The studied outcrops are in two limestone quarries near the city of Corumbá, in western Brazil. The Corcal and Laginha sections were selected for their vertical continuity, fresh rock exposures, accessibility, and availability of previous studies. These sites were visited in June 2018 and detailed lithostratigraphic logs were measured and described with a cm to dm resolution (Figures 2, 3). A total of 225 samples were collected for petrographic characterization and geochemical analyses. Outcrops conditions allowed for collecting samples bed by bed. In case of lenticular or amalgamated beds, the samples were collected considering a thicker interval, in order not to confuse the stratigraphic relationships. The average sample spacing resulted of ∼20 cm. Some intervals were completed using the lithostratigraphic logs of Adôrno et al. (2017) and Fazio et al. (2019), as highlighted in Figures 2, 3. This choice was taken for intervals that were already sufficiently detailed in these previous studies, or that were not possible to access.

Petrographic characterization was performed using a ZEISS petrographic microscope at the University of Brasília (UnB). Seventy-nine thin sections were studied, being 40 from the Corcal and 36 from the Laginha section (Supplementary Material). A plain white paper was placed underneath the thin section and used as a light diffuser to clarify textural features under the microscope (Delgado, 1977).

The classification of the carbonate lithofacies was done according to Wright (1992) and Flügel (2004) on the base of: 1) texture, considering the amount of matrix respect to other constituents; 2) composition, based on the mineralogy and the type of grains; 3) sedimentary structures. Therefore, the name represents the main textural features, and the adjectives refer to composition and structures (i.e., “massive calcimudstone”, “dolomitic sparstone”, “intraclastic breccia”).

The Wentworth (1922) grain size classification was used for the siliciclastic facies. Besides, “shale” and “mudstone” were differentiated, as the former displayed fissile or foliated structure and calcite filled veins, which were absent in the latter. Facies associations were defined according to the carbonate ramp model of Burchette and Wright (1992).

After the analysis at the optical petrographic microscope, which allowed for identifying constituents that needed a more detailed characterization, nine thin sections were selected among the representative lithofacies for SEM analyses. Prior to the analysis the sections were coated with carbon, and then analyzed at the University of Brasília (UnB) using a FEI QUANTA 450 scanning electron microscope with a BSE detector, 20 kV accelerating voltage, working distance of 10–12 mm, and in high-vacuum mode.

For carbon and oxygen isotopes analysis in carbonate (δ¹³C and δ¹⁸O), 116 samples from Corcal and 54 from Laginha were collected with 20–40 cm spacing. The respective lithofacies was always annotated alongside each sample, to constrain the interpretation of the results. Additionally, areas containing veins, fractures, and other post-depositional features, likely to interfere with the primary isotopic signal, were avoided. Fresh rock samples were powdered using a drill with a 3 mm-diameter bit. Prior to the analysis about 300 μg of material were put in glass vials, acclimatized at 72°C, and flushed with He to eliminate atmospheric gases. Afterwards, carbonate samples were reacted with concentrated phosphoric acid during 1 h, and the released CO₂ was analyzed with a Thermo^® Gasbench II connected to a Thermo^® Delta V^Plus mass spectrometer at the University of Brasília (UnB). The isotopic ratios are reported in per mil (‰), using the conventional δ notation, relative to the Vienna Pee Dee Belemnite (VPDB) standard. The analytical precision is ±0.1‰ for both δ¹³C and δ¹⁸O.

The Corcal section is 55 m thick and consists of m-thick bundles of shallow-water limestone with tabular to wavy amalgamated beds and cm-thick beds of mudstone and marl, alternating with m-thick intervals of shales (Figures 2, 4). Neither the base nor the top of the Tamengo Formation is exposed at this site. The lower and upper parts of the section were logged in the northern and southern side of the quarry, respectively. On the other hand, the middle part was completed using the log by Adôrno et al. (2017), as it was considered sufficiently detailed after verification in the field. The split of the profile in these three sub-sections was done to avoid parts that occurred tectonically disturbed by faults and folds, which deformed especially the most ductile shale beds. However, the stratigraphy was adequately exposed for allowing the recognition of the entire succession in continuity.

