Mean age of oceanic lithosphere drives eustatic sea-level change

since Pangea breakup

Jean-Pascal Cogn
a,

, Eric Humler
a,b,1
, Vincent Courtillot
a
a
Institut de Physique du Globe de Paris et Universit de Paris 7, 4 place Jussieu, 75252 Paris cedex 05 France
b
Laboratoire de Plantologie et Godynamique, Universit de Nantes, 4 rue de la Houssinire, 44322 Nantes cedex 3 France
Received 14 October 2005; received in revised form 14 March 2006; accepted 14 March 2006
Available online 19 April 2006
Editor: R.D. van der Hilst
Abstract
The Atlantic and Indian Oceans and the oceanic part of the Antarctic plate have formed at the expense of Panthalassa as a result
of Pangea breakup over the last 180 Myr. This major plate reorganization has changed the age vs. surface distribution of oceanic
lithosphere and has been a likely driver of sea-level change. Assuming that the age/surface structure of Panthalassa has remained
similar to the present-day global distribution from 180 Ma to Present, and using the isochron patterns preserved in the newly
formed oceans, we model resulting relative sea-level change. We find a first (slower) phase of sea-level rise (by 90 to 110 m),
culminating between 120 and 50 Ma, followed by a (faster) phase of sea-level drop. We show that this result is not strongly
sensitive to our hypothesis of constant mean age of Panthalassa, for which much of the information is now erased due to
subduction. When the effects of oceanic plateau formation and ice cap development are added, the predicted sea-level curve fits
remarkably well the first-order variations of observed sea-level change. We conclude that the changes in mean age of the oceanic
lithosphere (varying between 56 and 620.2 Myr), which are simply the expression of the Wilson cycle following Pangea breakup,
are the main control, accounting for 70%, of first-order changes in sea-level.
2006 Elsevier B.V. All rights reserved.
Keywords: oceans; sealevel; eustacy; global changes; Pangea breakup
1. Introduction the Pangea breakup
It is well known that over time intervals of hundreds
of millions of years, sea-level has fluctuated by several
hundred meters. The main evidence is the changing area
of marine sediment deposited on continents through
time, indicating that at certain periods continents were
flooded by seawater far more extensively than they are
today. It is often proposed that the most likely cause of
large-scale changes in sea-level is the variable volume
of ridge material [1], which can produce variations with
an amplitude of several tens of meters. If seafloor
spreading increases, then the ridge crest volume starts to
increase, displacing water and causing additional flood-
ing of continental areas. The critical quantity is the area
of seafloor produced per unit of time. This can be
changed either by an increase in spreading rate (for a
ridge crest of constant length) as proposed by Kominz
[2], or by an increase in the length of ridge crest, or by
some combination of the two. There are reasons,
Earth and Planetary Science Letters 245 (2006) 115122
www.elsevier.com/locate/epsl

Corresponding author. Equipe de Palomagntisme, Institut de

Physique du Globe de Paris et Universit de Paris 7, 4 place Jussieu,
75252 Paris cedex 05 France. Tel.: +33 1 44 27 60 93.
E-mail addresses: [email protected] (J.-P. Cogn),
[email protected] (E. Humler).
1
Tel.: +33 2 51 12 52 64.
0012-821X/$ - see front matter 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.epsl.2006.03.020
however, to question the foundations upon which this
linkage is based, because recent studies fail to show
evidence for large variations in either oceanic spreading
rate or global ridge length [35].
Various processes, other than mid-oceanic ridge
dynamics, can affect sea-level over time scales of tens
to hundreds of millions of years, such as ice cap growth,
ocean sediment volume changes, surface reduction of
continental lithosphere, ocean temperature variations,
continental breakup, true polar wander events, etc.
Some studies (e.g. [6,7]) emphasized that long-term sea-
level highstand during the Late Cretaceous was
preceded by a major change in plate motions the
breakup of Pangea. In particular, based on the models of
Heller and Angevine [8], Heller et al. [7] proposed
model curves of sea-level changes in the Mesozoic and
Cenozoic. Assuming (1) that the world consists of two
oceans, a rectangular Atlantic-type ocean and a
triangular Pacific-type ocean, including the Indian
Ocean (rectangular and triangular refer to the shape
of {dA/ dt}=f (t) curves), (2) that the entire Atlantic
rifted simultaneously, (3) that the production rate (km
2
yr
1
) remained constant in the Atlantic, and (4) that
subduction consumes as much young crust as old
throughout time, these authors demonstrated that this
mechanism produces a 100 m sealevel rise and fall
between 180 Ma and the Present, with a highstand
between 70 and 120 Ma.
