Professional Documents
Culture Documents
Temperatures and oxygen isotopic composition of Phanerozoic oceans
PII: S0012-8252(15)00060-4
DOI: doi: 10.1016/j.earscirev.2015.03.008
Reference: EARTH 2101
Please cite this article as: Veizer, Jan, Prokoph, Andreas, Temperatures and oxy-
gen isotopic composition of Phanerozoic oceans, Earth Science Reviews (2015), doi:
10.1016/j.earscirev.2015.03.008
This is a PDF le of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its nal form. Please note that during the production process
errors may be discovered which could aect the content, and all legal disclaimers that
apply to the journal pertain.
ACCEPTED MANUSCRIPT
Temperatures and oxygen isotopic composition of Phanerozoic oceans
PT
1*
Ottawa-Carleton Geoscience Center, University of Ottawa, Ottawa, Canada K1J 6E5
RI
2
Speedstat, 19 Langstrom Street, Ottawa, Canada K1G 5J5
SC
*corresponding author: jveizer@uottawa.ca, Department of Earth Sciences, University of Ottawa, 120
NU
University Avenue, Ottawa, Canada K1J 6E5, phone: 1 613 562 5800 x 6461
MA
Abstract
D
The temperature of ancient oceans is an important constraint for understanding the climate history of
TE
our planet. The classical oxygen isotope paleothermometry on fossil shells, while very proficient when
P
applied to the younger (Cenozoic) portion of the geologic record, is believed to yield only unreliable
CE
results for the Phanerozoic "deep time", either because the empirically well documented secular trend
to more negative 18O values with increasing age was generated by post-depositional recrystallization
AC
processes or, if primary, implies ecologically unpalatable hot early oceans. Here we present a
compilation of 18O measurements for 58532 low-Mg calcite marine shells that cover almost the entire
Phanerozoic eon, argue that the secular decline of about -6 is primary, propose that it reflects the
changing oxygen isotopic composition of sea water, and define a new baseline trend for 18O of paleo-
sea water; the latter providing a new template for calculation of ambient habitat temperatures of fossil
specimens. The resulting pattern for fossil taxa (foraminifera, brachiopods, belemnites and bivalves)
mimics their modern counterparts in temperature ranges and modes. This conceptual framework
enables application of actualistic concepts to ambient habitat temperatures of fossils and provides us
with a long overdue tool for interpretation of "deep time" geologic history.
ACCEPTED MANUSCRIPT
Keywords: paleotemperatures, oxygen isotopes, sea water, Phanerozoic, calcite fossils
1.0 Introduction
PT
The pioneering proposition that oxygen isotopes in marine shells may reflect temperatures of ancient
oceans (Urey et al., 1951) found rapid application in Quaternary studies (Emiliani, 1954). The
RI
application to Phanerozoic time scale, however, was almost immediately hampered by discovery of a 4
SC
- 8 18O decline in progressively older samples (Baertschi, 1957; Clayton and Degens, 1959), a
NU
trend similar to diagenetic overprint imposed during transformation of carbonate sediments into rocks
(Degens and Epstein, 1962; Gross, 1964). The observed secular trend was thereafter interpreted as
MA
mostly a diagenetic artifact rendering the oxygen isotope thermometry of limited utility for "deep
time" paleoceanographic research. This view, which reined for almost half a century, and is still
D
endorsed by some groups, is no longer tenable. Veizer et al. (1999) published oxygen isotope data for
TE
thousands of optically and texturally well preserved Phanerozoic low-Mg calcitic (LMC) fossils that not
P
only replicated the secular trend, but also yielded superimposed second order oscillations (Veizer et
CE
al., 2000) consistent with the pattern of climate deduced from geological indices. Moreover, these
same samples were utilized also for delineation of secular trends for Phanerozoic sea water for
AC
isotopes of carbon, radiogenic strontium (Veizer et al., 1999), sulfur (Kampschulte and Strauss, 1998),
calcium (Farkas et al., 2007), stable strontium (Vollstaedt et al., 2014) and Sr/Ca elemental ratios
(Steuber and Veizer, 2002). All of these trends replicated, or were replicated by, studies of
independent research groups. It is therefore inconceivable to argue that the atoms of oxygen, the
dominant structural unit of calcite crystals, must have been massively exchanged during diagenesis
while at the same time none of the other major and trace elements or isotopes were affected.
Accepting the reality of the oxygen isotope secular trend the principal question that arises is the
geological meaning of this overall gradient and its implication for environmental interpretation of the
The database assembled for this contribution (Table 1) contains compilation of 18O and 13C
measurements for 58532 low-Mg calcite marine shells that cover the last 512 Ma of the entire, 542 Ma
long, Phanerozoic eon. This compilation is a tenfold expansion of the dataset in Veizer et al. (1999)
PT
and two to threefold of earlier published archives (Prokoph et al., 2008; Shaviv et al., 2014). It
RI
summarizes experimental data published up to December 2013 and consists of 35171 benthic
SC
foraminifera, 13018 planktonic foraminifera, 5202 brachiopods, 2969 belemnites, 786 bivalves and
1386 "other" fossil samples. In addition, the new compilation provides better regional descriptions for
NU
both surface water (mixed layer) and deep-sea (>300m) benthic fossils, the former further
MA
differentiated into high- (~ 58-90 north/south), mid- (~ 35-58N/S) and low-paleolatitude (~ 26S
to 26N) subsets. The data are further grouped by taxa. Moreover, planktonic and benthic foraminifera
are grouped into ocean basins subsets (Table 1). The "other" fossil category includes intermediate
D
TE
water (below thermocline but above 300 m water depth) foraminifera, LMC fossils from the
Mediterranean considered to be an inland sea, and other calcitic and aragonitic fossils grouped by
P
paleolatitude. These "other" fossils are not discussed further in the text. The database (supplementary
CE
materials) is fully annotated, includes 1 standard deviation (1s) age uncertainty estimates, and is
AC
consistently referenced to the geologic time scale GTS2012 (Gradstein et al., 2012).
The deep-sea database (Appendix A-deepsea.xls) consists of data for benthic foraminifera compiled
mostly by Cramer et al. (2009). This compilation includes and replaces the Cenozoic summary of
Zachos et al. (2001) that was previously used in Prokoph et al. (2008). The regional structure of 5
subsets in Cramer et al. was retained for the present database. The Cretaceous benthic foraminifera
database of Friedrich et al. (2012), which contains most of the data in Prokoph et al. (2008) as well as
new data, was also included. Several hundred additional data that were not listed in Cramer et al.
(2009) and Friedrich et al. (2012) were added to the appropriate regional subsets. Note that over a
hundred measurements that were accidentally duplicated in Cramer et al. (2009) were removed from
ACCEPTED MANUSCRIPT
the new database. Most of these duplicates were from the North Atlantic (compare theOppoMcManus
in ODP980; Cramer et al., 2009). In summary, the number of deep-sea benthic foraminifera more than
PT
RI
Table 1: Number of Phanerozoic low-Mg carbonate fossils in the oxygen isotope database
SC
Material/Fossil Abbreviation Low-latitude# Mid-latitude High-latitude Deep Sea All
2008* 2013 2008* 2013 2008* 2013 2008* 2013 Veizer et al. (1999) 2008* 2013
NU
Belemnites Bel 164 675 975 2171 73 124 628 1212 2969
Trilobites N/A 4 4
Corals N/A 2
Others 1386
Total: Fossils 4823 16502 1569 4451 247 10222 11175 35171 5239 17814 58532
The surface-water (mixed surface layer) database (Appendix B-surface.xls) includes separate subsets
for high- (~ 58-90 north/south), mid- (~ 35-58N/S) and low-paleolatitudes (~ 26S to 26N).
