Laura SaviñaTime is slowly eroding the flat lands in the inland areas of the Brazilian states of Minas Gerais and São Paulo. The plateau containing the watershed of the São Francisco River, which originates in southwestern Minas Gerais and flows northeast to the state of Pernambuco, is gradually shrinking due to the retreat of the escarpments that form its margin. In the last 1.3 million years, this plateau has lost area to a neighboring region situated at lower altitudes, where the Doce River Basin lies. And that basin, in turn, has lost area to the Paraíba do Sul River Basin at the boundaries between the states of São Paulo, Minas Gerais and Rio de Janeiro. This gradual erosion of land, which has pushed the edges of the plateaus increasingly farther into the country’s interior, was recently brought to light by researchers from the states of Goiás and Minas Gerais in a paper published in the journal Geomorphology.
The study’s findings might, at first, cause concern for those who fear for the future of watersheds such as that of the São Francisco River. But that is not the case. First of all, the process of land erosion or denudation is extremely slow. And secondly, geological studies can, in a certain sense, be likened to stock market investments: past geological performance is not indicative of future results.
“The data indicate how the process has worked over the past 1.3 million years, but they do not enable us to engage in predictive speculation because, in the Earth sciences, there are low-frequency, high-intensity processes—such as major earthquakes—that would invalidate any prediction,” says Luis Felipe Cherem, a researcher at the Federal University of Goiás (UFG) and the paper’s first author, who collaborated with researchers from the federal universities of Ouro Preto (UFOP) and Minas Gerais (UFMG) and from the European Center for Research and Teaching in Environmental Geosciences, located in France.
The region’s most superficial features, from the coast to the São Francisco Basin well inside the continent, are largely the result of violent geological processes caused by plate tectonics, the same phenomenon that sets in motion the eternal dance of the continents across the globe. The first of these large movements, about 130 million years ago, broke up the supercontinent known as Gondwana and created Antarctica, South America, Africa, Australia, the Arabian Peninsula, India and the Atlantic Ocean.
Since that initial stage of formation of the South American coast, two additional tectonic events have affected the region in the past 65 million years, creating three elevations, according to the researchers at UFMG and UFOP. On the lowest elevation is the Paraíba do Sul River Basin, a plateau about 400 meters above sea level. Closer to the interior is the Doce River Basin, with an average altitude of 800 meters, and farther on, the São Francisco and Paraná river basins, which are 1,100 meters above sea level.
Luis Felipe Cherem/UFGTo conduct the work that revealed the gradual advance of the lower plateaus towards the continental interior, the researchers collected samples of river sediment from the Cristiano Otoni Ridge, an escarpment 30 kilometers wide and 250 to 350 meters high that separates the São Francisco Basin from the Doce River Basin. They also analyzed material obtained along the 65-kilometer length of the São Geraldo Ridge that divides the Doce River Basin from the Paraíba do Sul Basin. In both cases, they sought to sample material along the edges as well as on the back sides of the escarpments, the steep downward slopes that separate one elevation from another. The objective was to quantify the phenomenon known as denudation, over the last one million years.
Denudation is a process caused by constant erosion over time. Rain and wind decompose and displace the rocks closer to the surface, uncovering the terrain beneath. It is as if the surface of the region were being gradually lost, leaving the rock of the subsoil exposed.
In order to calculate the denudation rate, the researchers analyzed the river sediments at the top and base of each elevation. By comparing the material from these areas, it is possible to estimate how many millimeters are eroded per thousand years (or how many meters per million years).
As can be expected in erosive processes, and aided by the downward slope, the escarpments naturally undergo more denudation than the plateaus themselves. It was found that the plateau containing the São Francisco Basin, for example, has lost an average of 8.77 meters every million years. In the Doce River Basin, the loss was 15.68 meters during the same period. On the escarpments, that number is understandably higher: 17.5 meters every million years for the Cristiano Otoni Ridge, and 21.22 meters for the São Geraldo Ridge.