We attempt to refine the effects of such a mechanism,
using a different approach based on the direct measure-
ment of actual oceanic crust surfaces delineated by ocean
floor isochrons. Using the isochron patterns of Royer et al.
[9] and Mller et al. [10], we propose a model of evolution
of surface/age distribution, as new oceans formed during
breakup of Pangea, and the older Panthalassa was being
consumed. Our analysis is thus based on (1) a deter-
mination of the evolution as a function of time of oceanic
lithosphere surfaces in the Atlantic, Indian and circum-
Antarctic oceans basins based on the more recent isochron
database [9,10], and (2) a model of surface/age balance
through time of Panthalassa, excluding Indian and
circum-Antarctic oceans, associated to a direct measure-
ment of Panthalassa area at each reconstruction time. Note
that in doing so, we avoid strong hypotheses on variations
in spreading rates, which are still a matter of debate (e.g.
[25,11]). Finally, although the basic mechanism is
somewhat identical to the one proposed by Heller and
Angevine [8], we here propose a detailed analysis of
Atlantic, Indian and circum-Antarctic isochrons, rather
than relying on models of rectangular Atlantic-type and
triangular Pacific-type (including the Indian and Antarctic
basins) oceans.
2. Mechanism and surface balance
Let us first briefly recall Earth evolution since the
Late Carboniferous ( 250 Ma). At that time, continents
were assembled in a single unit called Pangea, sur-
rounded by a global ocean named Panthalassa, including
3 main units: MongolOkhotsk, Tethys, and paleo-
Pacific oceans (Fig. 1a). This configuration remained
quite stable for about 70 million years. At 190180 Ma,
the super-continent Pangea began to rift apart to form
the central Atlantic ocean, with North America rotating
away from Africa. Concurrently, the MongolOkhotsk
ocean began to vanish, and closure was achieved by
135 Ma [1214]. At about 160 Ma, the Somalia and
Southwest Indian basins opened, when a plate com-
posed of Madagascar, India, Antarctica and Australia
(East Gondwana) rotated away from East Africa (partly
west of Gondwana). East Gondwana broke up between
120 and 90 Ma, opening the rest of the Indian
ocean. The main episode of Tethys ocean closure began
at 100 Ma (e.g. [15]), during the fast northward drift
of India, which collided with Eurasia at 50 Ma [16].
Prior to this time, at about 130 Ma, the South Atlantic
ocean opened, and the whole Atlantic (central and
southern parts) continued opening up to the Present,
mainly at the expense of the paleo-Pacific, propagating
in the north Atlantic area and separating North America,
Greenland and Europe. The distribution of magnetic
lineations on the seafloor [9,10] can be used to
reconstruct the positions of the continents for the past
150 Myr. Because there is little seafloor older than this,
reconstructions prior to about 180 Ma are primarily
based on paleomagnetic measurements in continental
rocks [17,18] together with geological studies of
orogenic belts.
The basic idea underlying the way in which such a
history could have led to sea-level changes is the
following (e.g. [7]): when Pangea existed, say between
250 and 180 Ma, the global ocean, Panthalassa
(Fig. 1a), with its three main units, Tethys, Mongol
Okhotsk, and paleo-Pacific oceans (including now
vanished or partially subducted parts such as Izanagi,
Kula, Farallon, Cocos and Nazca plates), was an old, or
average ocean. The meaning of old or average is
discussed below. Importantly, this ocean started being
consumed due to the expansion of newoceans, such as the
Atlantic and Indian oceans, and to the expansion of the
Antarctic plate as well. Note that, in the present analysis,
all the now subducted Tethyan parts of the Indian ocean,
north of the northern margin of India, and between Africa
and Eurasia, are included in Panthalassa, not in Indian
ocean history. Therefore, this older ocean (Fig. 1d) started
116 J.-P. Cogn et al. / Earth and Planetary Science Letters 245 (2006) 115122
to be progressively replaced by a younger, hence
shallower, ocean, thus leaving less place for water than
was available above the older, deeper and disappeared,
subducted crust (Fig. 1e). Then, because new oceans
subsequently grow older and deeper, more space became
available again (Fig. 1f). Finally, the present-day situation
(Fig. 1g), in which the Pacific (including Nazca and
Cocos) represents approximately half of the global ocean,
and the others, the other half, was attained. At this stage,
both parts have a similar surface/age structure (Fig. 1g),
and globally, this structure is identical (in terms of space
available for seawater) to the one we model at the starting
age of 180 Ma (Fig. 1d). Thus, sea-level has come back to
its initial value at 180 Ma.