Where possible, subtropical (~ 26 to 35 N/S) measurements were differentiated from the low-
latitude data, but - because of their small numbers - they were retained in the low-latitude sets for
calculation of trends and visualization. The dataset includes LMC planktonic foraminifera, belemnites
ACCEPTED MANUSCRIPT
(Mn <300 ppm), brachiopods (Mn< 300 ppm) and bivalves (including oysters). In contrast to the
database of Prokoph et al. (2008) planktonic foraminifera were grouped into nine geographic regions
(Table 1), enabling, in addition to latitudinal patterns, also consideration of changes in ocean
circulation in response to the opening of the Atlantic, and the concomitant shrinking of the Pacific,
PT
oceans. The total of all belemnites, brachiopods, planktonic foraminifera and LMC bivalve data is
RI
more than triple of the 2008 compilation (Table 1). As noted above, the "other" fossil group listed in
SC
this appendix is not specifically discussed in the text. The database is fully annotated, includes 1s age
error estimates, and is consistently referenced to the geologic time-scale GTS2012 (Gradstein et al.,
NU
2012).
The compilation of 18O for modern taxa from variable latitudes (Appendix C-modern.xls), assembled
MA
here for comparison with the fossil data, includes measurements for deep-sea benthic foraminifera,
planktonic foraminifera from <300m water depth, and brachiopods, plus the references to the related
D
TE
data sources.
All references for the sources of fossil data that were not listed already in Prokoph et al. (2008) are
P
CE
Inspection of Figure 1 demonstrates the scarcity of shells secreted at cold temperatures, except for
the youngest portion of the record that is populated by the deep sea benthic foraminifera that evolved
only during the mid-Mesozoic radiation of calcareous species into pelagic realm, plus some younger
shallow water samples from high paleolatitudes. For the 4070 Paleozoic samples, however, practically
all measurements plot below the modern value for marine calcite of about 0 , with two-thirds
plotting even below -3.5 that could indicate temperatures in excess of 30C. Note that even this
massive compilation has almost a complete absence of measurements above the upper envelope of
ACCEPTED MANUSCRIPT
the secular trend. Yet at least some such relics should have been preserved if advancing diagenetic
18
recrystallization were the cause of the secular trend towards O depletion. The lower envelope of the
band is more diffuse, but even here all clear outliers, despite their apparent optical visibility (the main
body of data is too densely populated to be resolved in print) account for considerably less than 0.1 %
PT
of the database. These outliers were interpreted by Giles (2012) as high temperature episodes, but in
RI
our view they more likely represent post-depositional overprint of the isotope signal, except for the
SC
mid-Cretaceous episode at about 95 Ma ago that may have been real (Forster et al., 2007; Friedrich et
al., 2012). Irrespective of whether the secular band is viewed via its median or its upper envelope, the
NU
pattern suggests an overall 18O shift of -6 (+/-1) , with four superimposed major supercycles at
about 145 Ma frequency (Veizer et al., 2000). This is not to say that all samples in the dataset are
MA
18
pristine, but a diagenetic O depletion - as far as it is involved - shifted the oxygen isotopic
composition to more negative values from the Phanerozoic trend, not downwards towards it, resulting
D
in the diffuse lower envelope. While it is still entirely feasible to argue the integrity of some samples
TE
within the 4-5 width of the Phanerozoic band, the primary nature of the overall secular shift
P
should be beyond any reasonable dispute. This 18O decline continues, albeit at a shallower slope,
CE
further into the Precambrian (Veizer and Hoefs, 1976; Shields and Veizer, 2002; Jaffres et al., 2007;
AC
PT
RI
SC
NU
MA
Figure 1: 18O of Phanerozoic low-Mg calcitic fossils (n= 57146). The temperature estimates are based
D
TE
on the Visser et al. (2003) transfer function, assuming the present day 18O value of 0 SMOW
for seawater. Time scale after Gradstein et al. (2012). The data and their sources are listed in the
P
Supplementary Materials (Appendices A and B; references in Prokoph et al. (2008) and Appendix
CE
D).
AC
Biological evolution, constraints of field sampling, and the need for shells of low-Mg calcitic mineralogy
dictate that only few groups of organisms are available for construction of this record: brachiopods,
foraminifera, belemnites and bivalves, all having their own internal variability in addition to potential
systematic differences between groups. It is therefore essential to take these issues into account
because no single group of organisms provides a satisfactory record for the entire Phanerozoic. The
Paleozoic portion of the record is covered almost entirely by brachiopods while the other taxa
dominate the Mesozoic and Cenozoic time intervals. The multitude and interplay of parameters
involved in incorporation of oxygen isotopes into calcareous shells - imposed by analytical, biological
ACCEPTED MANUSCRIPT
and physical constraints, including issues such as temperature, pH and salinity of sea water - are well
summarized in Grossman (2012a,b). Such enumeration, however, is pertinent mostly to the below
discussed scatter within the band, not to the nature of the overall gradient, because the magnitude of
the latter is beyond capabilities of these second order phenomena. For example, the "ice correction"
PT
for a climate change as massive as a shift from an ice cap of the Last Glacial Maximum to no ice cover
RI
at all would yield only ~ 1.2 signal (Shackleton and Opdyke, 1973; Lhome et al., 2005). Note also
SC
that the width of the Phanerozoic band is not that dissimilar from modern situation where the global
aliasing of samples alone results in 18O spread of about 4 (Figure 2). If compounded further by
NU
aliasing over (multi)million years, which is the temporal resolution of our database, the spread should
be still greater, and fan out with any advancing diagenetic overprint. None of these features are
MA
evident in the Phanerozoic band (Veizer et al., 1999; Figure 1) that remains only marginally broader
than in the modern analogues (Figure 2). As a final qualification, we also point out that the temporal
D
resolution of the Phanerozoic time scale is not conducive to quantification of such second order
TE
causality. In general, the existing resolution implies causal relationship to processes operating at
P
similar or longer time scales. Our subsequent discussion will therefore concentrate on such temporal
CE
relationships, within and between the bands of fossil taxa, with a view for potential splicing of partial
AC
records.
Visual inspection of the four studied fossil taxa (Figure 2) that coexisted throughout the Cenozoic and
the Cretaceous confirms that they all carry oxygen isotope signals within the shape and limits of the
overall secular trend, but not necessarily across its full range. In particular, the belemnites appear to
18
crowd in the O enriched sections of the histograms. In contrast, foraminifera, brachiopods and
PT
RI
SC
NU
MA
D
P TE
Figure 2: Histograms of 18O values in modern to Cretaceous low-Mg calcitic shells. The underlying 4
CE
band is based on modern range of values for surface water taxa. The data and their sources
for modern samples are listed in Supplementary Materials (Appendix C) and for fossils in
AC
In order to enable discussion of variations in oxygen isotopic composition for the above taxa within
the context of secular evolution, it is essential to generate a temporally homogeneous time series.
Following the approach outlined in Prokoph et al. (2008) we constructed, as the first step, a
temporally homogeneous time series at equidistant 1 Ma intervals from the raw data via Gaussian
filtering.
