Geol Rundsch (1996) 85 : 452–465
Q Springer-Verlag 1996
ORIGINAL PAPER
M. A. Maslin 7 G. H. Haug 7 M. Sarnthein
R. Tiedemann
The progressive intensification of northern hemisphere glaciation
as seen from the North Pacific
Received: 26 May 1995 / Accepted: 24 November 1995
Abstract Ocean Drilling Project (ODP) site 882
(50722bN, 167736bE) provides the first high-resolution
GRAPE density, magnetic susceptibility, carbonate,
opal and foraminifera (planktonic and benthic) stable
isotopes records between 3.2 and 2.4 Ma in the Northwest Pacific. We observed a dramatic increase in ice
rafting debris at site 882 at 2.75 Ma, which is coeval
with that found in the Norwegian Sea, suggesting that
the Eurasian Arctic and Northeast Asia were significantly glaciated from 2.75 Ma onwards. Prior to
2.75 Ma planktonic foraminifera d 18O records indicate
a warming or freshening trend of 4 7C or 2‰ over 80 ka.
If this is interpreted as a warm pre-glacial Pliocene
North Pacific, it may have provided the additional
moisture required to initially build up the northern
hemisphere continental ice sheet. The dramatic drop in
sea surface temperatures (SST`7.5 7C) at 2.75 Ma
ended this suggested period of enhanced SST and thus
the proposed moisture pump. Moreover, at 2.79 and
2.73 Ma opal mass accumulation rates (MAR) decrease
in two steps by five fold and is accompanied by a more
gradual long-term decrease in CaCO3 MARs. Evidence
from the Southern Ocean (ODP site 704) indicates that
just prior to 2.6 Ma there is a massive increase in opal
MARs, the opposite to what is found in the North Pacific. This indicates that the intensification of northern
hemisphere glaciation was accompanied by a major
reorganisation of global oceanic chemical budget, posM. A. Maslin (Y) 1 7 G. H. Haug 7 M. Sarnthein
Geologisch-Paläontologisches Institut, Universität Kiel,
Olshausenstrasse 40, D-24098 Kiel, Germany
R. Tiedemann
GEOMAR Forschungszentrum für marine Geowissenschaften,
Wischhofstrasse 1-3, D-24148 Kiel, Germany
Present address:
1
Environmental Change Research Centre,
Department of Geography, University College London,
26 Bedford Way, London WC1H 0AP, UK
Fax: c44 171 380 7565
e-mail: MMaslin6geog.ucl.ac.uk
sibly caused by changes in deep ocean circulation. The
initiation of northern hemisphere glaciation occurred in
the late Miocene with a significant build up of ice on
southern Greenland. However, the progressive intensification did not occur until 3.5–3 Ma when the Greenland ice sheet expanded to include northern Greenland. Following this stage we suggest that the Eurasian
Arctic and Northeast Asia glaciated at 2.75 Ma, approximately 100 ka before the glaciation of Alaska
(2.65 Ma) and 200 ka before the glaciation of the North
East American continent (2.54 Ma).
Key words Paleoceanography 7 North Pacific 7
Ocean Drilling Project 7 Northern hemisphere
glaciation 7 Pliocene 7 Surface water productivity 7
Sea surface temperatures 7 Pacific moisture pump
Introduction
Ocean Drilling Project (ODP) site 882 in the North Pacific Ocean is located at the northern end of the Emperor Seamount Chain, on the western flank of the Detroit Seamount (50722bN, 167736bE) in a water depth of
3244 m (see Fig. 1; Rea et al. 1993; ODP Leg 145 1993).
This site provides the first high-resolution carbonate,
opal and foraminifera stable isotopes records between
3.2 and 2.4 Ma in the North Pacific. The sedimentation
rate at this site varied from 12 to 4 cm/1000 year between 3.2 and 2.4 Ma, with a major drop in the sedimentation rate occurring near 2.75 Ma (Tiedemann and
Haug 1995). These sedimentation rates are sufficient,
with the sampling density of 30 cm, to depict the oceanographic variability in the range of Milankovitch orbital forcing and to compare results with other high-resolution ODP records. High-resolution proxy records of
GRAPE density, magnetic susceptibility, CaCO3 and
biogenic opal for the past 6 Ma in conjunction with
magnetostratigraphy (Rea et al. 1993) have permitted
an astronomical age calibration (Tiedemann and Haug
1995) of the sediment record at site 882 and the determination of fluxes. This age model has been confirmed
453
water circulation and is also the terminus of the deepwater salinity conveyor belt (Stommel 1961; Broecker
et al. 1985) where deep water diffuses upward to the
surface (Craig et al. 1981; Broecker et al. 1988). This
ascending cold deep water is highly enriched in nutrients (Craig et al. 1981) and thus stimulates high surface-water productivity (Koblentz-Mishke et al. 1970,
1989). Intermediate water formation in the sub-Arctic
North Pacific is caused by the vertical mixing through
the pycnocline, especially in regions with high precipitation (Reid 1965), and this can enhance surface-water
productivity by efficient mixing and recycling of nutrients. The formation of in situ intermediate waters in
the North Pacific is also consistent with the known
deeper circulation of the Pacific which involves upwelling in subarctic regions, rather than overturn at intermediate depths (Reid 1965; Moriyasu 1972; Craig et al.
1981). However, Zahn et al. (1991a, b) suggest that intermediate water in the North Pacific could be formed
in the peripheral seas, i.e. the Sea of Japan and the Sea
of Okhotsk.
Observations and theories about the initiation of
northern hemisphere glaciation
Fig. 1 Major surface water currents of the North Pacific and the
location of site 882 (50721.79bN, 167735.99bE) and the other ODP
Leg 145 sites
largely by benthic and planktonic stable isotopes for
both the first 750 ka (Haug et al. 1995) and the period
between 3.2 and 2.4 Ma (Maslin et al. 1995a).
Modern oceanographic research has demonstrated
that the Northwest Pacific has a complicated surface
The glaciation of Greenland and the Arctic is believed
to have begun in the late Miocene (ODP Leg 151 and
152 1994). There is also strong evidence for the formation of sea ice and snow cover in the Arctic and subArctic regions in the late Miocene–early Pliocene
(Jansen et al. 1990; Jansen and Sjöholm 1991). The first
occurrence, however, of significant ice sheets in the
Northern Hemisphere was according to Shackleton et
al. (1984) at 2.55 Ma (using the new time scale of
Shackleton et al. 1990 and Hilgen 1991), although preceded by smaller increases in ice volume from 2.7 Ma
on (Backman 1979; Shackleton et al. 1984; Zimmerman
et al. 1985). These conclusions were based primarily on
relative coarse resolution records of %CaCO3 and d 18O
records from Deep Sea Drilling Project (DSDP) site
552 in the Atlantic Ocean (Shackleton et al. 1984). The
drop in %CaCO3 was explained by an increase in icerafted debris coeval with the drop in d 18O which indicates an increase in global ice volume (Shackleton et al.
1984).
Recent evidence, however, suggests that the initiation of major northern hemisphere glaciation was the
culmination of a longer-term high-latitude cooling
which began at 3.2 Ma (Ruddiman et al. 1986b; Tiedemann et al. 1994). Progressive and oscillatory enrichment of benthic foraminifera d 18O suggests that there
was significant deep-water cooling and an increase in
global ice volume after 3.2 Ma (Prell 1984; Hodell et al.
1987; Keigwin 1986; Ruddiman et al. 1986a, b; Sarnthein and Tiedemann 1989; Tiedemann et al. 1994).
There is strong evidence for this progressive cooling
of the northern hemisphere from the increase in the
percentage of cooler living planktonic foraminifera
454
(Loubere and Moss 1986; Raymo et al. 1986) and by
decreasing discoaster abundances (Backman and Pestiaux 1986). Sea-surface temperatures (SSTs), however,
at this time interval are not reconstructed easily because the assemblages are not analogous to modern
faunas (Ruddiman and Raymo 1988). Chapman (1993,
1995) has demonstrated, however, using evolutionary
equivalent species, that it is possible to reconstruct
Pliocene sub-tropical SSTs. His work on ODP site 659
has shown a major cooling of the tropical eastern Atlantic at 2.55 Ma.