Because of constant Earth radius and continental
lithospheric surfaces, it can be confidently assumed that
the total oceanic lithospheric area has remained constant
in the last 200 Myr. On the other hand, based on the
isochron maps of Royer et al. [9] and Mller et al. [10],
one can assume that the new oceans which formed
following Pangea breakup, the Atlantic and Indian
oceans (excluding their Tethyan parts), together with the
oceanic part of the Antarctic plate, have not been
significantly subducted in any part. Indeed, there are
Fig. 1. Left column: HammerAitoff projections centered on 180 Longitude illustrating the reduction of Panthalassa from 180 Ma (a) to the Present
(c) through 80 Ma (b); circled 1, 2 and 3: paleo-Pacific, Tethys and MongolOkhotsk oceans respectively; limits between these units (black dotted
lines) are arbitrary; (c) Present-day configuration with isochrons in Atlantic and Indian oceans, and oceanic part of the Antarctic plate, following
Royer et al. [9] and Mller et al. [10]; the Pacific area is intentionally left blank (see text). Right column: Schematic sketch of the proposed mechanism
for sea-level change between 180 Ma (d) and the Present (g), shown as seafloor topography and water depth (vertical) vs. crust age, as percentages of
the global ocean area A (horizontal); although schematic, the distributions in (d) and (g) reflect the present-day distribution of global oceanic surfaces;
in each box (d) to (g) the colored bars represent the percentage of oceanic crust of a given age range (width), vs. depth under sea-level (blue lines); the
color scale follows the isochron color scale of Mller et al. [10] (see Fig. 2b) modified for Gradstein et al. [36] magnetostratigraphic scale, with hot
(cold) colors for young (older) lithosphere; A: area of the oceans; subscripts o, p and n stand for Total, Panthalassa and New oceans respectively;
black and grey bars at the bottom of each box underline the relative amount of Panthalassa (black) and new oceans (grey) respectively; blue arrows
and dh indicate the sense of expected sea-level change between each step; t
1
and t
2
are intermediate ages between 180 Ma and the Present.
117 J.-P. Cogn et al. / Earth and Planetary Science Letters 245 (2006) 115122
exceptions to this simplifying first-order scheme, in the
northeastern Indian ocean (Sumatra subduction) and in
the western Central Atlantic (Caribbean arc), but they
are thought to be of secondary importance for the
present model. Therefore, the measurement of oceanic
crust segment areas delineated by isochron patterns in
these oceans (Fig. 1c) can be used as a proxy for the
history of disappeared old Panthalassa surfaces during
geological times since 180 Ma.
This measurement has been made by several authors,
since global isochrons became available (e.g. [3,4,19]),
demonstrating growth of these oceans by 150 million
km
2
(M km
2
) since the beginning of breakup. Using the
global paleogeographic reconstructions of Besse and
Courtillot [17,18], including Mller et al. [10] for major
continental plates, and Yang and Besse [20] for Asian
microplates (e.g. Fig. 1a and b), we have checked that the
area of Panthalassa was indeed reduced from 300 to
150 M km
2
between 180 Ma and the Present. This
surface decrease was achieved first by closure of the
MongolOkhotsk ocean at 135 Ma, then by closure of the
Tethys ocean, accompanied by reduction of the paleo-
Pacific ocean.
