CPClimate of the PastCPClim. Past1814-9332Copernicus PublicationsGöttingen, Germany10.5194/cp-13-1635-2017Latest Permian carbonate carbon isotope variability traces heterogeneous organic carbon accumulation and authigenic carbonate formationSchobbenMartinm.schobben@leeds.ac.ukschobbenmartin@gmail.comhttps://orcid.org/0000-0001-8560-0037van de VeldeSebastiaanhttps://orcid.org/0000-0001-9999-5586GliwaJanaLedaLucynaKornDieterStruckUlrichUllmannClemens Vinzenzhttps://orcid.org/0000-0002-5865-7289HairapetianVachikGhaderiAbbasKorteChristophNewtonRobert J.PoultonSimon W.WignallPaul B.School of Earth and Environment, University of Leeds, Woodhouse Lane, Leeds, LS2 9JT, UKMuseum für Naturkunde, Leibniz-Institut für Evolutions- und Biodiversitätsforschung, Invalidenstr. 43, 10115 Berlin, GermanyAnalytical, Environmental and Geochemistry, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, BelgiumInstitut für Geologische Wissenschaften, Freie Universität Berlin, Malteserstr. 74–100, 12249 Berlin, GermanyCollege of Engineering, Mathematics and Physical Sciences, Camborne School of Mines, University of Exeter, Penryn Campus, Penryn, Cornwall TR10 9FE, UKDepartment of Geology, Esfahan (Khorasgan) Branch, Islamic Azad University, P.O. Box 81595-158, Esfahan, IranDepartment of Geology, Faculty of Sciences, Ferdowsi University of Mashhad, Azadi Square, 9177948974, Mashhad, IranDepartment of Geosciences and Natural Resource Management, University of Copenhagen, Øster Voldgade 10, 1350 Copenhagen, DenmarkMartin Schobben (m.schobben@leeds.ac.uk, schobbenmartin@gmail.com)22November20171311163516594May201711May201719September201716October2017This work is licensed under the Creative Commons Attribution 3.0 Unported License. To view a copy of this licence, visit https://creativecommons.org/licenses/by/3.0/This article is available from https://cp.copernicus.org/articles/13/1635/2017/cp-13-1635-2017.htmlThe full text article is available as a PDF file from https://cp.copernicus.org/articles/13/1635/2017/cp-13-1635-2017.pdf
Bulk-carbonate carbon isotope ratios are a widely applied proxy for
investigating the ancient biogeochemical carbon cycle. Temporal carbon
isotope trends serve as a prime stratigraphic tool, with the inherent
assumption that bulk micritic carbonate rock is a faithful geochemical
recorder of the isotopic composition of seawater dissolved inorganic carbon.
However, bulk-carbonate rock is also prone to incorporate diagenetic signals.
The aim of the present study is to disentangle primary trends from diagenetic
signals in carbon isotope records which traverse the Permian–Triassic
boundary in the marine carbonate-bearing sequences of Iran and South China. By
pooling newly produced and published carbon isotope data, we confirm that a
global first-order trend towards depleted values exists. However, a large
amount of scatter is superimposed on this geochemical record. In addition, we
observe a temporal trend in the amplitude of this residual
δ13C variability, which is reproducible for the two studied
regions. We suggest that (sub-)sea-floor microbial communities and their
control on calcite nucleation and ambient porewater dissolved
inorganic carbon δ13C pose a viable mechanism to induce
bulk-rock δ13C variability. Numerical model calculations
highlight that early diagenetic carbonate rock stabilization and linked
carbon isotope alteration can be controlled by organic matter supply and
subsequent microbial remineralization. A major biotic decline among Late
Permian bottom-dwelling organisms facilitated a spatial increase in
heterogeneous organic carbon accumulation. Combined with low marine sulfate,
this resulted in varying degrees of carbon isotope overprinting. A simulated
time series suggests that a 50 % increase in the spatial scatter of organic
carbon relative to the average, in addition to an imposed increase in the
likelihood of sampling cements formed by microbial calcite nucleation to 1
out of 10 samples, is sufficient to induce the observed signal of carbon
isotope variability. These findings put constraints on the application of
Permian–Triassic carbon isotope chemostratigraphy based on whole-rock
samples, which appears less refined than classical biozonation dating
schemes. On the other hand, this signal of increased carbon isotope
variability concurrent with the largest mass extinction of the Phanerozoic
may provide information about local carbon cycling mediated by spatially heterogeneous
(sub-)sea-floor microbial communities under suppressed bioturbation.
Introduction
Carbon isotopes in carbonate rock are a pivotal tool for understanding the
ancient biogeochemical carbon cycle and a stratigraphic aid in
determining the age of sedimentary deposits. Individual fossil carbonate
shells are the preferred recorder of the isotope composition of marine
dissolved inorganic carbon (DIC). However, some deposits, such as
those of Precambrian age or those which were formed during biotic crises or
where shelly fossils are absent, are not suitable for such a single-component
approach . Under these circumstances,
bulk-rock samples are a widely used alternative recorder . This approach has attracted criticism due to the
complex multicomponent nature and high potential for diagenetic alteration,
which may result in a mixture of primary and secondary diagenetic signals.
Major sources of contamination that are known to affect the primary marine
carbon isotope signal of bulk-carbonate rock are degradation of organic
matter and methane oxidation .
Carbonate sediments are especially prone to chemical alteration during early
compaction and lithification when the highly porous sediment and unstable
carbonate polymorphs (e.g., aragonite and high-Mg calcite) are
subjected to cementation and dissolution . The potential for diagenetic alteration of biogenic
high-Mg calcite and aragonite results from elevated substitution of
foreign ions in the lattice and a higher degree of dislocations
.
On carbonate platforms subjected to rapid sea level fluctuations,
lithification might be controlled by interaction with meteoric fluids and
oxidized terrestrial organic matter . The latter mode of carbonate rock cementation with allochthonous
dissolved carbonate sources (meteoric water) in a relatively open system has
long been viewed as the predominant process of carbonate rock stabilization
. These notions were fuelled by an over-representative
focus on Pleistocene carbonate platforms (e.g., the Bahamas) and their
associated diagenetic stabilization under the influence of glacial–interglacial-induced sea level fluctuations . However, this mechanism is
less relevant for rock lithified under greenhouse conditions when sea level
changes are dampened .
Alternatively, carbonate rock alteration is driven by marine-derived fluids
that evolved through interaction with precursor sediments and ambient
chemical conditions . Foremost, the reactive
and biologically active upper section of the sediment column can be
envisioned as leaving an imprint on carbonate chemistry, including the carbon
isotope composition. This carbon isotope signal will be retained in
post-depositional carbonate cements and is controlled by organic carbon
(OC) fluxes, pH, alkalinity and the nature of the (sub-)sea-floor
microbial communities .
A more nuanced view on the nature of bulk-rock carbon isotope composition
would be to envision diagenetic overprinting on carbonate rock chemistry as a
spectrum where pure primary and strictly diagenetic end-member states are
considered as a continuum, with bulk-rock chemistry falling somewhere in
between these two extremes . In this case, the aim
should be to discern the degree to which an original imprint could have been
retained in the bulk-rock signal. Embracing both ends of this spectrum could
illuminate new aspects of marine sedimentary systems and ancient ocean
chemistry, as well as providing a more refined understanding of the
mechanisms that promote the diagenetic alteration of carbonate sediments.
Clear examples of the importance of interpreting diagenetic signals have been
given by the assertion that authigenic carbonate production might play a
primary role in the global biogeochemical carbon cycle
and in the isotopic imprint of terrestrial biomass on Precambrian carbonates
.
Background: Permian–Triassic carbonate carbon isotope records
Temporal trends in carbonate carbon isotope composition have
been recorded at varying timescales from million-year secular trends
to dramatic sub-million-year events, such as
at the Paleocene–Eocene Thermal Maximum . The sedimentary
record of the Permian–Triassic (P–Tr) boundary interval, marked by the
largest mass extinction of the Phanerozoic , is
also characterized by pronounced carbon isotope excursions, which are almost
exclusively recorded in bulk rock. However, chemical signatures from this
time period include records with a wide variety of amplitudes and shapes and
span centimetres to metres of rock sequence e.g.,. Equally diverse are the proposed drivers of these geochemical
signals, ranging from prolonged episodes of volcanism and the eustasy-controlled
erosion of organic-rich shelf sediments to abrupt blooms of
methane-producing microbes .
Variable-amplitude carbon isotope excursions in shelf to basin transects have
been linked to an intensified biological pump .
This notion contrasts with the viewpoint of collapsed primary productivity
(or a “Strangelove” ocean) as a driver for the demise of bottom-dwelling and
infaunal marine biota . The loss of these geobiological
agents resulted in a reduced thickness of the sedimentary mixed layer and a
global increase in the occurrence of laminated sediments . Contrary to disrupted primary productivity, these geological
features would also agree with a scenario of enhanced OC
remineralization and resulting widespread marine de-oxygenation
. These contrasting scenarios of global-scale
environmental deterioration are difficult to reconcile. Nevertheless, the
reduction of sediment mixing is an unequivocal feature that warrants
consideration when interpreting P–Tr chemical records.
An elevated oceanic carbonate inventory is a predicted aspect of an ocean
without pelagic calcifiers, which only started proliferating during the
Mesozoic and would effectively buffer short-term perturbations to P–Tr
marine carbon isotopes . This does not, however, exclude local departures from this
dynamic equilibrium or depth-related isotope differences, such as those
forced by the biological pump . High carbonate
ion concentrations are invoked to explain the advent of (microbial) carbonate
sea-floor structures (e.g., thrombolites, stromatolites and fan-shaped
structures) in the extinction aftermath and the ensuing Early Triassic
recovery phase . While conditions favoured the formation
of these structures, poorly buffered calcified metazoans (e.g., brachiopods
and corals) were proportionally more affected by the end-Permian mass
extinction . This biotic shift has been interpreted as a
prevalent carbonate factory turnover from skeletal to microbial
e.g.,.
A carbonate factory turnover, reduced sediment mixing and increased OC
sinking fluxes invite consideration of a scenario in which these individual
parameters act as synergistic effects on carbonate formation and diagenetic
stabilization. Moreover, the imprint of 12C-depleted carbon on
latest Permian carbonate rock produced by the introduction of authigenic carbonate has
been connected to a systematic carbon isotope offset between bulk rock and
brachiopod shells . Diagenetic and authigenic carbonate
sources can, however, result in a range of carbon isotope values relating to
the specific microbial community and sedimentary environment
. Hence, we hypothesize that spatial carbon isotope
variability at the P–Tr boundary interval relates to an increased importance
of microbially controlled calcite nucleation. Moreover, we postulate that
organic carbon sinking fluxes and subsequent in situ
remineralization by microbes determine the trajectory of carbonate rock
stabilization. Combined geographic differences in the OC sinking fluxes
(km scale) and sediment mixing (cm scale) might have generated
spatially heterogeneous dispersion of organic carbon. This spatial pattern of
OC distribution links to observed carbonate carbon isotope variability
by modulating the chance of sampling variable isotope signals along lateral
lithological transects. We adopt this conceptual model as a working
hypothesis for interpreting stratigraphic carbon isotope patterns in P–Tr
carbonate rock. In addition, however, this scenario likely has implications
for the interpretation of bulk-rock carbon isotope patterns during other
periods of the Precambrian and Phanerozoic.
To carry out this investigation, we studied carbonate-bearing sequences of
Permian to Triassic strata located in Iran and China using a compilation of
published and new data. The effect of microbial metabolism on sediments and
porewater can be numerically approximated by reactive transport models
. Similarly,
diagenetic models have proven useful in delineating trajectories of bulk-rock
Sr and Ca isotope stabilization .
For the purpose of our study, we will combine aspects of these models in an
effort to estimate the potential for the microbially mediated spatial
modification of bulk-rock carbon isotope signals.
Materials and methodsBulk-carbonate carbon isotope recordsMaterial selection and stable carbon isotope measurements
The P–Tr limestone sequences of Iran are particularly suited to study
spatial δ13Ccarb variations, as lateral lithological
homogeneity excludes a strong control from palaeoceanographic conditions or
selective preservation potential (see the Supplement). In addition to the
classical P–Tr sites of Shahreza, Zal and Ali Bashi in Iran, we
sampled several other sites for which we present the first
δ13Ccarb results. These sites include the Aras Valley
profile (39.015∘ N, 45.434∘ E) about 19 km WNW of the
towns of Dzhulfa (Azerbaijan) and Julfa (Iran), four parallel sections in the
Baghuk Mountains (section A: 31.563∘ N, 52.438∘ E; section 1:
31.567∘ N, 52.444∘ E; section C: 31.567∘ N, 52.443∘ E;
section J: 31.565∘ N, 52.441∘ E) 50 km NNW of Abadeh
and 140 km SSE of Esfahan and the Asadabad succession
(31.848∘ N, 52.181∘ E). Limestone, marlstone and calcareous shale
(>20%CaCO3) beds were collected bed by bed. Powder aliquots
were produced by micro-drilling fresh rock surfaces to avoid sampling of
obvious late diagenetic calcite veins and weathered surfaces.