The most recent benthic foraminifera d 18O results
from ODP site 659 in the tropical East Atlantic (Tiedemann et al. 1994) and site 846 in the equatorial East
Pacific (Shackleton et al. 1995), which show a gradual
enrichment of 18O of the ocean between 3.2 and
2.75 Ma. From 2.75 to 2.68 Ma there are three successive intervals with very heavy d 18O values, with minimal recovery in between. Subsequently, there were still
further increases in the average d 18O value and a
marked increase in the amplitude variation between
warm and colder periods. The major controversy at
present is how much of these d 18O shifts represent
cooling of the deep waters and how much represents
the first buildup of ice in the northern hemisphere. The
three major increases in benthic d 18O between 2.75 and
2.68 Ma, it is believed, were due primarily to a significant increase in global ice volume (Shackleton et al.
1995; Tiedemann et al. 1994).
Sarnthein and Tiedemann (1989), Tiedemann (1991)
and Raymo et al. (1992) demonstrated from benthic
d 13C and d 18O of a number of ODP cores in both the
Atlantic and Pacific Ocean, that the gradual global
cooling between 3 and 2 Ma was associated with an increased suppression of the North Atlantic Deep Water
(NADW) formation. They also suggested that the relative strength of NADW production during the late Pliocene was always greater than that observed during the
late Pleistocene glaciations. High-resolution benthic
isotopic data of Shackleton et al. (1995) and Tiedemann (pers. commun.) from the tropical Pacific and Atlantic demonstrate an in-phase relationship between
d 18O and d 13C during the Pliocene; thus, during the
colder periods of the Pliocene there is a marked increase in the suppression of deep-water formation, analogous to the late Quaternary glacial--interglacial contrast.
Several explanations have been put forward to explain the initiation of northern hemisphere glaciation.
One group of theories suggests changes in atmospheric
composition (i.e. CO2) or a change in total solar radiation. Suggestions such as changes in solar radiation are
not testable (Opik 1959), whereas changes in atmospheric CO2 content could be detected in the geological
record (Sarnthein and Fenner 1988). Increased volcanism during the latest Cenozoic (Kennett and Thunell
1975) has also been suggested as a possible cause of glaciation. It is now believed by some researchers that the
onset of glaciation may have caused the observed in-
crease in northern hemisphere volcanism (e.g. Rea et
al. 1995). Other theories include: virtual polar wander
(Ewing and Donn 1956; Schneider and Kent 1986),
uplift of the highlands of northern Canada (Flint 1957;
Emiliani and Geiss 1958; Birchfield et al. 1982) and
changes in land-sea distribution by sea floor spreading
(North et al. 1983). These theories are either too negligible in affect or too long term to have caused the sudden initiation of northern hemisphere glaciation.
Tectonic explanations have also been suggested
(Hay 1992), such as the emergence of the Panama Isthmus (Keigwin 1978, 1982; Keller et al. 1989; Mann and
Corrigan 1990) and the deepening of the Bering Straits
(Einarsson et al. 1967). The most recent dating of the
closure of the Pacific–Caribbean gateway (Keller et al.
1989) suggests that the Panama Isthmus began to
emerge gradually at 6.2 Ma and finally closed at 1.8 Ma.
Keller et al. (1989) also documented four major events
in the progressive closure of the Pacific–Caribbean gateway 6.2, 4.2, 2.4 (2.55 with new time scale) and
1.8 Ma. Keller et al. (1989) showed that there was an
increasing abundance of salinity-tolerant planktonic foraminifera in the Caribbean from 2.55 Ma onwards suggesting that the restriction of water flow between the
Pacific and Caribbean started at 2.55 Ma and finally
ceased completely at 1.8 Ma. However, progressive and
gradual closure from 2.55 Ma onwards is too late to
have been the cause of the intensification of northern
hemisphere glaciation. Possibly the lower sea level contributed to a more rapid closure of the Panama Isthmus. The reduced inflow of Pacific surface water to the
Caribbean increased the salinity of the Caribbean. This
would have increased both the salinity and strength of
the Gulf Stream, thus enhancing deep-water formation
(Mikolajewicz et al. 1993). Increased deep-water formation could have worked against the initiation of
northern hemisphere glaciation because greater heat
transport to the high latitudes would have tended to
prevent ice-sheet formation. A contrasting argument is
that the enhanced Gulf Stream could have pumped
moisture north, stimulating the formation of ice sheets
(Mikolajewicz et al. 1993). If the closure of the Panama
gateway did increase the strength of the Gulf Stream,
this should have in turn increased deep-water ventilation. However, the benthic d 13C record from site 659
over the past 5 Ma (Tiedemann 1991) does not show
any long-term increase in the d 13C between 3.5 and
2.0 Ma nor a step-like increase at approximately
2.55 Ma which might occur with a more abrupt closure
of the Panama gateway. Moreover, comparison with
the global benthic isotope records (Shackleton et al.
1995; Tiedemann et al. 1994) shows that there is a
strong warm period with a suppressed cool period at
2.6–2.5 Ma. Therefore, at present the evidence seems
ambiguous as to whether or not Panama Isthmus formation enhanced the glaciation of the northern hemisphere.
The tectonic changes of the Bering Sea were initially
dated between 3.5 and 3 Ma (Einarsson et al. 1967),
455
and more recently the submergence of the Bering Strait
has been dated at 3.2 Ma (Fyles et al. 1991), which is
too early to have caused the dramatic changes near
2.7 Ma, but they may have indeed contributed to the
long-term global cooling which started at approximately 3.2 Ma.
Ruddiman and Raymo (1988), Ruddiman et al.
(1989), and Ruddiman and Kutzbach (1991) have discussed how the initiation of northern hemisphere glaciation could have been caused by progressive uplift of
the Tibetan--Himalayan and Sierran–Coloradan regions. They have suggested that this uplift altered the
circulation of the atmospheric planetary waves such
that ablation was decreased, which allowed snow and
ice to build up in the northern hemisphere. Discussion
is ongoing as to whether orography (Charney and
Eliasson 1949; Bolin 1950), differential heating of land
and sea surfaces (Sutcliffe 1951; Smagorinsky 1953), or
a combination of both (Trenberth 1983), control the
structure and direction of the planetary waves. In contrast, Copeland et al. (1987) and Molnar and England
(1990) have suggested that the majority of the Himalayan uplift occurred much earlier between 20 and
17 Ma, whereas Harrison et al. (1992), from thermochronological results, have suggested that the Tibetan
Plateau reached its maximum elevation during the late
Miocene at approximately 7–8 Ma, and not during the
mid-Pliocene as suggested by paleobotany studies
(Mercier et al. 1987). Quade et al. (1989) also showed
from carbon and oxygen isotopes of a pedogenic sediment profile in Pakistan that the Asian monsoons underwent a strong intensification at 7.4 Ma, which they
present as further evidence of a late Miocene uplift,
thus invalidating the hypothesis of Ruddiman and Raymo (1988).
Raymo et al. (1988), Raymo (1991) and Raymo and
Ruddiman (1992) have taken the Himalayan debate
one stage further, suggesting there was a massive increase in tectonically driven chemical weathering in the
late Cenozoic. In two key areas, Tibet and southwest
North America, rapid uplift can be linked to increased
chemical weathering of the predominate silica rocks to
carbonates by extracting atmospheric CO2. This lows
atmospheric CO2, because the carbonates are transported into the oceans were they are taken up by marine organisms and deposited in deep sea sediments.
Raymo et al. (1988), Raymo (1991) and Raymo and
Ruddiman (1992) thus argue that lower atmospheric
CO2 would have caused a cooling of the global climate
and thus the onset of northern hemisphere glaciation.
Again, the recent earlier dates for the Himalayan uplift
suggest that if they are correct, then the Raymo model
was a cause for the long-term cooling of the late Cenozoic and thus a precursor to the onset of northern hemisphere glaciation. It cannot explain, however, the sudden step-like nature of the intensification of northern
hemisphere glaciation at approximately 2.7–2.5 Ma, unless we invoke a long-term nonlinear amplification.