The lower part of the section (from 0 to 18.60 m) consists of shallow-water limestones organized in 1–2 m thick bundles of wavy and amalgamated beds alternated with bundles of cm- to dm-thick tabular beds (Figures 2, 4). At the microscopic scale, the limestones in this interval consist mainly of calcimudstones, often recrystallized, with some extraclasts of quartz and mica, bioclasts, and intraclasts (Figure 5). Generally, they display a massive structure, but locally evidence of desiccation can be observed, such as birdseyes, anhydrate pseudomorphs, silicification, dolomitization, and micritization (Figures 5B,C,G). Alongside the calcimudstones, there are minor occurrences of packstones and grainstones formed by bioclasts, intraclasts, and siliciclastic extraclasts (Figures 5D,H–J). Grain size is from sand to granule and the intergranular space is filled with micrite or calcite cement. Occasionally recrystallization or dolomitization is present. The structure is massive, with lenticular veins and calcite-filled fractures. In few cases the grainstones display angular and variously elongated intraclasts, sometimes with combining edge, and local ferruginous matrix, suggesting possible brecciation by desiccation. The limestone succession is interrupted by few centimeter to decimeter thick beds of shales and marls, formed by clay to silt-size detrital grains, especially quartz, with massive to finely laminated structures and variable carbonate, occurring as cement or microscopic nodules (Figure 5E). Fazio et al. (2019) noted also the presence of gypsum pseudomorphs.

The middle part of the section (from 18.60 to 29.90 m) consists mainly of massive shales, interrupted by a 2.25 m thick and two decimeter thick limestone beds. The shales are siliciclastic mudstone, sometimes thinly laminated, with minor quartz grains and carbonate cement (Figure 4B). There are rare opaque grains, compaction-related segregation structures, calcite and quartz-filled fractures, and silicified nodules. These are also the lithological intervals yielding the fossils of C. werneri, indicating the upper Ediacaran age of the Tamengo Formation (Figure 5F; Adôrno et al., 2017). The limestone beds consist of calcimudstones and grainstones, similar to those in the lower part of the section.

The upper part of the section (from 29.90 to 54.40 m) consists of limestone arranged in sets of centimeter to decimeter thick beds, varying from tabular to wavy, interchanged with rare millimeter to centimeter thick shale beds, and, occasionally centimeter thick mudstone lenses (Figure 5A). Amorim et al. (2020) also described hummocky cross stratification in correlative outcrops of this interval. As this kind of sedimentary structure is generally highlighted by weathering on the surface of the rock, it is not visible in the Corcal section because the surface of the outcrop is too fresh, being it in an active quarry. At microscopic scale the limestones consist of calcimudstone and grainstone with intraclasts and bioclasts of Cloudina type organisms, as suggested by the tubular shapes (Figures 5H,I). At 34.65 m birdseye porosity was noted, and at 48.00 and 50.50 m there are tabular centimeter thick beds of light, greenish to whitish, clayey sediment, different from the other mudstones, described as possible volcanic tuffs. In addition, at 48.13 m there are pinkish centimetric calcite nodules, and a layer of intraclastic breccia was observed at 49.66 m. The uppermost limestone bed is a massive oolitic grainstone and, above, starts a massive interval of shales with carbonate lenses, described in detail by Fazio et al. (2019). SEM analysis revealed also the presence of fluorapatite in 4 different levels of the Corcal section: 6.50, 10.40, 2.55, and 18.42 m.

The Laginha section is about 100 m thick and displays both the lower and the upper boundaries of the Tamengo Formation. The succession yields a higher variety of lithofacies than the Corcal section (Figure 3).

