Introduction

Previous studies have shown that post-extinction recovery from the EPME was delayed1,2,3,4 until Middle Triassic due to anomalous oceanic conditions during the Early Triassic, elevated CO2 and surface temperatures5,6,7, widespread anoxia8,9,10,11,12,13 and high primary productivity14,15. However, discovery of an Early Triassic (Griesbachian to Early Smithian) rapid recovery of benthic faunas including bivalves, gastropods, brachiopods, and ostracods in Oman and the South China Block argues against persistent widespread anoxia16,17 and challenge the delayed-recovery hypothesis. Rather, a later period of widespread benthic anoxia and hot tropical temperatures at the Smithian-Spathian boundary created a new biocrisis with slowed rate of recovery6,7, making the pattern of post-extinction recovery more complex. Notably, most prior studies of the biotic recovery have focused on invertebrate benthic marine faunas2,3,4,16,17. Mesozoic marine reptiles, such as ichthyopterygians and sauropterygians, first occurred in the Early Triassic as a part of the transition from the ‘Paleozoic Fauna’ to the ‘Modern Fauna’, and dominated the Mesozoic seas18,19. Our collection of more than 80 well-preserved and multi-clade mid-Spathian marine reptile specimens from the Majiashan Section, Chaohu, South China documents that predatory marine tetrapods diversified in the Early Triassic and were well adapted to life in the sea, indicating a habitable marine environment in the mid-Spathian at Chaohu, which is earlier then previously suggested. The diverse predators, including fish and marine reptiles, and abundant invertebrates such as bivalves, ammonoids and arthropods, form the Chaohu Fauna, represent a nearly complete ecosystem. The first appearance of abundant and multi-clade Mesozoic marine reptiles in Chaohu provides the first glimpse of recovering biotic structure reaching a high trophic level. The lack of geochronologic data, however, has precluded constraining the age of the Chaohu Fauna and thus the timing of post-extinction recovery remains poorly known.

Deep-sea δ13Ccarb records of mid-Cretaceous to Cenozoic age reveal a persistent long-eccentricity (405-kyr) cyclicity, which records astronomical pacing of carbon cycling among Earth’s surface reservoirs20,21,22,23,24. Astronomical cyclicity is also recognized following the EPME25,26,27, when potentially the largest carbon cycle perturbation of the Phanerozoic was followed by a series of disturbances throughout the Early Triassic interval of protracted marine ecosystem recovery2,28. Astronomical cycles in carbon cycling, however, have not yet been recognized leaving ambiguous the timing of the post-extinction recovery and the role of astronomical pacing of climate and carbon cycling through a significant portion of the biotic recovery phase. Furthermore, constraining the drivers of major carbon cycle perturbation is fundamental to understanding climate and biosphere dynamics throughout Earth history29,30.

The nature of the mechanistic link between post-extinction carbon cycle perturbations and the Early Triassic biotic recovery has been the focus of much research invoking wholesale changes in ocean oxygenation and water column stratification13,15. Major oceanic restructuring is inferred to have occurred in response to global warming, likely driven by episodic release of exogenic carbon from Siberian Trap volcanism6,31. Conversely, the role of internal mechanisms and feedbacks in the Earth’s climate system, including cycling through the surface carbon reservoirs, has not been investigated in the context of the recovery of life after the largest extinction since the evolution of metazoans. In this study, we develop a new high-resolution C isotopic record of seawater dissolved inorganic carbon (δ13CDIC) for the end-Smithian through Spathian, utilize it to construct an astrononmical time scale, and in turn, apply it to infer major changes in ocean structure mechanistically linked to the timing of ecosystem recovery.