3. Model and results
The birth of new oceans decreases the average age of
the seafloor and should, therefore, decrease the global
volume of ocean basins. In order to quantify this effect, we
first need to know the surface/age structure of Pantha-
lassa, from at least 180 Ma to the Present. This appears to
be impossible for older and totally subducted oceans such
as MongolOkhotsk, rather speculative for Tethys [21]
which is completely subducted as well, and quite difficult
for the paleo-Pacific (e.g. [2224]), a large part of which
has also been subducted. In order to circumvent this
problem, we have based our reconstructions, on a first-
order, more conservative model. This model is based on
the generally accepted ideas of Sclater et al. [19] and
Parsons [25], that the currently observable seafloor age
distribution results from (1) a triangular {dA/ dt} distri-
bution arising from constant crustal production at the
global scale and (2) subduction of oceanic lithosphere
which is independent from the age of the crust. Because
neither the global ridge lengths [2,4], nor the spreading
rates [35] on these ridges have significantly varied in the
past, we assume that this seafloor age distribution can be
extrapolated to the last 180 Myr for Panthalassa. Prac-
tically, we modelled the average, or the old, whole
Panthalassa ocean (including Tethys with its north India
and north Africa margins, MongolOkhotsk and paleo-
Pacific oceans) as an average ocean having a maximum
crust age of 180 Ma at each time step, with a surface/age
balance similar to the present-day global one, and a
subduction rate independent from the lithospheric age. In
contrast, we directly determined the evolution, from
180 Ma to Present, of the oceanic basins volume gene-
rated by the new oceans (Atlantic, Indian, oceanic part of
Antarctic; AIA for short in what follows) from the
measurement of surface/age structure currently observ-
able in these oceans (Fig. 1c). Finally, we determined the
volumes of oceanic basins by using two different depth vs.
age relationships (i.e. [26,27]).
The balance of oceanic basin volume changes due to
Panthalassa subduction and newly generated volumes in
the AIA oceans, given a constant global oceanic surface
(Fig. 1d to g), produces the relative sea-level changes
illustrated in Fig. 2a and b. The red curves are derived
from the assumptions discussed above. The blue ones
include plateau formation and ice cap growth as explained
below. Figures a and b correspond to the two depth vs. age
models. The first, obvious fact in these figures is that our
simple model predicts sealevel rise during the Cretaceous,
stabilization between 100 and 50 Ma, and a fall since
50 Ma. The model curves (redand blue curves in Fig. 2a
and b) are in phase with the eustasy curve of Haq et al.
([28]; dashed black lines) and its first-order variation
(solid black lines). We further point out that even the
shapes (overall dissimetry) of the curves are in close
agreement, with a relatively slow rise of sea-level before
100 Ma and a sharper fall after 50 Ma (Fig. 2). This
confirms the results of Heller and Angevine [8] and Heller
et al. [7] and underlines the important fact that Pangea
breakup does have a eustatic effect which is in first order
agreement with observations as summarized by Haq et al.
[28].
However, amplitudes predicted by our original model
(red curves in Fig. 2) are not in close agreement with
observations. Indeed, modelling the effects of Pangea
breakup alone leads to only 90 to 110 m of relative
sea-level change (Fig. 2a and b). Although the largest
amplitude (Fig. 2b) is obtained when using the
geochemically based age vs. depth relationship of Hum-
ler et al. [27], it hardly exceeds 110 m. We thus follow
Harrison [29] and Heller et al. [7] in arguing that other
significant phenomena, not directly linked to oceanic
crust dynamics, might enhance the magnitude of these
changes: these are the emplacement of oceanic plateaus
between 120 and 70 Ma, which may have led to an
additional 60 m sea-level rise, and the growth of ice
caps in the past 40 Myr which may have caused a 80
to 60 m fall in sea-level [29]. Other minor effects (e.g.
sedimentary fluxes, continental crust thinning before
ocean opening, continental crust reduction in collision
118 J.-P. Cogn et al. / Earth and Planetary Science Letters 245 (2006) 115122
zones, etc) may be involved. Adding plateaus and ice
cap growth effects increases predicted model amplitudes
(blue curves in Fig. 2a and b) to 170 to 180 m
respectively, with an uncertainty on the order of 15 m.
It should be noticed, however, that these simple
calculations do not take into account any isostatic
effects (e.g. [29]) which could lower down sea level
amplitude by about 30%. Further error could arise from
our simplifying assumption of a constant average
surface/age structure of Panthalassa through time.
More detailed assumptions on the structure of the
paleo-Pacific may further improve the correlation.
