br.gif (641 bytes)


Geological and Paleontological Sites of Brazil - 095


Date: 01/04/2000

Augusto J. Pedreira
CPRM – Geological Survey of Brazil
Avenida Ulysses Guimarães, 2862 - Centro Administrativo da Bahia
41213-000 Salvador, Bahia, Brasil

© Pedreira,A.J. 2000. Sergi River Canyon, Brazil. In: Schobbenhaus,C.; Campos,D.A.; Queiroz,E.T.; Winge,M.; Berbert-Born,M. (Edit.) Sítios Geológicos e Paleontológicos do Brasil. Published 01/04/2000 on Internet at the address  [Actually]


(The above bibliographic reference of author copy rights is required for any use of this article in any media, being forbidden this use for any commercial purpose)

In the low course of the Sergi river, about seven kilometers northwest of the town of Santo Amaro, Bahia State, there is a deep canyon cut into sandstones that crop out along the western border of the Recôncavo Basin, of Mesozoic age. These sandstones, described for the first time in the forties, were named Sergi Formation. Their age is neo-jurassic and they constitute the main petroleum reservoir of that basin. The Sergi Formation deposited before the rifting that gave birth to the Recôncavo Basin, belonging to to Sequence of the Continent. It is underlain by the Aliança and Afligidos formations, of the same sequence. In the Sergi Formation were recognized two facies: a fluvial facies composed by medium scale cross bedded coarse grained sandstones, and an eolian facies that comprises fine grained sandstones with medium grained laminae, and large scale cross bedding. These can be of tabular or trough types and are in sets separated by horizontal beds that represent 1st. order bounding surfaces. The surfaces that separate cross bedded sets of different dips are interpreted as 2nd. order bounding surfaces. Between some horizontally bedded sets interpreted as interdunes, there are beds with convolute stratification that result from the deformation of water saturated sediments. The diverse sedimentary structures of the Sergi Formation sandstones are interpreted as composed and complex dunes separated by interdune areas, in a neo-jurassic sand sea. 


    Close to the western limit of the Recôncavo Basin, the oldest oil producing basin of Brazil, there is a north-south trending sandstone crest tens of kilometers long. In this crest, the Sergi river scoured a canyon more than a hundred meters deep, that runs since the Sergi station in the railroad that connects the towns of Cachoeira and Santo Amaro, until close to its confluence with the Subaé river. This region, known since long time ago, began to be geologically studied in the first half of the forties, in the beginning of the oil exploration in the basin, since the rocks that crop out there are its major oil reservoir (Lanzarini, 1996). In the escarpments of the canyon, the rocks show sedimentary structures indicative of eolian and fluvial processes that occurred more than 150 million years ago. Besides the geologic aspects, the tributaries of the Sergi river in both margins of its valley when enter in the canyon, form water falls that add beautiful landscapes to the geologic features.


    The Sergi River Canyon is located about seven kilometers northwest of the town of Santo Amaro (Bahia State , Brazil), ) limited by the coordinates 12o 27' S-12o 33' S and 38o 45' W- 38o 50' W, in the valley of the Sergi river (figure 1). This river, the main tributary of the Subaé river that crosses that town, has its source close to the town of Feira de Santana, discharging in it after traversing the canyon.
From the city of Salvador, the site is reached through the Salvador-Feira de Santana highway (BR-324), until its junction with the BR-240 road, that connects the towns of Santo Amaro and Cachoeira.
From the town of Santo Amaro, there are at least two alternatives to arrive at the canyon: the first one is to follow the road to Campinhos, until its junction to the village of Pedra, leaving the vehicle at the first railway crossing of the latter road and walking along the railroad (about 5km) northwestwards until the Cabeça do Cavalo bridge.. The second alternative is to follow the BR-420 road 22km towards the town of Cachoeira and entering in the village of Cepel, drive to the locality of Jambeiro; leaving the vehicle there, walk downhill until the Sergi station in the river valley and walk along the railroad towards the town of Santo Amaro.


    Along the fault that borders the Recôncavo Basin by the west, there are parallel belts of sandstone cuestas, forming long escarpments (Ghignone, 1979). That fault passes trough the towns of Amélia Rodrigues, Campinhos, Afligidos and Maragojipe (Figure 1).

fig1.jpg (85119 bytes)

Figure 1 - Location of the Sergi River Canyon

    Sandstones that compose one of these belts were described in the valley of the Sergi river, along a stretch of about 5km between the km 89 of the railroad and the Sergi station (Figure 2). They were named Sergi sandstone, as a member of the Brotas Formation, by the geologist E.F.Taylor, of the former National Petroleum Council (Taylor, 1948; after Baptista et al., 1984). According to the original definition, it is a white, red, purple and tan, coarse to medium grained, arkosian, friable sandstone, with quartz pebble zones and siliceous concretions with 5cm diameter; the sedimentary structures described then, were cross bedding. The thickness of the formation  ranges from 9 to more than 90 meters, in the western border of the basin, attaining a maximum of 440m in Itaparica island.

foto1.jpg (59781 bytes)

Figure 2 - Aerial photograph (Petrobrás/LASA, Flight 53-D, 1959, 1:25.000 scale), showing the regional geology: at the left side, the Precambrian basement, separated from the basinal sediments (Afligidos and Sergi formations) by a fault. The canyon is below the scale and its arrow indicates the north.