The lower part of the section (from−5 to 30 m) starts with a dolomitic interval, formed by 2.50 m thick massive bed of light grey grainstone, followed by ∼2.00 m of massive dark gray sandstone with dolomitic matrix mostly composed of very well-rounded quartz grains with approximately equal amounts of grains and matrix. Above, the is ∼0.5 m thick peloidal packstone with fenestral pores, filled with spathic calcite. The contact between the latter two beds is very irregular, indicating an erosional and possibly karstified unconformity (Figure 4C). This interval represents the uppermost part of the Bocaina Formation (Boggiani, 1998). The Tamengo Formation starts with a ∼15 m thick bed of massive polymictic breccia containing various types of intraclasts and extraclasts with composition of quartz, dolostone, limestone, chert, phosphorite, and granite type rocks. They are poorly sorted, with chaotic distribution and with a rounded to very angular shape. The matrix is dolomitic and dark gray, there are irregular voids filled with white spathic calcite (Figure 6A). The contact at the base is erosional. Above the thick polymictic breccia there is a 25 cm thick friable yellowish-brown, massive argillaceous sandstone, with sand-size quartz grains and lithoclasts within an argillaceous matrix (Figures 6B, 7A, 8A,B).

Upwards occurs a mainly carbonate succession, starting with 4.5 m of massive dolostones with numerous randomly oriented fractures, with millimeters to centimeters size and filled with white sparry calcite (Figure 8C). Above, there are 2.80 m of sparstones, arranged in sets of centimeter to decimeter thick tabular beds, dolomitic to calcitic, massive to laminated, containing stylolites and calcite-filled veins, locally brecciated and associated with fluorite (Figures 6C,D, 7D). The succession continues with alternating dolostones and limestones, consisting of laminated facies, with alternating crystalline and sandy laminae, the formers are calcitic to dolomitic, with sparse extraclasts of quartz and mica; and the latters are grain-supported, with calcitic to dolomitic matrix and sand-size extraclasts varying in composition from quartz to k-feldspar (Figures 6E,F, 7B, 8D–G). Authigenic pyrite and phosphatized grains are present, and SEM-EDS analyses revealed traces of zircon and possible anatase. Late compaction structures, such as stylolites and calcite-filled veins, occur frequently. In addition, there are massive wackestones with micritic matrix and abundant extraclasts, especially quartz and muscovite, ranging from fine to coarse sand-size; and grainstones formed predominantly by concentric ooids, occurring as single or composite, often micritized or completely recrystallized, with rare aggregate grains and cemented by calcite. At 22.75 m there is a 7 cm thick bed of intraclastic breccia, grain-supported, with micritic matrix and angular intraclasts, ranging in size from fine sand to granule, with the larger grains of tabular shape, and with abundant fine quartz grains (Figure 7C).

At 26.40 m occurs an irregular massive polymictic breccia, with maximum thickness of 5 m and erosional contact at the base, which laterally meets a light gray limestone in angular unconformity (Figure 7E). It is grain-supported, with limestone intraclasts and grains varying from quartz to dolomite, phosphorite, mudstone, k-feldspar, and pyrite, with micritic matrix. It is worth noting the absence of coarse granitoid clasts from the basement, observed in the breccia at the base of this unit (Figure 6G).

The middle part of the section (from 30 to ∼44 m) starts with 5.35 m of dark gray limestone, overlying the breccia with a very irregular, erosional, and perhaps karstified contact. This interval is massive, consisting of calcimudstones micritized to recrystallized, with fenestral pores and chert nodules; alternatively, they occur mottled, dolomitized, with quartz and opaque extraclasts, and calcite-filled fractures. Sometimes they are brecciated, with fractures and very angular intraclasts, with isopachous or dog tooth cement at the margins and syntaxial cement at the center (Figures 8H–L). The last 3.75 m thick interval of the middle section consists of dark gray limestones arranged in sets of centimeters to decimeters thick beds, varying from tabular to wavy. Microscopically they are oolitic grainstones, with local occurrences of micritic matrix and extraclasts (Figures 7F, 8M–O).