ACCEPTED MANUSCRIPT
The Gaussian filtering assumes that each data point x ( single 18O measurement) has a probability
distribution function p(t) around its stratigraphic age ti given with a Gaussian width . The estimated
measurement at time t is thus given by the sum of contributions from all measurements with their
respective weights:
PT
RI
(1)
SC
NU
Note, nevertheless, that it is only a few adjoining measurements that contribute significantly to the
MA
error in the stratigraphic age at a given time t. In general, the stratigraphic errors for sample older
than ~50 Ma became smaller compared to Prokoph et al. (2008) because the Ar-Ar-decay constants
and tie ages have been better calibrated and because additional geochronological constraints were
D
TE
included in Gradstein et al. (2012). The earlier 1s uncertainties, which were particularly large for old
ages, are now reduced and referenced to GTS2012. This Gaussian filtering was applied to all below
P
Considering that brachiopods are the only taxa spanning the entire Phanerozoic, this group was
AC
chosen to serve as a baseline to which all other taxa were normalized. We compared therefore the
oxygen isotope results for all one million year segments that contained samples of several taxa (Table
18
2). In this comparison, belemnites are consistently enriched in O relative to brachiopods, on average
suggest that in the tropics, a latitudinal belt of principal interest to our study, the habitats for
PT
High-Latitude Bivalves Belemnites Planktonic foraminifera Benthic foraminifera
RI
n 3 4 12 11
Standard error es 0.49 0.45 0.17 0.06
SC
Mid- Latitude
Difference to brachiopods -0.74 0.46 -0.63 1.48
NU
n 16 27 33 36
Standard error es 0.30 0.15 0.17 0.16
Low-Latitude
MA
Difference to brachiopods 0.36 0.96 0.33 2.15
n 12 15 23 18
Standard error es 0.15 0.15 0.33 0.25
D
The enrichment of benthic foraminifera relative to brachiopods appears to increase consistently from
CE
high- to low-latitudes (Table 2). This is almost certainly a reflection of colder and progressively deeper
AC
habitats for the benthic foraminifera, resulting in the increasing vertical temperature gradient equator-
In contrast to benthic foraminifera, the offset between brachiopods and planktonic foraminifera is
relatively minor and inconsistent. For the tropics, the difference is only 0.33 , as large as the
standard deviation, and the discrepancy may not be real. If so, their spliced low-latitude records may
The bivalve/brachiopod relationship appears to be similar to the one for the planktonic
foraminifera/brachiopods, but the bivalve database is as yet too limited for a clear judgement about
their potential utility for filling the gaps in the secular baseline.
ACCEPTED MANUSCRIPT
In summary, considering that the present day spread of oxygen isotope values in all three longitudinal
belts is at least 4 (Figure 2), and sustained throughout the Cenozoic and Mesozoic despite of
aliasing much longer time intervals, the error arising from the potential brachiopod/planktonic
foraminifera discrepancy in a spliced Phanerozoic reference frame for the tropics would be only about
PT
1/10 of the natural variability observable today within each oceanic latitudinal belt. The specific issues
RI
for each taxa are discussed below.
SC
5.0 Foraminifera
NU
Paleothermometry based on shells of foraminifera is a mature and well established research field (e.g.
MA
Lea, 2003; Ravizza and Zachos, 2003; Grossman, 2012a,b) and our discussion is therefore confined
only to recapitulation of major features of their secular trends (Figure 3) as an essential precondition
D
PT
RI
SC
NU
MA
D
P TE
CE
AC
Figure 3: Secular trends (2s bands) and mean values in 18O based on the Gaussian filtered 1 Ma
averages for low-Mg calcitic foraminifera from deep sea benthic and planktonic (low-, mid- and
high-latitude) habitats. Blank spaces indicate 1 Ma intervals without samples. For raw data see
increasing band of 18O values, by about 5 (Figure 3), reflecting some 20C cooling of deep ocean
waters from the mid-Cretaceous to the present (Savin, 1977; Zachos et al., 2001; Cramer et al.,
2009). Superimposed are second higher oscillations in ~ 40 Ma frequency range, with isotopic
PT
minimums in about Cenomanian - Turonian (90-100 Ma), early Eocene (50- 55 Ma) and Miocene (15
RI
Ma). The Atlantic, Pacific and Southern Oceans all follow consistently this overall cooling pattern with
SC
the existing regional differences (Cramer et al., 2009) smaller than the width of the band.
Considering that the above deep water masses are being generated principally in the high-latitude
NU
segments of the thermohaline conveyor belt, it is not surprising that the data for Cenozoic planktonic
foraminifera from high-latitude regions essentially mimic the benthic band, including its superimposed
MA
oscillations, albeit with a secular gradient of ~ 7.5 , about 2 - 3 greater than the deep ocean
samples.
D
TE
The 110 Ma long mid-latitude Cretaceous and Tertiary record of planktonic foraminifera (Figure 3) also
shows the above discussed cooling trend, including suggestions of the superimposed oscillations, of
P
similar magnitude as in the deep ocean realm. The overall cooling trend for the low-latitude record on
CE
the other hand arises mostly from the contrasting Cretaceous and Cenozoic patterns, with the
AC
apparent Cretaceous warming trend terminated by step cooling during the Santonian/Campanian (~
83 Ma ago) followed by an oscillating plateau with the "warmest" episode during the early Eocene (55
Ma).
In summary, the combined secular trends for the Cenozoic foraminifera (Figure 4) are consistent with
the traditional thermohaline conveyor belt scenario that posits transport and cooling of water masses
towards high latitudes, followed by their sinking to deeper domains and retro transport. For the
Cretaceous, however, the scenario may have been different, because here all pelagic trends (high-,
18
mid- and low-latitudes) define the same trough while the deep waters remain enriched in O, up to 3
, implying a vertical temperature gradient of about 10C. Such pattern argues for a strongly
ACCEPTED MANUSCRIPT
diminished meridional temperature gradient for oceanic surface layers and for generation of deep
water masses by a different mode and/or in different locations than in the Cenozoic. A scenario that
invokes halothermal oceanic circulation in the Cretaceous, superseded by thermohaline conveyor belt
during the Cenozoic (Hay and DeConto, 1999; Hay, 2009) may potentially explain the above
PT
conundrum. Cretaceous paleogeography, with its near equatorial Tethys bounded by large
RI
epicontinental seas, coupled with the exceptionally warm climate, may have provided the loci for
SC
generation of the dense hypersaline water masses that may have sourced the deep ocean. Such
thermohaline/halothermal dichotomy may be the reason for the exceptionally large secular gradient in
NU
18O of the high-latitude foraminifera.
MA
Figure 4: grayscale in print and online
D
P TE
CE
AC
Figure 4: Secular trends (running means at 1 Ma window size) in 18O for low-Mg calcitic foraminifera
based on the Gaussian filtered 1 Ma averages. Blank spaces indicate 1 Ma intervals without
samples. For raw data see Supplementary Appendices A and B. Time scale after Gradstein et al.
(2012).
ACCEPTED MANUSCRIPT
6.0 Brachiopoda
In contrast to foraminifera, utilization of brachiopod shells for paleoceanographic studies is still only a
developing field (Popp et al., 1986; Veizer et al., 1986, 1999; Grossman, 2012a, b). The shells of
articulate brachiopods, and particularly their secondary layers (Williams and Cusack, 2007), are mostly
PT
composed of texturally well preserved low-Mg calcite believed to have been secreted in oxygen
RI
isotopic equilibrium with ambient seawater (Lowenstam, 1961; Carpenter and Lohmann, 1995; Brand
SC
et al., 2003; Parkinson and Cusack, 2007; Takayanagi et al., 2013). The review of brachiopod data
(Figure 5) shows that samples from low-latitudes merge into the trend of low-latitude planktonic
NU
foraminifera, as do the mid-latitude samples, albeit the latter shifted slightly towards "colder" range.