These findings indicate that the process of denudation—a phenomenon with multiple causes that can occur at rates that vary from one region to another, is still ongoing. According to Cherem, these values are consistent with what one would expect to observe when comparing plateaus: those closer to the interior of cratons—the stablest part of tectonic plates—are generally more mature and undergo less denudation over time.
Cherem, Varajão and their colleagues arrived at these denudation rates by analyzing the presence of a certain variety of the chemical element beryllium in the rocks. Beryllium, the fourth element in the periodic table, has four protons in its nucleus. To measure the age of rocks, researchers assess the presence of beryllium-10, a version of the element that has six neutrons and tends to decay over time, losing one of its neutrons. For beryllium-10, the half-life—the time it takes for half the atoms in the sample to disintegrate—is estimated at 1.38 million years. By comparing the quantity of this element in a soil, therefore, it is possible to get an idea of the age of the sample. “The results obtained in Minas Gerais are similar to those observed in other, different margins, or zones of separation between two continents, around the world,” Cherem notes.
Earlier studies conducted in a nearby region but using different techniques had yielded different denudation rates. In 2010, researchers Silvio Hiruma of the Geological Institute of São Paulo, Claudio Riccomini of the Geosciences Institute of the University of São Paulo, and other colleagues published a paper in Gondwana Research indicating that the denudation rate could be much higher.
“Our data suggest that some parts of the Bocaina Ridge showed denudation of more than 3,000 meters in the last 60 million years, which would be around 50 meters per million years,” says Riccomini, who, along with researchers from France, has just published a new study on the topic in the Journal of Geophysical Research.
The discrepancy in denudation rates could arise from two factors. First, the technique used by the team at USP enabled them to analyze what occurred over a longer period of time—and denudation may moderate as plateaus mature. Second, the study conducted by Riccomini and his colleagues concentrates on Serra do Mar, which, though close to the area studied by Cherem and his colleagues, has a geological history that is distinct from what is seen in the more interior regions of the continent. “There is no contraposition or mutual negation, but rather complementarity in the search for a better understanding of the dynamics of the relief of southeastern Brazil,” the UFG researcher asserts.
Each sample has a history
According to Cherem, the number of samples analyzed instills confidence in the results. Nonetheless, it is possible that the denudation rates will vary a little as the number of samples increases. “I could indicate several places where the gains or losses of area in those basins vary a lot from one to the other,” says Allaoua Saad of the Federal University of Minas Gerais, a specialist in the geomorphology of southeastern Brazil.
Saadi acknowledges the quality of the study conducted by Cherem and Varajão, however. “The results presented in terms of denudation rates are the product of measurements at points chosen based on homogeneity criteria in the different basins,” says the UFMG geomorphologist. “What these measurements express is a belief that this represents a real measurement of long-term denudation that can be generalized to lead to the conclusions presented,” he comments.
In addition to the beryllium measurements used in the Geomorphology paper, Cherem affirms that other data, presented in his PhD dissertation, corroborate the idea that the escarpments in southeastern Brazil are retreating approximately 0.01 millimeter per year, causing the higher basins to lose area to the lower ones. In any case, he acknowledges that the geological mysteries of southeast Brazil are still far from being completely solved. “The escarpments are still there,” he says, “and they still have to be studied.”
CHEREM, L.F.S. et al. Long-term evolution of denudational escarpments in southeastern Brazil. Geomorphology. v. 173-4. pp. 118-27. 2012.
COGNE, N. et al. Post-breakup tectonics in southeast Brazil from thermochronological data and combined inverse-forward thermal history modeling. Journal of Geophysical Research. v. 117. 2012.
HIRUMA, S.T. Denudation history of the Bocaina Plateau, Serra do Mar, southeastern Brazil: relationships to Gondwana breakup and passive margin development. Gondwana Research. v. 18. pp. 674-87. 2010.