During the fieldwork we could not access to the upper part of the section, and thus we completed the log with that of Fazio et al. (2019). However, with the stratigraphic information available it was not possible to precisely link the middle to the upper part of the section and a 10 m gap is estimated between the two. According to Fazio et al. (2019), the upper part of the section comprises 30 m, with the lower half consisting of alternating limestones and laminated mudrocks, with a composition ranging from carbonate to siliciclastic. Above this interval, in conformable or para conformable contact, there are ∼24 m of pure siltstone, with millimeters to centimeters thick, plane parallel to wavy lamination, belonging to the Guaicurus Formation. The contact between the two formations in the Laginha section presents 1 m-thick non-cohesive yellowish beige siltstone with kaolinite and gypsum among its mineralogy (Fazio et al., 2019).

The Corcal section displays little dispersed δ¹³C values (Figure 9), ranging from 0.70‰ to 6.97‰. In the lower part of the section, the values increase gradually from ∼3.5‰ to ∼5‰, and then persist through the middle and upper part. There are minor positive peaks of 5.33‰, 6.97‰, and 5.75‰ at 10, 12.95, and 15.95 m, respectively, and a negative peak of 4.16‰ at 32.85 m, as well as few outliers. δ¹⁸O values range from −10.47‰ to −4.75‰. In the lower part of the section values are more dispersed but become steadier in the middle and upper part. The cross-plot between δ¹³C and δ¹⁸O shows no linear correlation or facies-related clustering (Figure 10).

The Laginha section yields δ¹³C values from −3.11‰ to 4.92‰ (Figure 9). In the Bocaina Formation the values show a linear increase from 1.53‰ to 3.14‰. Above the polymictic breccia, which was not analyzed for stable isotopes, values start at 0.79‰ and decrease to −3.00‰ at 21 m, then increase again to 0.13‰ at 27.35 m and 1.92‰ at 30.75 m. In this last interval the values are scattered, showing no clear trend. The middle part of the section has values around −1.00‰, with a distinct negative peak of −2.14‰ at 34.8 m. The upper part of the section has a lower sampling resolution, however, a rough trend can be observed with values starting around 4‰, increasing to 4.80‰ at 12 m, and then decreasing to around 2‰ from 17 m to the top of the Tamengo Formation. The average is −1.01‰ in the lower and middle part of the section, and 3.29‰ in the upper part. δ¹⁸O values range from −10.09‰ to −2.12‰, with average of −6.84‰. They are dispersed, showing no clear trend, within the lower and middle part of the section, and range between −8‰ and−6‰ in the upper part. The cross-plot between δ¹³C and δ¹⁸O shows no linear correlation or facies-related clustering (Figure 10).

The δ¹³C values have no clear correlation with δ¹⁸O in both studied sections, which indicates no significant alteration by early burial diagenesis even though some samples are recrystallized (Kaufman and Knoll, 1995; Derry, 2010). In addition, the isotopic values are not clustered according to the sedimentary facies. In the Corcal section there is little point-to-point variation and a clear increasing trend within the lower part, reaching values between 4‰ and 6‰ that occur steadily in the upper part. Overall, the δ¹³C record of the Corcal section is consistent with that of Boggiani et al. (2010), Rivera (2019), and the values agree with those of upper Ediacaran sections worldwide, such as in Northwester Canada (Macdonald et al., 2013), South China (Jiang et al., 2007; Ding et al., 2020), Namibia (Halverson et al., 2005), and Oman (Cozzi et al., 2004). In general, the δ¹⁸O values in the Corcal section are relatively heavier in the upper part of the section, as well as the δ¹³C. These general increasing trends are consistent with the interpretation of a relative rise in sea level from the lower to the upper section, suggested in Section 5.1.1, since a more active carbonate factory and a higher distance from the shore might increase the δ¹³C and the δ¹⁸O values (Immenhauser et al., 2003).