    In 1958, the Basin Study carried out by the Petróleo Brasileiro S.A .-Petrobrás (Viana et al.,1971), changed the rank of the Brotas Formation to group and of the Sergi sandstone to formation, maintaining the original stratotype. Its age, presumed to be Late Jurassic, is indicated by the presence of Tithonian non-marine ostracodes; the Tithonian in the Recôncavo Basin takes the local name of D. João Stage. Ghignone (1979) adds that the quartz pebbles form conglomerate lenses, occurring other ones of micaceous siltstone, both with irregular distribution, and that the Sergi Formation was recognized in Africa with the name of M ' Vily.
The Sergi Formation is part of the Continent Sequence (Asmus & Porto, 1980), having been deposited before the rifting that gave birth to the Recôncavo basin (Pre-rift Phase; Caixeta et al., 1994), together with the underlying Afligidos and Aliança formations (Figure 3). Its contact with the Aliança Formation is gradational or through channels with irregular bottom (Ghignone, 1979).

fig2.jpg (31974 bytes)

Figure 3 - Depositional model for the pre-rift alluvial, fluvial, eolian and lacustrine sediments of the D.João Stage, encompassing Afligidos, Aliança and Sergi formations. Modified from Medeiros & Ponte (1981).

    In the Sergi River Canyon, Medeiros & Ponte (1981) recognized two facies in the formation: 1) eolian facies comprising fine grained sandstones with coarse grained laminae, large scale cross bedding with faceted pebbles where the bedding laminae are tangent to the lower bed surface; and, 2) fluvial facies composed by conglomerates and coarse grained sandstones with medium scale cross bedding.
Studying in detail the eolian sediments above described, Lanzarini (1996) identified dunes and interdunes-sand sheets, interpreting the former as large scale bed forms 500 to 1000m long and tens of meters high. They are stacked barcanoids with strata limited by bounding surfaces due to the migration of foresets according to subcritical angle. These dunes may be associated to other forms of similar scale, composing draas in a Jurassic sand sea.
Using geoprocessing in order to minimize the distortions of field photogrammetry, Carrasco et al. (1996), interpreted the faciology of the Sergi Formation in the canyon as resulting of the piling up of dunes and interdunes. Besides the dunes described by Lanzarini (1996), they tracked three more facies interpreted as interdunes, deformations and bounding surfaces. The interdunes consist of sand deposits with almost horizontal lamination, intercalated to the dune deposits; their length is of a few hundreds of meters with thickness of 1.5 to 2.5m. The deformations are chaotic deposits subordinated to the dunes, developed by overburden on water saturated sediments below the water table. These deformations are characterized by convolute structures, without associated faults; their length ranges from 30 to 80m and their thickness in four meters. The bounding surfaces indicate the migration of dune fields (1st. order); are positioned between dunes indicating the passage of a migrating dune over a stationary one (2nd. Order); and form bands of cross lamination within the stratification cosets (3rd. order).


    The Sergi river canyon occupies a long extension however, the proposed site is restricted to the stretch between the Cabeça do Cavalo bridge and the Sergi station along the railway, because of its relatively easy access and the excellent exposures. Close to the western bridge head, already exist some railroad cuts where crops out the Sergi Formation, represented by friable sandstones with large scale tabular cross bedding (Figure 4).

foto2.jpg (67670 bytes)

Figure 4 - Railroad cut west of the Cabeça do Cavalo bridge.

    After crossing the bridge, the escarpments tabular show large scale cross beds indicative of dunes and 1st and 2nd order bounding surfaces, similar to those studied by Carrasco et al. (1996) (Figure 5).

foto3.jpg (53112 bytes)

Figure 5 - Large scale tabular cross bedding in the lower half of the outcrop, and horizontal bedding in the upper one.