Results

A high-resolution δ13Ccarb record is constructed spanning the end-Smithian through Spathian (~250.55 to 247.32 Ma, Olenekian, Early Triassic)32,33, from an intermediate- to deep-water 200 to 500 m) slope succession in Majiashan, Chaohu, South China (Fig. 1). Observed fluctuations in δ13Ccarb are interpreted to record secular changes in seawater δ13CDIC, an inference that is supported by the lack of correlation between δ13Ccarb and lithofacies changes, δ18Ocarb, δ13Corg or wt.% CaCO3 (see Supplementary Figs S4 and S5). The veracity of the new δ13Ccarb record is corroborated by similarity in long-term δ13Ccarb trends between the Majiashan section and other global time-equivalent δ13Ccarb records from South China and elsewhere2,34,35. The Majiashan δ13Ccarb time-series (Fig. 1) reveals 3 scales of variation: (i) an abrupt 4% positive shift across the Smithian-Spathian boundary (S-S boundary) observed in previous records31, (ii) a subsequent 106-yr-scale stepwise decrease from peak values of +2% to +4% in the early Spathian to minimum values of 0% to −3% in the middle-late Spathian; and (iii) low-amplitude (0.5–1%), shorter-term (105-kyr scale) oscillations superimposed on the middle-late Spathian portion of the longer-term trend (see Supplementary Fig. S6).

Figure 1: Stratigraphy and δ13Ccarb data values on column for Majiashan section (see Supplementary Table S1) with vertebrate fossil distribution.
figure 1

The pink horizontal line shows the stratigraphic level of the end-Smithian biocrisis, coincident with a rapid increase in δ13Ccarb; the green horizontal line shows the first occurrence of Triassic marine reptiles. Helong. Fm.: Helongshan Formation; D. Fm.: Dongma’anshan Formation. The large silhouettes to the left of the lithology are marine reptiles; the small silhouettes to the left of the lithology (and one medium silhouette) are fish.

Full size image

To evaluate potential astronomical cycles preserved in the record and their uncertainties, evolutive average spectral misfit (e-ASM) analysis36 was performed on the δ13Ccarb record following data preparation (see R_analysis file) by statistically comparing the observed oscillations with theoretical periods from the astronomical models La0437 and La10d38 (see Supplementary Fig. S8, Supplementary Table S2). The results show continuous short-term eccentricity, obliquity and precession over a range of sedimentation rates (see Supplementary Fig. S8, Supplementary Table S3). Using the ASM-calibrated astronomical cycles, frequency domain minimal tuning was applied to generate a “floating” ATS (see Supplementary Fig. S10). Following frequency-domain minimal tuning using the average eccentricity period of 115.30 kyr (E2/2 + E3/2 from Supplementary Table S2; see Supplementary Fig. S10), the Majiashan δ13Ccarb record exhibits power spectra peaks at periods of 457.20 kyr, 127.00 kyr, 46.79 kyr, 29.31 kyr, 22.07 kyr and 19.03 kyr (Fig. 2B, see Supplementary Fig. S11, Supplementary Table S5). A strong and continuous obliquity signal in the Evolutive Power Spectral Analysis (EPSA) of the tuned data (Fig. 2C, Supplementary Fig. S12), as well as a relatively persistent precession signal (Fig. 2C), provide independent confirmation of the short eccentricity minimal tuning implemented here.

Figure 2: Spectra analysis results of the Majiashan δ13Ccarb data following astronomical tuning and anchoring.
figure 2

