Brachiopods were amongst the first animals to successfully adopt external hard parts during the biomineralisation revolution of the Proterozoic-Paleozoic transition around 540 million years ago. Today, they are found in the oceans across a wide range of latitudes and water depths, even though their relative contribution to the composition of marine faunal assemblages has declined since the Mesozoic29. Modern brachiopods are characterized by a remarkable diversity in biomineralisation strategies and multi-layered shell structures, which are formed by amorphous calcium carbonate, low magnesium calcite, high magnesium calcite and phosphate30. This wide range of structure and composition is an evolutionary asset acquired, broadened, and preserved by brachiopods over more than 500 million years during which they adapted to environmental change and climatic upheavals that at times, as during the T-OAE, were catastrophic for global faunal diversity.
The early Toarcian carbon cycle perturbation
The most characteristic feature of the T-OAE carbon cycle perturbation is a pronounced negative CIE of several permil in magnitude commencing in the uppermost tenuicostatum ammonite zone1,3,4,7,13 (Fig. 2). The exact nature of this excursion is still debated, as its magnitude has been hard to constrain robustly. Belemnite calcite records6,25 usually show hardly any negative CIE, which can at least conceptually be explained by faunal turnover and habitat shifts of these mobile, marine predators6. Bulk organic matter3,4,7 and wood1,14,31 records on the other hand commonly feature a very prominent negative CIE, reaching amplitudes as large as 8‰. This amplitude is thought to be inflated due to coincident changes in organic matter sources for bulk organic matter2 as well as physiological effects in terrestrial floras1. To account for these effects, an amplitude correction has been proposed that limits the size of the negative carbon isotope excursion to 3–4‰2. The newly generated macrofossil dataset does not suffer from the above problems as it is derived from benthic organisms that are known for their consistent biomineralisation behaviour devoid of vital effects32,33. The negative CIE observed in Barranco de la Cañada has an amplitude of 4.2‰ when compared against a hypothetical linear increase between the carbon isotope maxima bounding the CIE (Fig. 2). This magnitude may nevertheless be somewhat biased. Potentially, strata relating to the most negative carbon isotope ratios were either not deposited or did not yield calcite fossils. The δ13C of ambient dissolved inorganic carbon may have also been somewhat affected by potential changes in surface productivity6. Nevertheless, because all other means of reconstructing the magnitude of the negative CIE are affected by the same complications, the new magnitude estimate is thought to be the most reliable thus far.
A rough estimate of the mass of 13C-depleted carbon that must be introduced into the exchangeable ocean-atmosphere carbon pool in order to drive the observed negative CIE can be derived from simple mass balance equations (see supplement, ref. 34). Assuming an exchangeable carbon pool with an average δ13C value of +2.6‰ and a size two to four times as great as that of pre-industrial times35, the observed negative CIE would require the addition of c. 3,500 to 7,000 Gt of carbon derived from methane (supplements). However, this estimate needs to be taken as a minimum for the amount of carbon injected into the ocean-atmosphere system. More realistically, one should conceive of the carbon cycle perturbations as resulting from multiple carbon sources, rather than a discrete, singular perturbation represented by the above calculations3,4,15,31,36.
Toarcian temperature evolution
The classification of the T-OAE as a major hyperthermal episode is not controversial (e.g. ref. 37), but the thermal evolution of the Earth surface system during the Toarcian is still poorly known. Large published belemnite datasets are likely affected by habitat and water mass changes6,38. Therefore, palaeotemperatures computed from the calcite of their fossil hard parts should not be taken as unequivocal reflection of sea surface water temperature. Data published on brachiopods5,18,26 on the other hand are hitherto sparse, lack resolution in the hyperthermal interval, and are partially based on terebratulids which are more prone than rhynchonellids to metabolically induced isotope disequilibrium33.