A recent suggestion is that changes in orbital forcing
may have been an important mechanism controlling the
gradual global cooling and the subsequent rapid intensification of northern hemisphere glaciation (Lourens
and Hilgen 1994; Maslin et al. 1995a). This suggestion
is an extension of the different periods of orbital forcing identified for the Pleistocene and late Pliocene by
Berger et al. (1993). Maslin et al. (1995a) suggested that
the observed increase in the amplitude variation of orbital obliquity, from 3.2 Ma onwards, may have increased the seasonality of the northern hemisphere,
thus initiating the long-term global cooling. The subsequent sharp rise in the amplitude of precession, and
thus insolation between 2.8 and 2.55 Ma, may have
forced the rapid glaciation of the northern hemisphere.
Essentially, however, the causes of the long-term global
cooling which started at 3.2 Ma and the significant increase in northern hemisphere glaciation near 2.75 Ma
are still unresolved.
The aim of this study was to investigate high-resolution Pliocene proxy records from the North Pacific to
see how it responded to the onset and progressive intensification of northern hemisphere glaciation. Moreover, we wish to present a picture of when each of the
separate northern hemisphere ice sheets first formed.
Methods
The calcium carbonate contents were determined by
the coulometric technique using a Coulomat 702 at Kiel
University. The biogenic opal content was separated
from the siliciclastic fraction in the carbonate-free fraction `2 mm by the density-separation technique (Bohrmann et al. 1990) using sodium-polywolframate as the
heavy-liquid solution. The carbonate-free clay fraction
was separated by means of the Atterberg separation
method (Müller 1967).
Mass accumulation rates (MAR; grams per square
centimetre per kyr) were calculated as the product of
sedimentation rates (centimetres per 1000 years), percentages of the individual compounds and dry-bulk
density values (grams per cubic centimetres; Rea et al.
1993). Mean sedimentation rates were interpolated linearly between age control points (Tiedemann and
Haug, in press), because the use of accumulation rates
removes the dilution effects by different sediment components which occur in the closed percentage system.
The d 18O and d 13C of the benthic and planktonic foraminifera were measured according to the standard
techniques at the University of Kiel (Maslin et al.
1995a). Three planktonic foraminifera species were
analysed, G. bulloides, N. pachyderma (r) and N. pachyderma (l). G. bulloides and N. pachyderma (r) occurred with a sufficient time resolution to have nearcontinuous records between 3.4 and 2.4 Ma, whereas N.
pachyderma (l) occurred sporadically and was used
only to confirm the pattern of the other two species.
To obtain complete and continuous proxy records
for the past 4.0 Ma, we adjusted the value measured at
456
each sampling depth to the composite depth record of
holes 882A and 882B. The two records were spliced using correlations between the magnetic susceptibility
and GRAPE records of holes 882A and 882B. This
composite depth scale was developed only for the upper 255.95 m (hole 882A), which is equivalent to a composite depth of 287.78 mbsf. Below this the composite
depth section remains uncertain because of only short
overlapping sections between holes 882A and 882B,
which ends at 270.4 mbsf (Rea et al. 1993).
GRAPE-density oscillations in the precession band to
the summer insolation at 657N (Berger and Loutre
1991). This calibration was based on the assumption
that GRAPE-density maxima are linked to minima in
biogenic opal and maxima in ice-rafted debris, and
hence to glacial stages (insolation minima) during the
past 2.75 Ma. Prior to 2.75 Ma, the relationship between the GRAPE density and climate was reversed,
which is confirmed by the benthic d 18O results from site
882. A detailed description of the astronomically calibrated age model is given by Tiedemann and Haug
(1995).
Age model
The time scale for sites 882 and 887 were based initially
on magnetostratigraphy. The ages for the magnetic reversal boundaries were derived from the orbitally
tuned time scale of Shackleton et al. (1990), Shackleton
et al. (1995) and Tiedemann et al. (1994), which modified Hilgen (1991). Using the astronomically dated magnetic reversals for initial age control, we found that the
fluctuations in the GRAPE density and magnetic susceptibility records were linked to variations in the
Earth’s orbit, as expected. Tiedemann and Haug (1995)
generated an astronomically calibrated stratigraphy for
site 882 for the past 4 Ma based on fine-tuning the
Fig. 2 Comparison of the
magnetic susceptibility, weight
percent opal (Haug 1995),
and weight percent calcium
carbonate of ODP site 882
with the benthic foraminifera
d 18O of site 659 (Tiedemann
et al. 1994) for the past 6 Ma
Results
In general, MAR of CaCO3 and opal have a pattern
similar to their weight percent (Haug et al. 1995; Haug
1995). The following description uses MAR of CaCO3
and opal (see Fig. 3). The carbonate MAR at site 882
for the past 5.9 Ma vary between 0 and 3 g/cm 2 per kyr
(Fig. 3). There are major cyclic variations in amplitude,
and many intervals with no carbonate preserved. In
general, Pleistocene interglacial periods and Pliocene
warm stages show good carbonate preservation (Fig. 2),
which is comparable to the pattern of carbonate preservation in the Atlantic.
457
Fig. 3 Comparison of biogenic
opal mass accumulation rates
(MAR), calcium carbonate
MAR and d 13C of G. bulloides and N. pachyderma (R
and L) as proxies for surface–
water paleoproductivity, and
magnetic susceptibility as an
indicator for terrigenous input
which is primarily ice-rafted
detritus at site 882 for the
time interval 2.4–3.2 Ma. The
shaded region indicates the
large increase in ice-rafted debris at site 882 (magnetic susceptibility), the dramatic drop
in opal MAR and the
GRAPE-density record (not
shown)
During the key period 3.2 to 2.4 Ma (Figs. 2 and 3)
there were large changes in the MAR of calcium carbonate. At least between 3.2 and 2.82 Ma the carbonate
MAR varies cyclically at approximately 41 ka (Tiedemann and Haug 1995), between high accumulation
rates (up to 4 g/cm 2 per kyr) and periods of very low or
no accumulation. The size of these variations decreases
gradually until 2.82 Ma, where subsequently there is
very little or no calcium carbonate accumulation for approximately 500 ka. This decrease in CaCO3 MAR parallels the gradual global cooling indicated by the benthic d 18O (Shackleton et al. 1995; Tiedemann et al.
1994). Comparison of the calcium carbonate MAR
peaks of site 882 with the benthic d 18O from site 846
indicates that they occur mostly during warm periods of
the pre-glacial Pliocene (Fig. 4). The only CaCO3 MAR
peaks which seem to be partially out of phase with the
global benthic d 18O and d 13C of site 846 (Shackleton et
al. 1995; Tiedemann et al. 1994) are at 2.91 and
2.97 Ma. If the CaCO3 MAR record is compared with
the site 882 Pliocene stable isotope records of C. wuellerstorfi, then the peaks correlate to light d 18O and heavy d 13C values (Maslin et al. 1995a). This suggests that
the CaCO3 MAR peaks at 2.91 and 2.97 Ma actually do
occur during warmer periods, and that there are minor
offsets in the tuned age model. The occurrence of peaks
in carbonate preservation during warm periods with
heavy benthic d 13C supports the link between enhanced
Atlantic deep-water formation and calcium carbonate
accumulation in the Northwest Pacific during the Pliocene, as first suggested by Haug et al. (1995). Between
2.72 and 2.65 Ma the carbonate accumulation increases
weakly to 0.4 g/cm 2 per kyr; subsequently, there are
only a few scattered smaller peaks until 1.7 Ma, after
that they follow the interglacial--glacial cycles (Haug et
al. 1995).
Despite the low concentrations of calcium carbonate
in large sections of the core planktonic foraminifera,
stable isotopes were obtained for site 882 between 3.1
and 2.5 Ma. The isotope records of N. pachyderma (r)
and G. bulloides are the most complete and have a high
degree of similarity in both their oxygen and carbon records. These records are in part confirmed by the data
from N. pachyderma (l), suggesting that these are a true
record of the surface-water conditions of the Northwest
Pacific, and not an individual species effect.
The planktonic d 18O records indicates cyclic, but
very light, values compared with the Holocene (Haug
et al. 1995), culminating near 2.75 Ma. At 2.75 Ma
there is a drop of at least 2.6‰ indicated by all three
458
Fig. 4 Comparison between
the calculatedinsolation for
657N (Berger 1984; Berger
and Loutre 1988), benthic
d 18O and d 13C isotope records
from the equatorial Pacific at
site 846 (Shackleton et al.