A total of 463 stable carbon isotope measurements were made at the University
of Copenhagen (UC) and the Museum für Naturkunde (MfN), Berlin. Glass
reaction vessels (Labco®) containing the sample powders were
flushed with helium, and carbonate was left to react with 50 (UC) or 30
(MfN) µL of anhydrous phosphoric acid (∼102%)
for at least 1.5 h. Carbon and oxygen isotope values were measured
from resulting CO2 using an Isoprime triple collector isotope ratio
mass spectrometer in continuous-flow set-up (UC) or a Thermo Finnigan Gasbench
(GS) II linked to a Thermo Finnigan MAT V isotope ratio mass spectrometer
(IRMS; MfN). Pure CO2 (99.995 %) calibrated against
international IAEA standards (NSB-18 and NSB-19) was used as a reference
gas. At the UC, results were corrected for weight-dependent isotope ratio
bias using multiple measurements of an in-house standard LEO (Carrara marble;
δ13C = 1.96 ‰) covering the entire range of
signal intensities encountered in the samples. External reproducibility was
monitored through the replicate analysis of the in-house standards LEO and Pfeil STD
(Solnhofen limestone) at the UC and MfN, respectively. Long-term accuracy was
better than 0.1 ‰ (2 SD; better than
0.2 ‰ 2 SD for oxygen isotopes). All carbon isotope
values are reported in ‰ relative to VPDB and in standard
δ notation.
Complementary data collection
Our new data are complemented by published bulk-carbonate δ13C
data from multiple P–Tr localities in Central Iran (Abadeh
,
Shahreza and north-western
Iran Ali Bashi mountains and Zal;. For a more global
perspective and a comparative approach, we additionally extracted published
data from the P–Tr Global Boundary Stratotype Section and Point (GSSP) at Meishan,
China . Data
were extracted from tables and supplementary files or provided by the authors
and where necessary read from figures with the open-source software
xyscan . This compilation adds 2077 data points to our
new dataset. The analytical uncertainty of the collected data, when given,
ranges between 0.02 and 0.2 ‰. We included replicate studies
on the same site, as these data are potentially valuable assets to test the
reproducibility of stable isotope investigations and to shed light on the
effect of the bulk-rock multicomponent nature on δ13Ccarb
composition.
Data projection
To facilitate a direct comparison of δ13Ccarb between
different sites with differences in total sediment thickness, we converted
the stratigraphic height to a dimensionless timeline. This approach preserves
a more direct connection to the sampled rock sequence rather than converting
to absolute ages or maintaining original stratigraphic heights. The
conversion to a dimensionless time grid was carried out by using the lower
and upper bounds of individual conodont assemblage biozones as tie points to
which a relative distance for an individual δ13Ccarb
data point is assigned (see the Supplement). The chronological scheme
used here divides the P–Tr interval into biostratigraphic units (A–K),
which enables the correlation of individual sequences in both geographic regions
(Table ).
a Meishan P–Tr GSSP. See the Supplement file
for references regarding the biostratigraphy studies considered to construct
this biozonation scheme.
To evaluate trends in the collected data, a sliding window with a bandwidth
equivalent to the dimensionless grid unit has been applied to the
δ13Ccarb data. This method ensures the extraction of
temporal trends at equivalent time resolution to the biozonation units
(Fig. ).
Compiled published and new data for multiple P–Tr rock sequences in
Iran (a) and China (b). The individual carbonate carbon isotope values are
placed on a dimensionless timeline that marries both geographic areas in the
most acceptable biochronological scheme (Table ). The solid
black line represents the subsampled median trend line. The dashed blue line
depicts the seawater δ13CDIC curve as obtained from the
time series simulation (Sect. ). The stratigraphic placement of
the sea level changes and the extinction horizon as well as the
biostratigraphic framework can be found in the Supplement.
To cancel out the potential for an uneven spread of data points, we also
applied a subsampling routine on the Iranian and Chinese datasets.
Subsampling was performed with the sliding window procedure, as described
above, before applying summary statistics. Subsequently, the median values
were calculated for each of the subsampled δ13Ccarb populations.
For simplification, the resulting median trend line and 95 % confidence
intervals (CIs) are calculated and visually weighted (Fig. )
visually weighted plot or watercolour
regression;. The latter is accomplished by
calculating a kernel density, and both the drawn median trend line and CI
interval are weighted by colour saturation based on this kernel density
value. In addition, the median trend line and the CI interval have been
assigned contrasting colours to further enforce the visual signal. Hence,
saturated and contrasting parts of the depicted regression curve (i.e.,
sections with high visual weight) depict areas with the highest fidelity of
the temporal geochemical pattern based on the subsampling routine. In
addition to the median trend lines, the same subsampling approach and visual
weighting has been applied to calculate and plot the interquartile range
(IQR; the value range containing 50 % of the sample population) and
95 % interpercentile range (IPR; the value range containing 95 % of the
sample population). Graphs are plotted with the open-source programming
platform R and with the aid of the R packages;
ggplot2 , reshape2 ,
plyr and gridExtra .
Visually weighted data plot for the Iranian (a) and South China
(b) P–Tr (sub)localities (Sect. ) depicting the median
trend line, the IQR (50 %) and the IPR (95 %) δ13C
value ranges. Trends with lowest fidelity are marked by a blurring of colours
and less contrasted colours (based on the CI of multiple subsample routines).
The dashed and the solid line represent the extinction horizon and P–Tr
boundary, respectively. The saturation level of the green tiles in the upper
two panels equals a more extended biozone thickness (0.12–32.00 m)
and longer duration (0.02–1.00 My). Grey tiles represent intervals
with no available biostratigraphic data. See Fig. for
colour legend of the stratigraphic units.
Reactive transport modellingGeochemical model formulation and biogeochemical reactions
Organic matter availability fuels in situ metabolic pathways linked
to calcite nucleation. Upon entering the sediment, organic matter is
microbially mineralized, which is coupled to the reduction of electron
acceptors. These electron acceptors are consumed in a well-defined sequence
based on their thermodynamic energy yield: O2, SO42- and
methanogenesis . Each of these microbial
metabolisms will imprint a specific carbon isotope signature on the porewater
DIC and thus create a potential source for diagenetic alteration of
carbonate C isotope signatures .
These mineralization processes can be described by mass balance equations
(Eqs. and ), which can subsequently be solved numerically
via the method of lines . The
numerical solutions of these equations solve the age–depth (t and
z) relationship of deposited sediments in terms of solids (Si),
solutes (Ci) in pore fluids and their respective reactions
between the solid and liquid phase. At the top of the sediment pile the
porewaters communicate with ocean water so that dissolved elements can
diffuse according to their concentration gradients. In addition to diffusion
transport processes, continued sedimentation supplies the sediment with
organic carbon and calcium carbonate.
φ∂Ci∂t=∂∂zφDi∂Ci∂z-φνCi+∑kνi,kRk+φα(z)(Cow-C(z))(1-φ)∂Si∂t=∂∂z(1-φ)Db(z)∂Si∂z-(1-φ)wSi+∑kνi,kRk
In Eqs. () and (), φ is porosity, Di is the
effective diffusion coefficient and Db and α(z) are parameters
associated with bioturbation (see below); νi,k is the stoichiometric
coefficient of species Ci in reaction Rk. Note that we express
reactions as mol per unit time per volume sediment and concentrations as mol
per volume porewater or volume solid phase. Therefore φ and
1-φ are introduced as unit conversions. The model includes three
different modes of transport: sedimentation (represented by downward
advection of solutes v and solids w), molecular diffusion and biological
transport (bioturbation). Since we only consider cohesive sediments, the only
advective transport is burial; i.e., new sediment is added on top of the
sediment column, and sediment at the bottom of the column is buried
(transported out of the column). Molecular diffusion for porewater solutes is
expressed by Fick's first law, where Di (the effective diffusion
coefficient) is calculated following the definition as given in
.
Di=D0(1-2lnφ),
where D0 is a function of temperature and salinity and has been calculated
with the R package marelac . Bioturbation is
implemented as two separate processes: bio-mixing and bio-irrigation
following conventional descriptions
. The bio-mixing is modulated over the depth
interval to account for the effects of sediment reworking by metazoans in the
top layer of the sediment pile. In this formulation it is assumed that
Db remains constant in a layer with thickness zb, after which it
attenuates with depth by following an exponential relation with coefficient
λDb and background bio-diffusivity (Db0) at the
seawater–sediment interface .
Db(z)=Db0 for z≤zbandDb(z)=Db0exp-(z-zb)λDb for z>zb
Bio-irrigation exchanges porewater with the overlying water via burrow
flushing. This is implemented via a non-local exchange process
.
Iirr(z)=α(z)(Cow-C(z))
The quantity α(z) represents the depth-dependent irrigation intensity,
and the solute concentrations of the bottom water and at depth are given by
Cow and C(z), respectively. The attenuation of bio-irrigation is
expressed by an exponential relation.
α(z)=α0exp-zxirr
In this formulation, α0 is the irrigation coefficient at the
sediment–water interface, and xirr is the attenuation depth coefficient.
Most of the macrofaunal activity takes place in the upper few centimetres of
the sediment (as animals are dependent on food resources that rain down via
the water column). Therefore, both bio-mixing and bio-irrigation processes
are most intense near the sediment–water interface.
Biogeochemical reaction equations and kinetic rate expressions.
Biogeochemical reactionsFormulationδ13CDICKinetic rate expressionPrimary redox reactions Oxic respirationCH2O+O2→HCO3-+H+-25 ‰[O2][O2]+KO2kmin[CH2O]Organoclastic sulfate reductionCH2O+12SO42-→HCO3-+12HS-+12H+-25 ‰[SO42-][SO42-]+KSO42-KO2[O2]+KO2kmin[CH2O]Organoclastic methanogenesisCH2O→12CH4+12CO2+15 ‰KSO42-[SO42-]+KSO42-KO2[O2]+KO2kmin[CH2O]Secondary redox reactions Aerobic oxidation of methane12CH4+O2→12CO2+H2O-45 ‰φkArOM[O2][CH4]Anaerobic oxidation of methaneCH4+SO42-→HCO3-+HS-+H2O-45 ‰φkAOM[SO42-][CH4]Canonical sulfur oxidationHS-+2O2→SO42-+H+–φkCSO[O2][HS-]Carbonate precipitation and dissolution Production diagenetic calciteCa2++2HCO3-→CaCO3+CO2+H2Oϵcarb-DICKO2[O2]+KO2RcarbDissolution calciteCaCO3+CO2+H2O→Ca2++2HCO3-ϵcarb-DICKO2[O2]+KO2Rcarb
Metabolism-produced carbon isotope values according to
, except organoclastic methanogenesis-sourced CH4
subsequently employed by AOM and ArOM, which is a conservative estimate
between -65 ‰ and near-quantitative sedimentary
OC conversion to CH4, thereby approaching the parent
OC-C isotope composition of -25 ‰.
The aim of this model is to give a parsimonious description of the potential
effect of early diagenetic reactions on the isotope signature of carbonates.
Therefore, we do not consider nitrification or metal cycling, as this would
increase the complexity of the model and calculations; in addition, this
would not fundamentally alter our conclusions. Organic matter can be
mineralized via different pathways: aerobic respiration, sulfate reduction
and organoclastic methanogenesis (in which organic matter acts both as electron
donor and acceptor; Table ). The reduction of sulfate produces
sulfide, which can be oxidized by oxygen via canonical sulfide oxidation
(CSO) . Methanogenesis produces methane,
which can be oxidized by sulfate in a process called anaerobic oxidation of
methane (AOM) or by oxygen (ArOM) . The mineralization
reactions are expressed via standard limitation–inhibition formulations
. The Monod constants (KO2 and KSO42-)
in these formulations determine the inhibition or limitation of a certain
electron acceptor. For example, at high concentrations of oxygen, oxygen
reduction will be efficient and other mineralization pathways will be
inhibited. At lower concentrations, oxygen reduction will
be less efficient, and other pathways will gain importance
. This
sequential alternation of limitation and subsequent inhibition of metabolic
pathways results in chemical zonation of the sediment profile
. The reoxidation reactions are given as second-order
rate laws in which the kinetic constants (kArOM, kAOM and kCSO)
determine the reaction rate of the respective reactions Table ;.
To mimic calcium carbonate recrystallization following an age–depth
relationship, dissolution and carbonate production reaction rates, expressed
as Eq. (), have been incorporated in the model with no net carbonate
dissolution or precipitation .
Rcarb(t)=0.4fdiaexp-t0.876fdia
In contrast to the models of we envision that
only a fraction of the whole carbonate rock determines the reactivity towards
dissolution and recrystallization. This fraction of 0.2 is considered to be
the diagenetic carbonate fraction (fdia) of the total rock volume
e.g.,. Specific metabolically produced
DICδ13C values are assigned for organoclastic
microbial sulfate reduction, AOM and organoclastic methanogenesis (see
Table ). These metabolic pathways have been considered to induce
calcite nucleation by changing the surrounding carbonate chemistry,
forming an organic matrix or destroying calcification-inhibiting organic
mucus . The peak activity of these metabolic pathways coincides with the
most reactive upper layers of the sediment column, i.e., highly porous and
high metastable calcite polymorph content ;
this is mathematically expressed by Eq. ().
