassium minerals of the mica group (illites) are lost by hydrolysis. Then, potassium is solubilized and subtracted from the solid phase while residual aluminum is fixed, thanks to in situ kaolinite neoformation or gibbsites precipitation (AlOH3) (8, 9). Measurements of Al/K usually are combined with mineralogical analyses such as the composition of clay minerals or the sandy fractions (with K-feldspar components), even if they are not abundant (4, 10). Bayon et al. supply no information about the textural or mineralogical nature of the marine sediments, which severely limits the relevance of their conclusions. Last, to use this method to assess marine sediments means implicitly assuming that a hypothetical process on the continent would be translated almost immediately into riverine fluxes and then into oceanic deposition, which is hardly believable. Bayon et al.’s observation that “Hf isotopes display significant downcore variations…which correlate well with the Al/K depth profile” could simply mean that, repeatedly, erosion was able to affect deeper horizons of the soils where the signatures of chemical weathering may have been more or less intense and more or less ancient.

Therefore, the assertion that “these data show that downcore fluctuations of both εHf and Al/K ratios…reflect variations in chemical weathering intensity within the Congo” is not very convincing because the erosion of modern superficial horizons of soils can interfere with that of the oldest horizons, so only strong chemical weathering could be apparent. The clear conclusion that emerges from this comparison (i.e., of big rivers of the world, including Congo) is that the degree of weathering of present-day river particulates is not consistent with most river dissolved loads derived from silicate weathering. This apparent disagreement is due to the fact that river suspended sediments integrate the whole weathering history of the rocks (recycling) within the drainage basin and not only the present-day weathering (4). The low values of εHf at 20,000 yr B.P. (dry climate and many grasses) and its small decrease, for the same reasons, during the Younger Dryas, seem coherent, contrary to the strong increase by 2500 yr B.P. in a landscape more open than in the previous period and with more irregular rains. Apart from a possible hypothesis of a massive colonization by the Bantus, the other logical explanation of such data would be the erosion of fossil soils horizons, evidence of the leaching during past climates.

If the authors’ statement that “chemical alteration in the Congo Basin [which is confused here with Central Africa sensu lato] has responded quickly to regional climatic changes” is acceptable on the scale of 103 years, it is not acceptable on the much shorter scale of 102 years, which is that of the presumed catastrophic event of Bantu colonization. Therefore, the assertion that “an abrupt trend toward higher Al/K and Hf isotope values indicates rapidly intensifying chemical weathering…centered at ~2500 years B.P.” seems to us exaggerated, especially when this episode is proposed to be the major weathering signal during the past 40,000 years. Moreover, the curves in Bayon et al. show a decrease of “weathering intensity” after 2000 yr B.P., whereas archaeological records—based on agricultural practices, iron-smelting industry, and ceramics—indicate that the very beginning of the Bantu colonization in the Congo Basin occurred about 2000 yr B.P. and reached a maximum only around 1000 yr B.P. (11, 12, 13). This is in complete contradiction of the suggestion of Bayon et al. that Bantu expansion was the main driver of this increase in weathering.


View Abstract

Related Content


Navigate This Article