Odekon M. Encyclopedia of paleoclimatology and ancient environments

Подождите немного. Документ загружается.

grains and facies found in warm-water settings. They are both

examples of carbonate shelves that are attached to continental

areas, as opposed to isolated shelves, banks and atolls that form

in offshore settings (for details see Tucker and Wright, 1990).

They differ, however, in that South Florida is in a semi-humid

subtropical setting and the Arabian Gulf is in an arid climate

setting (Figure C29). The climatic setting is seen to have an

overriding control on the nature of the sediments, particularly

those accumulating in the shallow, nearshore areas.

South Florida

South Florida is a shelf with a large, low-lying hinterland

(including the Everglades Swamps) set in a semi-humid, sub-

tropical climate fringed by coastlines stabilized by mangroves

(Figure C31a). Warm-water carbonate sediments accumulate

throughout this shelf area and deep boreholes in the region

show that this has been the case since Jurassic times. The plat-

form comprises four main elements; a semi-enclosed lagoon

(known as Florida Bay) that is partially isolated from open-

marine waters by a chain of islands (the Florida Keys), a back

reef area of normal marine waters that are protected from high

wave-energy by the rimmed margin, the reef and sand shoal

margin to the platform, and the forereef slope that descends

into the deep waters of the Strait of Florida (Figure C31f).

The semi-humid climate is seasonal, with a wet season from

July to December (100–150 cm annual precipitation) when

there is runoff of freshwater, reducing salinities in inner parts

of Florida Bay to 6% (normal marine waters have on average

34–38%). By contrast, in the warmer conditions of spring

and summer, temperatures increase and salinities up to 60%

are found in the restricted waters of Florida Bay. The reef areas

experience normal marine salinities of 35–38% and water tem-

peratures varying from 18–30



C. The region has a very low

tidal range, reaching 0.8 m in open-water settings and reducing

to zero in the semi-enclosed lagoons of Florida Bay.

Organisms that are fully marine cannot survive in the semi-

isolated shallow coastal waters of Florida Bay because of the

variable salinities, dissolved oxygen levels and fine carbonate

sediment suspended within the bay waters. Seagrass (e.g.,

Thalassia) beds (Figure C31b,e) and the calcareous organisms

that encrust the grass blades (foraminifera, serpulid worms

and bryozoa) are common, as are other skeletal organisms such

Figure C31 Nature and distribution of sediments in humid subtropical region of South Florida. Photographs (a)–(d) illustrate the control that

organisms have on carbonate sediment production and accumulation. (a) Mangrove roots stabilizing the shoreline of Florida Bay (foreground 4 m

across). (b) Thalassia seagrass bed. The grass leaves (20–30 cm high) act as a substrate for carbonate-secreting organisms whose skeletal debris is

trapped and stabilized within the grass beds. (c) Porites coral (branches 1 cm across) are common reef builders in the back reef zone. Note also

encrusted seagrass blades. (d) The shallow zone of the shelf margin barrier reef is mainly constructed by the elk horn coral, Acropora palmata

(branches 15–20 cm across). Map and cross section ((e)–(f)) illustrate nature and distribution of sediments discussed in text (after Bosence and

Wilson, 2003).

CARBONATES, WARM WATER 145

as mollusks, foraminifera and calcareous green algae ( Penicil-

lus and Udotea). When these organisms die, their skeletal deb-

ris accumulates as shelly muds that form bioclastic mudstones

and wackestones in the geological record (for more details on

these sediments, together with carbonate rock terminology,

see Tucker and Wright (1990) and Flugel (2004)). Shells are

concentrated on beaches and storm ridges as bioclastic gravels

and sands. The carbonate mud has the same mineralogical (cal-

cite and aragonite) and geochemical (Ca, Mg and Fe, etc.) com-

position as the skeletal organisms, which implies that most of

the mud is produced from the breakdown of skeletons of algae

and benthic invertebrates. Once formed, the mud is reworked

by burrowing organisms (mainly shrimps and worms) that

digest what they can from the mud and defecate the remainder

as sedimentary faecal pellets or peloids.

Back reef areas are of normal marine salinity and more ex-

posed to waves that generate well-oxygenated waters. These

conditions favor a more diverse community of carbonate-

secreting organisms such as corals, encrusting coralline algae,

calcified plate-like green algae (Halimeda), mollusks, echino-

derms and foraminifera. Corals are not abundant in this envir-

onment but they locally form small patch reefs constructed by

corals and coralline algae overgrowing each other to form a

rigid reef framework (Figure C31c,e). This framework forms

an attractive habitat for many other organisms, including those

that eat or bore into the coral such as sponges, worms, parrot-

fish, echinoids and mollusks. The coral debris they generate

forms much of the skeletal carbonate sand and gravel that accu-

mulates around the patch reefs. These, if lithified, will form

skeletal or bioclastic packstones or grainstones in the geologi-

cal record.