(A) 115.30-kyr-tuned time series (see Supplementary Table S4). (B) Multi-taper method spectral analysis (MTM) results, using three tapers and a time-bandwidth product of 2p. The calibrated periods of significant peaks are shown in kyr; these peaks achieve the 90% confidence level for both the MTM harmonic F-test and the AR1 red noise model (bold) or the AR1 noise model only (italics) (see Supplementary Fig. S11, Supplementary Table S5). (C) Evolutive power spectral analysis (EPSA) results, using a moving window of 1000 kyr, 3 tapers and a time-bandwidth product of 2p (see Supplementary Fig. S12). A linear trend was removed from each window prior to the EPSA. Letters E1-P mark the eccentricity, obliquity and precession cycles separately (see Supplementary Table S2). (D) Phases that are defined by the minima of the filtered 405-kyr components (in E). Horizontal dashed lines are phase boundaries. (E) Filtered δ13Ccarb time series for Majiashan section. Red and blue curves indicate short and long eccentricity cycles defined using bandpass filters of 0.0015–0.0028 cycles/m and 0.006–0.011 cycles/m, respectively. Black dotted lines show the instantaneous amplitude envelope of the δ13Ccarb short eccentricity signal. (F) Theoretical eccentricity from La10d from 249.0 Ma-247.0 Ma. Red and blue curves indicate short and long eccentricity cycles. Note that La10d eccentricity curves are reversed to illustrate relationship between δ13Ccarb maxima and eccentricity minima in Phase 6 marked by blue shading (see Supplementary Table S4). Radioisotopic dates from ref. 32,33. Correlation of latest Spathian interval constrained by radioisotopic ages shown by grey shading. The large silhouettes to the right of Panel F are marine reptiles; the small silhouettes to the right of Panel F (and one medium silhouette) are fish.

Full size image

The astrochronologic testing results provide a good basis for developing a nominal anchored astronomical time scale (ATS) and for dating the age of the first occurrence of Mesozoic marine reptiles, which invaded the marine environment and represent the start of the establishment of the new marine ecosystem dominated by the air-breathing tetrapods28. Given the lack of high-resolution radioisotopic geochronology and biostratigraphy to “anchor” the ATS, we rely upon an assumption based on late Mesozoic and early Cenozoic carbon cycling studies, i.e. δ13Ccarb minima correspond to eccentricity maxima20,21,22,23,24,39,40,41,42,43. Thus, the reversed eccentricity curve of La10d was tied at 247.95 Ma to a depth of 162.62 m (Fig. 2E,F). Using this approach, the S-S boundary34 in the Chaohu area, marked by an abrupt 4% positive shift in δ13Ccarb at the top of fossil-rich black shale (pink line in Figs 1 and 2D–F), is constrained to 250.21 Ma, which compares well with the radioisotopic date of 250.55 ± 0.51 Ma32 for the earliest Spathian. The duration of the 1.36-m thick black shale (18.46 m to 19.82 m, see Supplementary Fig. S2A) below the S-S boundary was estimated to be 19.21 kyr using an astronomically calibrated average sedimentation rate of ~7.08 cm/kyr (see Supplementary Table S4). Although the exact stratigraphic level of the Spathian-Anisian boundary is unknown in the Chaohu area, it can be inferred at the Majiashan section to be at ~171.70 ± 3.72 m above the S-S boundary using an end-Spathian age of 247.32 ± 0.08 Ma33. In this case, the astronomically tuned time series constrains the duration of the Spathian interval to 2.89 ± 0.08 My, which falls within the uncertainty of the U-Pb estimated duration of 3.23 ± 0.60 My32,33,44.

According to our excavation at the Majiashan section, the lowest stratigraphic level of occurrence of marine reptiles in the Chaohu Fauna occurs in the Middle Member of the Nanlinghu Formation, ~117.00 m above the base of the Helongshan Formation (Fig. 1), and is constrained to ~248.81 Ma (Fig. 2). Abundant ichthyosauromorphs occur in the interval 135.83 m to 149.26 m above the base of the Helongshan Formation with a duration of ~279.79 kyr (Fig. 1), including Chaohusaurus, which was fully adapted to the marine aqueous habitat and yet maintained a birth posture typical of terrestrial reptiles, and Cartorhynchus with possible amphibious habits and suction-feeding behavior45. The most primitive Eosauropterygia46, which shows similarity to Middle Triassic Eosauropterygia, occurs ~156.12 m above the base of the Helongshan Formation (Fig. 1) and is constrained to ~248.10 Ma (Fig. 2). Ichthyosaurs were also found in the upper Spathian (Fig. 2).