The newly gathered data from well-preserved fossils from the Barranco de la Cañada section now show that the beginning of the hyperthermal coincides precisely with the onset of the negative CIE (Fig. 2). The temperature increase is marked by a negative shift of 1.00 ± 0.09‰ in δ18O. Using the brachiopod oxygen isotope thermometer of Brand et al.39, this shift computes to a temperature increase of 3.5 ± 0.3 °C at the seafloor of the shallow epicontinental basin (Fig. 2; supplements), if confounding effects of sea-level change and salinity fluctuations can be discounted. This magnitude of temperature change is lower than previously inferred from brachiopod datasets from Portugal5 and Spain26, which can be explained by very different sample sizes. The earlier datasets are limited to a comparatively small number of analyses for which the inferred difference between coldest and warmest temperatures was taken to reflect T-OAE temperature change. Here, it is possible to quantify the increase of average temperatures from the earliest Toarcian into the T-OAE. This quantity is necessarily smaller than the difference between minimum pre-event and maximum event temperatures, but better reflects changes in the Earth surface system.
Besides temperature increase, more negative δ18O values could signify shallowing or partial freshening of the local seawater. Early Toarcian changes in sea-level have been recognized in Spanish21 and Portuguese basins1,40,41, generally pointing to an overall transgression and deepening. Inferring that sea-level is the primary driver of the observed oxygen isotope variability is therefore inconsistent with these reconstructions. Sea-level reconstructions neither suggest entirely parallel evolution in Portugal and Spain, nor a marked shallowing confined to the upper tenuicostatum and lower serpentinum zones. A gradual shallowing of a thermally stratified water body is also contradicted by the depositional environment of the studied sections21,27. Water mass stratification would likely have induced persistent bottom water anoxia for which there is no lithological or faunal indication in either of the studied successions.
Freshening has been proposed for UK and German basins42,43. However, the continuous presence of stenohaline organisms in both studied sections, particularly brachiopods44, makes such an interpretation unlikely. Salinity decrease in the European epicontinental sea has been interpreted to be much less noticeable in the Iberian region than the more northerly German and UK sites42,43. Substantial salinity fluctuations in the studied part of the western European epicontinental seas are also inconsistent with modelling results35,45 and thus appear much less plausible than a direct temperature forcing of the oxygen isotope signature.
Despite a gradual return to more positive δ13C values indicating the progressive drawdown of excess atmospheric greenhouse gases into sedimentary organic matter, temperatures remained stable and high. The first sign of a temperature reduction is only seen around the last occurrence of Soaresirhynchia, when δ13C values reached their global maximum. The temperature decrease thus lags several 100 kyr3,46 behind the minimum in δ13C values. A compatible pattern, albeit with less dense data coverage, is observed also in the Portuguese strata (Fig. 2). It therefore appears that at least regionally the T-OAE was associated with a climate shift into which the Earth system was locked.
Model constraints on palaeoenvironmental change
The newly assembled high-fidelity record of palaeoenvironmental proxy data permits the extraction of further Earth system parameters from inversion models such as GEOCARB and GEOCARBSULF47,48 and forward modelling (COPSE, refs. 49,50, supplements). Particular interest is in carbon cycle dynamics, their relation to climate change, and global weathering feedbacks which are intimately linked with atmospheric carbon dioxide levels and temperature.
Inverse modelling
Weathering fluxes can be approximated from palaeotemperature and pCO2 when combined with Toarcian-specific estimates of various tectonic forcings49,50 (Fig. 4; supplements). The combination of these weathering estimates and the measured δ13C data can then be used to derive corresponding estimates of the organic and carbonate carbon burial fluxes as long as inorganic and organic carbon pools are in isotopic equilibrium. At steady state, the fraction of carbon that is buried in reduced form, (i.e., organic carbon) increases with increasing carbonate δ13C. Carbonate δ13C values greater than +3‰ through much of the serpentinum zone can be attributed to such an increased organic matter burial flux. However, given the large-scale perturbation associated with external introduction of 13C-depleted carbon during the T-OAE (e.g., Refs. 3,7,14), it is highly likely that the system deviates from steady state at least at the low point of the negative CIE, compromising the reliability of carbon flux estimates at this point.