The rimmed margin of this platform comprises a barrier reef

(because the reefs are detached from the shoreline and act as a

barrier to the back reef and lagoonal areas inshore of them)

with intervening high-energy, shallow-water shoals of coarse

skeletal sand and gravel. These breaks in the arc of reefs are

probably due to the fact that Florida is near the northernmost

limit of both reef growth and warm-water carbonates

(Figure C29). The shallowest zone of the reefs is formed by

the branching coral Acropora palmata (Figure C31d,e), and

by some round shaped (or head) corals and encrusting coralline

algae. Deeper down the ocean-facing front of the reef, more

massive and rounded coral growth forms are found that give

way to deeper-water, flattened or plate-shaped corals. Lower

areas of this fore reef slope are characterized by coarse-grained

reef talus deposits formed of broken and bioeroded debris

from the reef. Similar reef morphologies and coral zones are

preserved in ancient reef limestones where they are called

coral boundstones and are composed mainly of fossil corals.

Reef corals are one of the most unequivocal biogenic indica-

tors of warm, shallow-water conditions in the Cenozoic and

Mesozoic.

Arabian Gulf

The southern coast of the Arabian (or Persian) Gulf offshore

from the Gulf Coast States (Abu Dhabi, Dubai and Qatar)

slopes gradually from the low relief desert of the coastal plain,

through the coastal waters and down to a maximum depth of

about 100 m over a distance of a couple of hundred kilometers

(Figure C32). This gently sloping shelf morphology is the dis-

tinctive feature of what is known as a carbonate ramp. The

shelf has no major reef systems or rimmed shelf margins as

are found in South Florida. The prevailing winds blow onshore

from the northwest, which make this a storm-wave dominated

coastline (Figure C32f). Due to the restricted opening to the

Indian Ocean through the Strait of Hormuz and the arid climate

setting, salinities in the Gulf are elevated to 40–45% and tem-

peratures range annually from 20 to 34



C. Details of the region

and its sediments are given in Purser (1973).

Wave-base, the water depth at which wave-generated cur-

rents affect the sea floor, has an important control on sediment

texture in gently sloping ramps and is used to subdivide ramps

into zones (Figure C32e). In the Gulf, the outer ramp areas

(below storm wave base) accumulate muds with a mixed com-

position (called marl); these are partly carbonate, of pelagic ori-

gin, and partly siliciclastic mud brought into the Gulf by rivers.

These marls pass up-slope to the south into muddy and then

clean skeletal and oolitic sands at the fair-weather wave base

at around 10–20 m (Figure C32e, f). Moderate energy waters

characterize this windward facing shore and hence areas of

clean skeletal sand are found in inner ramp environments.

These sands are commonly cemented at or just below the sea-

floor to produce extensive hardgrounds, another feature of

warm-water carbonates. Small patch reefs (Figure C32d) occur

but with a low diversity coral fauna because of the elevated

salinities and sometimes low winter temperatures.

The inner ramp in the Abu Dhabi area, prior to the extensive

coastal developments of today, used to be dominated by ooid

barrier beaches (with minor skeletal grains and peloids)

(Figure C32c). Aragonitic ooids are precipitated in the rela-

tively high salinity marine waters that have a moderate tidal

and wave energy regime. Ooid barrier islands are situated up

to 20 km away from the shoreline and broad, shallow lagoons

have developed behind the barriers (Figure C32c). The lagoons

have elevated salinities of 40–50% and so have a reduced

diversity of marine faunas, comprising gastropods and ostra-

cods that occur in lime muds and peloidal muddy sands. The

lime muds are aragonitic and both the absence of calcified

green algae (like those present in South Florida) and geochem-

ical evidence from the lagoon waters and the deposited muds

indicate that these muds were precipitated chemically within

the lagoon (Kinsman and Holland, 1969) rather than having a

skeletal origin such as those of Florida Bay.

The sabkhas are broad, saline, intertidal areas and coastal

plains (Figures C32a, b, e, f) that may be flooded by lagoon

waters during storms and high tides. The sabkhas have exten-

sive microbial populations that alternate with sediment-rich

layers to form stromatolites (Figure C32a, b). High aridity

in the area leads to a net evaporation of floodwaters, and

saline groundwaters in the sabkha, which results in the precipi-

tation of evaporite minerals such as dolomite, gypsum and

anhydrite.