MA
The high-latitude samples, on the other hand, generally plot at, or above, the upper envelope of the
foraminifera band; a pattern to be anticipated due to their colder habitats. While a firmer empirical
basis is desirable, the existing database suggests that splicing the records for low-latitude planktonic
D
TE
foraminifera and brachiopods may yield the sought for empirical baseline for the Phanerozoic oceans
(Figure 6). The considerably broader 2s band of the brachiopod baseline, compared to that of the
P
foraminifera, arises from the much smaller sample numbers coupled with increasing stratigraphic
CE
Based on this Phanerozoic trend for low-latitudes (Figure 6) we tentatively propose that the baseline
value for Phanerozoic seawater, the equivalent of the present day 0 SMOW, follows the trajectory
of equation 2.
18Opw()=-0.00003t2+0.0046t......... (2)
with pw being Phanerozoic sea water in SMOW and t being an age in Ma.
The multitude of higher order oscillation that must be superimposed on this general trend will have to
be resolved on case by case basis. These will include spatial variations in oxygen isotopic composition
of water similar to those in modern seawater (Schmidt et al., 1999), ice volume correction (Lhome et
ACCEPTED MANUSCRIPT
al., 2005), salinity correction (LeGrande and Schmidt, 2006), pH (Joachimski et al., 2005), chemistry
(Brand et al., 2013), changes of temperature with depth, non-equilibrium biological fractionation of
oxygen isotopes, and a host of other phenomena enumerated in Grossman (2012a,b). Accepting the
validity of the Phanerozoic baseline (Figure 6) as a starting proposition we may now be in a position to
PT
evaluate the utility of belemnites and bivalves for complementing the record and filling the gaps,
RI
particularly across the Paleozoic/Mesozoic transition.
SC
Figure 5: color in print and online
NU
MA
D
P TE
CE
AC
Figure 5: The 18O data for brachiopods (n= 5226) from low-latitude (black), mid-latitude (green) and
high-latitude (red) realms superposed on the low-latitude planktonic foraminifera 2s-band with 1
Ma averages. For raw data see the Supplementary Appendices A and B. Time scale after Gradstein
et al. (2012).
ACCEPTED MANUSCRIPT
Figure 6: grayscale in print and online
PT
RI
SC
NU
MA
D
P TE
CE
Figure 6: The baseline Phanerozoic secular trend and its 2s bands (shaded) generated by splicing the
AC
Gaussian filtered 1 Ma averages for brachiopods older than 117 Ma and low-latitude planktonic
foraminifera younger than 118 Ma. The calculated temperature curves are based on equation 2.
For raw data see Supplementary Appendices A and B. Time scale after Gradstein et al. (2012).
7.0 Belemnites
The belemnite 18O offset data in Table 2 suggest that they may have lived in colder habitats than
brachiopods. Previous authors (Mutterlose at al., 2010; Alberti et al., 2012; Price et al., 2013) argued
that this may have been due to their nektonic, rather than nektobenthic, mode of life, living as free
ACCEPTED MANUSCRIPT
swimmers mostly below the thermocline at depths of 100 to 400 meters. Yet, considered within the
Phanerozoic baseline framework (Figure 7), all Jurassic and early Cretaceous (older than 120 Ma)
belemnites plot around the baseline trend, with their modes shifting upwards, from low- to high-
latitudes. On the other hand, the mid- and high-latitude samples in the 120-80 Ma interval of the
PT
Cretaceous plot some 4 above the baseline trend and only the two low-latitude samples fall on it.
RI
The potential utility of low-latitude belemnites for filling the gaps in the brachiopod/foraminifera
SC
database is at this stage conditional on more and better resolved data in the 120-80 Ma age bracket.
NU
Figure 7: color in print and online MA
D
P TE
CE
AC
Figure 7: The 18O data for belemnites (n=2972) from low-latitude (black), mid-latitude (green) and
high-latitude (red) realms superposed on the baseline Phanerozoic secular trend from Figure 6. For
ACCEPTED MANUSCRIPT
raw data see the Supplementary Appendices A and B. Time scale after Gradstein et al. (2012).
8.0 Bivalves
The bivalve data (Table 2) yield an inconsistent picture. Plotted on the Phanerozoic baseline trend
PT
(Figure 8), the low- and mid-latitude samples fall on the secular trend, albeit with considerably higher
spread for the latter. The high-latitude samples plot, as anticipated, in the colder space. Tentatively, it
RI
appears that at least the low-latitude bivalve samples may be a suitable proxy for filling the gaps in
SC
the Phanerozoic baseline trend.
NU
Figure 8: color in print and online
MA
D
P TE
CE
AC
Figure 8: The 18O data for bivalves (n=786) from low-latitude (black), mid-latitude (green) and high-
ACCEPTED MANUSCRIPT
latitude (red) realms superposed on the baseline Phanerozoic secular trend from Figure 6. For raw
data see the Supplementary Appendices A and B. Time scale after Gradstein et al. (2012).
PT
9.0 Phanerozoic paleotemperatures
Utilizing the Phanerozoic baseline for seawater as given by equation 2 it is possible to recalculate the
RI
ambient paleotemperatures for all samples discussed in this study.
SC
The outcome is summarized as histograms for each taxa (Figure 9). In this interpretation, the
NU
dominant paleotemperature ranges for planktonic foraminifera are within the 18-26C range, in
agreement with a typical domain for Globigerina bulloides in the low- to mid-latitude Indian ocean,
MA
where it resides in normal salinity water masses at 75 m depth (Khare and Chaturvedi, 2012).
Because only a restricted number of planktonic species and localities are utilized in the mature
D
paleoceanographic studies, the preferred temperature range is relatively narrow. In contrast, the
TE
histograms for the other taxa contain multiple species and thus potentially more variable habitats,
P
their temperature ranges are therefore broader but still near symmetrically distributed around the
CE
preferred modes. Note also that the number of samples that plot beyond the below discussed modern
temperature threshold of about 30-32C (Figure 10) is minuscule, mostly arising from the
AC
diagenetically altered outliers discussed above (Figure 1). The preferred temperature ranges appear to
have been 14-32C for brachiopods, 14-28C for belemnites, and 14-28C for bivalves. The modes for
planktonic foraminifera, brachiopods, belemnites and bivalves are 23C, 24C, 19C and 22C,
respectively. As anticipated, the temperatures for the deep-sea benthic foraminifera span a colder
range, 0-22C, with a mode at 13C. This is a pattern of paleotemperature habitats that clearly
mimics its modern counterpart. The earlier claim for generation of the 18O secular trend by post-
depositional alteration that is precisely age calibrated all across the globe, and for ten thousands of
fossils, is not a credible proposition. Use of a different 18O temperature transfer function (e.g. Bemis
ACCEPTED MANUSCRIPT
et al., 1998), infilling of gaps in the secular trend, and future advances may lead to some
modifications of the picture presented in Figure 9, but the overall pattern is unlikely to change.
PT
RI
SC
NU
MA
D
P TE
CE
AC
Figure 9: Histogram of derived paleo-seawater temperatures for four surface water taxa and for
benthic foraminifera. The taxa and habitat specific seawater temperature for all Phanerozoic
samples is calculated by utilizing the 18O ( PDB) to T(C) transfer function of Visser et. al.
(2003) based on the Phanerozoic 18O trend (eq. 2), with a 0.27 adjustment for the SMOW to
PDB standard: T(C) =16.9-4(18O - 18O trend-0.27). For raw data see the Supplementary
Appendices A and B. Grey-shaded area marks present ocean water temperature range discussed
ACCEPTED MANUSCRIPT
below (Figure 10).