1995), the d 18O records from
N. pachyderma (R) and G.
bulloides and CaCO3 MAR at
site 882 for the time interval
2.4–3.2 Ma. The shaded regions represent global cold periods and are labelled according to the nomenclature of
Shackleton et al. (1995) and
Tiedemann et al. (1994)
species to much heavier values. This is due presumably
to a massive drop in SST as at approximately 2.75 Ma,
the GRAPE density and magnetic susceptibility indicate a significant increase in ice-rafted debris and thus
iceberg flux to site 882. The planktonic foraminifera
d 13C records also show a high degree of similarity with
the CaCO3 MAR record (see Fig. 3). Heavier planktonic d 13C values are associated with preservation
peaks in CaCO3. This may support the idea that the
CaCO3 MAR is controlled by surface-water productivity (Haug et al., 1995), if the increases in productivity,
which preferentially removes 12C, are indeed recorded
by the planktonic foraminifera (Ganssen and Sarnthein
1983).
Biogenic opal MARs were continuously high (3–6 g/
cm 2 per kyr) from before 3.2 to 2.725 Ma. There is an
irreversible twofold drop (by approximately a factor of
5) in both the opal MAR and weight percent records at
2.78 Ma (Fig. 3) and between 2.740 and 2.725 Ma
(Figs. 2 and 3), and this causes the dramatic drop in sedimentation rates from 13 to 4 cm/1000 years observed
at approximately 2.75 Ma (Tiedemann and Haug
1995). This is followed by very low opal MAR values
(~0.5 g/cm 2 per kyr) until 2.4 Ma. Between 2.4 and
2.25 Ma there are some minor cyclic peaks in the opal
MAR up to 1 g/cm 2 per kyr. From 2.25 and 1.6 Ma
there is little or no opal accumulation. Low but distinct
interglacial-to-glacial cycles (0.1–1.5 g/cm 2 per kyr) occur between 1.6 Ma until the present, with high opal
values linked to interglacial periods (Haug et al.
1995).
Discussion
Opal variations at site 882 between 3.2 and 2.4 Ma
The major change in the paleoceanography of the
Northwest Pacific culminated near 2.75 Ma, when there
was a dramatic increase in the amount of ice-rafted debris at site 882 (Fig. 3). This dramatic increase documented by the GRAPE density and magnetic susceptibility is accompanied by longer-term gradual decrease
459
in calcium carbonate MAR and a massive and irreversible drop in opal MAR.
The dramatic drop in the opal MAR is coeval with a
strong planktonic d 18O maximum, i.e. much colder seasurface temperature, suggesting a major re-organisation
of the surface-water and deep-water circulation. Various of the suggestions put forward to explain the
North Pacific-wide drop in opal MAR comprise (a)
changes in the surface-water circulation of the North
Pacific, (b) reduction in the upwelling of nutrient deep
water and there for reduced surface water productivity,
(c) changes in species of diatoms from elongated matforming to single individuals and (d) a re-organisation
of the global ocean silica chemical budget.
These concepts are discussed more extensively in
Maslin et al. (1995a). The general consensus is that the
reasons for the “order of magnitude” reduction of biogenic material reaching the sediment of the North Pacific, due primarily to the abrupt end of the high biogenic opal MAR which had lasted at least 1 million
years, is still unresolved. What is interesting to note is
that at the same time as the North Pacific underwent
this dramatic reduction, the biogenic flux to the Southern Ocean increased by an order of magnitude. Hodell
and Ciesielski (1990) state that at ODP site 704A in the
Southern Ocean across the Gauss/Matuyama boundary
(2.6 Ma) sedimentation rates increased from 1.8 cm/kyr
to 14.3 cm/kyr, and that this was due primarily to a massive increase in the surface-water productivity.
Calcium carbonate variations at site 882 between 3.2
and 2.4 Ma
Haug et al. (1995) have shown that high calcium carbonate accumulation rates at site 882 during the Pleistocene have generally co-varied with warm stages in the
interglacial–glacial cycles (see Fig. 2). This pattern is
similar to that found in the Atlantic (Olausson 1965;
Ruddiman et al. 1989), but is opposite to that found in
most of the rest of the Pacific Ocean (Hebbeln et al.
1990; Farrell and Prell 1991; Zahn et al. 1991b; Karlin
et al. 1992). The Atlantic-type pattern of calcium carbonate accumulation in the Northwest Pacific has also
been observed by Hovan et al. (1991) and Keigwin et
al. (1992). We suggest that the Pliocene record of site
882 indicates that increased calcium carbonate accumulation is associated with increased surface-water productivity, shown by the heavy planktonic foraminifera
d 13C, during warmer periods (see Fig. 3). This is because during interglacials or warmer periods in the
Pliocene North Atlantic deep-water formation was
stronger and thus there was strong upward diffusion of
nutrient-rich deep water in the Northwest Pacific, at the
terminus of the deep-water conveyer belt. This stimulated strong surface-water productivity, like the present, and was sufficiently large to overcome the strong
dissolution of the highly corrosive Pacific deep waters,
thereby allowing calcium carbonate to accumulate. In
contrast, the glacial periods or Pliocene colder periods
had greatly reduced deep-water formation, which
would suggest that the deep-water upwelling in the
North Pacific would have been much older and, due to
enhanced CO2 liberation, there would have been more
carbonate aggressivity. Moreover, reduced deep-water
circulation would also lead to reduced upwelling and
thus surface-water productivity in the Northwest Pacific. Deep-water dissolution therefore due to reduced
productivity would, during cold periods, be that dominant factor; thus, there is little or no calcium carbonate
accumulation in the glacial periods. This extension of
the Haug et al. (1995) theory has been taken further by
Rea et al. (1995) who have conjectured that variations
in deep-water upwelling and associated nutrients may
well help to explain the paleoceanography of the
Northwest Pacific as far back as the Oligocene.
Sea surface temperatures between 3.2 and 2.4 Ma and
the North Pacific moisture pump
The confidence in our age model between 4 and 2.4 Ma
is based first on the magnetic tie points of the Matuyama/Gauss boundary at 2.6 Ma and the Gauss/Gilbert at
3.6 Ma (Rea et al. 1993; Tiedemann and Haug 1995).
Secondly, both the two magnetic reversals and the lowresolution benthic d 18O at site 882 (Maslin et al., in
press) suggest that GRAPE density and carbonate
maxima are related to warm periods during the pre-glacial Pliocene, and this has facilitated the calibration of a
detailed orbital tuning age model (Tiedemann and
Haug 1995). Using the site 882 age model it is possible
to compare site 882 planktonic d 18O with the global
benthic C. wuellerstorfid 18O and d 13C records of the
Equatorial Pacific site 846 (Shackleton et al. 1995). Figure 4 indicates a high degree of similarity between the
two records, the most significant of which occurs between 2.7 and 2.9 Ma (Stage G14 to G4, using the nomenclature of Tiedemann et al. 1994). Between 2.85 and
2.75 Ma the planktonic foraminifera d 18O values show
a trend, overlying the warm–cold period variations, towards lighter values (`1‰ ca. 4 7C) and obtains its
lightest value at 2.76 Ma, indicating a very high surfacewater temperature or low salinity, or a combination of
both. Subsequently, at 2.75 Ma coeval with the onset of
the G6 glaciation, significant ice rafting occurred in the
Northwest Pacific Ocean and the planktonic foraminifera d 18O drops dramatically by 2.6‰.
If the major changes in the planktonic foraminifera
d 18O at site 882 between 2.85 and 2.65 Ma are assumed
to be only temperature driven (i.e. ignoring the ice volume and salinity effects) the sea-surface temperature
change between cold to warm periods was approximately 6.5 7C before 2.75 Ma with an overall increase
trend of 4 7C and a dramatic drop of no less than 10 7C
at 2.75 Ma (using the SST d 18O equation of Shackleton
1974).