Summary

From this review it can be seen that warm-water (>18–20



continental shelf seas today have extensive areas of carbonate

sediment accumulation. This, together with their relatively

rapid accumulation rates, explains why limestones that depos-

ited in warm-water shelf environments form the commonest

types of limestone in the geological record. Warm-water carbo-

nates have distinctive assemblages of component grains, both

skeletal and non-skeletal in origin. However, these vary back

through geological time so that the present is not always the

key to the past and care must be taken with paleoclimatic and

paleoenvironmental interpretations based on grain occurrences.

146 CARBONATES, WARM WATER

The two case studies of warm-water carbonate depositional

systems illustrate the range of sedimentary grain types and

sedimentary facies found in such environments today and

explain how these are controlled by their climatic setting.

Studies such as these provide the raw data for interpretation

of warm-water carbonates in the geological record.

Dan Bosence

Figure C32 Nature and distribution of sediments in arid subtropical region of the Arabian Gulf. (a) Intertidal flats showing alternating microbial

(dark) and sediment (light) layers: white patches within layering are precipitates of the calcium sulfate salt, anhydrite (scale in inches). (b) Dome-

shaped stromatolites in intertidal sabkha. These result from upward growth of microbial mats and sediment with lighter colored sediment

accumulating between domes. (c) Ooid sand beaches occur on barrier islands on the Abu Dhabi coast. Dark area by waters edge is microbial mat.

(d) Patch reefs offshore from the barrier islands have a low diversity compared with reefs of the nearby Indian Ocean because of the high salinity of

the Gulf waters. Map and cross-section ((e), (f)) of southeastern margin of the Arabian Gulf illustrate depth-related sediment types on carbonate

ramp. Note change in sediment type with storm wave base (swb) and fairweather wave base (fwwb) (modified from Bosence and Wilson, 2003).

CARBONATES, WARM WATER 147

Bibliography

Bosence, D.W.J., and Wilson, R.C.L., 2003. Carbonate depositional sys-

tems. In Coe, A. (ed.), The Sedimentary Record of Sea-level Change.

The Open University and Cambridge University presses, pp. 209–233.

Flugel, E., 2004. Microfacies of Carbonate Rocks. Berlin, New York:

Springer, 996pp.

Hallam, A., 1984. Pre-Quaternary sea-level changes. Annu. Rev. Earth

Planet. Sci., 12, 205–243.

Kinsman, D.J.J., and Holland, H.D., 1969. The co-precipitation of cations

with CaCO

IV. The co-precipitation of Sr

with aragonite between



C and 96



C. Geochem. Cosmochim. Acta, 33,1–17.

Kleypas, J.A., McManus, J.W., and Meñez, L.A.B., 1999. Environmental

limits to coral reef development: Where do we draw the line? Am. Zool.,

39, 146–159.

Lees, A., and Buller, A.T., 1972. Modern temperate water and warm water

shelf carbonates contrasted. Mar. Geol., 13, 1767–1773.

Purser, B.H. (ed.) 1973. The Persian Gulf: Holocene Carbonate Sedimenta-

tion in a Shallow Epicontinental Sea. Berlin: Springer, 471pp.

Riding, R., and Awramik, S.M. (eds.) 2000. Microbial Sediments. Berlin:

Springer, 332pp.

Tucker, M.E., and Wright, V.P., 1990. Carbonate Sedimentology. Oxford,

UK: Blackwell, 482pp.

Vail, P.R., Mitchum, R.M. Jr., Todd, R.G., Widmier, J.M., Thompson, S. III,

Sangree, J.B., Bubb, J.N., and Hatlelid, W.G., 1977. Seismic stratigraphy

and global changes of sea level. In Payton C.E. (ed.), Seismic

Stratigraphy – Applications to Hydrocarbon Exploration. American

Association of Petroleum Geologists Memoir, 26, pp. 49–212.

Cross-references

Carbonates, Cool Water

Coral and Coral Reefs

Evaporites

Greenhouse (warm) Climates

Icehouse (cold) Climates

Oxygen Isotopes

CENOZOIC CLIMATE CHANGE

The Cenozoic Era, or the last 65 million years, is a contrast of

two worlds. Climates during the first 15 million years of the

Cenozoic were generally warm and the Earth could be charac-

terized as a greenhouse world. Fossil remains of early Eocene

age (ca. 50 Ma) alligators found on Ellesmere Island (75



–



N) contrast sharply with the polar climates that exist there

today. This is not an isolated discovery or quirk of nature. An

ever-growing body of faunal, floral, and geochemical evidence

shows that the beginning of the Cenozoic Era was much war-

mer than the present time. What maintained such a warm cli-

mate and could it be an analog for future global warming?

