Make your own free website on

Geologic observation of natural and artificial recovery processes of Brazilian tropical forest destroyed by debris flow and bauxite mining

Akihisa Motoki

Department of Mineralogy and Igneous Petrology, Rio de Janeiro State University (DMPI/FGEL/CTC/UERJ), Rua São Francisco Xavier 524, Bloco A, MaracanE Rio de Janeiro, Brazil.

Proceedings of International symposium on application of natural materials for environmental goetechnology, 248-257.


Tropical forests have enormous amount of biomass that permits highly divergent species of animals and plants. In uncountable number of undiscovered species, it is expected possible presence of raw materials for new medicines to be applied to the diseases presently considered to be incurable. The plants of giant tropical forests, as Amazon Rain Forest, perform a great amount of photosynthesis, absorbing an immense quantity of atmospheric CO2. In this sense, tropical forests are watched with keen interests by certain groups as an effective method to prevent global worming, on going since the industrial revolution. The forests retain heavy rainwater preventing and/or mitigating natural disaster. So, the effectiveness of the forests becomes the centre of attention for the metropolis that is neighbouring maintains area, as Rio de Janeiro. In these forms, in recent years the importance of tropical forests is increasingly emphasised as a solution of environmental problems.

By means of forest destruction, a great amount of biomass transforms to atmospheric CO2. On the other hand, during its reconstruction, the same amount of CO2 is absorbed from the atmosphere to form the biomass of glowing plants. In recent years, Due to worldwide anxiety of global worming, in December 1988, the industrial countries agree to limit total CO2 emission, which has strong greenhouse effect, and registered the quantitative objectives of each country in the Kyoto Protocol. However, the estimation of CO2 absorption effect by forests is not took account in the protocol, due to unsure factors, such as forest fire.

Destruction and recovery processes of tropical forests are highly dependent on local geologic condition, and the elucidation of their mechanism is important not only for pluvial disaster mitigation but also global worming prevention. The present paper shows geologic observations of natural and artificial recovery process of Brazilian tropical forest, respectively of Tijuca Forest (Floresta da Tijuca), State of Rio de Janeiro, and Poços de Caldas bauxite mine, State of Minas Gerais (Fig. 1). By means of these examples, the author would like to emphasise the importance of tropical, in order to contribute to the forest maintenance and repairing.

Fig. 1 - Locality map of Rio de Janeiro and Pocos de Caldas.

Regional geology and rock weathering of Rio de Janeiro area

he refereed area corresponds to one of continental collision zone of Pan-African Orogenic Collage that took place between late Proterozoic and early Phanerozoic to form West Gondwana continent. The metamorphic belt of Rio de Janeiro area is called as Ribeira Metamorphic Belt, and formed at the last stage of above-mentioned continental collage process of 550Ma (Brito Neves, et. al., 1999; Campos Neto, 2000a; Trouw, et al., 2000).

The main rocks are: 1) biotite gneiss of pelitic origin and related migmatite, distributed at the Tijuca National Park forest area (Fig. 2A), a mountain area at the back of Rio de Janeiro city with relative height of 1000m; 2) augen gneiss of granitic origin, which forms great rock exposures and offers famous sightseeing points, as the Pão de Açúcar (so-called Sugar Loaf; Fig. 2B) and Corcovado; 3) later biotite granite (Fig. 2C) of dyke-like tabular rock bodies of 100m thick and several kilometres long (Penha, et al., 1979; Pires, et al. 1982; 1989). These old rocks are cut by early Cretaceous (130Ma) dyke swarm of basaltic composition, which is related to Pangea super continent break-up.




Fig. 2 - Main rock types and their weathering behaviour of Rio de Janeiro area: A) biotite gneiss at Tijuca Forest; B); augen gneiss at Pao de Acucar (Sugar loaf); C) biotite granit at Gavea rock (Pedra da Gavea).