PT
Insistence on a modern value for oxygen isotopic composition for past sea water (plus/minus the
superimposed ice volume effect) implies that the early to mid-Paleozoic oceans would have been hot,
RI
with temperatures mostly in ~ 30-60C range (Figure 1). The Paleozoic organisms that secreted shells
SC
for the present study lived in shallow tropical seas and the quoted temperature range would thus
NU
apply to the upper layer of tropical oceans. Yet, the 700000 measurements of modern ocean by the
Argo program (Figure 10) show an overall global span from zero to ~ 30-32C, a range that is
MA
abruptly capped at its upper limit by a disputed thermostatic regulation, potentially cloud formation
(Ramathan and Collins, 1991; Lohmann et al., 1995; Williams et al., 2009; Eschenbach, 2013). This
D
range is evidently also the optimal temperature regime for marine biological communities that involve
TE
higher life (Brock, 1985; Ravaux et al., 2013). The preference of life for optimal conditions is a rule of
P
nature, not an exception, and exceptions should not be therefore invoked to justify special pleadings
CE
invented solely for disposal of the unpalatable consequence of too hot oceans. For example, Royer et
al. (2004) argued that a pH correction on 18O, due to high carbon dioxide levels, can account for the
AC
discrepancy in temperature. In response, Shaviv and Veizer (2004) pointed out that any such
correction would only be marginal, even if applied twice as done by these authors (Ridgwell, 2005).
Moreover, the required acidification of the ancient oceans would have to be by whole units of pH if the
calibrations of Zeebe (2001) or Beck et al. (2005) are utilized, an unrealistic proposition in view of the
huge and multiple buffer systems. In another example, Trotter et al. (2008) claimed that carbonate
paleothermometer must be unreliable because their phosphate based conodont results yielded some
Ordovician temperatures that were reasonable. However, subsequent recalibration of the phosphate
thermometer by Puceat et al. (2010) resulted in temperatures that were as high as those based on
the carbonates. This albatross of the hot early to mid-Paleozoic oceans bedevils all scenarios that
ACCEPTED MANUSCRIPT
require modern type sea water oxygen isotopic composition. Only a disregard of all Paleozoic empirical
data more negative than about -3.5 (Grossman, 2012a, b) can dispense with this conundrum.
Considering that many of these old brachiopod samples differ in nothing, but their 18O, from their
isotopically more positive counterparts, we consider this to be a subjective and arbitrary treatment of
PT
the data.
RI
SC
Figure 10: grayscale in print and online
NU
MA
D
P TE
CE
AC
Figure 10: Histogram of about 700000 temperature measurements for modern oceans, up to 2000 m
Could the early Paleozoic oceans have been warmer than the above quoted limit of about 30-32C?
The estimates by the lately developed clumped, or del-47, technique (Eiler, 2007; 2011) yielded
ACCEPTED MANUSCRIPT
temperatures of 28 to 70C for the early to mid-Paleozoic (Came et al., 2007, 2008; Finnegan et al.,
2011; Wacker et al., 2012; Cummins et al., 2014) and 20 to 166C for the Mississipian (Henkes et al.,
2014) skeletons, clearly an unrealistic range for marine life. Note that these broad temperature ranges
were obtained from suites of coeval shell samples that were considered well preserved based on their
PT
original mineralogy, texture and chemical/isotopic signals. The lowest temperature for a given suite of
RI
samples was then accepted as the "best estimate", with the rest of samples believed to have been
SC
altered during their post-depositional history. Yet such rationalizations often appear to have been
based on somewhat arbitrary criteria. For example, utilizing trace elements as a parsing criterion in
NU
the set of Came et al. (2007) it can be shown that the hottest and the coldest Pennsylvanian
samples have almost identical Mn/Sr ratios; in their Silurian population the "coldest" sample has even
MA
higher Mn/Sr than the "warmest' one, opposite of the theoretical alteration trend. The same is true for
the Finnegan et al. (2011) set, providing the optically clearly recrystallized tabulate coral data are
D
exempt. Moreover, all these Sr and Mn concentrations are similar to those in not yet diagenetically
TE
affected modern counterparts (Brand et al., 2003) and may thus reflect the primary scatter. Note also
P
that even such lowest "clumped" temperatures are often warmer than those obtained by the standard
CE
The above qualifications notwithstanding, the lowest "clumped" estimates for suites of coeval samples
can provide the desirable upper limit for temperatures of the ancient oceans. This, in turn, enables
independent calculation of the oxygen isotopic composition of contemporary sea water. Unfortunately,
the approach often yields positive sea water 18O values, mostly for the "altered" but frequently also
for the "best" (lowest) temperature estimates, which are difficult to explain by processes operating
within time constraints of the studied events. The issue is therefore either ignored or disposed of by
ad hoc postulates, such as enormous glaciations coincident with warm oceans (Finnegan et al., 2011)
or by exceptional salinities for an open sea habitat (Brand et al., 2012). Note that these geologically
unrealistic scenarios (Gienne et al., 2014) are invoked solely to justify the apparent positive 18O
ACCEPTED MANUSCRIPT
composition of sea water.
The "clumped" technique is a promising tool, but its advantages as well as limitations have yet to be
defined. At this stage it is still the geological context that should be the primary defining criterion for
PT
interpretation of "clumped" data, not the other way around.
RI
Why do these fossils (Came et al., 2007, 2008; Finnegan et al., 2011; Brand et al., 2012; Wacker et
al., 2012; Grossman, 2012a, b; Henkes et al., 2013, 2014; Cummins et al., 2014) yield such elevated
SC
temperatures? The explanation (Passey and Henkes, 2012; Henkes et al., 2014) may relate to the
NU
fact that the respective isotope systematics operate at different spatial scales, the traditional oxygen
isotope thermometry at the grain size dimensions and the "clumped" one at the lattice bonding level.
MA
13
It may be therefore possible to alter the C-18O bonds, perhaps via solid state or self-diffusion
phenomena, without necessarily causing recrystallization of the grain domains. With increasing burial
D
the "bulk" shells may act as closed systems at the level of grains, retaining their original optical,
TE
13
chemical and isotopic attributes, while at the same time the C-18O bonding at lattice dimensions may
P
be reset to higher (or by uplift to lower) temperatures. Considering the many imperfections,
CE
crystallographic defects and heterogeneity of domains in biologically secreted shells, and the
geological time spans of up to 108 years, the process of "annealing" of such defect imperfections is a
AC
feasible proposition. Extrapolating from their experimental data, Passey and Henkes (2012) and
Henkes et al. (2014) argued that if post-depositional heating did not exceed about 100C the primary
temperature signal may be retained over time intervals of up to 109 years. For typical geothermal
gradients the signal may thus persist up to about 3 km burial depth. Note, nevertheless, that the
frequency of defects and lattice imperfections is particularly high in new shells and these will be
eliminated shortly after their deposition. Considering that the above heating experiments were
conducted on abiogenic and biogenic calcites that were already formed at, or were subjected to,
elevated temperatures, they may not be representative of processes at temperatures below 100C.
We contend that the inter- and intracrystralline annealing rates during the wet p/T regimes of the
ACCEPTED MANUSCRIPT
early diagenetic systems may compare favorably to those at elevated temperatures.
Considering that cold episodes, including glaciations, and analogous ecological niches and biological
communities, such as the open marine brachiopod-crinoid-coral one, persisted throughout most of the
PT
Paleozoic (Frakes et al., 2005; Blakey, 2008; Giles, 2012), the scenario of the hot oceans creates more
problems than it solves. It therefore is not a satisfactory, or at least not the full, explanation for the
RI
18
oxygen isotope secular trend. For the Precambrian, where the O depletion is still greater, the
SC
difficulties may only be compounded. If so, what could be the alternative?