Another reason for mismatch could be due to inaccuracy
of the eustasy chart [28,30] which is far from being
universally accepted (e.g. [31]), and we finally note that
the 100150 m amplitude of sealevel variations we
propose appear consistent with other proposed models
of Sahagian et al. [32] and Miller et al. [33]. Altogether,
therefore, we conclude that first-order sea-level changes
seem to be primarily driven by a mechanical effect
Fig. 2. (a) and (b) curves of relative sea-level (in m, left scale) from 180 Ma to the Present; black dotted curves: digitization of Haq et al. [28], in
Hardenbol et al. [30], relative sea-level curve; black solid curves: 3rd-order polynomial adjustment on the first-order sea-level variation; red curves:
model of sea-level variation due to Pangea breakup as proposed in this paper, using the depth vs. age models of (a) Stein and Stein [26] and (b)
Humler et al. [27]; blue curves: same as previous ones, but adding plateau and ice cap growth as suggested by Harrison [29] and Heller et al. [7]; the
grey area in (b) suggests the uncertainty on these models; these are relative curves, which explains the offset between experimental (black) and
modelled (red and blue) curves; in our model, we have taken the zero line of sea-level at an age of 180 Ma, whereas Haq et al. [28] chose 60 m of
height for the present-day zero sea-level. (c) Global weighted mean age of oceanic lithosphere (in Myr, right scale) as modelled from 180 Ma to the
Present; uncertainties are less than 0.2 Myr. Color bars follow the isochron color scale of Mller et al. [10] modified for Gradstein et al. [36]
magnetostratigraphic scale.
119 J.-P. Cogn et al. / Earth and Planetary Science Letters 245 (2006) 115122
(ocean lithosphere balance in surface/age distribution
through times), to which are added the effects of oceanic
plateau production and ice cap growth, without the need
to resort to changes in the rates of oceanic crust
production (e.g. [24,34]).
Furthermore, we point out that the mechanism of
Pangea breakup predicts that the oceanic lithosphere as a
whole should exhibit an episode of rejuvenation, due to
the opening of young oceans, followed by progressive
ageing of the new oceans. To test this idea, we have
computed the mean age of the oceanic lithosphere as the
sum of the ages of individual crustal segments bounded
between pairs of successive isochrons, weighted by the
area percentage of each of these segments, for each
reconstruction interval from 180 Ma to the Present. The
resulting weighted mean age varies between 56 and
62 Myr, and is clearly anticorrelated with sea-level
changes (Fig. 2c). We note that the 6 Myr amplitude of
mean age variation we obtain in measuring the oceanic
segments is in good agreement with the one proposed by
Heller and Angevine [8] on the basis of their model. Here
again, we observe a good phase correlation between the
two signals, with slow increasing sea-level correlated to
slow decreasing mean lithospheric age, and the sharper
sea-level fall correlated to sharper increasing age.
4. Robustness of the mean oceanic crust age vs.
sealevel relationship
We therefore conclude that a significant part of first
order sea-level change is primarily controlled by the mean
age of the oceanic lithosphere. We nowtest the robustness
of this inference viz. some of our underlying assumptions.
For instance, we have assigned Panthalassa, which is
largely unknown in this respect, a constant average
lithospheric age of 62 Ma through geological time since
180 Ma (the rest of the Global Ocean being measured via
the actual isochron patterns) i.e. the present-day value.
Because Tethys, MongolOkhotsk, and large parts of the
Pacific oceans have disappeared, and the overall surface
has decreased by 50% between 180 Ma and the Present
in going from Panthalassa to the Pacific, we have no
means to check the detailed age structure of this ocean in
the past. In other words, because some of our more
important conclusions are based on a strong assumption
on Panthalassa surface/age structure (thus, mean age), we
need to explore the sensitivity of our conclusions to this
assumption further.
We have therefore tested the effects of different models
of mean age evolution of Panthalassa crust (Fig. 3a), on
the shape of the global oceanic mean age curve (Figs. 2c
Fig. 3. (a) Curves of modelled Panthalassa crust mean age (in Myr) as a function of geological age (in Ma) following Models 0 to 4 (see text); grey
dotted XLBC(a) and continuous XLBC(b) curves: paleo-Pacific mean age deduced from Gordon and Jurdy [23] and Hall [35] respectively (after Fig.
5(a) and (b) of Xu et al. [24]). (b) Bottom: Global oceanic crust mean age (in Myr) as a function of geological age (in Ma) resulting fromModels 0 to 4
(see text); Top: first-order sea-level in meters (black dotted curve) following Haq et al. [28], and smoothed sea-level evolution from 180 Ma to the
Present (black curve), as a function of age (in Ma); grey area underlines period between 100 and 50 Ma; color bars in (a) and (b) as in Fig. 2.