Except of the basaltic rocks of very small amount, all these old rocks are of granitic composition, being mutually similar in their mineralogical and geochemistry. However, their weathering behaviours under tropical climate are highly variable (Motoki & Vargas, 1992). The biotite gneiss and related migmatite are easily weathered and form thick regolith of more than 30m, permitting development of dense tropical forest. The contact between latelitic regolith and non-weathered hard rock is relatively sharp, intercalating a thin transition zone of less than 1m thick. The regolith contains generally no boulders. The rock blocks, frequently angular ones, can be observed only in zones of debris-flow and mountainside collapse. In shallow depth of the gneissic mountain bodies, highly developed joints parallel to the surface are observed. The rockslides that take place along these fractures form sub-vertical rock-exposing walls of more than 100m high. Contrary to the expectation, the directions of these joints are generally not parallel to that of the gneiss structure. That is, these joints develop without relation to the gneiss structure, but parallel to the surface. So the rockslides take place even in oblique or perpendicular directions to the gneiss structure.

The granite shows contrastive weathering behaviours to those of biotite gneiss. Because of many large in-situ boulders of more than 10m scattered on the land surface, the forest development is not so well as that of the biotite gneiss zone. The boulder size increases underground, grading into fractured non-weathered rock body in more than 30m´s depth. The boundary between the regolith and non-weathered hard rock is not sharp, and many boulders are included in the regolith. The rock texture is homogeneous and no parallel fractures are present.

The augen gneiss presents intermediate weathering behaviours between those of the biotite gneiss and of the granite. The surface boulders are present but less in number, and their form is elliptic. Similar to the biotite gneiss, surface parallel fractures often occur and the rockslides make large rock exposures, and some of them, as Pão de Açúcar (Sugar Loaf), are famous points of tourism. The dykes of basaltic composition of more than 4m thick are of low cooling rate, so they are cause-grained being composed of gabbro. The gabbro and the granite are highly different in mineralogical and chemical composition, although, their weathering behaviours are similar.

Above-mentioned observations lead the conclusion that the weathering behaviours of the rocks of Rio de Janeiro area under the tropical climate are dependent rather their texture than the composition. In the rocks with strong mineral orientation, such as biotite gneiss and migmatite, the meteoric water permeates easily into the rocks up to a large depth along the mineral orientation plains, especially that of biotite. So, the rock weathering is rapidly advances. On the other hand, the rocks without mineral orientation, as granite and gabbro, has no water permeation plains, so the rock weathering proceeds slowly along the few fractures. Because of intermediate texture, the augen gneiss shows intermediate behaviours.

Natural recovery process of the Tijuca Forest

The Tijuca National Park (Parque Nacional da Tijuca) is the mountain area of 1024m high, situated in the back of Rio de Janeiro city, the second major seaport metropolis of Brazil. This area had been covered by Atlantic Forest (Mata atlâtica), but was deforested since 17th century and almost totally destroyed in 19th Century by means of coffee plantation installation. However, in late 19th Century, the second Brazilian emperor, Dom Pedro II, ordered to reforest the Tijuca area with objective of resolve water shortage problem. In obedience to the order, Major Archer and his 6 servants made best effort planting natural species during 13 years, and finally achieved the reforesting work in success.

On February 13, 1996, a strong pluvial impact of 300mm during 2 hours took place at Rio de Janeiro. This pluvial impact after a hundred years resulted immense urban disasters, such as landslide and inundation. The Tijuca Forest also attacked by the pluvial impact. Due to debris-flows and mountainside collapse, the many trees fell down opening forest destruction zones.

Most of the forest disasters occurred in biotite gneiss area in which are covered by thick regolith. In granite area, there were boulder falls, but no debris-flow took place. In the forest destruction zones, all of the trees fell downward, indicating presence of high-speed water flow on the surface. For this reason, it is considered that the forest destruction were not due to coherent landslide between regolith and non-weathered hard rock, but attributed to debris-flows. Provably, it was not easy for rainfall water to permeate in the regolith and reach up to the contact zone with the hard rock in a short time, because the regolith has high contents of clay minerals.

The debris-flows and mountainside collapses are classified from their mode of occurrence and form in 3 types: 1) valley bottom (Fig. 3A); 2) mountain slope (Fig. 3B); 3) rocky scarp (Fig. 3C). The valley bottom type debris-flows took place along valley bottom of with gentle dip, 15 ~ 25°, being most common type. The forest destruction zones are variable in size, but generally are narrow and long. In comparison with Japanese typical debris-flows, the dimension is smaller, and no alluvial fans, that are considered to be products of past debris-flows, were observed.