Solute diffusion tracts and the sedimentation of the carbon isotopes of
DIC and solid-phase carbonate are defined as 13C and
12C, respectively, and are calculated separately. Metabolically
introduced carbon (vmetabolism) by reaction (Rk) and its respective
effect on porewater DIC carbon isotope composition is formulated by
Eqs. () and () for the heavy and light isotope, respectively.
vmetabolism=Rk11+rkvmetabolism=Rkrk1+rk
We assign a 13C/12C ratio (r) to the metabolism-specific
produced DIC from the organic carbon remineralization reactions
according to the data in Table . The mathematical formulation of
metabolism-introduced DIC ensures that the isotope composition of
ambient porewater DIC is altered and ensures mass balance in the model carbon input and output.
Numerical solutions and parameters
The model includes nine state variables: organic matter (CH2O),
solid carbonate (13Ccarb and 12Ccarb), DIC
(13CDIC and 12CDIC), dissolved oxygen (O2),
dissolved sulfate (SO42-), dissolved methane (CH4) and
dissolved sulfide (HS-). We numerically solved Eqs. () and
() on the open-source programming platform R
with a finite difference approach by expanding the spatial derivatives of
partial differential equations over a sediment grid .
Through the
application of the R package ReacTran , a
sediment grid 10 m thick with 20 000 layers of identical thicknesses was
generated. The finite difference approach transforms the partial
differentials into ordinary differentials which are subsequently integrated
by a stiff solver routine vode within the R
package DeSolve . Boundary conditions were
chosen to reflect Permian oceanic conditions based on the current state of
research (Sect. ) and complemented by current knowledge of
shallow marine environments. The upper boundary conditions were set to
concentrations for normal modern bottom water conditions: 0.28 µmolcm-3O2 and zero for the reduced species (CH4 and
HS-). A lower seawater SO42- value of 4 µmolcm-3 (compared to the modern value of 28 µmolcm-3)
was chosen based on evidence of a reduced global marine S reservoir in the
latest Permian to Early Triassic . On
the other hand, DIC was set at 4.5 µmolcm-3 (higher
than the modern value of 2.2 µmolcm-3) based on model
calculations of a Late Permian ocean without pelagic calcifiers
. The solid phases entering the model at the top of the
sediment stack are set at 730 µmolcm-2yr-1 for both the
OC flux (FOC) and the carbonate flux (Fcarb) in the baseline
model, which are typical average shelf sedimentation values
. For the baseline condition,
sedimentation rates (v and w) are fixed at 0.2 cmyr-1 and the
δ13C composition is set at +5 ‰ (VPDB) based
on primary carbon isotope values from pristine preserved brachiopod calcite
from the Ali Bashi and Meishan sections, as well as sites in northern Italy
. For the bio-mixing and
bio-irrigation parameters, the values for the Palaeozoic proposed by
were chosen: Db0=5cm2yr-1, zb=2cm, α0=50yr-1, xirr=1cm.
This will approximate the sedimentary conditions of a sea floor inhabited by a
Permian benthic community. The relation between the isotope composition of
DIC and solid-phase carbonate is given by the following thermodynamic
relation (Eq. ) established by in which we
assume temperature (T) to be constant at 25 ∘C.
ϵcarb-DIC=1.85+0.035[T(∘C)-20]
The lower boundary conditions for both solid and solute species were set at a
“no gradient” boundary to ensure that materials are only transported by
burial to deeper parts of the sediment column. The kinetic parameters are
summarized in Table and are based on published models.
Parameter values for the kinetic constants of the reactive transport model.
Bulk-carbonate carbon isotope signal imprinting and its sensitivity towards environmental parameters
The previously defined parameters provide a baseline model to test the
sensitivity towards certain boundary conditions in bringing forward carbon
isotope compositional changes in the porewaters. Two pathways of carbonate
rock formation and stabilization are considered to be of importance in
capturing these porewater carbon isotope compositional changes.
The first is diagenetic carbonate alteration, which is envisioned as ongoing
recrystallization and dissolution with depth after burial, as formulated in
Eq. (), and with both processes operating in equilibrium
. Under this mechanism carbonate is constantly
exchanged between the solid (carbonate precipitate) and aqueous phase (porewater
DIC) with ongoing burial, thereby to some degree buffering the
isotopic perturbations caused by microbial mineralization. This pathway of
carbonate sediment progression can be viewed as the classical interpretation
of limestone stabilization: dissolution of less stable components and
subsequent occlusion of the produced pore spaces by cements
.
The second is authigenic carbonate addition, which is regarded to be a process of
near-instantaneous precipitation of carbonate crusts at (or close to) the
sea floor by microbial mat communities. The latter mechanism would simulate
the sea-floor crust formation commonly observed for the P–Tr transition,
which is recognized as a time of unusual carbonate sedimentation; e.g.,
stromatolites, thrombolites and microbially induced cementation
. However, the mechanism directly
responsible for this type of biologically induced calcification is ambiguous
and might be related to extracellular polymeric substances and the
orientation of functional groups on this organic matrix (creating nucleation
sites), destruction of organic compounds that inhibit calcification or a
direct consequence of metabolism and associated changes in ambient carbonate
chemistry . We further assume
that cryptic forms of carbonate precipitates, e.g., rock-binding cements
e.g.,, might have escaped detection when selecting
carbonate rock for carbon isotope studies.
These two pathways are henceforward referred to as diagenetic carbonate
alteration and authigenic carbonate addition, respectively. Note that
porewater chemistry and its control on carbonate formation are not explicitly
formulated in the model and are instead expressed by a reaction rate that can
vary with age (Eq. ) . This
formulation is justifiable due to the uncertainties in the ultimate control of
biological activity in steering calcification, as cited above.
Diagenetic carbonate alteration is determined by taking the evolved
solid-phase δ13C at 10 m of sediment depth where carbon
isotope exchange between porewater and solid carbonate reaches equilibrium
see alsofor a comparative Sr and Ca isotope
equilibrium with depth. Although transient seawater
chemistry changes can occur over the time needed to complete carbonate
stabilization (i.e., before reaching equilibrium at 10 m of depth), we consider
the adopted isotope value to be a good approximation of a partially
recrystallized carbonate rock, as diffusion and the majority of metabolic
processes are confined to only a small interval of the sediment column (upper
∼1m of the sediment pile). Authigenic carbonate
addition is regarded to take place in the upper 0.1 m of the sediment
column through precipitation from porewater DIC that is most severely
perturbed by metabolic activity and largely unbuffered by exchange with
carbonate sediments through dissolution and recrystallization. The evolved
bulk-rock carbon isotope value is then calculated by using the mass balance of the primary
and authigenic carbonate (fauth) components.
An important parameter inducing changes in the vertical structure of a
diagenetic redox profile, and ultimately carbonate carbon isotope composition
during carbonate rock formation and stabilization, is the FOC. By
systematically changing this parameter we test the sensitivity of the
diagenetic baseline model to the OC sinking flux and the consequential
organic matter remineralization trajectories, i.e., the dominant microbial
communities involved in local in situ C-cycling. The effect
of the OC sedimentary regime on the carbonate isotope system is probed
by looking at the C isotope offset, the primary carbonate precipitate
(settling through the water column) and the bulk-rock end-member
(Δ13Cprimary-bulk). In addition to the OC flux, we
performed sensitivity experiments for the sedimentation rate (w and v), the
concentrations of bottom water DIC and sulfate and the fraction
of diagenetic and/or authigenic carbonate
(fdia and/or fauth) incorporated in the
bulk rock. For each separate sensitivity experiment we return the changing
parameter to the initial value, as defined for the baseline model
(Sect. ), except for the parameter of interest and the OC
flux. In addition, we tested the influence of bio-diffusion (Db0) and
bio-irrigation (α0) under varying depth profiles of sediment mixing
on Δ13Cprimary-bulk.
ResultsFirst-order temporal trends
By compiling δ13Ccarb data and unifying them in a
biochronological framework, we can distinguish first-order temporal features
of the combined Iranian and the Meishan P–Tr (sub-)localities in China
(Fig. ). The subsampling routine prevents modifications through unequal
sampling intensity; hence this temporal pattern is unbiased by the sampling
strategy applied in individual studies. A gradual decline towards
4 ‰ lower δ13Ccarb can be discerned
starting from the middle Changhsingian, with minimum values
(0 ‰) reached in the earliest Triassic and a return to
2 ‰ higher δ13Ccarb commencing above the
I. isarcica zone (unit I). Nonetheless, disparity in regional trends
is visible as a double-peaked negative excursion marking the P–Tr
transitional beds of Meishan, whereas such a signature is absent in
time-equivalent boundary beds deposited at the Iranian sites.
By comparing the combined subsampled median trend lines of individual studies
(constructed following the same statistical routine as outlined in
Sect. ), it is possible to determine whether the first-order
isotope trend is a consistent feature of the analysed stratigraphic
sequences. A large portion of the compared individual datasets is marked
with a high coefficient of determination (r2; Fig. ). We can
conclude that the geochemical signal obtained is reproducible and, as such,
observable in rock collected during separate sampling campaigns over several
decades. However, the comparative study also suggests that the correlation
coefficients are consistently weaker in the Meishan data relative to the
Abadeh data. This lack of reproducibility might stem from limited
stratigraphic coverage; see , who report on a
composite dataset of two sub-localities, Meishan D and Z, only encompassing
Changhsingian strata. Sampling campaigns of reduced stratigraphic coverage
might not be able to capture the first-order trend. Nevertheless,
region-specific differences in signal reproducibility mark an overall greater
disparity between individual Meishan δ13Ccarb records and
might point to a higher-order of δ13C variability.
A comparative analysis of first-order temporal trends in
δ13Ccarb from individual studies for the Abadeh and Meishan
(sub-)localities. The saturation level of the individual tiles measures the
correlative strength of stratigraphic trends obtained during individual
sampling campaigns based on the coefficient of determination (r2).
Residual carbon isotope variability
Subsampled and time-sliced median interpolations are depicted in
Fig. . Trends with the highest confidence are highlighted by a
well-defined white line, and the width of the CI is represented by a blue
colour that becomes less intense with increasing data spread. Combined, these
graphing features result in a more blurred image with a larger spread between
the median values of individual subsampling routines (Sect. ).
From these figures we can discern three features in the second-order
δ13C excursions: (1) the residual δ13C
variability is seemingly random or stochastic and defined δ13C
excursions are unreproducible across lithological successions from different
geographic locations (Iran) or studies targeting the same site (e.g.,
Meishan section; Fig. ). (2) There is a returning temporal pattern towards
decreased fidelity of the median trend line (blurred white trend line)
connected with increased CI dispersion (i.e., residual δ13C
variability seen as less saturated blue colour tones) across the extinction
horizon and ranging into the Early Triassic at both geographic locations.
Increased maximum δ13C value ranges from less than 1 and
2 ‰ (IQR) or 2 and 3 ‰ (IPR) in the
pre-extinction beds to more than 2 and 3 ‰ (IQR) or 5 and
8 ‰ (IPR) around the extinction horizon for Iran and China,
respectively, further enforces the observation of a globally significant peak
in residual δ13C variability. This pattern seems to recover
after the P–Tr boundary, with generally higher confidence median trend lines
and a trend to smaller maximum δ13C value ranges of less than
2 and 3 ‰ (IQR) or 5 and 8 ‰ (IPR) for Iran
and China, respectively. (3) In addition, a significant regional offset can be
discerned in δ13C value ranges; the Chinese profile
displays systematically higher residual δ13C variability.
Studies focussing on boundary events (e.g., the P–Tr transition) tend to
channel sampling effort at a focal point around the presumed faunal turnover
and horizon of palaeoenvironmental change . Sampling effort is
known to have a profound impact on studies that use fossil data to track
animal diversity through time . By correcting
for sampling effects through subsampling (Sect. ), we cancelled
out the potential effect of over-representative data accumulation on the
temporal trend of stochastic residual δ13Ccarb variability.
Sampling effort might also have affected fossil collecting, thereby biasing
conodont biozonation schemes, which heavily relies on the first-appearance
datum of certain fossil species . As such, it
would be conceivable that longer-duration biozones could capture a more
temporally variable marine DICδ13C compared to
shorter-duration biozones. When evaluating the average thickness and duration
of our chosen biochronological units, there is an apparent decrease in unit
thickness and a shorter duration when approaching the extinction
interval (Fig. ). A comparison of biozone stratigraphic thickness
and duration points to a relationship between higher
δ13Ccarb dispersion and smaller and shorter-duration units
(Fig. ). This inverse relationship suggests that
δ13Ccarb variability is not controlled by the increased
potential sample size and the inherent risk of sampling a more temporally
variable isotope signature. Hence, we exclude the applied sampling strategy
and the biochronological framework as causal factors behind the
observed temporal trend of stochastic δ13Ccarb variability.
Model response to organic carbon accumulation
In the above, we have conceptualized a model that links the sedimentation of
organic material to carbonate carbon isotope alteration. An important
variable in this model is the amount of organic carbon that arrives at the
sea floor, which controls the redox depth profile and the respective
importance of metabolic pathways as reaction per unit area. Low organic
carbon fluxes (500 µmolcm-3yr-1) yield aerobic respiration
and microbial sulfate reduction as important biochemical reactions
(Fig. ). On the other hand, the importance of this metabolic
pathway is reduced under high OC accumulation regimes
(> 1000 µmolcm-3yr-1), which are signified by intense
methane production through organoclastic methanogenesis (Fig. ).
Consequently, the dominant organic carbon remineralization pathway determines
the final evolved carbon isotope composition of the equilibrated bulk-rock
end-member, e.g., towards lower δ13Ccarb under low
OC accumulation and higher δ13Ccarb under high
OC loading of the sea floor (Figs. –).