On the other hand, the Laginha section displays δ¹³C values more variable and generally 1‰–2‰ lower than those in the Corcal, especially in the lower and middle parts. However, there is no distinct relationship between δ¹³C and sedimentary facies, lithology, or diagenetic features. Moreover, this variability can be observed laterally, as well as vertically. Boggiani et al. (2010) measured two profiles, one on the East and one on the West side of the Laginha quarry, which yielded significantly different values, despite representing the same stratigraphic interval (Figure 12). The profile of the Laginha section in this study corresponds to the West side of Boggiani et al. (2010), however, the data overlap with both profiles (Figure 11).

The different facies and δ¹³C values of the Tamengo Formation in the two studied sections make also difficult to correlate them precisely. This becomes even more difficult as Ediacaran guide fossils, such as C. lucianoi and C. werneri, were found only in the Corcal section, but lack in the Laginha (Adôrno et al., 2017). This is due, in part, to the fact that the lower part of the Tamengo Fm. occurs only in the Laginha section, however, the upper part of this section should display δ¹³C values that match those of the Corcal, which is not the case. The greater variability of δ¹³C values in the Laginha section indicates a stronger influence by local environmental or diagenetic factors in this section, respect to Corcal (Klein et al., 1999; Frank and Bernet, 2000; Spangenberg et al., 2014; Swart and Oehlert, 2018). Unraveling these factors is a complex work, which is beyond the goal of this study. However, it is important to stress that the δ¹³C signal of the Laginha section is not representative of the general Ediacaran seawater and thus this section is not suitable for chemostratigraphic correlations. Previous studies chose the Laginha as reference section for correlation since it displays the entire Tamengo Formation, from the base to the top (Boggiani et al., 2010; Oliveira et al., 2019; Amorim et al., 2020). Our results show that the validity of these correlations must be re-assessed and that the Corcal section is a more suitable reference for stratigraphic correlation, despite not encompassing the entire unit.

The new δ¹³C record of the Corcal section allows for enhancing the correlation of the Corumbá Group with the Arroyo del Soldado Group in Uruguay, which presumably deposited along the same continental margin (Gaucher et al., 2003). Considering the facies similarity, Gaucher et al. (2003) proposed that the Tamengo Formation could be correlated with the Polanco Formation of the Arroyo del Soldado Group. Adôrno et al. (2019a) revised this correlation with a biostratigraphic approach and showed that the Tamengo Formation is equivalent to the upper part of the Yerbal Formation and the lower to middle part of the Polanco Formation. The δ¹³C curve of the Polanco Formation shows three positive excursions, one in the lower and two in the upper part (Figure 12; Gaucher, 2014). By integrating the biostratigaphic data of Adôrno et al. (2019a) with the δ¹³C data of Gaucher (2014) and of the Corcal section, the upper part of the Tamengo Formation turns equivalent to the lower part of the Polanco Formation (Figure 12). According to this correlation, the upper Polanco, the Barriga Negra, and the Cerro Espulitas formations of the Arroyo del Soldado Group should be equivalent to the uppermost part of the Tamengo Formation and to the lower part of the Guaicurus Formation, suggesting that the uppermost part of the Ediacaran is very condensed in the Corumbá Group. However, it is also possible that an unconformity occurs between the Tamengo and the Guaicurus formations (Adôrno et al., 2019a; Fazio et al., 2019).

Previous studies proposed that the Tamengo Formation deposited as a ramp carbonate platform onto a passive margin (Boggiani et al., 2010; Oliveira et al., 2019; Amorim et al., 2020). According to this model the Laginha section represents the shallower part of the carbonate platform, whereas the Corcal section the mid- and outer parts. The clastic lithofacies and the cross and hummocky stratification indicate an environment controlled mainly by waves and storms (Dott and Bourgeois, 1982; Seguret et al., 2001; Dumas and Arnott, 2006; Amorim et al., 2020). However, here we identified desiccation features in the lower part of the Corcal section, indicating that the environment reached also back- to inner ramp conditions (Tucker and Wright, 1990; Burchette and Wright, 1992; Tucker and Dias-Brito, 2017). Consequently, the carbonate ramp of the Tamengo Formation shifted from a back-ramp to an outer ramp setting, represented by the C2 and C8 facies, respectively (Table 1). This implies an environment more dynamic than previously thought, strongly controlled by sea level fluctuations, as well as by waves and storms.