Fig. 3 - Forest destruction zones of the Tijuca Forest: A) valley bottom type, 96A; B); mountain slope type, 96C; C) rocky scarp type, 96D.

One example is debris-flow 96A, of 160m long and 8m wide in its upper stream and 20m wide in its lower reaches. The maximum valley dip is 25° and the mean one is 20°, that is, lower than stability angle. Due to the low inclination, regolith blanket before the debris-flow was relatively thick, being of 5m. In the forest destruction zones, many angular boulders of 1~2m in size were observed. In general, such rock blocks are rarely found in biotite gneiss area. Detailed field observations revealed that this debris flow is constituted by 4 segments: from the upper stream to the lower reaches, of 20, 20, 30, and 80m long; 8, 12, 15, 20m wide. At the top of each segment, there is an outcrop of 10m wide, dipping 35~40° (Fig. 4A). Above-mentioned angular boulders might be generated there by means of local collapse and rockslide. At the lower most reaches of the debris-flows, there is terminal deposit area with abundant clay minerals and organic materials of 10m long dipping less than 5°.



Fig. 4 - A) top outcrop of the 96A after 3 months; B); terminal deposit of the 96C area after 2 yars.

In the fieldwork 3 months after the debris-flow, some broadleaf plants of 30m tall are found. However, the zone was not grassy, exposing directly the soil. After 2 years, the zone was overgrown with weeds, but the trees were only of 5m tall, being lower than those of adjacent area, of 20m. After 4 years, the trees were enough tall and the natural forest recovery was concludes. Due to the low dip angle, the valley bottom type debris-flow zones are abundant in ground water, surface water, remnant soil, and organic materials, which make possible such rapid natural recovery.

The mountain slope type corresponds to events of intermediate characters between debris-flow and mountainside collapse. The dip angle is the same or higher than the stability angle, being, 30~35 degrees. Generally, the forest destruction zones are wide and composed by only one segment. This type takes place on general mountain slope and not along streams. The debris-flow 96C, an example of this type, is 160m long and 40m wide, and its slop dips in 35 degrees. The regolith before the debris-flow was 2m thick, but the trees on the regolith were 20m tall. The angular rock blocks that take place in the valley-bottom type are present but less in number, and small rock exposures crop out everywhere on the forest destruction slope. On the top of the debris-flow, there is a large outcrop of 20m wide and 5m high, dipping 40°, and rockslide was observed its surface. The deposit area present at its lower most is large, more than 100m wide, and the surface is sub-horizontal, dipping less than 3 degrees.

The 96D, another example, is larger, 300m long and 80m wide, with terminal deposit area of 200m long. The top outcrop is 40m high and 80 wide, dipping 40~60°, forming a rocky scarp type forest destruction zone.

In the fieldwork after 3 months of the event, almost no natural forest recovery was observed, except of lower most terminal deposit area (Fig. 4B). After 4 years, only almost small recovery was noted, being comparable to that of the valley bottom type after 3 months. The natural recovery is very slow then, even today, after 16 years, no notable advance is observed. On the other hand, the terminal deposit area is almost completely recovered. In the mountains slope type events, the original dip is higher than the stability angle. So, surface materials such as weathered soil slip down from the surface in normal condition and water and organic materials cannot be retained. So the natural recovery is very difficult. Before the debris-flow, the roots of the trees had retained the soil. Without trees, the soil cannot be retained, and without soil, the trees cannot glow.

The rocky scarp type events took place at high-angle mountains side near the top. They correspond rather to scarp collapse than debris-flow. They are composed only of one segment, more than 40m wide and their slope angle is high, being more than 50°. The regolith before the event was only less than 50cm thick, and had been retained exclusively by plant roots. The collapse plains show large outcrops. At the Tijuca forest, there are many rock exposures formed by past collapse events. The lower most terminal deposit have relatively high angle surface, dipping 25°. The fieldwork 3 months after the event revealed that the deposit have been submitted to erosion since their formation. On the rock exposure, obviously no soil is present and the natural recovery process is not in progress.