Diagenetic depth profiles of Late Permian sea-floor sediments of
(a) oxidized and reduced solutes under a low shelf OC flux 500 µmolcm-3yr-1, (b)DIC and carbonate δ13C forced
with a low shelf OC flux 500 µmolcm-3yr-1,
(c) oxidized and reduced solutes under a high OC flux 1200 µmolcm-3yr-1, (d)DIC and carbonate δ13C forced
with a high OC flux 7000 µmolcm-3yr-1.
(e) Situational sketch of pre-extinction bulk-carbonate accumulation with a
normal OC flux and active benthic fauna and (f) situational sketch of
the post-extinction sedimentation with elevated OC accumulation,
removal of metazoan benthic fauna and consequentially reduced sediment mixing
(artwork by Mark Schobben; http://cyarco.com).
By systematically changing FOC we see a defined relationship between
bulk-rock C isotope alteration and the predominant mode of organic
matter remineralization, e.g., microbial sulfate reducers, AOM or
methanogenesis (Fig. ). The sensitivity experiments further reveal
that the fraction of diagenetic precipitate incorporation and bottom
DIC have only a limited effect on the difference between the primary
carbonate precipitate and the bulk-rock end-member C isotope value
(Δ13Cprimary-bulk). In contrast, marine sulfate levels
modulate the total range of observed δ13C (∼2–3.5 ‰) and introduce a switch in the system, which narrows
the range of OC accumulation over which the maximum deviation in
Δ13Cprimary-bulk occurs (Fig. ). In a similar
fashion, elevated sedimentation (v and w) causes a shift in peak
Δ13Cprimary-bulk values. However, this shift in peak values
causes a modest increase in the total attained C isotope alteration
with varying OC accumulation under higher sedimentation rates (v and
w).
Sensitivity experiment for diagenetic carbonate alteration through
equilibrium recrystallization with depth designed to investigate the forcing
effect of (a) the fraction of diagenetic carbonate incorporated (fdia),
(b) sulfate content of overlying water, (c)DIC content of overlying
water and (d) sedimentation rate (v and w) on the carbon isotope offset
between the diagenetic end-member rock and the primary calcite over a range
of FOC. Panels (e) and (f) depict sensitivity experiments for
bio-irrigation (α0) and bio-diffusion (Db0), respectively, on
diagenetic altered carbonate for changing modes of sediment reworking and
irrigation by biota under a normal OC flux (730 µmolcm-3yr-1). An increasingly suppressed alteration of the end-member carbon isotope signal is observed for the pre-extinction
and modern depth profiles compared
with sediment that is not inhabited by metazoans (post-extinction).
A systematic study of OC accumulation results in comparable
trajectories of diagenetic C isotope modification on bulk rock through
authigenic carbonate addition (Fig. ). However, authigenic
sea-floor cementation results in a 10 ‰ range of attainable
bulk-rock end-member C isotope values, which is a wider range than
obtained for diagenetic carbonate alteration. This style of carbonate
cementation is also strongly controlled by seawater sulfate concentration.
Lower than modern marine dissolved sulfate values (28 µmolcm-3) allow for a large range of δ13C end-member
values to be generated under a smaller range of FOC (Fig. ).
As opposed to diagenetic carbonate alteration, heightened sedimentation (v
and w) only causes a shift in the switch from positive to negative
Δ13Cprimary-bulk in the end-member rock but does not change
the total range of C isotope alteration. However, both elevated bottom
water DIC and a decreased size of the authigenic fraction diminishes
the maximum range of Δ13Cprimary-bulk.
Both bio-mixing and bio-irrigation by pre-extinction Permian benthic fauna
(Db0 = 5 cm2yr-1, zb=2cm, α0=50yr-1, xirr=1cm) cause minimal modulation of
Δ13Cprimary-bulk. The influence of modern benthic fauna
(Db0>5cm2yr-1, zb=3cm, α0>50yr-1, xirr=2cm) would suppress
δ13C modifications caused by OC-steered diagenetic carbonate
alteration and authigenic carbonate addition. On the other hand, the absence
of benthic fauna (post-extinction situation) allows for an unconstrained
impact of the previously cited OC-controlled trajectories of C isotope modification during bulk-rock formation and stabilization under
P–Tr environmental conditions.
Sensitivity experiment for authigenic carbonate addition through sea-floor
cementation designed to investigate the forcing effect of (a) the fraction
of authigenic carbonate (fauth) incorporated, (b) sulfate content of
overlying water, (c)DIC content of overlying water and
(d) sedimentation rate (v and w) on the carbon isotope offset between the
diagenetic end-member rock and the primary calcite over a range of FOC.
Panels (e) and (f) depict sensitivity experiments for bio-irrigation
(α0) and bio-diffusion (Db0), respectively, on authigenic
carbonate addition for changing modes of sediment reworking and irrigation by
biota under a normal OC flux (730 µmolcm-3yr-1). An
increasingly suppressed alteration of the end-member carbon isotope signal is
observed for the pre-extinction and modern
depth profiles compared with sediment that is not
inhabited by metazoans (post-extinction).
Simulation of a virtual carbon isotope time series
In order to understand the obtained temporal patterns in residual
δ13Ccarb variability, a series of reactive transport models
have been solved to steady state under varying OC sedimentation
regimes. The virtual time series approach is an amalgamation of multiple sets
of individual reactive transport model runs (50 model iterations = 1 set = 1
time unit) sliding across a timeline of length 100 and with time increments
of length one (Fig. ).
By performing sensitivity tests (Sect. ), we have established
that there is a systematic relation between OC arriving at the seabed and
the magnitude of bulk-rock C isotope alteration, which is expressed as
Δ13Cprimary-bulk (Fig. ). Hence, the initial
FOC can be estimated because the Δ13Cprimary-bulk can
be approximated by the C isotope offset between calcite from
well-preserved Permian brachiopods and time-equivalent bulk-rock samples
(∼1.0 ‰) and equates to
802 µmolcm-3yr-1. The FOC is subsequently modulated
for each set (i.e., time unit) based on the observation that OC accumulation
increases by a factor of 4 across the P–Tr transition
and by linearly scaling this parameter to the observed residual
δ13Ccarb variability (IQR) as obtained from
Fig. , yielding a continuous range for FOC (802≤FOC≤3206µmolcm-3yr-1).
Schematic depiction of the work flow behind the virtual carbon
isotope time series. Model sets consist of 50 separate reactive transport
model solutions driven by a randomly selected FOC. The randomly
generated organic OC flux values translate into a density distribution
with a right-skewed tail. Increased FOC and a widening of the
value range is used to simulate increased spatial OC heterogeneity.
With regard to examining the effects of spatially distinct OC sinking
fluxes and/or benthic fauna on the lateral distribution of OC, we
forced individual model runs (within one time unit) with slightly deviating
FOC values. The maximum attained FOC variability is constrained by
adopting a log-normal skewed density distribution R package
emdbook; around the former established
FOC, which represents the population's median value (Fig. ),
and by randomly selecting a value from the created distribution. A log-normal
skewed density distribution has been chosen to represent natural variation in
the FOC-population, which would be signified by variation that is skewed
towards heavy sea-floor OC loading in rare instances, whereas most
variation is smaller in magnitude and can modulate the flux towards both
smaller and higher contributions. Temporal variations in this parameter are
attained by modulating the width of the sample population ranging up to a
factor of 1.5. This lateral OC dispersion parameter is also
linearly scaled to the observed residual δ13Ccarb
variability (IQR) as obtained from Fig. . This results in
sample population variations of 524 ≤ median absolute deviation ≤ 3889 µmolcm-3yr-1 when combined with the initial
FOC range for homogeneously and more heterogeneously dispersed sedimentary
OC, respectively.
The initial carbon isotope composition of Fcarb is based on the median
δ13Ccarb values of the first time unit of the pooled
Iranian dataset (Fig. ). This value is subsequently corrected by
-1.0 ‰ based on the systematic isotope offset
between bulk carbonate and pristine carbonate from brachiopod shells
in order to obtain the C isotope value of Fcarb.
The systematic relationship of OC sedimentation with
Δ13Cprimary-bulk, as obtained from Fig. ,
yielded the primary δ13C of calcium carbonate particles
arriving at the seabed for successive time units.
The processes governing increased authigenic sea-floor cementation in
post-extinction strata are largely speculative. However, undisturbed
substrates are a feature required to enable the growth of calcifying microbial
mats, regardless of the mechanism inducing the global-scale proliferation of
these sea-floor precipitates e.g.,. Considering
this universal control, it is justified to assume a direct link between
bioturbation intensity and authigenic sea-floor cementation. This assumption
links both forms of carbonate rock formation and stabilization: equilibrium
recrystallization and authigenic sea-floor cementation. The probability
(p) of producing authigenic sea-floor cements has been assumed to
vary linearly in the range 0.01≤p≤0.1 following the upper two
terciles of the range attained by the previously defined OC dispersion
parameter. This definition determines the response of introducing authigenic
sea-floor cement δ13C to the sample pool.
In order to further simulate the mass extinction of marine metazoans across
the end-Permian extinction on sedimentary conditions, the parameters that
define bio-mixing and bio-irrigation (Sect. ) are set to zero
from time unit 60 onward (equivalent to biozone G; Fig. and
Table ). In addition, it has been postulated that sedimentation
rates increased over the P–Tr transition , so the
sedimentation rate (of v and w) was modulated at time unit 70 (equivalent
to biozone H; Fig. and Table ). However, a
conservative 2-times sedimentation increase is adopted, as the previously
cited estimates largely hinge on calculations based on the duration of the
respective stages, which are subject to continuous modification
. The fdia and/or fauth for both trajectories of carbonate
rock stabilization is kept constant at 0.2. The final bulk-rock carbon
isotope value is obtained by using the mass balance from the calculated primary and
diagenetic and/or authigenic carbonate components.
A total of 5050 reactive transport models have been solved in order to match
the length and resolution of the cumulative “real” δ13C time
sequence from sites in Iran (Fig. ). The resulting virtual time
series of model runs forced by previously established value ranges for the
OC sinking flux, sediment homogeneity and authigenic mineral addition
is depicted in Fig. . Although these steady-state solutions will
not always be truly representative of a dynamic sedimentary environment
(e.g., sedimentation rate changes and sediment reworking), we conclude that
the conceptualized model (Sect. ) and selected parameter values
can explain trends in residual carbon isotope variability.
Subsampled virtual carbon isotope time series with model results
plotted as “visually weighted” trends. Each time unit incorporates 50 model
iterations with a randomly chosen OC flux from a log-normal skewed
density distribution. The time series simulates an increase of 4 times the
average OC flux, a 1.5 times increase in the spatial heterogeneity of
OC accumulation and increased likelihood of producing sea-floor
authigenic carbonate to a maximum of 1 in 10. Bioturbation (expressed as
bio-mixing and bio-irrigation) is modulated at the extinction horizon to
mimic the effect of a die-off among benthic organisms. In addition,
sedimentation (v and w) is increased twofold over the P–Tr transition
(see main text for sources).
DiscussionStochastic carbon isotope signatures of the Permian–Triassic boundary beds
Global comparisons of P–Tr carbon isotope records have uncovered disparate
δ13C trends and excursions (Fig. ). These
deviations are often encountered on refined stratigraphic and lateral scales,
complicating interregional high-resolution correlations . Subaerial exposure and projected trajectories of bulk-carbonate
stabilization under the influence of meteoric fluids and high water–rock
ratios might be invoked to explain this disparity . Other studies have pointed to stratigraphic variations in
predominant mineral rock composition as a source of
δ13Ccarb modulations, citing the mineral-specific isotope
offset between aragonite, dolomite and low-Mg calcite
.
In this study, we provide an alternative explanation for disparate
second-order δ13Ccarb excursions involving the observed
stochastic residual δ13C variability on a refined spatial
scale (Fig. ). We regard this test to be of importance, as a
transgression marks the boundary beds and Lower Triassic deposits
(Fig. ). Furthermore, subaerial exposure and consequential early
rock stabilization with the high water-to-rock ratios (open system) of the
studied localities is unlikely to have left an isotopic overprint. Under such
circumstances, regionally extensive carbonate sediment dissolution through contact
with an undersaturated solution could have sourced the underlying sediment
pile with saturated fluids for cementing . However,
evidence of such exposure in the form of karstification is absent in the
studied locations . A recent study on the Meishan
section suggests that recrystallization (i.e., zoned dolomite crystals) and
the modification of the C isotope composition of certain stratigraphic levels
was forced by a short-lived regression and consequential exposure to meteoric
water . However, the assignment of such petrological indices to
specific diagenetic environments is often fraught with uncertainties, and it is
not uncommon for alleged diagnostic features to occur in a range of diagenetic
environments . Although a complete exclusion of isotope
resetting by meteoric water is not possible, it is possible that a large
portion of the global carbonate archive lithified without being subjected to
meteoric fluids, therefore justifying the exploration of alternative modes of
isotope resetting.