460
If the global change in ocean d 18O, according to the
benthic isotopes of site 846, are taken into account,
then this would decrease some of the estimated temperature changes at site 882. If we assume that all the variation of the site 864 benthic d 18O record is due to ice
volume (ignoring possible deep-water cooling) the
North Pacific SST estimates must be reduced by approximately 2.5 7C (0.6‰) before 2.75 Ma and the estimated drop at 2.75 Ma reduced by 3.5 7C (0.8‰). In
contrast, the possible salinity changes at site 882 during
the pre-glacial Pliocene are much harder to quantify.
During globally cooler periods the following processes
would have lowered surface-water salinity:
1. Reduced upwelling of saline deep water – approximately a 0.5–1‰ reduction in salinity (Moriyasu
1972; Craig et al. 1981)
2. Increased influence of the cold Oyashio current and
thus the reduction of the influence of the warmer
Kuroshio current – approximately a 0.5-1‰ reduction in salinity (Craig et al. 1981)
3. Increased precipitation (unknown effect)
4. Increase in ice rafting after 2.75 Ma (unknown effect, but ice rafting in the North Atlantic during the
last glacial maximum (LGM) caused a drop in salinity between 1 and 2 ‰; Duplessy et al. 1991, 1992;
Maslin 1993; Maslin et al. 1995b)
The following would have raised surface-water salinity:
1. Increase in global ice volume [between 0.1 and
0.5‰, which is approximately equivalent to a 10- to
50-m drop in sea level (Fairbanks 1989)]
2. Reduced evaporation due to lower SST (unknown
effect).
If it is assumed that during a cold stage the effect of
the decreased evaporation is countered by the increased precipitation, we suggest that the possible salinity changes from warm to cool periods at site 882 could
have been between 0.5 and more than 2‰. This represents a minimum oxygen isotope shift in the surface water of approximately 0.25‰, which increases the estimated SST change by at least 1 7C (Duplessy et al. 1991;
1992; Maslin 1993; Maslin et al. 1995b).
In summary, considering the possible influences on
the planktonic foraminifera d 18O record, the SST variation between warm and cold periods before 2.75 Ma
would have been at least 5 7C. Whereas the SST drop at
2.75 Ma was no less than 7.5 7C. This estimate could be
much larger if meltwater from ice rafting was a major
influence on the surface waters. Unfortunately, it is not
possible to determine the cause of the trend between
2.85 and 2.75 Ma in the planktonic foraminifera d 18O
record; it could represent a major warming of a maximum of 4 7C, a significant freshening of the Northwest
Pacific (ca. 0.5–1‰ change in salinity) or a combination
of both.
The CLIMAP Project Members (1976, 1981) have
shown that the SST during the LGM compared with
the present in the Northwest Pacific dropped by a fairly
minor 1–2 7C. The SST variations estimated at site 882
for the late Pliocene are more comparable to the LGM
anomalies in medium to high latitudes, e.g. off the East
of Japan (–8 7C) and at the same latitude (507N) in the
North Atlantic (-10 and -12 7C). If these temperature
shifts are compared with those of the late Pliocene at
site 882, it seems that the late Pliocene Northwest Pacific was much more thermodynamic than during the
last glacial--interglacial cycle. We suggest that the very
high Northwest Pacific SSTs and the possible observed
warming trend prior to 2.75 Ma would have produced
enhanced evaporation and thus a mechanism to transport additional moisture to the growing northern hemisphere continental ice sheets. This, however, would
have been a self-destructive feedback, because as soon
as the ice sheets were big enough to calve into the
North Pacific at 2.75 Ma the SST were dramatically reduced, cut the enhanced moisture pump. The major
SST drop after 2.75 Ma recorded at site 882 may support the suggestion that there could have been at
2.75 Ma a shift from a Kuroshio-current-dominated
North Pacific to a more contemporary system with a
sub-polar gyre with a strong Oyashio current and Alaskan stream.
Timing of the initiation and intensification of northern
hemisphere glaciation
The earliest recorded onset of continental wide glaciation in Antarctica is approximately 35 Ma (Shackleton
et al. 1984; Ruddiman and Raymo 1988). In contrast, it
has been shown that 6 Ma is the earliest recorded glaciation in the northern hemisphere (ODP Leg 151 1994;
Jansen et al. 1990; Jansen and Sjöholm 1991). This late
Miocene initiation of the northern hemisphere glaciation has been recorded in both the Norwegian Sea
(Jansen et al. 1990; Jansen and Sjöholm 1991) and in
the Arctic (ODP Leg 151 1994). Site 882 magnetic susceptibility record also provides evidence for late Miocene ice rafting in the Northwest Pacific. A major in-
Fig. 5 Timing of the intensification of northern hemisphere gla- P
ciation. Map shows the extent of the northern hemisphere ice
sheets during the last glacial maximum (CLIMAP 1981). The separate ice sheets are identified by colour and number. The present
climatic fronts are also shown to represent the possible atmospheric circulation of the pre-glacial Pliocene. The graph shows
the estimated time at which each ice sheet was large enough to
encroach on adjacent coast and release icebergs compared with
its relative effect on the sea level, estimated from the change in
the global d 18O signal (Tiedemann et al. 1994; Shackleton et al.
1995). Iceland (in orange) seems to have had repeated glacial episodes from the mid-Pliocene onwards, but no clear dates are yet
available for when this became sustained (Einarsson and Albertsson 1988; Geirsdottir and Eiriksson 1994). Southern Greenland
(ODP Leg 151 and 152 1994); Northern Greenland (ODP Leg
151 and 152 1994); Eurasian Arctic (Jansen et al. 1990; Jansen and
Sjöholm 1991) and Northeast Asia (Maslin et al. 1995a and this
study); Alaska and the Northwest coast of America (Rea et al.
1993, and this study); Northeast America (Shackleton et al. 1984;
Ruddiman and Raymo 1988; Raymo, pers. commun.)
461
462
tensification of northern hemisphere glaciation occurred between 3.2 and 2.5 Ma. Evidence from the Norwegian Sea and Iceland suggests that the first minor increase in Arctic, Iceland and Scandinavian glaciation
occurred at 3 Ma, whereas a more pronounced and sustained increase occurred at 2.75 Ma (Jansen et al. 1990;
Jansen and Sjöholm 1991; Einarsson and Albertsson
1988; Geirsdottir and Eiriksson 1994). This is coeval
with the magnetic susceptibility record of site 882,
which demonstrates a minor increase in ice rafting at
3 Ma (see Fig. 4), and a major increase at 2.75 Ma. This
confirms that there was some ice cover in the circumArctic continents from approximately 3 Ma, and this
significantly expanded at 2.75 Ma.
Evidence from northern Alaska (pollen, plant macrofossil and marine vertebrate) suggests that there
were three major marine transgressions in the late Pliocene (Brigham-Grette and Carter 1992; Kaufman and
Brigham-Grette 1993). These were dated using amino
acid geochemistry, paleomagnetic studies, vertebrate
and invertebrate paleontology and strontium isotopes
to between 2.48 and 2.7 Ma (2.62 and 2.86 Ma using the
new time scale), which represents three warm periods
between 104 and G12. They also found that these periods were too warm for permafrost and even seasonal
sea ice. During the waning stages of these transgressions the terrestrial conditions were cool enough to
support a herbaceous tundra vegetation with scattered
larch trees. They found no evidence for any glaciation
of northern Alaska between 2.62 and 2.86 Ma, suggesting that the ice rafting observed at site 882 was predominantly from the Eurasian Arctic (via the Chukchi and
Bering seas) and Northeast Asia. This has been confirmed by McKelvey et al. (1995), who has traced the
major source of Northwest Pacific late Pliocene ice rafting to the Kamchatka peninsula and the northern coast
of the Sea of Okhotsk.
Magnetic susceptibility and GRAPE density records
from site 887 in the Gulf of Alaska (Fig. 1) indicates a
dramatic increase in the supply of ice-rafted debris
(Rea et al. 1993) only one obliquity cycle earlier than
2.6 Ma (2.6 Ma is controlled by the Matuyama/Gauss
boundary). This suggests that Alaska glaciated 100 ka
years after the Eurasian Arctic and Northeast Asia
(Fig. 5).