Artificial forest recovery at Poços de Caldas bauxite mine

Bauxite miming is frequently considered to be a most harmful mineral explotation from environmental viewpoints because all of the surface materials in a large area are removed during explotation (Fig. 5). However, the CBA (Companhia Brasileira de Alumínio) mine, Poçs de Caldas, State of Minas Gerais, adopts the A horizon preservation methodEin order to resolve the problem, moreover, contributes to pulp production.

Fig. 5 - Bauxite explotation, CBA mine, Pocos de Caldas.

Bauxite ores normally occur in tropical rainforest zone near the equator. But, Poçs de Caldas is present in subtropical zone and on a highland of 1100m, of temperate zone climate. The ore formation even in such climate condition might be due to the mother rock, phonolite and nepheline syenite. These rocks are very vulnerable to chemical weathering because of absence of quartz, presence of nepheline and high contents of alkaline feldspar.

Above-mentioned area is underlain by felsic peralkaline undersaturated rocks with high K2O in subcurcular are of 30km x 25km, forming Poçs de Caldas Alkaline Intrusive Complex Rock Body (Complexo Alcalino Intrusivo de Poçs de Caldas). The basement rocks are late Precambrian (630Ma) metamorphic rocks, high K2O granites, alkaline feldspar syenite related to continental collision (Campos Neto and Caby, 1999; Campos Neto, 2000b). Above-mentioned metamorphic rocks include ultra-high pressure metamorphic zone (Parkinson, et al., 2001), that is, the oldest one of the world and the firstly discovered in the American Continents.

This alkaline complex intrusive rock body was considered to be the largest volcanic caldera edifice (Bjørnberg, 1956; 1959; Ellert, 1959; Ellert, et al., 1959), of villes-type (Fraenkel, et al., 1984; Loureiro and Santos, 1988). But, the volcanological considerations based on detailed fieldwork data of structural geology, contact outcrop observations and landform studies lead to another conclusion. That is, this rock body does not correspond to a volcanic edifice but to a denudated shallow intrusive rock body, and the present morphological characteristics are attributed to differential weathering and erosion (Motoki and Oliveira J.L.S., 1987; Motoki, et al., 1988; Motoki, 1988). The intrusive age is defined to be about 85Ma by means of Rb-Sr dating of mineral isochron (Kawashita, et. al., 1994). A volcanic conduit (Oliveira, J.I., 1986) of the present intrusive body holds uranium ore (Loureiro and Santos, 1988). The Morro do Ferro, that means “iron hillE is famous because of the highest natural radioactivity of the world.

The bauxite deposits are originated from various types of the alkaline rocks, and the ore bodies are present in phonolite and of nepheline syenite. From the surface to the interior, the regolith is divided to the A-horizon (black soil with high organic materials) of 50cm, B-horizon (yellow clay mineral-rich layer) of 1m, C-horizon (weathered rock) of 5 to 10m, and D-horizon (not weathered hard rock). All of the C-horizon and part of the B-horizon are explotation objects. The main minerals are gibbsite and amorphous aluminum hydroxides.

For bauxite extraction, the A-horizon and overlying forest must be removed. After the mining, natural forest recovery does not enter in action because the D-horizon is exposed directly on the surface. Such situation occurs at some areas within the alkaline complex body because of predatory explotation made by incontinent landowners.

The CBA mine operates the forest recovery in consideration. Before the mining, the A-horizon is removed and transported to the deposit area. After the total explotation, the A-horizon material at the deposit area returns to the original area and cover it again (Fig. 6A). The recovered areas are used according to the desire of their landowners. In many cases, the areas are reforested and produce woods (Fig. 6B). Due to abundance of organic material in recovered soil, artificial forest recovery is performed in a short period. The CBA provides various species of plants, either natural ones or imported ones. Since the decade of 1950, in mean 60.000 trees have been planted by year



Fig. 6 - Artifitial forest recovery at CBA mine, Pocos de Caldas gA-horizon preservationh: A) A-horizon material covering the original area again; B) reforesting on recovered area.