Systematically 12C-depleted bulk rock relative to brachiopod
calcite further counters the notion of a variable
predominance of an aragonite precursor for Permian carbonate rock
. Nonetheless, secondary calcite and dolomite
additions might be a causal factor behind observed δ13C
variability. However, these additions can also be the result of in
situ microbially steered mineral formation, which incorporates metabolically
produced DIC isotope signatures . As
such, our conceptualized model is not mutually exclusive with regards to
carbonate polymorph compositional changes, but it does not require rapid
global secular marine ion inventory changes. Moreover, our model outcome
predicts δ13C variation driven solely by small biologically
steered mineral contributions (fdia and/or fauth=0.2; Fig. ),
whereas the recorded δ13C fluctuations are up to 10 times
higher in amplitude than otherwise predicted for thermodynamically driven
mineral-specific isotope compositional offsets . Note also that the mineral-specific isotope offset would
require bulk-rock samples to display effectively complete (100 %) mineral
compositional changes to a pure carbonate polymorph composition. In contrast,
our model predicts a mechanism that can introduce temporal variation in
spatially heterogeneous carbonate polymorph assemblies and associated isotope
compositional differences through post-depositional processes driven by spatially
and temporally variable OC accumulation comparable to modern ranges
. Depleted marine dissolved sulfate concentrations
would have exacerbated the sensitivity of carbonate rock towards only small
deviations in this allochthonous C source (Figs.
and ). Combined with a less well-mixed sediment and microbially
steered cementation, this model can account for the entirety of the observed
residual δ13Ccarb variability.
The carbon isotopic composition of the Permian–Triassic DIC reservoir
The first-order P–Tr negative carbon isotope trend can be explained by
changes in the sources and sinks of the long-term (> 100 ky) carbon
cycle, e.g., a reorganization of organic carbon burial and volcanic
CO2 outgassing .
Carbon-cycle partitioning between the deep and shallow ocean through the
marine biological pump have been proposed to explain the depth-gradient
isotope differences or a rapid
(< 100 ky) C isotope excursion . On the
one hand, the stochastic residual δ13Ccarb variability on
a confined centimetre–metre stratigraphic scale is inherently difficult to
reconcile with transient perturbations in the dynamic C-cycle
equilibrium enforced by repeated biological reordering of DIC between
the shallow and deep ocean (i.e., consecutive events of global productivity
increase and wholesale collapse). On the other hand, if lateral variations in
marine DIC-C isotope composition account for
δ13Ccarb variability, water column mixing would likely
erase spatial compositional differences on smaller scales (<km scale). For example, residual δ13Ccarb
variability recorded at Meishan (Fig. ) on a centimetre to
metre scale cannot be reconciled with the heterogeneity of the overlying
water mass, and less well-mixed sediments become a more acceptable
alternative. Although the geographically disparate C isotope signals of
different sites in Iran (>km scale) could instead be accounted for
by the heterogeneity of water masses, the temporal trend in the amplitude of the
residual δ13Ccarb variability is similar for both China and
Iran (Fig. ). An overarching control of the here-invoked
OC-steered diagenetic mechanism is the most parsimonious model to
explain this geographical coherence.
Sea-floor methane clathrate dissociation, coal and organic-rich sediment
combustion through igneous intrusions and ocean Ni fertilization
stimulating methane production by methanogens are viable sources of transient
(< 100 ky) isotope-depleted carbon contributions to the
atmosphere–ocean system . Although all of the cited
mechanisms are conceivable triggers for defined second-order
δ13Ccarb excursions, they are less likely to create
spatially divergent isotope trends and repeated temporally distinct
fluctuations; instead, they essentially reflect bed-to-bed variation. The
existence of transient carbon isotope excursions is also unlikely
considering the predicted elevated DIC levels of an ocean without
pelagic calcifiers, resulting in conservative behaviour of the
DIC pool towards transient C isotope perturbations i.e.,
increasing the system response time;.
Nonetheless, second-order (transient and depth-gradient) C isotope signals of
a primary origin might still be imprinted in the isotope records. However,
rock formed under the previously outlined marine depositional conditions
(with intensified spatially heterogeneous trajectories of C isotope
alteration) increases the chance of sampling spatially variable and
diagenetically modulated δ13C signals, and these would obscure
primary signals.
Perspectives on the Permian–Triassic carbonate archive: stratigraphy, biological evolution and the global biogeochemical carbon cycle
These findings put renewed constraints on the application of whole-rock
δ13C as a high-resolution stratigraphic tool. Diagenesis
forced by a recrystallization process under the influence of meteoric
fluids, high fluid-to-rock ratios and oxidized terrestrially derived organics
is generally considered to be the main driver of bulk-rock C isotope
alteration . Rocks
without petrological evidence of meteoric fluids percolating
through voids and interacting with sediment grains, e.g., palaeokarsts,
blocky calcite and pendant cements, are usually regarded as pristine and
minimally isotopically altered . Regarding the observed P–Tr residual whole-rock
δ13C variability and model predictions, we have shown that
under certain marine and sedimentary situations marine carbonate rock
diagenesis can introduce significant δ13C fluctuations to the
whole-rock isotope composition. The insight gained here undermines the idea
of using second-order δ13C fluctuations as stratigraphic
markers, which are often confined to limited stratigraphic intervals (e.g.,
individual limestone beds) without a proper understanding of the potential
effects of diagenetic carbon isotope modification. At least for the data
presented here, we can demonstrate that shifts in bulk-rock
δ13C smaller than classical biozonation schemes have to be
considered with caution when they coincide with the sedimentological features of
reduced bioturbation (e.g., lamination) and likely do not serve as a
globally universal marker. Instead, long-term first-order trends
cross-cutting formational boundaries (see Supplement) have a
higher likelihood of representing secular variation in seawater DIC isotope composition and are largely unaffected by marine diagenetic
processes (Fig. ). Extrapolation of our model results to Mesozoic
and Cenozoic bulk-carbonate records signified by comparatively reduced
δ13C variability might reflect buffering of highly variable
diagenetic C isotope contributions under elevated marine sulfate
levels and physical sediment mixing by benthic fauna.
The carbonate chemical signal towards increased δ13C
variability supports the global significance of three parameters during
the deposition of P–Tr carbonate rock: reduced sediment mixing, authigenic
precipitates and heterogeneous OC accumulation. Although we have
postulated a connection between sediment lamination and C isotope
variability, the relation of these observations to the end-Permian mass
extinction is unconstrained. The cataclysmic mechanism that drove the
extinction of benthic fauna is still a matter of debate, but bottom water
anoxia is one of the main contenders. This could have been caused by
decreased seawater oxygenation instigated by stagnating ocean circulation
. Alternatively, a higher OC sinking
flux might have led to increased O2 drawdown through aerobic respiration
and H2S production by sulfate-reducing microbes, leading to
widespread marine euxinia .
However, δ13C variability complies most compellingly with at
least continuing OC accumulation over the studied interval and
thereby largely undermines scenarios of reduced primary productivity
e.g.,. On the other hand, the proliferation of
post-extinction authigenic sea-floor precipitates (e.g., thrombolites,
stromatolites and fan-shaped structures) likely represents a symptom of
reduced sediment disturbance and elevated carbonate saturation
. Since some of the more conspicuous
post-extinction sea-floor structures have been connected with microbial
communities and distinct metabolism-mediated isotope signatures
e.g.,, they might be equally important constituents of
the observed spatial δ13C variability. Geographic differences
observed as systematically higher residual δ13C variability at
the Chinese localities are suggestive of a regionally distinct hydrographic
and depositional setting. This regional feature could be linked with evidence
of episodic Late Permian euxinic conditions in the South China basin
. Comparatively higher and more spatially variable
OC loading of the sea floor in the South China basin under these
circumstances might be triggered by enhanced primary productivity or
OC sinking fluxes . On the other hand, the
depositional sites found in Iran lack evidence for bottom water anoxia and
favour other drivers behind the marine faunal extinction, e.g., thermal stress
or ocean acidification .
In addition to capturing regional biological expressions of the global faunal
disruption, our approach illuminates potentially important feedbacks of the
Earth system that control the global-scale carbon cycle. Recently, attention
has increased on the potential significance of authigenic carbonate
production for the global carbonate reservoir and the
addition of authigenic carbonate to Permian and Triassic carbonate-bearing
sequences . The numerical
exercise highlights the link between local sedimentary OC-cycling and
bulk-rock diagenetic stabilization, hinting at the potentially significant
portion of remineralized OC sequestered by authigenic carbonates.
Additionally, bulk-rock residual δ13C variability might serve
as an indicator of infaunal biological activity, effectively preserving an
ecological signal, which could in turn be an indicator of local sedimentary
C-cycling. This sub-cycle is strongly controlled by bioturbation and
controls the amount of buried organic matter and the supply of
electron acceptors . These fluxes drive
carbon removal from the exogenic carbon reservoir and potentially require a
revision of generally accepted constraints on these budgets, including during
events with postulated major C-cycle perturbations, such as the latest
Permian extinction.
Conclusions
The δ13Ccarb record straddling the P–Tr
boundary of Iran and South China can be decomposed in a first-order temporal
trend with a negative trend marking the transition beds and superimposed
second-order residual δ13C variability. The primary goal of
the current study was to delineate the nature of the second-order
δ13C fluctuations, which harbour a reproducible temporal
pattern towards increased variability across the extinction beds. By
investigating the origin of these second-order geochemical signals, it is
possible to better constrain the limits of carbon isotope chemostratigraphy,
most notably the time resolution on which it can be applied. Having
established that second-order C isotope variability traces local
effects rather than the global carbon reservoir, we discerned that
mineralogical heterogeneity and meteorically steered cementation are
unsatisfactory candidates to explain this spread in δ13C.
Hence, the possibility of metabolism-steered carbonate addition is more
likely, as the anaerobic microbial pathway introduces isotopically
depleted (or enriched) carbon to the porewater and the same microbes
are often connected with the production of diagenetic and authigenic carbonates.
Moreover, organic carbon degradation occurs in a part of the sediment column
signified by high reactivity of the primary carbonate sediments. By testing
these assumptions with reactive transport models and with the construction
of a virtual time series, we can recreate the observed variability in the end-member bulk-rock δ13C composition by inducing the heterogeneous
spatial distribution of sedimentary OC, i.e., a reduction of physical
sediment mixing or spatially divergent OC sinking fluxes. Authigenic
sea-floor cementation, possibly mediated by microbial mat communities, is
another source that might have introduced diagenetic overprints in the P–Tr
boundary beds. On the other hand, these mechanisms can still preserve
long-term trends in oceanic DIC carbon isotope composition. These
findings strengthen the notion that bulk carbonate δ13C is a
fruitful source of information that records secular trends in carbon cycle
evolution and potentially informs us about the biosphere, notably
infaunal animal activity, sea-floor microbial communities and OC
delivery to the sea floor.
Data visualization, statistical data treatment and the reactive transport model have been written
in the R language. The R scripts and the data-mined and newly produced 13C
values are available on the GitHub repository:
https://github.com/MartinSchobben/carbonate
(10.5281/zenodo.888754).
The Supplement related to this article is available online at https://doi.org/10.5194/cp-13-1635-2017-supplement.
MS designed the study. MS, CK, VH, AG and LL collected material in the field
and prepared samples for isotope analysis. US, CK and CVU performed C isotope analysis. MS and
SvdV constructed the numerical model. MS and JG data mined the complementary published C isotope
results. All authors provided intellectual input and contributed to writing the paper.
The authors declare that they have no conflict of interest.
Acknowledgements
We thank Sylvain Richoz (Lund University, Sweden) and Yoshitaka Kakuwa
(University of Tokyo, Japan) for providing us with
published carbon isotope data. We acknowledge support provided by the Aras Free
Zone Office and Adel Najafzadeh (Tabriz, Iran) to sample the Ali Bashi, Aras
Valley and Zal sites and Bo Petersen (University of Copenhagen, Denmark) for
analytical assistance in the laboratory. This project was funded by the Deutsche
Forschungsgemeinschaft (DFG; projects KO1829/12-1, KO1829/12-2, KO2011/8-1,
KO1829/18-1 and FOR 2332). Martin Schobben is currently funded by a DFG Research
Fellowship (SCHO 1689/1-1), Sebastiaan van de Velde is supported by a PhD fellowship from the
Research Foundation Flanders (FWO), Clemens Vinzenz Ullmann acknowledges funding from NERC grant
NE/N018508/1 and Simon W. Poulton acknowledges support from a Royal Society Wolfson
Research Merit Award. The publication of this paper was funded by the Open
Access Fund of the Leibniz Association.