How did we go from the warm, greenhouse conditions to the

cold poles that characterize today’s world? To address these

and other questions, it is necessary to consider more than a

qualitative estimate of planetary temperatures. Quantitative

temperature estimates (both magnitudes and rates of change)

are required to depict how the Earth’s climate has changed

through time. One of the most powerful tools used to recon-

struct past climates during the Cenozoic Era is the analysis of

oxygen isotopes in the fossil shells of marine organisms. The

calcium carbonate shells of the protist foraminifera are the most

often analyzed organisms because the different species are dis-

tributed throughout surface (planktonic) and deep (benthic)

marine environments.

Oxygen isotopes as a paleothermometer

The stable isotopes of oxygen used in paleoceanographic

reconstructions are

O and

O. During the 1940s, Harold

Urey at the University of Chicago predicted that the

ratio in calcite (CaCO

) should vary as a function of the tem-

perature at which the mineral precipitated. His prediction

spurred on experiments by himself and others at the University

of Chicago, where they measured

O ratios in CaCO

precipitated in a wide range of temperatures, leading to the

development of a paleotemperature equation based on oxygen

isotopes.

The temperature during the precipitation of calcite can be

estimated by measuring the d

O value in calcite-secreting

organisms (foraminifera, corals, and mollusks) and the d

value of the water in which the organisms live. The various

paleotemperature equations all follow the original proposed

by Sam Epstein and his colleagues (University of Chicago):

T ¼ 16:5  4:3 ðd

calcite

 d

water

þ 0:14 ðd

calcite

 d

water

ð1Þ

where T and d

water

are the temperature (



C) and oxygen

isotope value of the water in which the organism lived, and

calcite

is the oxygen isotope value of calcite measured in

the mass spectrometer.

Equation 1 shows that the changes in d

calcite

are a func-

tion of the water temperature and d

water

value. A one-to-

one relationship between d

calcite

and d

water

values

dictates that a change in the d

water

term will cause the exact

change in the measured d

calcite

value. However, an inverse

relationship between d

calcite

and T changes dictates that for

every 1



C increase in temperature, there is a 0.23% decrease

in the measured d

calcite

value. These relationships enable

us to interpret d

calcite

changes generated from foraminifera,

corals, and mollusks. Most researchers follow a convention

that plots d

calcite

values with the axis reversed (higher values

to the left or bottom) so that d

O records mimic temperature

changes (e.g., colder to the left or bottom).

Temporal variatio ns

Variations in the amount of water stored on land through time,

usually in the form of ice, can have a significant effect on the

mean ocean d

water

value, and hence, the marine d

calcite

record. At present, high-latitude precipitation returns to the

oceans through summer ice/snow melting. During glacial per-

iods, snow and ice accumulate into large ice sheets. Because

the difference in ice sheet and mean ocean values is large

ice

= –35 to –40% vs. d

water

mean ocean = 0%),

ice sheet fluctuations are reflected in mean oceanic d

water

values. This relationship can be illustrated by examining how

the mean ocean d

water

value increased during the Last

Glacial Maximum (LGM) relative to the present (Figure C33).

During the LGM, water stored in continental ice lowered global

sea level by 120 m, removing 3% of the ocean’s volume. Thus,

the mean ocean d

water

value increased by 1.2% during the

LGM relative to the present (Figure C33; the marine d

calcite

record is also affected by mean ocean temperature).

Cenozoic d

O records

The first Cenozoic d

O syntheses based on foraminiferal

O records were produced during the mid-1970s. Nicholas

Shackleton and James Kennett produced a composite benthic

148 CENOZOIC CLIMATE CHANGE

O record for the Cenozoic using cores from a region to the

south of Australia. A second group led by Samuel Savin gener-

ated low-latitude planktonic and benthic foraminiferal d

syntheses. Both records are important in understanding Ceno-

zoic climate changes. Benthic foraminiferal records best reflect

global temperature and ice volume changes. Additional advan-

tages of the benthic foraminiferal composite include the follow-

ing: (a) deep-ocean temperatures are more uniform than surface

water with respect to horizontal and vertical gradients; (b) deep

ocean d

water

values are less variable than large surface water

changes; (c) the deep ocean approximates high-latitude surface

water conditions from which deep waters originated during

the Cenozoic (i.e., Antarctica, northern North Atlantic); and

(d) many benthic foraminiferal taxa are long-lived, therefore

one species can be used to construct records spanning several

millions of years, in contrast to planktonic taxa, which have

shorter durations, such that records have to be spliced together

from several species.