    The 1st order bounding surface is represented by the horizontal stratifications separating the cross bedding cosets, that are interpreted as migrating dunes. The 2nd order one is the inclined line from the left to the right in the lowest coset, that separates two sets of different dips. From this place are seen the escarpments that form the canyon (Figure 6).

foto4.jpg (59086 bytes)

Figure 6 - Escarpments of the Sergi river canyon

    The tributaries of the left and right banks of the Sergi river form water falls several heights, whose flow increases or decreases according to the season of the year (Figure 7).

foto5.jpg (57258 bytes)

Figura 7 - Small water fall formed by a tributary of the Sergi river left bank, photographed in middle march.

    The escarpments in the right bank of the river show large scale tabular cross bedding separated by 1st order bounding surfaces (Figure 8) that represent draas in a sand sea (Lanzarini, 1996).

foto6.jpg (78524 bytes)

Figure 8 - Two sets of large scale tabular cross bedding, separated by a first order bounding surface.

    A large tributary of the right bank forms inside the canyon the Urubu water fall. To reach it is necessary climb down from the railroad and ford the Sergi river, following the trail shown in the Figure 9, that crosses the forest that hides the water fall.

foto7.jpg (61681 bytes)

Figure 9 - Trail to the Urubu water fall, among moss covered sandstone boulders.

    The Urubu water fall is about 20m high and in its foot there is a very deep small lake. In the Figure 10, at the right of the water fall a bed shows large scale tabular cross bedding. The paleowinds that formed these cross beds blew alternatively from left to right and vice-versa. This type of cross bedding may be found in longitudinal dunes (seif), in the "horns" of bachanoid dunes and in the arms of star (rhourd) dunes.

foto8.jpg (69064 bytes)

Figure 10 - Urubu water fall

    In the right bank escarpment of the canyon one may see the outcrop studied by Carrasco et al.(1996) (Figure 11), with convolute stratifications interpreted as the result of water saturated sediments deformation, large scale trough cross bedding and horizontal bedding interpreted as dunes and interdunes, as well as 1st, 2nd and 3rd order bounding surfaces (Figure 5). This escarpment is upstream of the river that forms the Urubu water fall.

foto9.jpg (75090 bytes)

Figure 11 - Escarpment about 50m high, in the southern wall of the canyon. The sector between the lines of vegetation is about fifteen meters thick.


    Presently there are no protective actions in the site. However, as the area is sparsely populated and rather far from urban centers, they may be restricted to the improvement of the access in the side of Sergi station and to the Urubu water fall. In this one, the stone walls are carved with the names of visitors and, owing to the friability of the sandstones, generally the names are deeply carved, being impossible their remotion.


Asmus,E.H.; Porto,R. 1980. Diferenciação nos estágios iniciais da evolução da margem continental brasileira: possíveis causas e implicações. In: SBG, Congresso Brasileiro de Geologia, 31, Camboriú, Anais, 1: 225-233.

Baptista,M.B.; Braun,O.P.G.; Campos,D.A.C. 1984. Léxico Estratigráfico do Brasil. Brasília, CPRM, 560 p. (Convênio DNPM/CPRM).

Caixeta,J.M.; Bueno,G.V.; Magnavita,L.P.; Feijó,F.J. 1994. Bacias do Recôncavo, Tucano e Jatobá. Boletim de Geociências da Petrobrás, 8 (1): 163-172.

Carrasco,B.N.; Fonseca,L.E.N.; Durães,E.M. 1996. Fotointerpretação de fácies e elementos arquiteturais eólicos no afloramento do Canyon do Sergi, Bacia do Recôncavo, Brasil. In: SBG, Congresso Brasileiro de Geologia, 39, Salvador, Anais, 1: 141-144.

Ghignone,J.I. 1979. Geologia dos sedimentos fanerozóicos do estado da Bahia. In: Inda, H.A .V., org., Geologia e Recursos Minerais do Estado da Bahia; Textos Básicos. Salvador, SME/CPM, p. 23-154.

Lanzarini,W.L. 1996. Geometria das unidades genéticas fluviais e eólicas das formações Aliança e Sergi na borda oeste da Bacia do Recôncavo. In: SBG, Congresso Brasileiro de Geologia, 39, Salvador, Anais, 1: 328-331.

Medeiros,R.A.; Ponte,F.C. 1981. Roteiro Geológico da Bacia do Recôncavo. Salvador, Petrobrás/SEPES/DIVEN/SEN-BA, 63 p. (Acordo Petrobrás-SENAI).

Taylor,E.F. 1948. Estado da Bahia. In: Brasil. Conselho Nacional do Petróleo. Relatório de 1946. Rio de Janeiro, p. 148-172.

Viana,C.F.; Gama Jr.,E.; Moura,J.A.; Fonseca,J.R.; Alves,R.J. 1971. Revisão estratigráfica da bacia do Recôncavo/Tucano. Boletim Técnico da Petrobrás, 14 (3/4): 157-178.