This has also important implications for constraining the dwelling conditions of the Ediacaran organisms C. lucianoi and C. werneri, of which fossils have been recovered in the Corcal and in the correlative Sobramil section (Adôrno et al., 2017; Walde et al., 2015; Oliveira et al., 2019; Amorim et al., 2020). The occurrence of these fossils is strongly facies dependent, as Cloudina type fragments have been observed in the packstone and grainstone facies, whereas C. werneri occurs only in the shales. This indicates shifting ecological conditions in response to sea level variations, with C. werneri proliferating in a deeper environment, with lower energy and a softer and finer substrate, during transgressive phases. On the other hand, C. lucianoi became dominant in shallower and higher energy conditions, therefore with more light and nutrients availability, during high stands (Sanders and Pons, 1999; Walde et al., 2015; Oliveira et al., 2019; Amorim et al., 2020; Li et al., 2021).

The “classical” facies of the Tamengo Formation are those observed in the Corcal section, which occur also in other outcrops of the region along the Paraguay river, such as in the sections of Porto Sobramil (also known as Claudio or Saladeiro), Cacimba, Porto Figueiras, Goldfish, Ladario-Corumbá escarpment, and Horii mine (Boggiani et al., 2010; Walde et al., 2015; Adôrno et al., 2019b; Oliveira et al., 2019; Rivera, 2019; Amorim et al., 2020). Alongside the same lithofacies, these sections display also consistent and rather homogenous δ¹³C values (Boggiani et al., 2010; Rivera, 2019; Caetano Filho, 2020; and this work). These facies are different from those observed in the Laginha section, and therefore the Laginha was considered from a more proximal setting than the Corcal and other sections along the Paraguay river (Boggiani et al., 2010; Oliveira et al., 2019; Amorim et al., 2020). However, previous studies focus only on the upper part of the Tamengo Formation and do not consider the thick breccia layers in the Laginha section. In fact, as discussed in the facies interpretation, these breccias, are meters to tens of meters thick, homogeneous, and polymictic, indicating a deposition by mass sediment flows. Therefore, it was necessary a steep topography for triggering the flow and sufficient space to accommodate tens of meters of sediment at once. These conditions contrast with a shallow inner carbonate ramp. Besides, the more variable facies sequence of the Laginha section indicates a more dynamic environment, influenced by other factors alongside waves and sea level variations.

A steep topography and a rapidly opening accommodation space can be found in setting dominated by extensional tectonic (e.g., Sanders and Höfling, 2000; Eyles and Januszczak, 2007). In this context can occur also basement uplift, which provides coarse, angular, and fresh clasts such as those occurring in the lower part of the Laginha section. Warren et al. (2019) propose a model for the basin evolution of the Itapucumi Group, in Paraguay, which outcrops about 250 km to the South of Corumbá and is correlative to the Corumbá Group. According to this model, at the time of deposition of the Tamengo Fm (ca 555 to 542 Ma) the basin was in a thermal subsidence regime, during a post-rift stage, but still influenced by extensional tectonic. Therefore, the carbonate platform of the Tamengo Formation formed onto an irregular margin, with a complex topography, controlled by extensional faults (i.e., Figure 13A, Sanders and Höfling, 2000; Ding et al., 2021). In this context, the Corcal section corresponds to a more stable portion of the ramp, not significantly influenced by the tectonic. On the other hand, the Laginha section sat close to or into the margin of a graben, with a steep topography, and next to a horst that provided the terrigenous supply. Therefore, the Corcal section was not necessarily further from the shore than the Laginha, but just laid onto a flatter and smoother topography (Figure 13A). The upper part of the Tamengo Formation consists of mid-ramp carbonate facies in both studied sections. Consequently, according to this interpretation, it can represent a period of tectonic quiescence, during which the relative sea level rose. This could have favored the progradation of the carbonate platform, which leveled the topography (e.g., Figure 13B, Tucker and Wright, 1990; Sanders and Pons, 1999; Schlager, 2005; Łabaj and Pratt, 2016).