120 J.-P. Cogn et al. / Earth and Planetary Science Letters 245 (2006) 115122
and 3b). Model 0 (black line in Fig. 3a) is the one
presented above, where the mean age of Panthalassa crust
is assumed constant through times at 62 Ma, the present-
day value of the global oceanic crust. Models 1 (blue line,
Fig. 3a) and 2 (red line) involve linear decrease and
increase, respectively, of the mean age of Panthalassa
crust, from 180 Ma to the Present, with an empirical
amplitude of 6 Myr. Models 3 (dotted red curve, Fig. 3a)
and 4 (dotted blue curve) involve smoothly variable mean
ages of Panthalassa, starting at 62 Myr at 180 Ma (same as
Present), with respectively a decrease or an increase
arbitrarily fixed at 90 Ma. Here again, the amplitude is
empirically fixed at 6 Myr. These five, simple models are
sufficient to test the response of global sea level variations
to reasonable changes in Panthalassa age distribution.
We have next computed changes in the global weighted
mean age of oceanic lithosphere predicted from these
models of Panthalassa mean ages. Whichever model is
used, the creation of new AIA oceans at the expense of the
older one induces a significant, asymmetric dip in the
curves of global mean age vs. time. Notwithstanding the
model used, the minimum occurs between 100 and 50 Ma
(grey area in Fig. 3b), i.e. the time of sealevel highstand
(black curves, Fig. 3b), and the decrease in mean oceanic
crust age before 100 Ma is slower than the rise after 50 Ma.
These two features, which correlate with the smooth in-
crease and sharper decrease of sea level during these pe-
riods, are robust whichever Panthalassa age model is used.
As could be guessed from inspection of Fig. 3, the
(anti)correlation coefficient between observed first order
sea level change and the prediction of Model 0 is high and
significant, at 0.94. Models 1 to 4 lead to correlation
coefficients ranging from 0.44 to 0.97. Of course, the
minimum correlation is formed with Model 4, in which
assumed Panthalassa age structure fully opposes the
observed sea level change. Yet, even in this worse case
scenario, the main features of Cretaceous younging
followed by faster Cenozoic ageing of the global crust,
and related sea level variations are preserved.
Indeed, models in which Panthalassa shows null to
positive increases in mean age in the past (Models 0, 2
and 3, Fig. 3) better correlate with sealevel change. This
is actually supported by detailed studies of oceanic
structure in the Pacific. We have plotted the paleo-Pacific
reconstructions of Xu et al. [24] for the last 65 Myr in
Fig. 3 (grey curves), computed after the previous recon-
structions of Gordon and Jurdy [23] (XLBC(a) dotted
grey curve, Fig. 3) and Hall [35] (XLBC(b), continuous
grey curve, Fig. 3). Although it shows some scatter
around our model, the more recent XLBC(b) curve
appears highly consistent with our study in the 065 Ma
period, which supports the mechanism of ocean
lithosphere rejuvenation due to Pangea breakup pro-
posed here.
5. Conclusions
We believe that the main result from this study is that
Pangea breakup controls a significant part of sea-level
variations since 180 Myr, in line with the models of Heller
and Angevine [8] and Heller et al. [7]. We based our
results on direct computation of oceanic lithosphere sur-
faces encompassed between isochron pairs in newly
formed oceans (Atlantic, Indian and oceanic part of An-
tarctic), assuming a model of surface/age structure for
Panthalassa. Two key assumptions were that: (1) the
newly formed oceans accreted at the expense of Pantha-
lassa; (2) due to the difficulty of modelling the age
structure of the subducted oceanic segments in Pantha-
lassa, including the vanished MongolOkhotsk and
Tethys oceans, and the largely consumed paleo-Pacific
plates (including all or parts of Kula, Izanagi, Farallon and
Cocos), we assumed an average age/surface distribution
within this ocean similar to the present-day global one.
The surface/age structure balance between this modelled
Panthalassa and the observed, actual surface distribution
of the Atlantic, Indian and Antarctic oceans/plates bet-
ween 180 Ma and the Present is a major forcing mecha-
nism which accounts for approximately half of first order
sea-level change, due to the rejuvenation of the global
oceanic lithosphere. There is no need to resort to a pulse in
oceanic production in mid-Cretaceous times, consistent
with the conclusions of Heller et al. [7]. Regardless of the
model of age evolution of Panthalassa oceanic crust,
rejuvenation of global lithosphere ages occurs between
100 and 50 Ma. This rejuvenation appears to be well
correlated with first-order sealevel change in the last
180 Myr, and is therefore a candidate as a first-order
forcing mechanism, accounting for 90 to 110 m
( 70%). However, first order sea-level change also re-
sults from the addition of two other important mechan-
isms, oceanic plateau production and ice caps deve-
lopment, contributing respectively +60 m in the 120
70 Ma period, and 60 to 80 m since 40 Myr. First-
order sea-level change is probably a good proxy of mean
oceanic lithospheric age, which has itself evolved in the
past as a result of Wilson cycles of supercontinent
fractionation and amalgamation.