Modelling results for carbon cycle perturbation, temperature change, weathering and nutrient dynamics as well as intensity of ocean anoxia. Measured δ13C values are shown next to temperature reconstructions using temperature equations of Anderson and Arthur69 and Brand et al.39. The coarse features of the δ13C values are matched by a transient injection of thermogenic CO2 as well as CH4 in the COPSE forward biogeochemical model. *: The representation of the global ocean anoxic fraction depicted is defined as the fraction of the ocean surface area below which the oxygen saturation in the oxygen minimum zone would be below 10%70.
In order to sustain high marine organic carbon burial, limiting nutrients supporting primary production have to be in high supply, and areas subject to ocean anoxia due to the increased oxygen consumption of organic carbon remineralization expanded51. Increased fluxes of elements to the ocean derived from continental weathering are well documented for the T-OAE12,36,37,52 even though timing and magnitude of these fluxes are not unequivocal. The ocean anoxic fraction and the quantities of limiting nutrients, which ultimately derive from such rock weathering, can be approximated using carbon burial flux estimates (supplements).
Important modulations of the Earth system’s response to climate perturbations arise on the basis of two distinct limiting factors on the silicate weathering flux53,54: kinetic factors (i.e., CO2 supply and temperature) versus substrate- related factors (i.e. supply of fresh rock through uplift). If silicate weathering is limited by CO2 then it is intimately connected with atmospheric carbon dioxide and thus planetary temperature via a homeostatic negative feedback55. By contrast, if supply of silicates is insufficient to counterbalance any increase in atmospheric CO2, this will prolong any associated hyperthermal greenhouse interval (and, in this case, any associated OAE). The nature of silicate weathering limitation has consequently been proposed as a key controlling factor on the duration of other hyperthermal intervals56.
Forward biogeochemical modelling
“Forward” biogeochemical models allow reproduction of the observed carbon cycle and climate perturbation of the T-OAE independent of the analysed data. We employ the COPSE (Carbon, Oxygen, Phosphorous, Sulphur, Evolution49,50) box model in order to mimic a carbon cycle perturbation driven by large igneous province (LIP)-derived CO2 and hydrate-derived CH4 (supplements; Fig. 4). Geochemical data extracted from fossils can be matched closely by this approach. Modelling these data provides further constraints on types, timing and magnitude of carbon injection and yields valuable information on the weathering feedback and nutrient supplies at the time36,52,57.
Within the COPSE model, the greenhouse perturbation leads to an increase in the input of phosphorous to the ocean via weathering, which boosts production and therefore the respiratory oxygen demand imposed by remineralization of sinking organic matter. This increases marine anoxia, leading to a reduction in marine phosphate burial which further amplifies organic carbon production, thus creating a short term positive feedback. Over longer timescales, the increase in production also leads to enhanced marine organic carbon burial. This increase in organic carbon burial ultimately causes a compensating increase in the global atmosphere-ocean oxygen reservoir, which may be of sufficient magnitude to create a long-term negative feedback that contributes to the system coming out of the OAE58. This sequence of events is reproduced by the model, and is sufficient to reproduce the coarse features of the δ13C curve when coupled to the carbon isotope equations (see supplement). However, the associated change in the global oxygen reservoir is modest, and CO2 consumption by weathering, relative to the magnitude of the initial greenhouse input, is the main controlling factor on the duration of the OAE.
The timing of anoxia lags behind that of the greenhouse gas input (Fig. 4), consistent with the lag between the lower limit of the δ13C curve and the maximum value of the marine organic carbon burial flux implied by the inversion model. As observed in the fossil δ18O data, the hyperthermal interval also lasts longer than the negative CIE, due to the time needed for weathering to consume the CO2 introduced via the perturbation. The duration of the OAE is ultimately dictated by the time it takes to consume the excess CO2 by silicate weathering, again raising the above issue of supply versus kinetic silicate weathering limitation. Figure 4 represents this using a formulation at which silicate weathering cannot exceed twice its current value (supplements).