Low-latitude planktonic foraminiferal d

O records are good

proxies for tropical sea-surface temperatures. Tropical tempera-

tures are an important component of the climate system

because they influence evaporation, and hence, total moisture

in the atmosphere. A comparison of planktonic and benthic

foraminiferal d

O values allows assessment of equator-to-pole

as well as vertical temperature gradients during the Cenozoic,

and thus, determination of planetary temperature changes.

Much of the climatic change in the last 65 million years has

been ascribed to poleward heat transport or greenhouse gas

fluctuations. General circulation models indicate that each

mechanism would produce different temperature patterns that

can be validated using the planktonic and benthic d

O records.

The first benthic d

O syntheses generated, as well as more

recent compilations, show the same long-term patterns. After

the Cretaceous-Tertiary (K/T) boundary event, deep-water

O values remained relatively constant for the first 7 million

years of the Paleocene (Figure C34a). At 58 million years (Ma)

ago, benthic foraminiferal d

O values began a decrease over

the next 6 Myr, which culminated during the early Eocene with

the lowest recorded value (–0.5%) of the Cenozoic. Following

this minimum at 52 Ma, d

O values increased by 5.5%, with

maximum values (5%) occurring during the glacial intervals

of the Pleistocene. The first part of this long-term change

was a gradual increase of 2% through the end of the Eocene

(52–34 Ma). The remainder of the increase was accomplished

through large steps at the Eocene/Oligocene boundary

(33.5 Ma), during the middle Miocene (ca. 15–13 Ma), and

late Pliocene (ca. 3.2–2.6 Ma). After 2.6 Ma, the amplitude

of the high-frequency signal increased to >1%, reaching

1.8% over the past 800 thousand years.

Planktonic and benthic foraminiferal d

O values co-varied

in general during the early Cenozoic (65–34 Ma). Planktonic

values averaged about –1% between 65 and 58 Ma, before

decreasing to –2.5% during the early Eocene, which were

the lowest values of the Cenozoic (Figure C34a). From 52 to

33 Ma, planktonic foraminiferal values increased by 2%.In

spite of a break in the latest Eocene record, it appears that the

tropical ocean differed from the deep ocean across the

Eocene/Oligocene boundary. For much of the Oligocene

(33–25 Ma), planktonic foraminiferal d

O values remained

unusually high, averaging –0.5%. Begi nni ng around the

Oligocene/Miocene boundary (25 Ma), planktonic foraminif-

eral d

O values began a long-term decrease, culminating in

the Pleistocene, with average values of –1.5%. In contrast,

the benthic d

O record permanently changed during the middle

Miocene d

O shift and late Pliocene increase.

Apportioning the d

O changes recorded by the benthic and

planktonic foraminifera between temperature and ice volume

changes requires knowledge of or reasonable estimates for

one of these parameters. One promising tool that may help

discriminate between each effect is the Mg/ Ca ratio mea-

sured in benthic foraminifera. Initial studies by a group at

Cambridge University using Mg/Ca ratios confirmed the

Figure C33 The growth and shrinkage of large ice sheets can have a significant effect on the d

O composition of the ocean. Removal of 3% of

the ocean’s water during the Last Glacial Maximum lowered sea level by 120 m. The d

O difference between the ocean and the ice is 40%,

causing a whole ocean d

O change of 1.2%. The reverse process occurs during the melting of large ice sheets. If the Antarctic and Greenland

ice were to melt, then sea level would rise 70 m. The volume of water stored in these ice sheets is equivalent to 2% of the water in the

ocean. Therefore, the mean d

O value of the ocean would decrease by 0.7–0.8%.

CENOZOIC CLIMATE CHANGE 149

long-term temperature changes during the Cenozoic that were

calculated using the d

O record and other climate proxies. If

verified, this record implies that small ice sheets grew during

the middle and late Eocene and fluctuated in size throughout

the Oligocene to Miocene. A t present, the Mg/Ca record

lacks the resolution for key intervals and still requires verifi-

cation of interspecies offsets before it can be applied unequi-

vocally to isolate the ice volume-induced d

water

component

in the foraminiferal d

O records. For the discussion that fol-

lows, glaciological evidence is used to estimate the ice

volume/d

water

variations.