Acknowledgements
We thank J. Besse, L.E. Ricou, F. Fluteau, J. Dyment,
Y. Gaudemer, J. Marcoux and L. Barrier for helpful
discussions and comments on this work. S. Gilder and S.
121 J.-P. Cogn et al. / Earth and Planetary Science Letters 245 (2006) 115122
Tait kindly edited an early draft of this manuscript. A first
version of this paper largely benefited from constructive
criticisms of R.L. Larson, R.V. Demicco, R.D. Mller and
C. Lithgow-Bertelloni. This is contribution 2123 of IPGP.
References
[1] J.D. Hays, W.C. Pitman, Lithospheric plate motion, sea level
changes and climatic and ecological consequences, Nature 246
(1973) 1822.
[2] M. Kominz, Oceanic ridge volumes and sea-level change an
error analysis, AAPG Memoir 36 (1984) 109127.
[3] D.B. Rowley, Rate of plate creation and destruction: 180 Ma to
present, Geol. Soc. Amer. Bull. 114 (2002) 927933.
[4] J.P. Cogn, E. Humler, Temporal variation of oceanic spreading
and crustal production rates during the last 180 My, Earth Planet.
Sci. Lett. 227 (2004) 427439.
[5] J.P. Cogn, E. Humler, Trends and rhythms in seafloor generation
rate, Geochem. Geophys. Geosyst. 7 (2006) Q03011,
doi:10.1029/2005GC001148.
[6] T.R. Worsley, D. Nance, J.B. Moody, Global tectonics and Eustasy
for the past 2 billion years, Mar. Geol. 58 (1984) 373400.
[7] P.L. Heller, D.L. Anderson, C.L. Angevine, Is the middle
Cretaceous pulse of rapid sea-floor spreading real or necessary?
Geology 24 (1996) 491494.
[8] P.L. Heller, C.L. Angevine, Sea-level cycles during the growth of
Atlantic-type oceans, Earth Planet. Sci. Lett. 75 (1985) 417426.
[9] J.Y. Royer, R.D. Mller, L.M. Gahagan, L.A. Lawyer, C.L.
Mayes, D. Nrnberg, J.G. Sclater, A global isochron chart, Tech.
Rep., vol. 117, University of Tex. Inst. for Geophys, Austin, 1992.
[10] R.D. Mller, W.R. Roest, J.Y. Royer, L.M. Gahagan, J.G. Sclater,
Digital isochrons of the world's ocean floor, J. Geophys. Res.
102 (1997) 32113214.
[11] R.V. Demicco, Modelling seafloor-spreading rates through time,
Geology 32 (2004) 485488, doi:10.1130/G20409.1.
[12] L.P. Zonenshain, M.I. Kuzmin, J.M. Natapov, Geology of the
USSR: A Plate Tectonic Synthesis, Geodyn. Series, vol. 21,
AGU, Washington, D.C, 1990, p. 242.
[13] V.A. Kravchinsky, J.P. Cogn, W. Harbert, M.I. Kuzmin,
Evolution of the MongolOkhotsk ocean with paleomagnetic
data from the suture zone, Geophys. J. Int. 148 (2002) 3457.
[14] J.P. Cogn, V.A. Kravchinsky, N. Halim, F. Hankard, Late
JurassicEarly Cretaceous closure of the MongolOkhotsk
Ocean demonstrated by new Mesozoic paleomagnetic results
from the Trans-Bakal area (SE Siberia), Geophys. J. Int. 163
(2005) 813832.
[15] L.E. Ricou, Tethys reconstructed: plates, continental fragments
and their boundaries since 260 Ma from Central America to
South-eastern Asia, Geodin. Acta 7 (1994) 169218.
[16] P. Patriat, J. Achache, IndiaEurasia collision chronology has
implications for crustal shortening and driving mechanism of
plates, Nature 311 (1984) 615621.
[17] J. Besse, V. Courtillot, Revised and Synthetic apparent polar
wander path of the African, Eurasian, North American and Indian
plates, and True polar wander since 200 Ma, J. Geophys. Res. 96
(1991) 40294050.