Edited by: Arne Winguth
Reviewed by: two anonymous referees
ReferencesAlgeo, T. J. and Twitchett, R. J.: Anomalous Early Triassic sediment fluxes
due to elevated weathering rates and their biological consequences, Geology,
38, 1023–1026, 10.1130/G31203.1, 2010.Algeo, T. J., Kuwahara, K., Sano, H., Bates, S., Lyons, T., Elswick, E.,
Hinnov, L., Ellwood, B., Moser, J., and Maynard, J. B.: Spatial variation in
sediment fluxes, redox conditions, and productivity in the Permian-Triassic
Panthalassic Ocean, Palaeogeogr. Palaeoclimatol. Palaeoecol., 308,
65–83, 10.1016/j.palaeo.2010.07.007,
2011.Algeo, T. J., Henderson, C. M., Tong, J., Feng, Q., Yin, H., and Tyson, R. V.:
Plankton and productivity during the Permian–Triassic boundary crisis: An
analysis of organic carbon fluxes, Global Planet. Change, 105, 52–67,
10.1016/j.gloplacha.2012.02.008,
2013.Alroy, J., Aberhan, M., Bottjer, D. J., Foote, M., Fürsich, F. T.,
Harries, P. J., Hendy, A. J. W., Holland, S. M., Ivany, L. C., Kiessling, W.,
Kosnik, M. A., Marshall, C. R., McGowan, A. J., Miller, A. I., Olszewski,
T. D., Patzkowsky, M. E., Peters, S. E., Villier, L., Wagner, P. J., Bonuso,
N., Borkow, P. S., Brenneis, B., Clapham, M. E., Fall, L. M., Ferguson,
C. A., Hanson, V. L., Krug, A. Z., Layou, K. M., Leckey, E. H.,
Nürnberg, S., Powers, C. M., Sessa, J. A., Simpson, C., Tomašových, A.,
and Visaggi, C. C.: Phanerozoic trends in the global diversity of marine
invertebrates, Science, 321, 97–100, 10.1126/science.1156963, 2008.Arp, G., Reimer, A., and Reitner, J.: Photosynthesis-induced biofilm
calcification and calcium concentrations in Phanerozoic oceans, Science,
292, 1701–1704, 10.1126/science.1057204,
2001.Auguie, B.: gridExtra: Miscellaneous Functions for “Grid” Graphics,
available at: http://CRAN.R-project.org/package=gridExtra (last access: November 2017), r package
version 2.0.0, 2016.Banner, J. L. and Hanson, G. N.: Calculation of simultaneous isotopic and
trace element variations during water-rock interaction with applications to
carbonate diagenesis, Geochim. Cosmochim. Ac., 54, 3123–3137,
10.1016/0016-7037(90)90128-8,
1990.Bathurst, R. G. C.: Cementation, in: Developments in sedimentology 12:
Carbonate sediments annd their diagenesis, 10, 415–457, Elsevier,
10.1016/S0070-4571(08)70904-X,
1975.Bathurst, R. G. C.: Microfabrics in Carbonate Diagenesis: A Critical Look at
Forty Years in Research, Springer New York, New York,
10.1007/978-1-4684-9421-1_1, 1993.
Baud, A., Magaritz, M., and Holser, W. T.: Permian-Triassic of the Tethys:
Carbon isotope studies, Geologische Rundschau, 78, 649–677, 1989.
Baud, A., Cirilli, S., and Marcoux, J.: Biotic response to mas extinction: the
lowermost Triasic microbialites, Facies, 36, 238–242, 1997.Berg, P., Rysgaard, S., and Thamdrup, B.: Dynamic modeling of early diagenesis
and nutrient cycling. A case study in an Arctic marine sediment, Am.
J. Sci., 303, 905–955, 10.2475/ajs.303.10.905,
2003.Berner, R. A.: An idealized model of dissolved sulfate distribution in recent
sediments, Geochim. Cosmochim. Ac., 28, 1497–1503,
10.1016/0016-7037(64)90164-4, 1964.Berner, R. A.: Examination of hypotheses for the Permo-Triassic boundary
extinction by carbon cycle modeling, Proc. Natl. Acad.
Sci. USA, 99, 4172–4177,
10.1073/pnas.032095199, 2002.Berner, R. A., Scott, M. R., and Thomlinson, C.: Carbonate alkalinity in the
pore waters of anoxic marine sediments, Limnol. Oceanogr., 15,
544–549, 10.4319/lo.1970.15.4.0544,
1970.Birgel, D., Meister, P., Lundberg, R., Horath, T. D., Bontognali, T. R. R.,
Bahniuk, A. M., de Rezende, C. E., Vasconcelos, C., and McKenzie, J. A.:
Methanogenesis produces strong 13C enrichment in stromatolites of Lagoa
Salgada, Brazil: a modern analogue for Palaeo-/Neoproterozoic
stromatolites?, Geobiology, 13, 245–266, 10.1111/gbi.12130, 2015.Boetius, A., Ravenschlag, K., Schubert, C. J., Rickert, D., Widdel, F.,
Gieseke, A., Amann, R., Jørgensen, B. B., Witte, U., and Pfannkuche, O.:
A marine microbial consortium apparently mediating anaerobic oxidation of
methane, Nature, 407, 623–626, 10.1038/35036572, 2000.
Bolker, B. M.: Ecological Models and Data in R, Princeton University Press,
2008.Bolker, B.: emdbook: Ecological Models and Data in R,
available at: https://CRAN.R-project.org/package=emdbook (last access: November 2017), r package version
1.3.9., 2016.Bontognali, T. R. R., Mckenzie, J. A., Warthmann, R. J., and Vasconcelos, C.:
Microbially influenced formation of Mg-calcite and Ca-dolomite in the
presence of exopolymeric substances produced by sulphate-reducing bacteria,
Terra Nova, 26, 72–77, 10.1111/ter.12072,
2014.Boudreau, B. P.: On the equivalence of nonlocal and radial-diffusion models
for porewater irrigation, J. Mar. Res., 42, 731–735,
10.1357/002224084788505924, 1984.Boudreau, B. P.: A method-of-lines code for carbon and nutrient diagenesis in
aquatic sediments, Comput. Geosci., 22, 479–496,
10.1016/0098-3004(95)00115-8, 1996.Boudreau, B. P.: Diagenetic models and their implementation. Modelling
transport and reactions in aquatic sediments, 3, Springer, New York,
10.I007/97S-3-642-60421-S, 1997.Boudreau, B. P. and Meysman, F. J. R.: Predicted tortuosity of muds, Geology,
34, 693–696, 10.1130/G22771.1, 2006.Brand, U. and Veizer, J.: Chemical diagenesis of a multicomponent carbonate
system-1: Trace elements, J. Sediment. Res., 50, 1219–1236,
10.1306/212F7BB7-2B24-11D7-8648000102C1865D,
1980.Brand, U. and Veizer, J.: Chemical diagenesis of a multicomponent carbonate
system-2: Stable isotopes, J. Sediment. Petrol., 51, 987–997,
10.1306/212F7DF6-2B24-11D7-8648000102C1865D, 1981.Brand, U., Jiang, G., Azmy, K., Bishop, J., and Montañez, I. P.:
Diagenetic evaluation of a Pennsylvanian carbonate succession (Bird Spring
Formation, Arrow Canyon, Nevada, U.S.A.) – 1: Brachiopod and whole rock
comparison, Chem. Geology, 308–309, 26–39,
10.1016/j.chemgeo.2012.03.017,
2012a.Brand, U., Posenato, R., Came, R., Affek, H., Angiolini, L., Azmy, K., and
Farabegoli, E.: The end-Permian mass extinction: A rapid volcanic CO2 and CH4 climatic catastrophe, Chem. Geology, 322,
121–144, 10.1016/j.chemgeo.2012.06.015,
2012b.Brosse, M., Bucher, H., and Goudemand, N.: Quantitative biochronology of the
Permian-Triassic boundary in South China based on conodont unitary
associations, Earth Sci. Rev., 155, 153–171,
10.1016/j.earscirev.2016.02.003, 2016.Brown, P. N. and Hindmarsh, A. C.: Reduced storage matrix methods in stiff ODE
systems, Appl. Math. Comput., 31, 40–91,
10.1016/0096-3003(89)90110-0,
1989.Bundeleva, I. A., Shirokova, L. S., Pokrovsky, O. S., Bénézeth, P.,
Ménez, B., Gérard, E., and Balor, S.: Experimental modeling of
calcium carbonate precipitation by cyanobacterium Gloeocapsa sp.,
Chem. Geology, 374–375, 44–60, 10.1016/j.chemgeo.2014.03.007, 2014.Busenberg, E. and Plummer, L. N.: Thermodynamics of magnesian calcite
solid-solutions at 25 ∘C and 1 atm total pressure, Geochim.
Cosmochim. Ac., 53, 1189–1208, 10.1016/0016-7037(89)90056-2,
1989.Canfield, D. E.: Sulfate reduction and oxic respiration in marine sediments:
implications for organic carbon preservation in euxinic environments,
Deep
Sea Res., 36, 121–138, 10.1016/0198-0149(89)90022-8,
1989.Canfield, D. E., Jørgensen, B. B., Fossing, H., Glud, R., Gundersen, J.,
Ramsing, N. B., Thamdrup, B., Hansen, J. W., Nielsen, L. P., and Hall, P. O.:
Pathways of organic carbon oxidation in three continental margin
sediments, Mar. Geology, 113, 27–40, 10.1016/0025-3227(93)90147-N,
1993.Cao, C., Wang, W., and Jin, Y.: Carbon isotope excursions across the
Permian-Triassic boundary in the Meishan section, Zhejiang Province,
China, Chinese Sci. B., 47, 1125–1129, 10.1360/02tb9252, 2002.Cao, C., Love, G. D., Hays, L. E., Wang, W., Shen, S., and Summons, R. E.:
Biogeochemical evidence for euxinic oceans and ecological disturbance
presaging the end-Permian mass extinction event, Earth Planet. Sci.
Lett., 281, 188–201, 10.1016/j.epsl.2009.02.012,
2009.Chen, J.-S., Chu, X.-L., Shao, M.-R., and Zhong, H.: Carbon isotope study of
the Permian-Triassic boundary sequences in China, Chem. Geology: Isotope
Geoscience Section, 86, 239–251, 10.1016/0168-9622(91)90052-X,
1991.Clarkson, M. O., Kasemann, S. A., Wood, R. A., Lenton, T. M., Daines, S. J.,
Richoz, S., Ohnemueller, F., Meixner, A., Poulton, S. W., and Tipper, E. T.:
Ocean acidification and the Permo-Triassic mass extinction, Science, 348,
229–232, 10.1126/science.aaa0193, 2015.Contreras, S., Meister, P., Liu, B., Prieto-mollar, X., Hinrichs, K.-u., and
Khalili, A.: subsea floor redox zonation on the Peruvian shelf, Proc. Natl. Acad. Sci., 110, 1–6,
10.1073/pnas.1305981110, 2013.Dale, A. W., Boyle, R. A., Lenton, T. M., Ingall, E. D., and Wallmann, K.: A
model for microbial phosphorus cycling in bioturbated marine sediments:
Significance for phosphorus burial in the early Paleozoic, Geochim.
Cosmochim. Ac., 189, 251–268, 10.1016/j.gca.2016.05.046, 2016.Dickens, G. R., O'Neil, J. R., Rea, D. K., and Owen, R. M.: Dissociation of
oceanic methane hydrate as a cause of the carbon isotope excursion at the end
of the Paleocene, Paleoceanography, 10, 965–971, 10.1029/95PA02087,
1995.Emrich, K., Ehhalt, D. H., and Vogel, J. C.: Carbon isotope fractionation
during the precipitation of calcium carbonate, Earth Planet. Sci.
Lett., 8, 363–371, 10.1016/0012-821X(70)90109-3,
1970.
Erwin, D. H.: The great Paleozoic crisis. Life and Death in the Permian,
Columbia University Press, New York, 1993.Fantle, M. S. and DePaolo, D. J.: Sr isotopes and pore fluid chemistry in
carbonate sediment of the Ontong Java Plateau: Calcite recrystallization
rates and evidence for a rapid rise in seawater Mg over the last 10 million
years, Geochim. Cosmochim. Ac., 70, 3883–3904,
10.1016/j.gca.2006.06.009,
2006.Fantle, M. S. and DePaolo, D. J.: Ca isotopes in carbonate sediment and pore
fluid from ODP Site 807A: The Ca(aq)2+-calcite equilibrium
fractionation factor and calcite recrystallization rates in Pleistocene
sediments, Geochim. Cosmochim. Ac., 71, 2524–2546,
10.1016/j.gca.2007.03.006,
2007.Flügel, E.: Microfacies of Carbonate Rocks, Springer-Verlag, Berlin
Heidelberg, 2nd edn., 10.1007/978-3-662-08726-8, 2010.Fossing, H., Berg, P., Thamdrup, B., Rysgaard, S., and Munk, H.: A model
set-up for an oxygen and nutrient flux model for Aarhus Bay (Denmark), 483,
National Environmental Research Institute,
available at: http://www.2.dmu.dk/1_Viden/2_Publikationer/3_fagrapporter/rapporter/FR483.PDF (last access: November 2017),
2004.Froelich, P. N., Klinkhammer, G. P., Bender, M. L., Luedtke, N. A., Heath,
G. R., Cullen, D., Dauphin, P., Hammond, D., Hartman, B., and Maynard, V.:
Early oxidation of organic matter in pelagic sediments of the eastern
equatorial Atlantic: suboxic diagenesis, Geochim. Cosmochim. Ac.,
43, 1075–1090, 10.1016/0016-7037(79)90095-4,
1979.Garcia-Pichel, F., Al-Horani, F. A., Farmer, J. D., Ludwig, R., and Wade,
B. D.: Balance between microbial calcification and metazoan bioerosion in
modern stromatolitic oncolites, Geobiology, 2, 49–57,
10.1111/j.1472-4669.2004.00017.x, 2004.Grice, K., Cao, C., Love, G. D., Böttcher, M. E., Twitchett, R. J.,
Grosjean, E., Summons, R. E., Turgeon, S. C., Dunning, W., and Jin, Y.:
Photic zone euxinia during the Permian-Triassic superanoxic event, Science,
307, 706–9, 10.1126/science.1104323, 2005.