Evidence from ODP 610 and 609 (Ruddiman and
Raymo 1988; Raymo, pers. commun.) suggest that the
first increase of ice rafting into the North Atlantic occurred at 2.75 coeval with the Eurasian ice sheet. However, evidence from these sites and DSDP site 552
(Shackleton et al. 1984) suggests that a significant
North American ice sheet did not occur until 2.54 Ma,
when there was a significant increase in the ice rafting
debris recorded in the North Atlantic Ocean (oxygen
isotope stage 100). There may have been, therefore, a
delay of 200 kyr between the intensification of the glaciation of the Arctic and Northeast Asia and the major
buildup of ice on the North East America continent
(Fig. 5). This suggests that the initiation and intensifica-
tion of northern hemisphere glaciation was driven by
the Arctic and North Pacific, as opposed to the Laurentide-ice-sheet-dominated scenario of the LGM and
much of the late Pleistocene (Ruddiman and McIntyre
1981).
Conclusions
The following conclusions were reached as a result of
this work:
1. Minor increases in ice rafting have been detected in
the Norwegian Sea and the Northwest Pacific from
3 Ma onwards.
2. A dramatic increase in ice rafting occurred in the
Northwest Pacific and the Norwegian Sea at
2.75 Ma, suggesting that the Eurasian Arctic and
Northeast Asia was significantly glaciated after
2.75 Ma. In the Northwest Pacific this major change
was accompanied by a dramatic drop in SSTs
(`7.5 7C) and opal MARs (fivefold decrease) and is
overlain by a more gradual long-term decrease in
CaCO3 MARs. This dramatic change was proceeded
by 80 ka of unusually warm or fresh Northwest Pacific surface waters which, if interpreted as warmth,
may have provided the moisture for northern hemisphere ice sheet expansion before 2.75 Ma.
3. Data from site 887 in the Northeast Pacific suggest
that Alaska glaciated at 2.65 Ma, 100 ka years after
the Eurasian Arctic and Northeast Asia, whereas
sites 609, 610 and 552 in the Atlantic Ocean indicate
that the Northeast American continent first glaciated at 2.75 Ma, but increased dramatically at
2.54 Ma, approximately 200 ka after the Eurasian
Arctic and Northeast Asia. Hence, the initiation and
intensification of northern hemisphere glaciation
was driven by the Arctic and North Pacific, as opposed to the Laurentide-ice-sheet-dominated scenario of the late Pleistocene.
Acknowledgements The authors are grateful to N. J. Shackleton
for providing the site 846 isotope data and to E. Jansen, M. Chapman and J. Thiede for providing data and insight. We gratefully
acknowledge the cooperation of H. Erlenkeuser and H. Cordt,
who supervised the operation of the spectrometer in Kiel with
special care. We thank E. Heinrich, J. Hennings, and C. Hohnemann for technical assistance, and Catherine Pyke of the Department of Geography, UCL, Drawing Office. We thank the
“Deutsche Forschungsgemeinschaft” for supporting this study.
We thank Torsten Bickert and Wolfgang Berger for their careful
and thoughtful reviews.
References
Backman J (1979) Pliocene biostratigraphy of DSDP Sites 111
and 116 from the North Atlantic Ocean and the age of the
northern hemisphere glaciation. Stockh Contr Geol 32 : 115–
137
463
Backman J, Pestiaux P (1986) Pliocene discoaster abundance variations at DSDP site 606 : biochronology and palaeoenvironmental implications. In: Kidd RB, Ruddiman WF, Thomas E
et al. (eds) Init Rep DSDP, 94, U.S. Government Printing Office, Washington, pp 903–910
Berger WH (1989) Global maps of ocean productivity. In: Berger
WH, Smetacek VS, Wefer G (eds) Productivity of the oceans:
present and past. Dahlem Konferenzen. Wiley, New York, pp
429–455
Berger A, Loutre MF (1991) Insolation values for the climate of
the last 10 million years. Quat Sci Rev 10 : 297–317
Berger WH, Bickert T, Schmidt H, Wefer G (1993) Quaternary
oxygen isotope record of pelagic foraminiferas: site 806, Ontong Java Plateau. In: Berger WH, Kroenke LW, Mayer LA et
al. (eds) Proc ODP Sci Results 130. Ocean Drilling Program,
College Station, Texas, pp 381–395
Birchfield GE, Weertman J, Lunde AT (1982) A model study of
the role of high latitude topography in the climate response to
orbital insolation anomalies. J Atm Sci 39 : 71–87
Bohrmann G, Henrich R, Thiede J (1990) Miocene to Quaternary
paleoceanography in the northern north Atlantic: variability
in carbonate and biogenic opal accumulation. In: Bleil U,
Thiede J (eds) Geological history of the polar oceans: Arctic
versus Antarctic. Kluwer, Netherlands, pp 647–675
Bolin B (1950) On the influence of the earth’s orography on the
general character of the westerlies. Tellus 2 : 184–195
Brigham-Greete J, Carter LD (1992) Pliocene marine transgressions of northern Alaska: circumartic correlation and paleoclimate interpretations. Arctic 45 (1): 74–89
Broecker WS, Peteet DM, Rind D (1985) Does the ocean-atmosphere system have more than one stable mode of operation.
Nature 315 : 21–26
Broecker WS, Ree M, Bonani G, Wolfli W, Oeschger H, Klas M,
Mix A, Curry W (1988) Preliminary estimates for the radiocarbon age of deepwater in the glacial ocean. Paleoceanography 3 : 659–669
Chapman M (1992) Late Pliocene planktonic foraminifera: palaeoceanography and faunal evolution. PhD thesis, University
of East Anglia, England, 361 pp
Chapman M (1995) Surface circulation patterns and late Pliocene
expansion of northern hemisphere ice sheets. Terra Nova abstracts, EUG8, Strasbourg
Charney JG, Eliassen A (1949) A numerical method for predicting the perturbations of the middle-latitude westerlies. Tellus
1 : 38–54
CLIMAP Project Members (1976) The surface of the ice-age
earth. Science 191 : 1131–1137
CLIMAP Project Members (1981) Seasonal reconstructions of
the Earth’s surface at the glacial maximum. Geol Soc Am Map
and Chart Series, MC-36
Copeland P, Harrison TM, Kidd WSF, Ronghua X, Yuquan Z
(1987) Rapid early Miocene acceleration of uplift of the Gagdese Belt, Xizang (southern Tibet), and its bearing on accomodation mechanisms of the India--Asia collision. Earth Planet Sci Lett 86 : 240–252
Craig H, Broecker WS, Spencer D (1981) GEOSECS Pacific Expedition, (vol 4), Sections and profiles. NSF, Washington
Duplessy J-C, Labeyrie L, Arnold M, Paterne M, Duprat J, Weering T van (1992) Changes in surface water salinity of the
North Atlantic Ocean during the last deglaciation. Nature
358 : 485–488
Duplessy J-C, Labeyrie L, Juillet-Leclerc A, Maitre F, Duprat J,
Sarnthein M (1991) Surface salinity reconstruction of the
North Atlantic Ocean during the last glacial maximum. Oceanological Acta 14 (4): 311–324
Einarsson T, Albertsson K. J (1988) Glaciation of Iceland. Phil
Trans R Soc Lond B 318 : 227–234
Einarsson T, Hopkins DM, Doell RR (1967) The stratigraphy of
Tjornes, northern Iceland, and the history of the Bering Land
Bridge. In: Hopkins DM (ed) The Bering Land Bridge, California. Stanford University Press, Stanford, Calif., pp 312–
325
Emiliani C, Geiss J (1958) On glaciations and their causes. Geol
Rundsch 46 : 576–601
Ewing M, Donn WL (1956) A theory of ice ages. Science
123 : 1061–1066
Fairbanks RG (1989) A 17 000 year glacio-eustatic sea level record: influence of glacial melting rates on the Younger Dryas
event and deep-ocean circulation. Nature 342 : 637–642
Farrell JW, Prell WL (1991) Pacific CaCO3 preservation and d 18O
since 4 Ma: paleoceanic and paleoclimatic implications. Paleoceanography 4 : 485–498
Flint RF (1957) Glacial and Pleistocene Geology. Wiley, New
York
Fyles JG, Marincovich L, Matthews JV Jr, Barrendrest R
(1991) Unique mollusc find in the Beaufort formation (Pliocene) Meighen Island, arctic Canada. Curr Res Part B, Geol
Surv Canada Pap 91-B: 105–112
Ganssen G, Sarnthein M (1983) Stable isotope composition of foraminifera: the surface and bottom waters record of coastal
upwelling. In: Suess AE, Thiede J (eds) Coastal upwelling, its
sediment record, part A. Plenum Press, New York
Geirsdottir A, Eiriksson J (1994) Growth of an intermittent ice
sheet in Iceland during the late Pliocene and early Pleistocene. Quat Res 42 : 115–130
Haug G, Maslin MA, Sarnthein M, Tiedemann R, Stax R (1995)
Evolution of Northwest Pacific sedimentation patterns since
6 Ma: site 882. Proc ODP Sci Rep 145. Ocean Drilling Project,
College Station, Texas : 293–314
Haug G (1995) The evolution of Northwest Pacific Ocean over
the last 6 Ma: ODP LEG 145. PhD thesis, University of Kiel,
Kiel, Germany, pp 200
Harrison TM, Copeland P, Kidd WSF, Yin A (1992) Raising Tibet. Science 255 : 663–670
Hay W (1992) The cause of the late Cenozoic northern hemisphere glaciations: a climate change enigma. Terra Nova 4 :
305–311
Hebbeln D, Wefer G, Berger WH (1990) Pleistocene dissolution
fluctuations from apparent depth of deposition in core
ERDC-12/P, west equatorial Pacific. Mar Geol 92 : 165–176
Hilgen FJ (1991) Extension of the astronomically calibrated (polarity) timescale to the Miocene/Pliocene boundary. Earth
Planet Sci Lett 107 : 349–368
Hodell DA, Ciesielski PF (1990) Southern Ocean response to the
intensification of northern hemisphere glaciation at 2.4 Ma.