In the forests in reconstruction, many plants are growing, so the forests absorb a great amount of CO2 form the atmosphere. Once the forest recovery is completed, the plants stop to grow and the forests stop absorb CO2. Each plant can absorb CO2 by means of photosynthesis, on the other hand, the biomass of dead plants returns to atmosphere as CO2. In stable forests, total CO2 balance must be zero. This argument is applicable also to natural stable forests.

The recovered forests can provide woods. After the deforesting by cutting, new reforesting process must be performed. If so, the atmospheric CO2 absorption continues. Since the decade of 1950, the CBA mine has absorbed CO2 by this way. Contrary to the popular knowledge, stable forests, such as Amazon Rainforest has no contribution to global warming prevention for they do not absorb CO2. However, exhausted bauxite mines do contribute during artificial recovery process.

Recommended reforesting methods

Because of difference in behaviours of surface soil retention, each type of destruction area of the Tijuca Forest has specific recovery method.

he valley bottom type debris flows leave gentle slopes that can retain surface soil in natural condition. Because of the good condition for natural spontaneous recovery, in principle, no civil engineering works or artificial reforesting is necessary.

The mountains slope type debris flows leave steep slopes that cannot retain surface soil, and small civil engineering works are necessary to begin spontaneous forest recovery. Artificial reforesting is recommended but not essential. Without them, natural recovery can take almost 100 years.

The scarp type collapses leave large rock exposures so natural recovery is almost impossible. It is necessary to plant precursor species that cover on the rock surface for minimum condition to retain surface soil on the rocks.

Above-mentioned recovery methods are applicable only to the destruction zones appeared in the forests. Because of seeds and plant body fragments provided by neighbour forest zone, these areas are spontaneously reforested if water, soil and organic materials are present on the surface. However, in case of total deforesting hills, spontaneous forest recovery does not enter in action even more than 50 years later. So, artificial reforesting works are necessary.

The forest recovery after predatory bauxite mining is difficult because of rock outcrops left on the surface. In this sense, the application of A-horizon preservation methodEand artificial reforesting is desired. This method can be applied to all of mines of surface ore.

From the viewpoint of global warming prevention, reforesting areas are much more important than stable ones, because stable forests do not absorb atmospheric CO2. The countries that cannot reduce CO2 emission according to the Kyoto Protocol can compensate the deficiency by means of forest construction or reconstruction in their territory or that of the other countries. The reforesting must take into account when CO2 reduction amount is computed.


The present paper is a results of helpful suggestion of the Commission 111, Japan Society of Progress in Sciences (JSPS) and its chairman Dr. Hideo Minato. The author is highly grateful to them.


Bjørnberg, A.J.S., 1959. Rochas clásticas do Planalto de Poços de Caldas. Bol. Fac. Fil. Ciênc. Let. Univ. São Paulo, 237, Geologia 18, 65-132.

Brito Neves, B.B.; Campos Neto, M.; Fuck R.A. 1999. From Rodinia to Western Gondwana: an approach to the Brasiliano-Pan African cycle and orogenic collage. Episode, 22-3, 155-166.

Campos Neto, 2000a. Orogenic systems from sotywestern Gondwana: An approach to Brasilinao-Pan African cycle and orogenic collage in southeastern Brazil. In: Cordani, U.G., Milani, E.J., Thomáz Filho, A., and Campos, D.A. (Eds.), Tectonic evolution of south America, 335-365, Rio de Janeiro, Brazil.

Campos Neto, 2000b. Terrane accretion and upward extrusion of high-pressure granulites in the Neoproterozoic nappes of southeast Brazil: Petrologic and structural constraints. Tectonics, 19-4, 669-687.

Campos Neto, M. and Caby, R., 1999. Neoproterozoic high-pressure metamorphism and tectonic constraint from the nappe system south of the São Francisco Craton, southeast Brazil. Precambrian Research, 97, 3-26.

Ellert, R.; Bjørnberg, A.J.S.; Coutinho, J.M.V., 1959. Mapa geológico do maciço alcalino de Poços de Caldas, Brasil. Bol. Fac. Filos. Ciênc. Let. Univ. São Paulo.

Ellert, R., 1959. Contribuição à geologia do maciço alcalino de Poços de Caldas. Bol. Fac. Filos. Ciênc. Let. Univ. São Paulo, 237, Geologia 18, 1-64.