[18] J. Besse, V. Courtillot, Correction to Apparent and true polar
wander and the geometry of the geomagnetic field in the last
200 Myr, J. Geophys. Res. 108 (2003) 2469, doi:10.1029/
2003JB002684.
[19] J.G. Sclater, B. Parsons, C. Jaupart, Oceans and continents:
similarities and differences in the mechanisms of heat loss,
J. Geophys. Res. 86 (1981) 11,53511,552.
[20] Z.Y. Yang, J. Besse, New Mesozoic apparent polar wander path
for south China: tectonic consequences, J. Geophys. Res. 106
(2001) 84938520.
[21] C. Heine, R.D. Mller, C. Gaina, Reconstructing the lost Eastern
ocean basin: convergence history of the SE Asian margin and
marine gateways, AGU Monogr. Ser. 149 (2004) 3754.
[22] D.C. Engebretson, A. Cox, R.G. Gordon, Relative motions
between oceanic and continental plates in the Pacific basin, Geol.
Soc. Am. Spec. Pap. 206 (1985) 159.
[23] R.G. Gordon, D.M. Jurdy, Cenozoic global plate motions,
J. Geophys. Res. 91 (1986) 12,38912,406.
[24] X. Xu, C. Lithgow-Bertelloni, C.P. Conrad, Global reconstruc-
tions of Cenozoic seafloor ages: implications for bathymetry and
sea level, Earth Planet. Sci. Lett. 243 (2006) 552564,
doi:10.1016/j.epsl.2006.01.010.
[25] B. Parsons, Causes and consequences of the relation between area
and age of the ocean floor, J. Geophys. Res. 87 (1982) 289303.
[26] C.A. Stein, S. Stein, A model for the global variation in oceanic
depth and heat flow with lithospheric age, Nature 359 (1992)
123129.
[27] E. Humler, C.H. Langmuir, D. Daux, Depth versus age: new
perspectives from the chemical compositions of ancient crust,
Earth Planet. Sci. Lett. 173 (1999) 723.
[28] B.U. Haq, J. Hardenbol, P.R. Vail, Chronology of fluctuating sea
levels since the Triassic, Science 235 (1987) 11561166.
[29] C.G.A. Harrison, Long-term eustasy and epeirogeny in con-
tinents, in: Geophysics Study Committee (Ed.), Sea-level
Change, Washington D.C., U.S. National Research Council,
Geophysics Study Committee, National Academy Press, 1990,
pp. 141158.
[30] J. Hardenbol, J. Thierry, M.B. Farley, P.C. de Graciansky, P.R.
Vail, Mesozoic and Cenozoic sequence chronostratigraphic
framework of European basins, in: P.C. de Graciansky, J.
Hardenbol, T. Jacquin, P.R. Vail (Eds.), Mesozoic and Cenozoic
Sequence Stratigraphy of European Basins, SEPM Spec. Pub.,
vol. 60, 1998, pp. 313.
[31] A.D. Miall, C.E. Miall, Sequence stratigraphy as a scientific
enterprise: the evolution and persistence of conflicting para-
digms, Earth-Sci. Rev. 54 (2001) 321348.
[32] D. Sahagian, O. Pinous, A. Olferiev, et al., Eustatic curve for the
Middle JurassicCretaceous based on Russian platform and
Siberian stratigraphy: zonal resolution, Am. Assoc. Pet. Geol.
Bull. 80 (1996) 14331458.
[33] K.G. Miller, M.A. Kominz, J.V. Browning, J.D. Wright, G.S.
Mountain, M.E. Katz, P.J. Sugarman, B.S. Cramer, N. Christie-
Blick, S.F. Pekar, The Phanerozoic record of Global Sea-level
change, Science 310 (2005) 12931298.
[34] R.L. Larson, Latest pulse of the Earth: evidence for a mid-
Cretaceous superplume, Geology 19 (1991) 547550.
[35] R. Hall, Cenozoic geological and plate tectonic evolution of SE
Asia and the SW Pacific: computer-based reconstructions, model
and animations, J. Asian Earth Sci. 20 (2002) 353434.
[36] F.M. Gradstein, J.G. Oggs, A.G. Smith, et al., A Geologic Time
Scale 2004, Cambridge University Press, Cambridge, England,
2004, p. 589.
122 J.-P. Cogn et al. / Earth and Planetary Science Letters 245 (2006) 115122