Gruszczyński, M., Małkowski, K., Szaniawski, H., and Cheng-Yuan, W.:
The carbon biogeochemical cycle across the Permian–Triassic boundary
strata and its implications: Isotope record from the Changhsingian Stage at
Meishan, south China, Acta Geologica Polonica, 53, 167–179,
2003.Heindel, K., Richoz, S., Birgel, D., Brandner, R., Klügel, A., Krystyn,
L., Baud, A., Horacek, M., Mohtat, T., and Peckmann, J.: Biogeochemical
formation of calyx-shaped carbonate crystal fans in the subsurface of the
Early Triassic seafloor, Gondwana Research, 10.1016/j.gr.2013.11.004,
2013.Hermann, E., Hochuli, P. A., Bucher, H., Vigran, J. O., Weissert, H., and
Bernasconi, S. M.: A close-up view of the Permian–Triassic boundary based
on expanded organic carbon isotope records from Norway (Trøndelag and
Finnmark Platform), Global Planet. Change, 74, 156–167,
10.1016/j.gloplacha.2010.10.007,
2010.Heydari, E., Hassandzadeh, J., and Wade, W. J.: Geochemistry of central
tethyan Upper Permian and Lower Triassic strata, Abadeh region, Iran,
Sediment. Geol., 137, 85–99, 10.1016/S0037-0738(00)00138-X,
2000.Heydari, E., Wade, W. J., and Hassanzadeh, J.: Diagenetic origin of carbon and
oxygen isotope compositions of Permian-Triassic boundary strata, Sediment.
Geol., 143, 191–197, 10.1016/S0037-0738(01)00095-1,
2001.Heydari, E., Arzani, N., and Hassanzadeh, J.: Mantle plume: The invisible
serial killer – Application to the Permian–Triassic boundary mass
extinction, Palaeogeogr. Palaeoclimatol. Palaeoecol., 264,
147–162, 10.1016/j.palaeo.2008.04.013,
2008.Heydari, E., Arzani, N., Safaei, M., and Hassanzadeh, J.: Ocean's response to
a changing climate: Clues from variations in carbonate mineralogy across the
Permian-Triassic boundary of the Shareza Section, Iran, Global Planet.
Change, 105, 79–90, 10.1016/j.gloplacha.2012.12.013,
2013.
Hoffman, A., Gruszczynskj, M., Malkowskj, K., and Wski, H. S.: Should the
Permian/Triassic boundary be defined by the carbon isotope shift?, Acta
Geologica Polonica, 48, 141–148, 1998.Hofmann, R., Buatois, L., MacNaughton, R., and Mángano, M.: Loss of the
sedimentary mixed layer as a result of the end-Permian extinction,
Palaeogeogr. Palaeoclimatol. Palaeoecol., 428, 1–11,
10.1016/j.palaeo.2015.03.036,
2015.Horacek, M., Richoz, S., Brandner, R., Krystyn, L., and Spötl, C.:
Evidence for recurrent changes in Lower Triassic oceanic circulation of the
Tethys: The δ13C record from marine sections in Iran,
Palaeogeogr. Palaeoclimatol. Palaeoecol., 252, 355–369,
10.1016/j.palaeo.2006.11.052,
2007.Hsiang, S. M.: Visually-Weighted Regression,
available at: http://www.fight-entropy.com/2012/08/watercolor-regression.html (last access: 2016),
2012.Huang, J., Luo, G., Bai, X., and Tang, X.: Organic fraction of the total
carbon burial flux deduced from carbon isotopes across the Permo-Triassic
boundary at Meishan, Zhejiang Province, China, Frontiers of Earth Science in
China, 1, 425–430, 10.1007/s11707-007-0052-z,
2007.Irwin, H., Curtis, C., and Coleman, M.: Isotopic evidence for source of
diagenetic carbonates formed during burial of organic-rich sediments,
Nature, 269, 209–213, 10.1038/269209a0,
1977.
James, N. and Choquette, P.: Limestones: introduction, in: Diagenesis, 5,
159–161, Geoscience Canada, 1983.Jin, Y., Wang, Y., Wang, W., Shang, Q., Cao, C., and Erwin, D.: Pattern of
marine mass extinction near the Permian-Triassic boundary in South China,
Science, 289, 432–436, 10.1126/science.289.5478.432,
2000.Kaiho, K., Kajiwara, Y., Chen, Z.-Q., and Gorjan, P.: A sulfur isotope event
at the end of the Permian, Chem. Geol., 235, 33–47,
10.1016/j.chemgeo.2006.06.001,
2006.Kakuwa, Y. and Matsumoto, R.: Cerium negative anomaly just before the Permian
and Triassic boundary event – The upward expansion of anoxia in the water
column, Palaeogeogr. Palaeoclimatol. Palaeoecol., 229, 335–344,
10.1016/j.palaeo.2005.07.005,
2006.Kershaw, S., Li, Y., Crasquin-soleau, S., Feng, Q., Mu, X., Collin, P.-Y.,
Reynolds, A., and Guo, L.: Earliest Triassic microbialites in the South
China block and other areas: controls on their growth and distribution,
Facies, 53, 409–425, 10.1007/s10347-007-0105-5, 2007.Kershaw, S., Crasquin, S., Collin, P.-Y., Li, Y., Feng, Q., and Forel, M.-B.:
Microbialites as disaster forms in anachronistic facies following the
end-Permian mass extinction: a discussion, Aust. J. Earth
Sci., 56, 809–813, 10.1080/08120090903002623,
2009.Knauth, L. P. and Kennedy, M. J.: The late Precambrian greening of the
Earth, Nature, 460, 728–32, 10.1038/nature08213,
2009.Knoll, A. H., Bambach, R. K., Canfield, D. E., and Grotzinger, J. P.:
Comparative Earth history and Late Permian mass extinction, Science, 273,
452–457, 10.1126/science.273.5274.452, 1996.Knoll, A. H., Bambach, R. K., Payne, J. L., Pruss, S., and Fischer, W. W.:
Paleophysiology and end-Permian mass extinction, Earth Planet.
Sci. Lett., 256, 295–313, 10.1016/j.epsl.2007.02.018,
2007.
Korte, C. and Kozur, H. W.: Carbon isotope stratigraphy across the
Permian/Triassic at Jolfa (NW-Iran, Peitlerkofel (Sas de Pütia, Sass de
Putia, Pufels (Bula, Bulla), Tesero (all three Southern Alps Italy) and
Gerennavár (Bükk Mts., Hungary), J. Alpine Geol., 47,
119–135, 2005.Korte, C., Kozur, H. W., Joachimski, M. M., Strauss, H., Veizer, J., and
Schwark, L.: Carbon, sulfur, oxygen and strontium isotope records, organic
geochemistry and biostratigraphy across the Permian/Triassic boundary in
Abadeh, Iran, Int. J. Earth Sci., 93, 565–581,
10.1007/s00531-004-0406-7,
2004a.Korte, C., Kozur, H., and Mohtat-Aghai, P.: Dzhulfian to lowermost Triassic
δ13C record at the Permian/Triassic boundary section at Shahreza ,
Central Iran, Hallesches Jahrbuch Geowissenschaften, Reihe B, 18, 73–78,
2004b.
Korte, C., Kozur, H. W., and Partoazar, H.: Negative carbon isotope excursion
at the Permian/Triassic boundary section at Zal, NW-Iran, Hallesches
Jahrbuch Geowissenschaften, Reihe B, Beiheft 18, 69–71,
2004c.Korte, C., Jasper, T., Kozur, H. W., and Veizer, J.: δ18O and
δ13C of Permian brachiopods: A record of seawater evolution and
continental glaciation, Palaeogeogr. Palaeoclimatol. Palaeoecol.,
224, 333–351, 10.1016/j.palaeo.2005.03.015,
2005.Korte, C., Pande, P., Kalia, P., Kozur, H. W., Joachimski, M. M., and
Oberhänsli, H.: Massive volcanism at the Permian-Triassic boundary and
its impact on the isotopic composition of the ocean and atmosphere, J. Asian Earth Sci., 37, 293–311, 10.1016/j.jseaes.2009.08.012,
2010.Kristensen, E., Penha-Lopes, G., Delefosse, M., Valdemarsen, T., Quintana,
C. O., and Banta, G. T.: What is bioturbation? the need for a precise
definition for fauna in aquatic sciences, Mar. Ecol. Progress Series,
446, 285–302, 10.3354/meps09506, 2012.Krull, E. S. and Retallack, G. J.: δ13C depth profiles from
paleosols across the Permian-Triassic boundary: Evidence for methane
release, Bull. Geol. Soc. Am., 112, 1459–1472,
10.1130/0016-7606(2000)112<1459:CDPFPA>2.0.CO;2,
2000.Kump, L. R. and Arthur, M. A.: Interpreting carbon-isotope excursions:
carbonates and organic matter, Chem. Geol., 161, 181–198,
10.1016/S0009-2541(99)00086-8,
1999.Leda, L., Korn, D., Ghaderi, A., Hairapetian, V., Struck, U., and Reimold,
W. U.: Lithostratigraphy and carbonate microfacies across the
Permian–Triassic boundary near Julfa (NW Iran) and in the Baghuk Mountains
(Central Iran), Facies, 60, 295–325, 10.1007/s10347-013-0366-0, 2014.Li, M., Ogg, J., Zhang, Y., Huang, C., Hinnov, L., Chen, Z.-Q., and Zou, Z.:
Astronomical tuning of the end-Permian extinction and the Early Triassic
Epoch of South China and Germany, Earth Planet. Sci. Lett., 441,
10–25, 10.1016/j.epsl.2016.02.017,
2016.Li, R. and Jones, B.: Diagenetic overprint on negative δ13C excursions
across the Permian/Triassic boundary: A case study from Meishan section,
China, Palaeogeogr. Palaeoclimatol. Paleoecol., 468, 18–33,
10.1016/j.palaeo.2016.11.044,
2017.Liu, X.-C., Wang, W., Shen, S.-Z., Gorgij, M. N., Ye, F.-C., Zhang, Y.-C.,
Furuyama, S., Kano, A., and Chen, X.-Z.: Late Guadalupian to Lopingian
(Permian) carbon and strontium isotopic chemostratigraphy in the Abadeh
section, central Iran, Gondwana Research, 24, 222–232,
10.1016/j.gr.2012.10.012,
2013.Luo, G., Kump, L. R., Wang, Y., Tong, J., Arthur, M. A., Yang, H., Huang, J.,
Yin, H., and Xie, S.: Isotopic evidence for an anomalously low oceanic
sulfate concentration following end-Permian mass extinction, Earth
Planet. Sci. Lett., 300, 101–111, 10.1016/j.epsl.2010.09.041,
2010.Marshall, J. D.: Climatic and oceanographic isotopic signals from the
carbonate rock record and their preservation, Geology Magazine, 129,
143–160, 10.1017/S0016756800008244, 1992.Mayhew, P. J., Bell, M. A., Benton, T. G., and McGowan, A. J.: Biodiversity
tracks temperature over time, Proc. Natl. Acad.
Sci. USA, 109, 15141–5,
10.1073/pnas.1200844109, 2012.Melim, L., Swart, P., Munnecke, a., Eberli, G., and Westphal, H.: Questioning
carbonate diagenetic paradigms: evidence from the Neogene of the Bahamas,
Mar. Geol., 185, 27–53, 10.1016/S0025-3227(01)00289-4, 2002.Meyer, K., Kump, L., and Ridgwell, A.: Biogeochemical controls on photic-zone
euxinia during the end-Permian mass extinction, Geology, 36, 747,
10.1130/G24618A.1, 2008.Meyer, K., Yu, M., Jost, A. B., Kelley, B., and Payne, J.: δ13C
evidence that high primary productivity delayed recovery from end-Permian
mass extinction, Earth Planet. Sci. Lett., 302, 378–384,
10.1016/j.epsl.2010.12.033,
2011.Meysman, F. J., Middelburg, J. J., Herman, P. M., and Heip, C. H.: Reactive
transport in surface sediments. II. Media: an object-oriented problem-solving
environment for early diagenesis, Comput. Geosci., 29, 301–318,
10.1016/S0098-3004(03)00007-4,
2003.Meysman, F. J. R., Boudreau, B. P., and Middelburg, J. J.: When and why does
bioturbation lead to diffusive mixing?, J. Marine Res., 68,
881–920, 10.1357/002224010796673885, 2010.Meysman, F. J. R., Risgaard-Petersen, N., Malkin, S. Y., and Nielsen, L. P.:
The geochemical fingerprint of microbial long-distance electron transport in
the seafloor, Geochim. Cosmochim. Ac., 152, 122–142,
10.1016/j.gca.2014.12.014, 2015.Müller, P. J. and Suess, E.: Productivity, sedimentation rate, and
sedimentary organic matter in the oceans-I. Organic carbon preservation,
Deep Sea Res. Part A, 26, 1347–1362,
10.1016/0198-0149(79)90003-7,
1979.Munnecke, A. and Samtleben, C.: The formation of micritic limestones and the
development of limestone-marl alternations in the Silurian of Gotland,
Sweden, Facies, 34, 159–176, 10.1007/BF02546162, 1996.Payne, J. L., Turchyn, A. V., Paytan, A., Depaolo, D. J., Lehrmann, D. J., Yu,
M., and Wei, J.: Calcium isotope constraints on the end-Permian mass
extinction, Proc. Natl. Acad. Sci. USA, 107, 8543–8548, 10.1073/pnas.0914065107,
2010.Prokoph, A., Shields, G. A., and Veizer, J.: Compilation and time-series
analysis of a marine carbonate δ18O , δ13C ,
87Sr/86Sr and δ34S database through Earth history,
Earth Sci. Rev., 87, 113–133, 10.1016/j.earscirev.2007.12.003,
2008.Pruss, S. B., Bottjer, D. J., Corsetti, F. A., and Baud, A.: A global marine
sedimentary response to the end-Permian mass extinction: Examples from
southern Turkey and the western United States, Earth Sci. Rev., 78,
193–206, 10.1016/j.earscirev.2006.05.002,
2006.Rampino, M. R. and Caldeira, K.: Major perturbation of ocean chemistry and a
“Strangelove Ocean” after the end-Permian mass extinction, Terra Nova, 17,
554–559, 10.1111/j.1365-3121.2005.00648.x,
2005.R Core Team: R: A Language and Environment for Statistical Computing,
available at: https://www.r-project.org (last access: November 2017), 2016.Reimers, C. E. and Suess, E.: The partitioning of organic carbon fluxes and
sedimentary organic matter decomposition in the ocean, Mar. Chem., 13,
141–168, 10.1016/0304-4203(83)90022-1,
1983.Reitner, J., Peckmann, J., Reimer, A., Schumann, G., and Thiel, V.:
Methane-derived carbonate build-ups and associated microbial communities at
cold seeps on the lower Crimean shelf (Black Sea), Facies, 51, 66–79,
10.1007/s10347-005-0059-4,
2005.Renne, P. R., Black, M. T., Zichao, Z., Richards, M. A., and Basu, A. R.:
Synchrony and causal relations between Permian–Triassic boundary crises and
siberian flood volcanism, Science, 269, 1413–1416,
10.1126/science.269.5229.1413, 1995.Retallack, G. J. and Jahren, A. H.: Methane Release from Igneous Intrusion of
Coal during Late Permian Extinction Events, The Journal of Geology, 116,
1–20, 10.1086/524120, 2008.Riccardi, A., Kump, L. R., Arthur, M. A., and D'Hondt, S.: Carbon isotopic
evidence for chemocline upward excursions during the end-Permian event,
Palaeogeogr. Palaeoclimatol. Palaeoecol., 248, 73–81,
10.1016/j.palaeo.2006.11.010,
2007.Richoz, S., Krystyn, L., Baud, A., Brandner, R., Horacek, M., and Mohtat-Aghai,
P.: Permian–Triassic boundary interval in the Middle East (Iran and N.
