An enormous bare rock wall in the Brejões mountain range presents a striking picture in one of the most beautiful landscapes in Chapada Diamantina National Park, in the Brazilian state of Bahia. On a horizontal band of this wall—seen in the photo at right—a team led by geologist Renato Paes de Almeida, a professor at the Geosciences Institute, University of São Paulo (IGc-USP), identified signs of a fast-flowing river that crossed the region 1.5 billion years ago. The river may have traversed a vast, level terrain, differing considerably from the large, winding rivers that now flow through the plain in deep channels, framed by vegetation. Unlike that setting, the ancient river likely crossed barren land and consisted of many shallow, endlessly braided channels interrupted by broad sandbars (see infographic).
Almeida and his colleagues used a new mathematical model to identify the marks that the primitive river had left on the rocks. The model, proposed by the USP group and published in the journal Geology in 2016, helped reconstruct the contours and behavior of rivers that existed before plants grew on the continents. “It helped us envision the features of these rivers,” the researcher says. “After we did the theoretical work, we went into the field to search rock outcrops for structures suggested by the model.”
One of the places they visited was the Brejões range, an outcrop well known by geologists. Bearing in mind the new features suggested by the model, the researchers began to observe details that had previously escaped notice.
Records of the activity of primitive plains rivers such as those now found in Bahia are rarely seen. In most cases, the marks of large waterways that existed more than 440 million years ago are engraved in rocks that formed the steepest segments of their channels, in mountainous regions close to the headwaters. Segments of the middle course or near the mouth of the rivers began to leave more abundant records only after the emergence of environments dominated by vascular plants that can grow on dry land. This transformation occurred around 434 million years ago, in the Silurian Period.
Prior to that time, the continents were practically deserts, and macroscopic life forms were restricted to the oceans. The initial advance of carpets of moss-like vegetation into continental interiors began to mold the contours of the rivers. Later, as the soil became stabilized by plant roots, the rivers developed more solid, well-defined banks. Their channels deepened, and more sand and pebbles were transported to coastal regions. On the great plains in moist climate regions, the rivers developed meandering paths, like the present-day Mississippi River in the United States.
Recent decades have seen significant advances in knowledge about the transformations undergone by the rivers, from the emergence of the mosses about 460 million years ago to the appearance of the first trees 390 million years ago. One motivation behind such advances was economic in nature, driven by the knowledge that ancient rivers surrounded by vegetation gave rise to oil reserves. But from the 1970s to the present, there has been practically no change in our understanding of the course of pre-vegetation rivers.
Geology books today still teach that before vegetation existed, the rivers were of the sheet-braided type, commonly seen in mountainous areas or in glacial regions with steep terrain. These waterways likely resembled the present-day rivers of Iceland, where erosion is substantial and almost no plants grow. The flowing water spreads into multiple channels that are generally shallow, bifurcating as sandbars build up. “Rivers at the foot of mountains have scarcely changed in appearance since before the Silurian,” explains Almeida. “The old pre-vegetation models were based on small, steep-gradient rivers and an arid climate. Researchers never wondered about the characteristics of rivers in the segments far from the mountains, or of the rivers on the plains where it rained as much as it does in the Amazon today.”
Almeida and his colleagues developed their mathematical model by modifying one that had been proposed in the 1990s by physicist and geologist Chris Paola of the University of Minnesota, in the United States. In Almeida’s model, the equation indicates the amount of sand and mud that a river without vegetation would deposit in a given stretch of its course, from the areas near the headwaters at the top of a mountain to the farthest reaches of the interior of a plain.
Based on the solutions given by the equation, Almeida concluded that rivers without vegetation are like braided rivers only in their steepest sections. These rivers begin at the foot of mountains with an extensive network of shallow intersecting channels, where the current is still strong enough to transport coarse, pebbly sand. But they quickly lose the ability to transport heavy sediments as they advance into lower-gradient regions. “In these sections, they stop being typical braided rivers,” Almeida says. “They have fewer, relatively wider channels that are deep and winding, because they transport and deposit only fine sand and mud.”