The actual threshold at which silicate weathering becomes supply-limited is uncertain, and the formulation in the model relates the theoretical kinetically limited rate to a maximum weathering rate parameter. Some calculations place an upper limit on this maximum value of about 10 times the present rate, on the basis of crustal composition56, whereas lower estimates of c. 3 times the present flux have been derived within Precambrian modelling based on the maximum amount of CO2 consumable by rock flour59. Further empirical data may help determining such controls on the length of OAEs, including the Toarcian, with better reliability, and our model highlights the importance of such data.
Environmental extremes and the faunal response of Soaresirhynchia
Soaresirhynchia is commonly associated with the later part of the T-OAE, and found in the lower serpentinum ammonite zone in western Tethyan sections where black shales are not extensively developed21,22,60. A singular Soaresirhynchia occurrence in the lowermost Toarcian of the Lusitanian Basin has previously been reported by Almeras61 which is here confirmed by rare finds in the tenuicostatum zone in Fonte Coberta (Fig. 2; supplements). These finds may indicate that Soaresirhynchia already partially took hold in the Lusitanian Basin before the T-OAE. This earlier occurrence coincides with the end of the Pliensbachian-Toarcian boundary perturbation, an event that has been described as a significant extinction event and thermal perturbation in its own right9,11,20. The sediments in Fonte Coberta are interpreted to represent deeper water environments than at Barranco de la Cañada21,28. In Fonte Coberta, Soaresirhynchia is found together with two other brachiopod genera showing very similar shell ultrastructure and geochemical composition, pointing at comparable biomineralisation styles, and suggesting that physicochemical conditions at Fonte Coberta aligned better with the ecological requirements of Soaresirhynchia than in basins east of the Iberian Massif (Fig. 3). Indeed, this morphologically simple but variable brachiopod is thought to be a representative of a group of deep water brachiopods that invaded the shallow shelf seas during times of environmental perturbation62.
The reduced isotopic heterogeneity in its shell calcite compared with other Toarcian rhynchonellids from shallow water habitats (Fig. 3) suggests that this genus grew comparatively slowly. Such an interpretation is further supported by the low Sr/Ca and Mg/Ca ratios in its shell (Fig. 3), which in modern brachiopods are preferentially observed for species from habitats below the photic zone30,33. In addition, the reduced shell organic matter content as compared to the shallow water brachiopods at Barranco de la Cañada signifies that Soaresirhynchia secreted shell calcite at very little energy expense63 and thus that metabolic rates were low.
Soaresirhynchia has the coarse shell fibres with isometric cross sections typical of the rhynchonellid superfamily Basilioloidea64, which is thought to thrive in warm waters65. The association of Soaresirhynchia with low palaeolatitudes with only rare finds beyond 35°N in palaeolatitude22,60 further suggests that this genus relied on warm waters. Indeed, its disappearance in the Barranco de la Cañada section is marked by the first indications for a return to cooler palaeotemperatures (Fig. 2). A thermophilic character of this genus, however, seems at odds with its supposed deeper water origin. One might rather hypothesize that Soaresirhynchia could tolerate a wide temperature range and thrive where competition with less well adapted species was reduced or absent26.
The advent of Soaresirhynchia, occupying a large part of western Tethyan shallow shelf seas and ranging far beyond its earliest Toarcian appearance in the Lusitanian Basin has been ascribed to the pioneering repopulation after severe benthic turnover of the T-OAE22,66. However, considering the protracted hyperthermal conditions along with major changes in ocean chemistry caused by expanded anoxia elsewhere in the global ocean and increased continental weathering fluxes36,67, one should not equate this repopulation phase with a return to normal environmental conditions. Besides the ability of Soaresirhynchia to withstand high temperatures it could also survive in seawater characterized by substantial changes in ion content, having been depleted in sulphate52,68 and enriched in nutrients36,37 (Fig. 4) as a consequence of the T-OAE. The observed faunal turnover in brachiopods across the T-OAE may serve as an analogue for similar scenarios of faunal changes that are possible outcomes of ongoing alterations in climate and oceanic nutrient balance51. Shallow marine environments might become occupied by similarly unlikely survivors, calling for particular study and conservation of deeper water faunas.
Source: Ecology - nature.com