Oman): Progressive environmental change from detailed carbonate carbon
isotope marine curve and sedimentary evolution, J. Asian Earth
Sci., 39, 236–253, 10.1016/j.jseaes.2009.12.014,
2010.Ridgwell, A.: A Mid Mesozoic Revolution in the regulation of ocean chemistry,
Mar. Geol., 217, 339–357, 10.1016/j.margeo.2004.10.036, 2005.Riding, R. and Liang, L.: Geobiology of microbial carbonates: metazoan and
seawater saturation state influences on secular trends during the
Phanerozoic, Palaeogeogr. Palaeoclimatol. Palaeoecol., 219,
101–115, 10.1016/j.palaeo.2004.11.018,
2005.Ries, J. B., Anderson, M. A., and Hill, R. T.: Seawater Mg/Ca controls
polymorph mineralogy of microbial CaCO3: A potential proxy for
calcite-aragonite seas in Precambrian time, Geobiology, 6, 106–119,
10.1111/j.1472-4669.2007.00134.x,
2008.Rothman, D. H., Fournier, G. P., French, K. L., Alm, E. J., Boyle, E. A., Cao,
C., and Summons, R. E.: Methanogenic burst in the end-Permian carbon
cycle, Proc. Natl. Acad. Sci. USA, 111, 5462–7, 10.1073/pnas.1318106111, 2014.Saltzman, M. R.: Late Paleozoic ice age; Oceanic gateway or pCO2?,
Geology, 31, 151–154, 10.1130/0091-7613(2003)031<0151:LPIAOG>2.0.CO;2, 2001.Sarmiento, J. L. and Gruber, N.: Calcium carbonate cycling, in: Ocean
Biogeochemical Dynamics, chap. 9, 1–44, Princeton University Press, New
York, available at: http://press.princeton.edu/titles/8223.html (last access: 2016), 2004.Schobben, M.: carbonate project, 10.5281/zenodo.888754, available at: https://github.com/MartinSchobben/carbonate, https://github.com/MartinSchobben/carbonate,
last access: 2017.Schobben, M., Joachimski, M. M., Korn, D., Leda, L., and Korte, C.:
Palaeotethys seawater temperature rise and an intensified hydrological cycle
following the end-Permian mass extinction, Gondwana Research, 26, 675–683,
10.1016/j.gr.2013.07.019,
2014.Schobben, M., Stebbins, A., Ghaderi, A., Strauss, H., Korn, D., and Korte, C.:
Flourishing ocean drives the end-Permian marine mass extinction,
Proc. Natl. Acad. Sci., 112, 10298–10303,
10.1073/pnas.1503755112, 2015.Schobben, M., Ullmann, C. V., Leda, L., Korn, D., Struck, U., Reimold, W. U.,
Ghaderi, A., Algeo, T. J., and Korte, C.: Discerning primary versus
diagenetic signals in carbonate carbon and oxygen isotope records: An example
from the Permian–Triassic boundary of Iran, Chem. Geol., 422,
94–107, 10.1016/j.chemgeo.2015.12.013,
2016.Schobben, M., Stebbins, A., Algeo, T. J., Strauss, H., Leda, L., Haas, J.,
Struck, U., Korn, D., and Korte, C.: Volatile earliest Triassic sulfur
cycle: A consequence of persistent low seawater sulfate concentrations and a
high sulfur cycle turnover rate?, Palaeogeogr. Palaeoclimatol. Palaeoecol., 10.1016/j.palaeo.2017.02.025, 2017.Schönbrodt, F.: Visually weighted/Watercolor Plots, new variants:
Please vote!, available at: http://www.nicebread.de/ (last access: 2016), 2012.Schrag, D. P., Higgins, J. A., Macdonald, F. A., and Johnston, D. T.:
Authigenic carbonate and the history of the global carbon cycle, Science,
339, 540–543, 10.1126/science.1229578, 2013.Soetaert, K. and Meysman, F.: Reactive transport in aquatic ecosystems: Rapid
model prototyping in the open source software R, Environ. Model.
Softw., 32, 49–60, 10.1016/j.envsoft.2011.08.011, 2012.Soetaert, K., Herman, P. M. J., and Middelburg, J. J.: A model of early
diagenetic processes from the shelf to abyssal depths, Geochim.
Cosmochim. Ac., 60, 1019–1040, 10.1016/0016-7037(96)00013-0,
1996.Soetaert, K., Petzoldt, T., and Setzer, R.: Package deSolve: solving initial
value differential equations in R, J. Stat. Softw., 33,
1–25, 10.18637/jss.v033.i09, 2010.Soetaert, K., Petzoldt, T., and Meysman, F.: marelac: Tools for Aquatic
Sciences, available at: https://CRAN.R-project.org/package=marelac (last access: November 2017), r
package version 2.1.6, 2016.Song, H., Tong, J., Algeo, T. J., Horacek, M., Qiu, H., Song, H., Tian, L., and
Chen, Z. Q.: Large vertical δ13CDIC gradients in Early
Triassic seas of the South China craton: Implications for oceanographic
changes related to Siberian Traps volcanism, Global Planet. Change,
105, 7–20, 10.1016/j.gloplacha.2012.10.023,
2013.Song, H., Tong, J., Algeo, T. J., Song, H., Qiu, H., Zhu, Y., Tian, L., Bates,
S., Lyons, T. W., Luo, G., and Kump, L. R.: Early Triassic seawater sulfate
drawdown, Geochim. Cosmochim. Ac., 128, 95–113,
10.1016/j.gca.2013.12.009,
2014.Svensen, H., Planke, S., Polozov, A. G., Schmidbauer, N., Corfu, F.,
Podladchikov, Y. Y., and Jamtveit, B.: Siberian gas venting and the
end-Permian environmental crisis, Earth Planet. Sci. Lett., 277,
490–500, 10.1016/j.epsl.2008.11.015,
2009.Ullrich, T. S.: xyscan,
available at: http://rhig.physics.yale.edu/~ullrich/software/xyscan/ (last access: 2016),
2016.Van Cappellen, P. and Wang, Y.: Cycling of iron and manganese in surface
sediments: A general theory for the coupled transport and reaction of carbon,
oxygen, nitrogen, sulfur, iron, and manganese, Am. J. Sci., 296, 197–243, 10.2475/ajs.296.3.197,
1996.van de Velde, S. and Meysman, F. J.: The influence of bioturbation on iron and
sulphur cycling in marine sediments: a model analysis, Aquat. Geochem.,
22, 469–504, 10.1007/s10498-016-9301-7,
2016.Veizer, J., Ala, D., Azmy, K., Bruckschen, P., Buhl, D., Bruhn, F., Carden, G.
A. F., Diener, A., Ebneth, S., Godderis, Y., Jasper, T., Korte, C., Pawellek,
F., Podlaha, O. G., and Strauss, H.: 87Sr/86Sr, δ13C and
δ18O evolution of Phanerozoic seawater, Chem. Geol., 161,
59–88, 10.1016/S0009-2541(99)00081-9,
1999.Visscher, P. T., Reid, P. R., and Bebout, B. M.: Microscale observations of
sulfate reduction: Correlation of microbial activity with lithified micritic
laminae in modern marine stromatolites, Geology, 28, 919–922,
10.1130/0091-7613(2000)28<919:MOOSRC>2.0.CO,
2000.Wacey, D., Wright, D. T., and Boyce, A. J.: A stable isotope study of
microbial dolomite formation in the Coorong Region, South Australia,
Chem. Geol., 244, 155–174, 10.1016/j.chemgeo.2007.06.032,
2007.Wang, Y., Sadler, P. M., Shen, S. Z., Erwin, D. H., Zhang, Y. C., Wang, X. D.,
Wang, W., Crowley, J. L., and Henderson, C. M.: Quantifying the process and
abruptness of the end-Permian mass extinction, Paleobiology, 40, 113–129,
10.1666/13022,
2014.
Wickham, H.: Reshaping Data with the reshape Package, J. Stat.
Softw., 21, 1–20, 2007.Wickham, H.: ggplot2: Elegant Graphics for Data Analysis, Springer-Verlag New
York, available at: http://ggplot2.org (last access: 2017), 2009.
Wickham, H.: The Split-Apply-Combine Strategy for Data Analysis, J.
Stat. Softw., 40, 1–29, 2011.Wignall, P. B. and Twitchett, R. J.: Oceanic anoxia and the End Permian mass
extinction, Science, 272, 1155–1158, 10.1126/science.272.5265.1155, 1996.Winguth, A. M. E., Shields, C. A., and Winguth, C.: Transition into a Hothouse
World at the Permian-Triassic boundary-A model study, Palaeogeogr.
Palaeocl., 440, 316–327,
10.1016/j.palaeo.2015.09.008, 2015.Woods, A. D., Bottjer, D. J., Mutti, M., and Morrison, J.: Lower Triassic
large sea-floor carbonate cements: Their origin and a mechanism for the
prolonged biotic recovery from the end-Permian mass extinction, Geology, 27,
645, 10.1130/0091-7613(1999)027<0645:LTLSFC>2.3.CO;2, 1999.
Xie, S., Pancost, R. D., Huang, J., Wignall, P. B., Yu, J., Tang, X., Chen, L.,
Huang, X., and Lai, X.: Changes in the global carbon cycle occurred as two
episodes during the Permian-Triassic crisis, Geology, 35, 1083–1086,
10.1130/G24224A.1, 2007.Xu, D.-Y. and Yan, Z.: Carbon isotope and iridium event markers near the
Permian-Triassic boundary in the Meishan section, Zhejiang Province, China,
Palaeogeogr. Palaeoclimatol. Palaeoecol., 104, 171–176,
10.1016/0031-0182(93)90128-6,
1993.Yin, H., Jiang, H., Xia, W., Feng, Q., Zhang, N., and Shen, J.: The
end-Permian regression in South China and its implication on mass
extinction, Earth Sci. Rev., 137, 19–33,
10.1016/j.earscirev.2013.06.003, 2014.Zachos, J., Pagani, M., Sloan, L., Thomas, E., and Billups, K.: Trends,
rhythms, and aberrations in global climate 65Ma to present, Science, 292,
686–693, 10.1126/science.1059412, 2001.Zeebe, R. E. and Westbroek, P.: A simple model for the CaCO3 saturation
state of the ocean; the “Strangelove,” the “Neritan,” and the
“Cretan”
ocean, Geochem. Geophys. Geosyst., 4, 1–26,
10.1029/2003GC000538, 2003.Zhao, M.-Y., Zheng, Y.-F., and Zhao, Y.-Y.: Seeking a geochemical identifier
for authigenic carbonate, Nature Comm., 7, 10885,
10.1038/ncomms10885, 2016.Zuo, J.-X., Tong, J.-N., Qiu, H.-o., and Zhao, L.-S.: Carbon isotope
composition of the Lower Triassic marine carbonates, Lower Yangtze Region,
South China, Science in China Series D: Earth Sciences, 49, 225–241,
10.1007/s11430-006-0225-8,
2006.