Discussion

Persistent eccentricity and obliquity cyclicity in the EHA of the tuned δ13Ccarb data (see Supplementary Fig. S12) indicates astronomically-paced redistribution of carbon between surficial C reservoirs. Seven phases of δ13Ccarb are identified from the minima of the filtered 405-kyr components (long-term eccentricity) (Fig. 2D). The good alignment between the amplitude envelope of the filtered 100-kyr components and the 405-kyr components in Phases 5 through 7 shows the robustness of the tuning results (Fig. 2E). The overall weaker short eccentricity and variable phase relationship observed in Phases 1 through 4 suggests a fundamental change in the response of seawater δ13C to astronomical forcing in the mid-Spathian, coincident with the onset of increased short-term δ13CDIC volatility and a 2% decrease in δ13CDIC, the last of a stepped long-term decrease following the S-S boundary positive excursion. The amplification of obliquity (Fig. 2C) begins between the two abrupt δ13C shifts that define the longer-term decrease in the Spathian.

The response of the ocean carbon reservoir to astronomical forcing is largely dependent on the oceanic structure, which in turn responds to climate change23. Astronomically paced changes in ocean ventilation (oxygenation) lead to periodic releases of 12C-enriched carbon to the surface ocean. For greenhouse times, such as the Early Triassic47, the greater water column stratification of greenhouse oceans makes the ocean C budget more sensitive to orbital variations20,23,24. Together the shift in eccentricity phasing and obliquity amplification likely record a major restructuring of oceanic circulation (cf. refs 24 and 25) linked to climate amelioration during the Spathian13,31. A strongly stratified water column would have supported an expanded oxygen minimum zone and build up of a large dissolved organic carbon (DOC) reservoir in the deep ocean (cf., ref. 48). Turnover from such an oceanic state to a better and deeper ventilated ocean with major restructuring would have released large amounts of this 12C-enriched carbon pool to surficial C reservoirs and accelerated the rate of carbon cycling thus amplifying the response of seawater δ13CDIC to eccentricity and obliquity forcing/pacing20,21. In this context, we interpret the overall higher δ13Ccarb values (+2 to +4%) of the first 800 kyr (Phase 2–3) of the Spathian as recording strong water column stratification, high productivity, enhanced organic matter burial and likely build up of the deep ocean DOC reservoir promoted by the expanded oxygen minimum zone13. The subsequent change in the nature of eccentricity and obliquity cyclicity, along with the final abrupt shift to values that fluctuate around a stable mean (~−2%) characteristic of pre-perturbation thermocline δ13CDIC values in the Majiashan slope environment31, constrains the timing of termination of O2-limited oceanic conditions to the mid-Spathian. The observation of obliquity amplification within phase 4, and continuing into phases 5 through 6, is consistent with the development of a high-latitude, oxygenated, intermediate to deep-water source perhaps in response to cooling and/or eustatic change (cf. ref. 49, 50, 51) that would have led to enhanced ventilation of the deep ocean. The near loss of long-term eccentricity δ13Ccarb cyclicity in Phase 7, and weakening of obliquity, could capture the return to a fully stable well-oxygenated ocean sustained by vigorous circulation (cf. ref. 24).

The ATS indicates that the stratigraphically lowest marine reptile fossil in the Chaohu Fauna, and around the world52, occurs in the later portion of Phase 4 and has an age of ~248.81 Ma. This is the oldest known Mesozoic marine reptile found to date despite concerted but unsuccessful search efforts to find older fossils. Thus, the timing of reptilian invasion into the oceans was likely not much earlier than 248.81 Ma. This estimate constrains the post-extinction ecosystem recovery period to ~3.35 Myr after the PTB (at ~252.16 Ma, as recognized in ref. 32). The main reptile fossil beds that yield Cartorhynchus, a basal and potentially amphibious ichthyosauriform45, as well as abundant multi-clade marine reptile fossils, occur within Phases 5 and 6 (Fig. 2D–F). The Chaohu Fauna is the oldest marine fauna with tetrapod as top predators and a high marine ecosystem complexity, as far as we know.

