The Codó Formation is an important geological unit in Brazil, representing the only record of Neoaptian rocks exposed along the Brazilian equatorial margin. This unit consists of bituminous black shales, limestones and evaporites, which are particularly well represented in the south and east margins of the São Luís-Grajaú Basin, around the towns of Codó and Grajaú, State of Maranhão. These areas were investigated in order to: 1. improve the depositional system, discussing the hypothesis that the Codó Formation was produced in a lacustrine setting; and 2. reconstruct the paleohydrological conditions with basis on the integration of facies, stratigraphy, petrography and isotope (C, O,Sr and S) data. Hence, the field data presented herein confirmed a lacustrine system for the Codó area, where prevailed stable, well-stratified, saline lakes characterized by periods of closure, anoxia and salt precipitation in the central saline lakes. On the other hand, ephemeral conditions with development of a sabkha/saline pan complex prevailed in the Grajaú area, where salts precipitated mostly in the marginal portions of the system (i.e., marginal saline pans and mudflats).
Studies focusing facies and stratigraphy also revealed that in both areas the Codó Formation is arranged into several shallowing-upward cycles formed by progradation of marginal into central lake deposits. Three types of cycles were distinguished, referred to here as lower, intermediate and higher rank cycles. The lower rank cycles correspond to millimetric interbeddings of: a) bituminous black shale and evaporite; b) bituminous black shale and calcimudstone; c) bituminous black shale and peloidal wackestone-packstone; d) grey/green shale and calcimudstone; e) grey/green shale and peloidal wackestone-packstone; f) grey/green shale and ostracodal wackestone/grainstone; h) ostracodal wackestone/grainstone and/or calcimudstone with cryptomicrobial mats and ooidal/pisoidal packstone. These are attributed to seasonal deposition with basis on their regular nature forming very thin cycles resembling varves.
The intermediate rank cycles average 1.7 m thick and are formed by complete and incomplete cycles. Complete cycles show an upward transition from central to intermediate and then marginal facies associations, and include two types: C1 cycles with central lake deposits consisting of evaporites and black shales; and C2 cycles with central lake deposits formed by gray/green shale. Incomplete cycles are those formed by successions lacking at least one of the facies associations, consisting of either central and intermediate lake deposits (cycles I1) or intermediate and marginal lake deposits (cycles I2).
The higher rank cycles average 5.2 m thick and consist of four depositional units, which display shallowing-upward successions formed by both complete and incomplete, intermediate rank cycles that vary their distribution upward in the section, and are bounded by sharp surfaces. Unit 1, the lowermost one, averages 2.7 m in thickness, being entirely composed by thin I1 cycles. Unit 2 averages 5.2 m thick, and displays all of the aforementioned intermediate cycles, especially complete ones. Unit 3, averaging 2.6 m thick, consists of 80% of cycles I2. Finally, unit 4, which averages 2.2 m in thickness, displays only incomplete cycles, though its uppermost part was not preserved due to erosion during the development of the Aptian sequence boundary.
The detailed sedimentological characterization and the stratal stacking patterns of the intermediate and higher rank cycles support a genesis linked to syn-sedimentary tectonic activity, particularly suggested by high facies variability, limited lateral extension, as well as frequent and random thickness changes of the intermediate-rank cycles. Additionally, the matching between the four higher rank cycles with four stratigraphic zones having different styles of soft-sediment deformation structures previously described in the literature as resulting from seismic activities, is a further argument to corroborate this interpretation. Therefore, the several episodes of lake shallowing recorded in the intermediate and higher rank cycles of the Codó Formation are attributed to fluctuations in the lake water level, triggered by seismic pulses alternating with sediment deposition. The petrographic analysis of the evaporites from the Codó Formation allowed to better defining both the lake-sabkha-saline pan depositional system and the post-depostional histories. Seven evaporite morphologies were recognized: 1. chevron (selenite) gypsum; 2. nodular/lensoidal gypsum/anhydrite; 3. acicular gypsum; 4. mosaic gypsum; 5. brecciated gypsum/gypsarenite; 6. pseudo-nodular anhydrite/gypsum; and 7. rosettes of gypsum. Despite of this large variety of evaporite phases, the chevron gypsum, the nodular/lensoidal gypsum/anhydrite and the brecciated gypsum/gypsarenite record the preservation of primary features. The association of these morphologies with deposits displaying cyclic horizontal bedding, attributed to lake level fluctuations eventually culminated with subaerial exposure, reinforces this interpretation. Even acicular gypsum and mosaic gypsum, which replaced the chevron and brecciated gypsum/gypsarenite, respectively, formed under the influence of the depositional surface. Burial phases of gypsum are only recorded in the pseudo-nodular anhydrite/gypsum, attributed to salt mobilization induced by halokinesis. In addition, rosettes of gypsum, which crosscut the other evaporite morphologies, diagenetic in origin, have probably formed as the latest evaporite phase of the study area, under the influence groundwater and/or surface weathering. In the present research, isotope studies aiming paleoenvironmental purposes were motivated by both confirmation of strong depositional influence for at least great part of the evaporites from the Codó Formation (i.e., primary and eodiagenetic gypsum), and the low diagenetic modification recorded for the limestones. Results of these approaches show that expansion/contraction cycles in both studied areas were accompanied by significant changes in isotope values. The wide dispersion of Sr and S isotope data within individual depositional cycles reinforces the lack of significant diagenetic modification as suggested by the petrographic analysis, and confirms the utility of these isotopes as environmental tools. Additionally, a non-marine brine source is suggested by 87Sr/86Sr ratios ranging from 0.707824 to 0.709280, which are higher than those from late Aptian seawater (i.e., between 0.70720 and 0.70735). The δ34S varies from 16.12 to 17.89 %o(V-CDT) in the Codó area, which is also in disagreement with late Aptian marine values (ranging from 13 to 16 %o(V-CDT)). Both geochemical tracers were influenced by facies characteristics, and thus a model is provided where expansion of saline pan/lake systems led to decreasing 87Sr/86Sr values due to the inhibition of the 87Sr from clay minerals originated during the internal draining of mudflats. During expansion peaks, the 87Sr/86Sr values were lower due to submergence of mud flats and introduction of external 87Sr-depleted waters related to weathering of Permian to Neocomian marine limestones and evaporites, as well as Triassic to Neocomian basaltic rocks. Furthermore, the sulphur isotope values decrease in the southern margin of the basin from 14.79 to 15.60 %o(V-CDT) probably due to increased evaporation in shallower water settings. While the studies of Sr and S isotopes emphasized the evaporites of the Codó Formation, the analysis of C and O isotopes were carried out on the carbonates. The data revealed a wide distribution of dominantly low δ13C and δ18O values, ranging from –5.69‰ to –13.02‰ and from –2.71‰ to –10.80‰, respectively. It was also observed that these ratios vary according to seismically-induced shallowing-upward cycles, in general becoming lighter in their bases, where central lake deposits dominate, and progressively heavier upward, where marginal lake deposits are more widespread. In addition to confirm a depositional signature for the analysed samples, this behavior led to introduce a seismic-induced isotope model. Hence, lighter isotope ratios appear to be related with flooding events promoted by subsidence, which resulted in the development of a perennial lake system, while heavier isotope values are related to ephemeral lake phases favored through uplift and/or increased stability. Furthermore, the results show that a closed lake system dominated, as indicated by the overall good positive covariance (i.e., +0.42 to +0.43) between the carbon and oxygen isotopes, though open phases are also recorded by negative covariance values of –0.36.
|