This interpretation can explain the more variable facies and stable isotopes records of the Laginha section. A steep and tectonically active setting can be more susceptible to changes in input of continental material, local currents, and even burial fluids, which might have contributed to the dolomitization of the lower part of the section (Caetano Filho, 2020). It can also explain the absence of C. werneri remains in the Laginha section, as a steep topography and the input of coarse material might have precluded the installation of quite conditions and fine substrate necessary for these organisms. Therefore, the Laginha section represents an exceptional setting for the Tamengo Formation, strongly influenced by local factors both in the lithological and in the stable isotopes’ records.

The Tamengo Formation contains allochemical and bioclastic constituents typical of Ediacaran carbonate ramps (e.g., Sumner and Grotzinger, 1993; Narbonne et al., 1994; Walter et al., 1995; Grotzinger and James, 2000; Gaucher et al., 2003; Jiang et al., 2003b; Lindsay et al., 2005; Jiang et al., 2011; Xiao et al., 2016; Vaziri and Laflamme, 2018; Warren et al., 2019; Álvaro et al., 2020). On the other hand, the detail facies analysis presented here allows for unraveling more specific characteristics of its depositional environment.

Differently from some of the most renown coeval carbonate successions, such as the Nama Group in Namibia, the Little Dal Group in northwestern Canada, the Buah Formation in Oman, and even the closest Bambuí Group in Brazil, the Tamengo Formation consists of totally clastic lithofacies, with no biogenic buildups (Batten et al., 2004; Cozzi et al., 2004; Grotzinger et al., 2005; Alvarenga et al., 2014; Uhlein et al., 2019). Microbial laminations have been observed only sporadically as reworked intraclasts in the breccia facies. Cloudina type skeletal fragments are frequent, but no specimens have been found in place. However, the abundance of micritic matrix and intraclasts indicates the presence of a developed biogenic carbonate factory. Moreover, recent evidence of microalgae, preserved in shale facies, indicate a diversified biotic assemblage (Diniz et al., 2021). Amorim et al. (2020) explain this apparent paradox suggesting that frequent storms and currents might have prevented the formation of extensive biogenic carbonate buildups, or that the shallower part, in which the carbonate producers dwelled, was not preserved. We agree that the high energy might have contributed, but consider unlikely the non-preservation of carbonate reefs and buildups, as they generally have the higher potential of preservation (e.g., Tucker and Wright, 1990; Schlager, 2005).

The Tamengo Formation yields lithofacies like those of the Doushantuo and Dengying formations, which form an extensive, well preserved, and well-studied Ediacaran carbonate platforms system onto the Yangtze block, in South China. There, the sediments of the carbonate ramp are preserved with lateral continuity, from the inner shelf to the basin (Jiang et al., 2011; Ding et al., 2019; Ding et al., 2021). This indicates that Ediacaran carbonate platforms were heterogeneous and the presence of subtidal stromatolites, thrombolites, or even metazoan reefs is not necessarily a prerogative. Even though the availability of the outcrops in the Corumbá region is not as high as that in South China, we can infer that the Tamengo Formation was a “Doushantuo- Dengying type” carbonate platform, characterized by a productive carbonate factory, but with environmental conditions that did not allow the deposition of significant buildups with a biogenic framework. This could be due to particularly high energy, but also to temperature, salinity, or nutrients conditions (Tucker and Wright, 1990; Schlager, 2005). In the case of the Tamengo Formation, the presence of extraclasts in several layers suggest the possibility of runoff that episodically decreased the salinity and increased the trophic level of the environment. Besides, especially in the Laginha section, a steep topography might have limited the space with favorable ecological conditions for growing biogenic reefs. It is interesting to note that also the upper Ediacaran Dengying Fm. formed in a setting controlled by extensional tectonic (Ding et al., 2021).