Previous studies of Early Triassic benthic faunas and vertebrate fossil localities with limited marine ecosystem complexity53,54 supported the hypothesis of a delayed biotic recovery2,4. The appearance of marine reptiles45,46 and the Triassic middle fish fauna (TMFF)55 in Chaohu in the middle Spathian marks the initiation of the establishment of a new Meso-Cenozoic marine ecosystem and thus indicates a high degree of biotic recovery following the end-Permian Mass Extinction at this time52. The temporal relationship between the occurrence of the oldest marine reptile fossil and obliquity amplification (Phase 4) and the appearance of multi-clade marine reptiles during the subsequent change in the nature of the eccentricity modulation (Phases 5–6), temporally linked to the final phase of a two-step decrease in seawater δ13CDIC, indicates a mechanistic link between the appearance and diversification of the marine reptiles and the major change in ocean structure and carbon cycling. We attribute the biotic recovery and initiation of a new marine ecosystem to the final breakdown of ocean stratification and the onset of a high-latitude deep-water source, which would have enhanced ventilation of the deep ocean. Together these changes in ocean structure would have stimulated surface ocean primary productivity promoting the diversification of the marine reptiles. They further suggest that reptiles may have first invaded Panthalassa in the middle Spathian in response to this reorganization.

Marine reptiles introduced a new pattern of shallow marine nutrient circulation in the Early Triassic that may have played a critical role in the post-extinction build-up of the marine ecosystem, by feeding at various depths and defecating near the sea surface52. The post-extinction recovery of the forest ecosystem in the middle Spathian is contemporaneous with that of the marine ecosystem56,57 as would be anticipated if changes in ocean circulation were driven by climate amelioration. Ultimately, stabilization of oceanic conditions in the Early Triassic, which involved a change in the response of carbon cycling to astronomical forcing, likely accelerated the building of the new Meso-Cenozoic marine ecosystem.

In summary, the first occurrence of the Mesozoic marine reptiles is approximately coincident with a shift in astronomically-paced marine carbon cycling, and a ~2% shift in seawater DIC, which indicates that the South China seas remained inhabitable to large air-breathing marine predators following the end-Permian Mass Extinction until a major rearrangement of oceanic circulation. This change led to intensified ventilation and restored well-oxygenated conditions, which is mechanistically linked to astronomically driven changes in C cycling between Earth’s surface reservoirs. The distribution of diversified marine reptiles along the coasts of the Panthalassan Sea in the late Spathian54 was likely fueled by the climate amelioration, changing ocean structure, and increased surface ocean primary productivity in the middle Spathian. The same factors may have encouraged reptiles to invade the sea. Ultimately, stabilization of oceanic conditions in the Early Triassic may have involved climate-life interactions in which a change in the astronomical pacing of climate and carbon cycling among the Earth’s surface carbon reservoirs accelerated the building of the new Meso-Cenozoic marine ecosystem.

Methods

Carbonate samples were collected and analyzed at a spacing of ~5–10 kyr from the Majiashan section, Chaohu spanning the late Smithian to Spathian in the Isotope Lab of Nanjing Institute of Geology and Paleontology, Chinese Academy of Sciences with an analytical precision of <0.03%. δ18Ocarb, δ13Ccarb δ13Corg with an analytical precision of <0.1% and wt.% CaCO3 of microdrilled samples were generated in the Stan Margolis Stable Isotope Lab, UC Davis. The δ13Ccarb data were prepared and analyzed in the R software package “Astrochon” (Meyers, 2014; R Core Team, 2015).

Additional Information

How to cite this article: Fu, W. et al. Eccentricity and obliquity paced carbon cycling in the Early Triassic and implications for post-extinction ecosystem recovery. Sci. Rep. 6, 27793; doi: 10.1038/srep27793 (2016).