Fraenkel, M.O.; Santos, R.C.; Loureiro, F.E.V.L.; Muniz, W.S., 1984. Jazida de urânio no planalto de Poços de Caldas - Minas Gerais. Principais Depósitos Mineris do Brasil, Vol. 1, DNPM, 89-103.

Kawashita, K., Mariques, M.M., and Ulbrich, H.H.G.J. 1984. Idades Rb/Sr de nefelina sienitos do anel norte do Maciço Alcalino de Poços de Caldas, MG-SP. Resumos do XXXIII Cong. Bras. Geol., 244-245.

Loureiro, F.E.L.; Santos, R.C., 1988. The intra-intrusive uranium deposits of Poços de Caldas, Brazil. Ore Geol. Rev., 3, 227-240.

Motoki, A., 1988. An outline about problems of volcanic caldera hypothesis of the Poços de Caldas Alkaline Complex Rock Body, Minas Gerais - São Paulo, Brazil. An. VII Congr. Latinoamer. Geol. 1, 309-323.

Motoki, A.; Vargas T.; Chianello E.; Corrêa, F.J.G;, Oliveira, J.L.S; Klotz M., 1988. Nível de denudação atual do Complexo Alcalino de Poços de Caldas, MG-SP. An. XXXV Congr. Bras. Geol. 6, 2633-2648.

Motoki, A; Oliveira, J.L.S. 1987. Reconsiderações vulcanológicas sobre a hipótese de caldeira vulcânica no Complexo Alcalino de Poços de Caldas, MG. Parte I : Rochas sedimentares como corpos capturados, fragmentados e afundados no magma fonolítico. An. IV Simp. Geol. Minas Gerais , 420-433.

Motoki, A.; Vargas, T., 1992. Forma de intrusão de cropos graníticos presentes nos altos picos da Floresta da Tijuca, RJ. Bol. Res. Exp. 37° Congr. Bras. Geol. 1:377-379.

Motoki, A.; Vargas, T.; Marques, M.V., 1996. Descrição geológica e mecanismos de ocorrência dos deslizamentos da encosta gnáissica da Floresta da Tijuca, Rio de Janeiro, formados durante o impacto pluvial concentrado de fevereiro de 1996. Bol. Res. Exp. XXXIX Congr. Bras. Geol. 4, 95-99.

Oliveira, J.I., 1986. Um modelo geológico de mineralizações em caldeira de subsidência. Brasil Mineral , 28, 45-47.

Parkinson, C.D.; Motoki, A.; Onishi, C.T.; Maruyama, S., 2001. Ultrahigh-pressure pyrope-kyanite granulites and associated eclogites in Neoproterozoic nappes of southeast Brazil. Fluid/slab/mantle Interactions and ultrahigh-Pminerals, 87-90.

Penha, H.M.; Ferrari, A.L.; Ribeiro, A.; Amador, E.S.; Pentagna, F.V.P.; Junho, M.C.B.; Benner, T.L., 1979. Projeto carta geológica do Estado do Rio de Janeiro, Folha Petrópolis. Relatório final e mapa, DRM/UFRJ, Niterói.

Pires, F.R.M.; Brauer, E.H.; Crescêncio Jr., F; Gonzaga, G.G.; Torres, M.G. 1989. Lito-estratigrafia dos gnaisses na Floresta da Tijuca, Rio de Janeiro, RJ. Bol. Res. Simp. Reg. Sudeste. 153-154.

Pires, F.R.M.; Valença, J.G.; Ribeiro, A. 1982. Multistage generation of granite in Rio de Janeiro, Brazil. An. Acad. Bras. Ciências, 54/3:563-574.

Trouw, R.; Heilbron, M.; Ribeiro, A.; Paciullo, F.; Valeriano, C.M.; Almeida, J.C.H.; Tupinambá, M.; Andreis, R.R., 2000. The central segment of the Ribeira Belt. In: Cordani, U.G., Milani, E.J., Thomáz Filho, A., and Campos, D.A. (Eds.), Tectonic evolution of south America, 287-310, Rio de Janeiro, Brazil.