In: Bleil U, Thiede J (eds) Geological history of the polar
oceans: Arctic versus Antarctic, Kluwer Academic Publishers,
Netherlands, pp 707–728
Hodell DA, Müller DW, Ciesielski PF, Mead GA (1987) Synthesis of oxygen and carbon isotopic results from site 704 : implications for major climatic--geochemical transitions during the
late Neogene. In: Ciesielski PF, Kristofferson Y et al. (eds)
Proc ODP Sci Results 114. Ocean Drilling Program, College
Station, Texas, pp 475–481
Hovan SA, Rea DK, Pisias NG (1991) Late Pleistocene continental climate and oceanic variability recorded in northwest Pacific sediments. Paleoceanography 6 : 349–370
Jansen E, Sjöholm J, Bleil U, Erichsen JA (1990) Neogene and
Pleistocene glaciations in the northern hemisphere and late
Miocene – Pliocene global ice volume fluctuations: evidence
from the Norwegian Sea. In: Bleil U, Thiede J (eds) Geological history of the polar oceans: Arctic versus Antarctic. Kluwer Academic Publishers, Netherlands, pp 677–705
Jansen E, Sjöholm J (1991) Reconstruction of glaciation over the
past 6 Myr from ice-borne deposits in the Norwegian Sea. Nature 349 : 600–603
Karlin R, Lyle M, Zahn R (1992) Carbonate variations in the
northeast Pacific during the late Quaternary. Paleoceanography 7 : 43–61
Kaufman D, Brighamn-Grette J (1993) Aminostratigraphic correlations and paleotemperature implications, Pliocene-Pleistocene high-sea-level deposits, northwestern Alaska. Quat Sci
Rev 12 : 21–33
464
Keigwin LD (1978) Pliocene closing of the Isthmus of Panama,
based on biostratigraphic evidence from nearby Pacific Ocean
and Caribbean cores. Geology 6 : 630–634
Keigwin LD (1982) Pliocene paleoceanography of the Caribbean
and east Pacific: role of panama uplift in late Neogene times.
Science 217 : 350–353
Keigwin LD (1986) Pliocene stable isotope record of DSDP site
606 : sequential events of 18O enrichment beginning at 3.1 Ma.
In: Kidd RB, Ruddiman WF, Thomas E et al. (eds) Init Rep
DSDP, 94, U.S. Govt Printing Office, Washington, pp 911–
920
Keigwin LD, Jones GA, Froelich PN (1992) A 15 000 year paleoenvironmental record from Meiji Seamount, far northwestern Pacific. Earth Planet Sci Lett 111 : 425–440
Keller G, Zenker CE, Stone SM (1989) Late Neogene history of
the Pacific-Caribbean gateway. J South Am Earth Sci 2 : 73–
108
Kennett JP, Thunell RC (1975) Global increase in Quaternary explosive volcanism. Science 187 : 497–503
Koblents-Mishke OI, Volkovinskiy VV, Kabanova YG (1970)
Plankton primary production in the world ocean. In: Wooster
W (ed) Scientific exploration of the South Pacific. National
Academy of Sciences, Washington, pp 183–193
Loubere P, Moss K (1986) Late Cenozoic climatic change and the
onset of northern hemisphere glaciation as recorded in the
northeast Atlantic Bull Geol Soc Am 97 : 818–828
Lourens LJ, Hilgen FJ (1994) Chapter 9 : Long-period orbital variations and their relation to third-order Eustatic cycles and the
onset of major glaciations – 3.0 million years ago. In: Astronomical forcing of Mediterranean climate during the last 5.3
million years. PhD thesis, University of Utrecht, Utrecht, Holland, pp 199–206
McKelvey B, Chen W, Anzulus R (1995) Provenance of PliocenePleistocene ice rafted debris, Leg 145. Proc ODP Sci Rep 145.
Ocean Drilling Project, College Station, Texas : 195–204
Mann P, Corrigan J (1990) Model for late Neogene deformation
in Panama. Geology 18 : 558–562
Maslin MA (1993) A study of the paleoceanography of the N.E.
Atlantic in the late Pleistocene. PhD dissertation, Cambridge
University, UK
Maslin MA, Haug GH, Sarnthein M, Tiedemann R, Erlenkeuser
H, Stax R (1995a) Northwest Pacific site 882 : the initiation of
northern hemisphere glaciation. Proc ODP Sci Rep 145.
Ocean Drilling Project, College Station, Texas : 315–332
Maslin MA, Shackleton N, Pflaumann U (1995b) Surface water
temperature, salinity and density changes in the N.E. Atlantic
during the last 45 000 years: Heinrich events, deep water formation and climatic rebounds. Paleoceanography 10 : 527–544
Mercier JL, Armijo R, Tapponnier P, Carey-Gailhrdis E, Lin T
(1987) Change from late Tertiary compression to Quaternary
extension in southern Tibet during the India–Asia collision.
Tectonics 6 : 275–304
Michel E (1991) L’ocean ou dernier maximum glaciaire: le cycle
du carbone et la circulation. Contraintes isotopiques et modelisation. PhD thesis, Univ. de Paris-Sud, Centre d’Orsay, Orsay France
Mikolajewicz U, Maier-Reimer E, Crowley TJ, Kim KY (1993)
Effect of Drake and Panamanian gateways on the circulation
of an ocean model. Paleoceanography 8 (4): 409–427
Mix AC, Le J, Shackleton NJ (in press) Benthic foraminiferaa
stable isotope stratigraphy of site 846 : 0–1.8 Ma. In: Mayer L,
Pisias N, Janecek T et al. (eds) Proc ODP Sci Results 138.