NU
11.0 Changing oxygen isotopic composition of sea water
MA
The alternative approach is to accept that world ocean temperatures were buffered within ~ 0 to 30-
32C range and it was the oxygen isotopic composition of sea water that did evolve. Indeed, the band
D
TE
of empirical data (Figures 1, 6) falls well within these temperature constraints, with the rate of change
declining throughout the Phanerozoic. The overall Phanerozoic gradient for oxygen isotopic
P
composition of sea water would then be about 6 SMOW, following a nonlinear trend. The changes
CE
in the oxygen isotopic composition of sea water of this magnitude, at million-year time scales, can
AC
only be accomplished by exchange of oxygen between water and silicate rocks (Perry et al., 1978;
Muehlenbachs and Clayton, 1976; Gregory and Taylor, 1981). Interaction with rocks at temperatures
18
in excess of 350C, as in the ridge hydrothermal systems, leads to O enrichment of sea water and
opposite happens at temperatures below 350C. The maximal model rate of change of about 1
per 108 years (Walker and Lohmann, 1989) is consistent with the overall 6 gradient observed in
carbonates. Somewhat surprisingly, the sign of the trend demands that the relative importance of
high- to low-temperature water-rock interactions would have to increase in the course of the
Phanerozoic. Originally it was argued (Holmden and Muehlenbachs, 1993) that the chemical and
isotopic properties of ancient ophiolites demonstrate that this high/low temperature relationship must
ACCEPTED MANUSCRIPT
have been essentially constant throughout geologic history, thus precluding any larger change in
isotopic composition of sea water. However, these ancient examples represent rock buffered systems
where the imposed changes on chemical and isotopic properties of the rocks are not distinctive
PT
From the modeling perspective alone it is not difficult to generate a scenario where changing hot/cold
RI
rock-water interactions would replicate the observed oxygen isotope trend in carbonates. The difficulty
SC
arises with an identification of the geologically realistic physical process(es) that could drive the
model. For example, why is it that the apparent relative impact of high-temperature hydrothermal
NU
systems becomes more important, and low- temperature declines, over time despite the belief that
MA
the planetary thermal regime has been declining in the course of geologic history? Moreover, why is
the maximal rate of change apparently confined to the early/mid Paleozoic rather than to the early
Precambrian?
D
TE
Perry et al. (1978) suggested that the ubiquitous pillow basalts and pyroclastic materials in the
18
Archean were subjected to large scale alteration by cold sea water. This could indeed explain the O
P
CE
depletion in the early Precambrian, but why was the major transition delayed until the early/mid
Paleozoic? The suggestion of Walker and Lohmann (1989) that the early oceans may have been
AC
shallower and oceanic ridges partly subaerial, resulting in more cold water alteration during the
Archean, has the same timing mismatch as that of Perry et al. Wallmann (2004) proposed that
blanketing by pelagic sediments may have sealed the low-temperature hydrothermal systems to
access by sea water. This explanation suffers the opposite timing problem because the pelagic biota in
pre-Jurassic times was scarce or entirely absent. Lately, Kasting et al. (2006) proposed that because of
the planetary temperature regime that declined with time, the ridge crests in the distant past may
have been less deeply submerged, resulting in a reduced pressure, penetration depths and
temperatures within hydrothermal circulation systems. This, in turn, could have reduced the amount
the high- temperature hydrothermal systems, may have been attained only in the early Paleozoic,
potentially coincident with the exceptionally high sea level stands of that time (Miller et al., 2005). A
thick continuous blanketing of submarine volcanics by radiolarite ooze (Tolmachova et al., 2001) may
PT
have been a contributing factor.
RI
In conclusion, the hypothesis of evolving oxygen isotopic composition of sea water over geologic
SC
history generates far more consistent template for interpretation of oxygen isotope data in ancient
sediments than does the suggestion of the hot oceans. We emphasize, however, that we argue this
NU
point only for the overall secular trend and do not dispute that temperature plays also a role,
MA
particularly for the scatter of data within the band.
D
12.0 Summary
TE
A collation of 58532 18O measurements on low-Mg calcitic shells of fossil foraminifera, brachiopods,
P
belemnites and bivalves documents a consistent trend, of about -6 , throughout the Phanerozoic.
CE
This internal consistency provides a template for proposing new baseline values for evolution of
AC
oxygen isotopic composition of ancient seawater that, in turn, enable calculation of ambient habitat
temperatures for fossil taxa. The preferred paleotemperature ranges are 14-32C (mode 24C) for
brachiopods, 14-28C (19C) for belemnites, 14-28C (22C) for bivalves, 18-26C (23C) for
planktonic foraminifera, and 0-22C (13C) for benthic foraminifera; an overall pattern mimicking that
The history of chemical and isotopic research convincingly demonstrates that the original ideas of
fixed ocean water compositions were invariably superseded by recognition that the system is dynamic,
variable and evolving, a realization strenuously resisted for oxygen isotopes. It is only with the
acceptance of such paradigm shift that a modern-like actualistic outcome for ambient habitat
ACCEPTED MANUSCRIPT
temperatures of ancient fossils emerges, thus providing us with a new, long overdue, tool for
PT
Acknowledgements
We acknowledge technical support of Ms. Patricia Wickham and Kern Lee, as well as the infrastructure
RI
support of the Department of Earth Sciences, University of Ottawa.
SC
NU
MA
List of references
D
Alberti, M., Fuersich, F.T., Pandey, D.K., 2012. The Oxfordian stable isotope record (18O, 13C) of
TE
belemnites, brachiopods, and oysters form the Kachchh basin (western India) and its potential for
P
Beck, W.C., Grossman, E.L., Morse, J.W., 2005. Experimental studies of oxygen isotope fractionation in
the carbonic acid system at 15o, 25o, and 40oC. Geochim. Cosmochim. Acta 69, 3493-3503.
Bemis, B.E., Spero, H.J., Bijma, J., Lea, D.W., 1998. Reevaluation of the oxygen isotope composition of
Blakey, R.C., 2008. Gondwana paleogeography from assembly to breakup - a 500 m.y. odyssey. In:
Fielding. C.R., Frank, T.D., Isbell, J.L. (eds.) Resolving the Late Paleozoic Ice Age in Time and Space.
ACCEPTED MANUSCRIPT
Geol. Soc. Am. Special Paper 441, 1-28.
Brand, U., Logan, A., Hiller, N., 2003. Geochemistry of modern brachiopods: Applications and
PT
Brand, U., Posenato, R., Came, R.E., Affek, H., Angiolini, L., Azmy, K., Farabegoli, E., 2012. The end-
Permian mass extinction: A rapid volcanic CO2 and CH4-climatic catastrophe. Chem. Geol. 322323,
RI
121144.
SC
Brand, U., Azmy, K., Bitner, M.A., Logan, A., Zuschin, M., Came, R., Ruggiero, E., 2013. Oxygen
NU
isotopes and MgCO3 in brachiopod calcite and a new paleotemperature equation. Chem. Geol. 359,
23-31. MA
Brock, T.D., 1985. Life at high temperatures. Science 230, 132-138.
Came, R.E., Eiler, J.M., Veizer, J., Azmy, K., Brand, U., Weidman, C.R., 2007. Coupling of surface
D
TE
temperature and atmospheric CO2 concentrations during the Paleozoic era. Nature 449,
doi:10.1038/nature06085.
P
CE
Came, R.E., Brand, U., Guo, W., Veizer, J., Azmy, K., Eiler, J., 2008. Application of carbonate "clumped
isotope" thermometry to marine brachiopods from icehouse to greenhouse periods in the Paleozoic
AC
era - Preliminary results. Geol. Soc. Am. Annual Mtg., Abstracts 40/6, 402.