The model also explains why records of the middle course of plains rivers prior to the Silurian are rarely found. Since these rivers deposited little sediment on the more distant, flatter plains, nearly the only marks found are those left by the initial segments of their course, where there was greater accumulation of sand and other sediments.
This hypothesis introduces important ideas that could help lead to a better understanding of the evolution of ecosystems on Earth. If the model is correct, the flow of sediment from the mountains to the sea would have been much less prior to the Silurian than it is today. One consequence is that the planet’s beaches would have had finer-grained sand, potentially affecting the life of animals and plants in shore areas where the waves break. A further consequence is that the seawater near the coastline would have been less muddy, thus allowing sunlight—essential to most marine life—to penetrate to greater depths than it does today.
Almeida and geologists André Marconato of USP, Bernardo Tavares Freitas of the University of Campinas (Unicamp), and Bruno Boito Turra of the Geological Survey of Brazil visited the outcrops of the Tombador Formation in Chapada Diamantina, in search of traces of the middle course of a plains river prior to the Silurian. The area is full of steep canyons where one can see deposits from different river segments that existed between 1.6 billion and 1 billion years ago and were part of the same sedimentary basin.
The geologists measured the structures engraved in the rocks, took high-resolution photos and made detailed drawings of the outcrops. After examining the textures and shapes of these structures, they inferred the configurations of the channels and sandbars, as well as the direction of water flow and the amount of sediment transported. “It’s not intuitive,” Almeida says. “Based on the vertical sections, we had to construe a three-dimensional structure.”
Bernardo Freitas used rappelling ropes to descend rock faces in the area of the Brejões range, and identified the channel of a river that greatly resembles the one suggested by Almeida’s model for plains rivers prior to the Silurian. A horizontal band about 100 meters wide by 15 meters high preserved the configuration of the intertwined channels between sandbars and the contours of the layers of mud deposited on the banks during flooding on the plain.
Almeida’s model suggests that the plains rivers that existed before terrestrial vegetation emerged bore more resemblance to segments of the Irawaddy River, which crosses the plains of Myanmar in Southeast Asia. “The truth is that there is no present-day analog to pre-vegetation rivers,” he says.
The model used by Almeida and his colleagues also suggests that, after the rivers cross a long stretch of plains and approach the ocean, they only transport very fine suspended particles of mud. In these regions, they may develop a more meandering configuration. Geologist Maurício Martinho dos Santos, a former PhD student of Almeida and now a postdoctoral researcher at the Rio Claro Institute of Geosciences and Exact Sciences at São Paulo State University (Unesp), recently identified a rare occurrence of fine sediments deposited by a meandering river that existed before the Cambrian, more than 540 million years ago, in an outcrop in Scotland. The discovery, shared with Geraint Owen of Swansea University in the United Kingdom, was announced in the journal Precambrian Research in January 2016. “Meandering rivers are the rule today, but they used to be the exception,” Almeida notes.
“This is an interesting hypothesis that needs to be tested in other sedimentary basins,” says geologist Mario Luis Assine, a professor at Unesp Rio Claro who is an expert in ancient and present-day fluvial systems. “The model used by Almeida and his colleagues proposes to explain why meandering rivers were rare in the Silurian. There is a possibility that more of them existed and that we just haven’t yet correctly identified them,” Assine points out, adding, “We don’t see the rivers. We only see their sedimentary deposits and we make interpretations on that basis.”
Architecture of pre-vegetation alluvial systems of the Tombador Formation (Mesoproterozoic, Northeastern Brazil (nº 2011/50280-4); Grant Mechanism: Regular Research Grant; Principal Investigator: Renato Paes de Almeida (IGc-USP); Investment: R$166,64.99.
ALMEIDA, R. P. et al. The ancestors of meandering rivers. Geology. V. 44 (3), p. 203-6. March 2016.