Ocean Drilling Program, College Station, Texas
Molnar P, England P (1990) Late Cenozoic uplift of mountain
ranges and global climate change: chicken or egg? Nature
346 : 29–34
Moriyasu S (1972) Deep waters in the western North Pacific. In:
Stommel H, Yoshida K (eds) Kuroshio: physical aspects of the
Japan current. University of Washington Press, Seattle, pp
387–409
Müller G (1967) Methods in sedimentary petrology. In: Engelhardt WV Füchtbauer H, Müller G (eds) Sedimentary petrology, part 1. Schweizerbarth’sche Verlagsbuchhandlung, Stuttgart
North GR, Mengel JG, Short DA (1983) Simple energy balance
model resolving the seasons and continents: applications to
the astronomical theory of ice ages. J Geophys Res 88 : 6576–
6586
ODP Leg 145 Scientific Party (1993) Paleoceanography record of
North Pacific quantified. EOS 73 (36): 406–411
ODP Leg 151 Scientific Party (1994) Exploring Arctic history
through scientific drilling. EOS 75 (25): 281–286
Olausson E (1965) Evidence of climatic changes in north Atlantic
deep sea cores, with remarks on isotopic paleotemperature
analysis. Progr Oceanogr 3 : 221–252
Opik E (1959) Climate and the changing sun. Sci Am 198 : 85–92
Quade J, Cerling TE, Bowman JR (1987) Development of Asian
monsoon revealed by marked ecological shift during the latest
Miocene in northern Pakistan. Nature 342 : 163–165
Prell WL (1984) Covariance patterns of foraminifera d 18O: an
evaluation of Pliocene ice-volume changes near 3.2 million
years ago. Science 226 : 692–694
Raymo ME (1991) Geochemical evidence supporting T.C. Chamberlin’s theory of glaciation. Geology 19 : 344–347
Raymo ME, Ruddiman WF (1992) Tectonic forcing of late Cenozoic climate. Nature 359 : 117–122
Raymo ME, Ruddiman WF, Clement BM (1986) Pliocene-Pleistocene paleoceanography of the North Atlantic DSDP site
609. In: Kidd RB, Ruddiman WF, Thomas E et al. (eds) Init
Rep DSDP, 94, Washington, pp 895–902
Raymo ME, Hodell D, Jansen E (1992) Response of deep ocean
circulation to initiation of northern hemipshere glaciation (3–
2 Ma). Paleoceanography 7 : 645-672
Raymo ME, Ruddiman WF, Froelich PN (1988) Influence of late
Cenozoic mountain building on ocean geochemical cycles.
Geology 16 : 649–653
Rea DK, Basov IA, Janecek TR, Palmer-Julson A et al. (1993)
Proc ODP Init Rep 145. Ocean Drilling Project, College Station, Texas
Rea DK, Basov IA, Krissek LA, the Leg 145 Scientific Party
(1995) Scientific results of drilling the North Pacific Transect.
Proc ODP Sci Rep 145. Ocean Drilling Project, College Station, Texas : 577–598
Reid JL (1965) Intermediate waters of the Pacific Ocean. Johns
Hopkins Press, Baltimore
Ruddiman WF, McIntyre (1981) Oceanic mechanisms for amplification of the 23 000-year ice volume cycle. Science 212 : 617–
627
Ruddiman WF, Raymo ME (1988) Northern hemisphere climate
regimes during the past 3 Ma: possible tectonic connections.
Phil Trans R Soc Lond B 318 : 411–430
Ruddiman WF et al. (1989) Late Miocene to Pleistocene evolution of climate in Africa and the low-latitude Atlantic – overview of Leg 108 results. In: Ruddiman WF, Sarnthein M, Baldauf J et al. (eds) Proc ODP Sci Results 108. Ocean Drilling
Program, College Station, Texas, pp 463–487
Ruddiman WF, Kutzbach JE (1991) Plateau uplift and climatic
change. Sci Am 264 : 3, 66–75
Ruddiman WF, McIntyre A, Raymo M (1986a) Paleoenvironmental results from North Atlantic sites 607 and 609. In: Kidd
RB, Ruddiman WF, Thomas E et al. Init Rep DSDP, 94, U.S.
Govt Printing Office Washington, pp 855–878
Ruddiman WF, Raymo M, McIntyre A (1986b) Matuyama
41 000-year cycles: North Atlantic Ocean and northern hemisphere ice sheets. Earth Planet Sci Lett 80 : 117–129
Ruddiman WF, Sarnthein M, Baldauf J et al. (1989) Proc ODP
Sci Results 108. Ocean Drilling Program, College Station,
Texas
Sarnthein M, Fenner J (1988) Global wind-induced change of
deep-sea sediment budgets, new ocean production and CO2
reservoirs ca. 3.3–2.35 Ma BP. Phil Trans R Soc Lond
318 : 487–504
465
Sarnthein M, Tiedemann R (1989) Toward a high-resolution stable isotope stratigraphy of the last 3.4 million years: sites 658
and 659 off northwest Africa. In: Ruddiman WF, Sarnthein M,
Baldauf J et al. (eds) Proc ODP Sci Results 108. Ocean Drilling Program, College Station, Texas, pp 167–187
Sarnthein M, Pflaumann U, Ross R, Tiedemann R, Winn K
(1992) Transfer function to reconstruct ocean paleoproductivity: a comparison. In: Summerhayes CP, Prell WL, Emeis K-C
(eds) Upwelling systems: evolution since the early Miocene.
Geol Soc Spec Publ 64 : 411–427
Schneider DA, Kent DV (1986) Influence of non-dipole field on
determination of Plio-Pleistocene true polar wander. Geophys
Res Lett 13 : 417–474
Shackleton NJ (1974) Attainment of isotopic equilibrium between ocean water and the benthic foraminiferaa genus Uvigerina: isotopic changes in the ocean during the last glacial. In:
Les méthodes quantitatives d’étude des variations due climat
au cours du Pleistocene, report no. 219 for the Colloques Internationaux de Central National de la Recherce Scientifique,
Gif sur Yvette, Paris (CNRS)
Shackleton NJ, Opdyke ND (1973) Oxygen isotope and paleomagnetic stratigraphy of equatorial Pacific core V28-238. Quat
Res 3 : 39–55
Shackleton NJ et al. (1984) Oxygen isotope calibration of the onset of ice-rafting and history of glaciation in the North Atlantic region. Nature 307 : 620–623
Shackleton NJ, Berger A, Peltier WR (1990) An alternative astronomical calibration of the lower Pleistocene time based on
ODP site 677. Trans R Soc Edinburgh Earth Sci 81 : 251–261
Shackleton NJ, Hall MA, Pate D (1995) Pliocene stable isotope
stratigraphy of ODP site 846. In: Mayer L, Pisias N, Janecek T
et al. (eds) Proc ODP Sci Results 138. Ocean Drilling Program, College Station, Texas : 337–355
Smagorinsky J (1953) The dynamical influence of large-scale heat
source and sinks on the quasi-stationary mean rotations of the
atmosphere. Q Jlr Met Soc 79 : 342–366
Stommel H (1961) Thermohaline convection with two stable regimes of flow. Tellus 13 : 224–230
Sutcliffe RC (1951) Mean upper-air contour patterns of the
northern hemisphere: the thermal-synoptic viewpoint. Q Jlr
Met Soc 77 : 435–440
Tiedemann R (1991) Acht Millionen Jahre Klimageschichte von
Nordwest Afrika und Paleaozeanographie des angrenzenden
Atlantiks: Hochaufusende Zeitreihen von ODP sites 658–661.
Ber. 46. Geol-Paleaont Inst, Universtät Kiel, Kiel, Germany)
190 pp
Tiedemann R, Haug G (1995) Astronomical calibration of cycle
stratigraphy for site 882 in the Northwest Pacific. Proc ODP
Sci Rep 145. Ocean Drilling Project, College Station, Texas : 283–292
Tiedemann R, Sarnthein M, Shackleton NJ (1994) Astronomic timescale for the Pliocene Atlantic d 18O and dust flux records
of ODP site 659. Paleoceanography 9 : 619–638
Trenberth KE (1983) Interactions between orographically and
thermally forced planetary waves. J Atm Sci 40 : 1126–1153
Warren BA, Owens WB (1988) Deep currents in the Central subarctic Pacific Ocean. J Phys Oceanogr 18 : 529–551
Zahn R, Pederson TF, Bornhold BD, Mix AC (1991a) Water
mass conversion in the glacial subarctic Pacific (547N, 1487W):
physical constraints and the benthic-planktonic stable isotope
record. Paleoceanography 6 : 543–560
Zahn R, Rushdi A, Pisias NG, Bornhold BD, Blaise B, Karlin R
(1991b) Carbonate deposition and benthic d 13C in the subarctic Pacific: implications for changes of the oceanic carbonate
system during the past 750 000 years. Earth Planet Sci Lett
103 : 116–132
Zimmerman H et al. (1985) History of Plio-Pleistocene climate in
the Northeast Atlantic, Deep Sea Drilling Project Hole 552A.
In: Schnitker D, Roberts D et al. (eds) Init Rep DSDP, 81,
U.S. Govt Printing Office, Washington, pp 861–875