Carpenter, S.J., Lohmann, K.C., 1995. 18O and 13C values of modern brachiopod shells. Geochim.
Clayton, R.N., Degens, E.T., 1959. Use of carbon isotope analyses of carbonates for differentiating
fresh-water an marine sediments. Am. Assoc. Petrol. Geol. Bull. 43, 890-897.
Cramer, B.S., Toggweiler, J.R., Wright, J.D., Katz, M.E., Miller, K.G., 2009. Ocean overturning since the
Late Cretaceous: Inferences from a new benthic foraminiferal isotope compilation. Paleoceanography
constraints on Silurian ocean temperature and seawater 18O. Geochim. Cosmochim. Acta 140, 241-
258.
Degens, E.T., Epstein, S., 1962. Relationship between O18/O16 ratios in coexisting carbonates, cherts,
PT
and diatomites. Am. Assoc. Petrol. Geol. Bull. 45, 534-542.
RI
Eiler, J.M., 2007. "Clumped-isotope" geochemistry - The study of naturally-occurring, multi-substituted
SC
isotopologues. Earth Planet. Sci. Letters 262, 309-327.
NU
Eiler, J.M., 2011. Paleoclimate reconstruction using carbonate-clumped isotope thermometry.
http://wattsupwiththat.com/2012/02/12/
P
CE
http://wattsupwiththat.com/2013/10/06/.
AC
Farkas, J., Bohm, F., Wallmann, K., Blenkinsop, J., Eisenhauer, A., van Geldern, R., Munnecke, A., Voigt,
S., Veizer, J., 2007. Calcium isotope budget of Phanerozoic oceans: Implications for chemical evolution
of seawater and its causative mechanisms. Geochim. Cosmochim. Acta 71, 5117-5134.
Finnegan, S., Bergmann, K., Fischer, W.W., Jones, D.S., Fike, D.A., Hughes, N.C., Tripati, A., Eiler, J.M.,
2011. Constraints on the durations and magnitude of Late Ordovician-Early Silurian glaciation. Science
331, 903-906.
Forster, A., Schouten, S., Moriya, K., Wilson, P.A., Sinninghe Damste, J.S., 2007. Tropical warming and
intermittent cooling during Cenomanian/Turonian oceanic anoxic event 2: Sea surface temperature
ACCEPTED MANUSCRIPT
records from equatorial Atlantic. Paleoceanography 22, doi:10.1029/2006PA001349.
Frakes, L.A., Francis, J.E., Syktus, J.I., 2005. Climate Modes of the Phanerozoic. Cambridge University
Press. 288 p.
PT
Friedrich, O., Norris, R.D., Erbacher, J., 2012. Evolution of middle to late Cretaceous oceans - A 55
m.y. record of Earth's temperature and carbon cycle. Geology 40, 107-110.
RI
Ghienne, J.-F., Desrochers, A., Vandenbroucke, T., Achab, A., Asselin, E., Dabard, M.-P., Farley, C., Loi,
SC
A., Paris, F., Wickson, S., Veizer, J., 2014. A Cenozoic-style scenario for the end-Ordovician glaciation.
NU
Nature Communications, NCOMMS-13-12521B, 2014.
Giles, P.S., 2012. Low-latitude Ordovician to Triassic brachiopod habitat temperatures (BHTs)
MA
determined from 18O [brachiopod calcite]: A cold hard look at ice-house tropical oceans.
Gradstein, F.M., Ogg, J.G., Schmitz, M.D., Ogg, G.M., 2012. The Geologic Time Scale 2012, Elsevier,
Gregory, R.T., Taylor, H.P., 1981. An oxygen isotope profile in a section of Cretaceous oceanic crust,
Samail Ophiolite, Oman: Evidence for 18O buffering of the ocean by deep (5km) seawater-
AC
Gross, M.G., 1964. Variations in the 18O/16O and 13C/12C ratios of diagenetically altered limestones
Grossman, E.L., 2012a. Applying oxygen isotope paleothermometry in deep time. In: Ivany, L.C. and
Huber B.T. (eds.) Reconstructing Earth's Deep-Time Climate - The State of the Art in 2012. Paleont.
Grossman, E.L., 2012b. Oxygen isotope stratigraphy. In: Gradstein, F.M., Ogg. J.G., Schmitz, M.D.,
Ogg, G.M. (eds.) The Geological Time Scale 2012. Elsevier, 181-206.
ACCEPTED MANUSCRIPT
Hay, W.W., 2009. Cretaceous oceans and ocean modeling. In: Hu, X.M., Wang, C.S., Scott, R. W.,
Wagreich, M., Jansa, L. (eds.) Cretaceous Oceanic Red Beds: Stratigraphy, Composition, Origins,
Hay, W.W., DeConto, R.M., 1999. A comparison of modern and Late Cretaceous meridional energy
PT
transport and oceanology. In: Barrera, E., Johnson, C. (eds.) Evolution of the Cretaceous
RI
Ocean/Climate System. Geol. Soc. Am. Special Paper 332, 283 - 300.
SC
Henkes, G.A., Passey, B.H., Wanamaker, A.D., Grossman, E.L., Ambrose, W.G., Caroll, M.L., 2013.
Carbonate clumped isotope composition of modern marine mollusks and brachiopod shells. Geochim.
NU
Cosmochim. Acta 106, 307-325. MA
Henkes, G.A., Passey, B.H., Grossman, E.L., Stanton, B.J., Perez-Huerta, A., Yancey, T.E., 2014.
Temperature limits for preservation of primary calcite clumped isotope temperatures. Geochim.
D
Holmden, C., Muehlenbachs, K., 1993. The 18O/ 13C ratio of 2-billion-year-old seawater inferred from
P
Jaffres, J.B.D., Shields, G.A., Wallmann, K., 2007. The oxygen isotopic evolution of seawater: A critical
AC
review of a long-standing controversy and an improved geological water cycle model for the past 3.4
Joachimski, M., Simon, L., van Geldern, R., Lecuyer, C., 2005. Boron isotope geochemistry of Paleozoic
brachiopod calcite: Implication for a secular change in the boron isotope geochemistry over the
Kampschulte, A., Strauss, H., 2004. The sulfur isotopic evolution of Phanerozoic seawater based on
the analysis of structurally substituted sulfate in carbonates. Chem. Geol. 204, 255-286.
Kasting, J., Howard, M.T., Wallmann, K., Veizer, J., Shields, G., Jaffres, J., 2006. Paleoclimates, ocean
depth, and the oxygen isotopic composition of seawater. Earth Planet. Sci. Letters 252, 82-93.
ACCEPTED MANUSCRIPT
Khare, N., Chaturvedi, S.K., 2012. Tracing the signature of various frontal systems in stable isotopes
(oxygen and carbon) of the planktonic foraminiferal species Globigerina bulloides in the Southern
Lea, D.W., 2003. Elemental and isotopic proxies of past ocean temperatures. In: Holland, D., Turekian,
PT
K.K. (eds.) Treatise of Geochemistry. Vol. 6., Elsevier, 1-26.
RI
LeGrande, A.N., Schmidt, G.A., 2006. Global gridded data set of the oxygen isotopic composition in
SC
seawater. Geophys. Res. Letters 33, L12604, doi:10.1029/2069GL026011.
NU
Lhome, N.G., Clarke, K.C., Ritz, C., 2005. Global budget of water isotopes inferred from polar ice
Lohmann, U., Roeckner, E., Collins, W.D., Heymsfield, A.J., McFarquhar, S.M., Barnett, T.P., 1995. The
role of water vapor and convection during the Central Equatorial Pacific Experiment from observations
P
Lowenstam, H.A., 1961. Mineralogy, O18 /O16 ratios, and strontium and magnesium contents of
AC
recent and fossil brachiopods and their bearing on the history of the oceans. J. Geol. 69, 241-260.
Miller K.G., Kominz M.A., Browning G.V., Wright J.D., Mountain G.S., Katz, M.E., Sugarman, P.J., Cramer
B.S., Christie-Blick N., Pekar, S.F., 2005. The Phanerozoic record of global sea-level change. Science
310, 1293-1298.
Muehlenbachs, K., Clayton, R.N., 1976. Oxygen isotope composition of the oceanic crust and its
Mutterlose, J., Malkoc, M., Schouten, S., Sinninghe Damste, J.S., 2010. TEX86 and stable 18O
Parkinson, D., Cusack, M., 2007. Stable oxygen and carbon isotopes in extant brachiopod shells: Keys
2522-2531.
PT
Passey, B.H., Henkes, G.A., 2012. Carbonate clumped isotope bond reordering and geospeedometry.
RI
Earth Planet. Sci. Letters 351-352, 223-236.
SC
Perry Jr., E.C., Ahmad, S.N., Swulius, T.M., 1978. The oxygen isotope composition of 3800 M.Y. old
NU
metamorphosed chert and iron formation from Isukasia, West Greenland. J. Geol. 86, 223 239.
Popp, B.N., Anderson, T.F., Sandberg, P.A., 1986. Brachiopods as indicators of original isotopic
MA
compositions in some Paleozoic limestones. Geol. Soc. Am. Bull. 97, 1262-1269.
Price, G.D., Twitchett, R.J., Wheeley, J.R., Buono, J., 2013. Isotopic evidence for long term warmth in
D
TE
Prokoph, A., Shields, G.A., Veizer, J., 2008. Compilation and time-series analysis of marine carbonate
P
CE
18O, 13C, 87Sr/86Sr and 34S databases through Earth history. Earth Sci. Reviews 87, 113-133.
Puceat, E., Joachimski, M.M., Bouilloux, A., Monna, F., Bonin, A., Montreuil, S., Moriniere, P., Henard,
AC
S., Mourin, J., Dera, G., Quesne, D., 2010. Revised phosphate-water fractionation equation reassessing
paleotemperature derived from biogenic apatite. Earth Planet. Sci. Letters 298, 135-142.
Ramathan, V., Collins, W., 1991. Thermostatic regulation of ocean warming by cirrus clouds deduced
Ravaux, J., Hamel, J., Zbinden, M., Tasiemski, A.A., Boutet, I., Leger, N., Tanguy, A., Jollivet, D.,
Schillito, S., 2013. Thermal limit for metazoan life in question: In vivo heat tolerance of the Pompeii
Ravizza, G.E., Zachos, J.C., 2003. Record of Cenozoic ocean chemistry. in: Holland, H.D, Turekian, K.K.
ACCEPTED MANUSCRIPT
(eds.) Treatise on Geochemistry. Vol. 6., Elsevier, pp. 551-581.
Ridgwell, A., 2005. A Mid Mesozoic revolution in the regulation of ocean chemistry. Marine Geol. 217,
339-357.
PT
Royer, D.L., Berner, R.A., Montanez, I.P., Tabor, N.J., Beerling, D.J., 2004. CO2 as a primary driver of
RI
Savin, S.M., 1977. The history of Earth's surface temperature during the past 100 million years. The
SC
Ann. Rev. Earth Planet. Sci. 5, 319-355.
NU
Schmidt, G.A., Bigg, G.R., Rohling, E.J., 1999. Global seawater oxygen-18 database.
http://data.giss.nasa.gov.o18data/
MA
Shackleton, N.J., Opdyke, N.D., 1973. Oxygen isotope and paleomagnetic stratigraphy of equatorial
Pacific core V28-238: Oxygen isotope temperature and ice volume on a 105 year and 106 year time
D
TE
Shaviv, N., Veizer, J., 2004. Comments on "CO2 as a primary driver of Phanerozoic climate" by Royer
P
CE
Shaviv, A., Prokoph, A., Veizer, J., 2014. Is the Solar System's Galactic Motion Imprinted in the
AC
Shields, G.A., Veizer, J., 2002. Precambrian marine carbonate isotope database: Version 1.1.
Steuber, T., Veizer, J., 2002, Phanerozoic record of plate tectonic control of seawater chemistry and
Takayanagi, H., Asami, R., Abe, O., Miyajima, T., Kitagawa, H., Sasaki, K., Iryu, Y., 2013. Intraspecific
Trotter, J.A., Williams, I.S., Barnes, C.R., Lecuyer, C., Nicoll, R.S., 2008. Did cooling trigger Ordovician
PT
Urey, H.C., Lowenstam, H.A., Epstein, S., McKinney, C.R., 1951. Measurement of paleotemperatures
RI
and temperatures of the Upper Cretaceous of England, Denmark, and the southeastern United States.
SC
Bull. Geol. Soc. Am. 62, 399-416.
NU
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., Strauss, H., 1999. 87Sr/86Sr, 13C and
MA
18O evolution of Phanerozoic seawater. Chem. Geol. 161, 59-88.
Veizer, J., Fritz, P., Jones, B., 1986. Geochemistry of brachiopods: Oxygen and carbon isotopic records
D
TE
Veizer, J., Godderis, Y., Francois, L.M., 2000. Evidence for decoupling of atmospheric CO2 and global
P
CE
Veizer, J., Hoefs, J., 1976. The nature of O18/O16 and C13/C12 secular trends in sedimentary
AC
Visser, K., Thunell, R., Stott, L., 2003. Magnitude and timing of temperature change in the Indo-Pacific
Vollstaedt, H., Eisenhauer, A., Wallmann, K., Bhm, F., Fietzke, J., Liebetrau, V., Krabbenhft, A.,
Farka, J., Tomaovch, A., Raddatz, J., Veizer, J., 2014. The Phanerozoic 88Sr/86Sr record of
seawater: New constraints on past changes in oceanic carbonate fluxes. Geochim. Cosmochim. Acta
128, 249-265.
Wallmann, K., 2001. The geological water cycle and the evolution of marine 18O values. Geochim.
ACCEPTED MANUSCRIPT
Cosmochim. Acta 65, 2469-2485.
Wallmann, K., 2004. Impact of atmospheric CO2 and galactic cosmic radiation on Phanerozoic climate
change and the marine 18O record. Geochem. Geophys. Geosystems 5: doi:
PT
10.1029/2003GC000683.
Wacker, U., Fiebig, J., Munnecke, A., Joachimski, M.M., Schoene, B.R., 2012. Clumped isotopes applied
RI
to Silurian brachiopod shells, Gotland/Sweden. Goldschmidt Conf. Abstracts.
SC
Walker, J., Lohmann, K.C., 1989. Why the oxygen isotopic composition of sea water changes with
NU
time. Geophys. Res. Letters 16, 323-326.
Williams, A., Cusack, M., 2007. Chemicostructural diversity of brachiopod shells. Treatise of
MA
Invertebrate Paleontology, Brachiopoda. Vol. 6, 2396-2521.
D
Williams, I.N., Pierrehumbert, R.T., Huber, M., 2009. Global warming, convective threshold and false
TE
Zachos, J., Pagani, M., Sloan, L., Thomas, E., Billups, C., 2001. Trends, rhythms, and aberrations in
P
Zeebe, R.E., 2001. Seawater pH and isotopic paleotemperatures of Cretaceous oceans. Palaeogeog.
AC