The greenhou se world

The oldest unequivocal evidence for ice sheets on Antarctica,

ice-rafted detritus (IRD) deposited by icebergs in the ocean,

places the first large ice sheet in the latest Eocene or earliest

Oligocene. Thus, it is reasonable to assume that ice sheets were

small to absent and that surface and deep-water temperature

changes controlled much if not all of the d

O change prior to

34 Ma. The modern Antarctic and Greenland ice sheets lock

up 2% of the total water in the world’s ocean. If melted, these

ice sheets would raise global sea level by 70–75 m and mean

Figure C34 Planktonic and benthic foraminiferal d

O composite records, representing the tropical surface and deep ocean conditions. The

thick line through both records was generated using a 1 million year Gaussian filter. Temperature estimates based on planktonic and

benthic records and ice volume estimates discussed in the text are shown in (b).

150 CENOZOIC CLIMATE CHANGE

ocean d

water

value would decrease to –0.9% PDB. One can

then apply Equation 1 to the benthic and planktonic foraminif-

eral d

O records to estimate deep- and surface-ocean tempera-

tures for the first half of the Cenozoic (ca. 65–34 Ma).

During the early to middle Paleocene, deep-water temperatures

remained close to 10



C(Figure C34b). The 1% decrease

between 58 and 52 Ma translates into a deep-water warming of



C, reaching a high of 14



C. This is in sharp contrast to modern

deep-water temperatures, which range between 0 and 3



C. Fol-

lowing the peak warmth at 52 Ma, the 2% increase in benthic

foraminiferal values indicates that the deep waters cooled by



C and were 7



C by the end of the Eocene. If small ice sheets

existed during the Paleocene and Eocene, then temperature esti-

mates would be on the order of 1



C warmer than those calculated

for the ice-free assumption.

Tropical surface water temperatures warmed from 22 to 24



based on Equation 1, at the beginning of the Cenozoic to 28



during the early Eocene (52 Ma; Figure C34b). The higher esti-

mate is similar to temperatures in the equatorial regions of the

modern oceans. Planktonic foraminiferal d

O values recorded a

long-term increase of 2% (–2.5 to –0.5%) through the remainder

of the Eocene. Just prior to the Eocene/Oligocene boundary, tro-

pical surface water temperatures were 21



C, ending the long-

term tropical cooling of 7



Cfrom52to34Ma.

The icehouse world of the last 33 million years

As mentioned above, southern ocean cores contain IRD at and

above the Eocene/Oligocene boundary. Widely distributed

IRD and glacial tills on parts of the Antarctic continental mar-

gin that represent the period from Oligocene to Recent mark

the onset of large ice sheets. Whether these sediments represent

persistent or periodic ice cover is uncertain. At least some ice

was present on Antarctica during the Oligocene to early Mio-

cene. Glaciological studies show that Antarctic ice sheet has

been a fixture since the middle Miocene (15 Ma). Our record

of Northern Hemisphere ice sheets suggests that they were

small or non-existent prior to the late Pliocene. For the pur-

pose of estimating surface and deep temperatures, an ice

volume estimate slightly lower than the modern ice volume will

be applied for the interval that spans from the Oligocene into the

middle Miocene (33–15 Ma). For the interval between 15 and

3 Ma, average ice volumes were probably similar to today’s

ice sheets. From 3 Ma, ice volumes varied between the modern

and LGM. Using these broad estimates for ice volumes,

mean ocean d

water

values for those three intervals were –

0.5, –0.22, and 0.4% PDB, respectively. The 0.4% estimate

reflects the average between the maximum and minimum

conditions during the Plio-Pleistocene. As discussed below,

the largest portion of the high frequency signal is controlled

by ice volume changes.

The benthic foraminiferal d

O increase at the Eocene/

Oligocene boundary occurred rapidly (100 kyr; Figure C35a).

At the peak of the Eocene/Oligocene boundary event, benthic

foraminifera recorded d

O values similar to modern values.

Using the ice volume assumption from above, deep-water tem-

peratures approached modern deep-ocean temperatures (3



C).

This marks an important transition from the relatively warm

oceans of the Paleocene and Eocene to the cold deep waters

of the Oligocene to present. This switch to a cold ocean where

bottom waters formed at near-freezing temperatures heralded

the development of the psychrosphere, where warm surface

waters are found above the cold deep waters. Following

the Eocene/Oligocene boundary, deepwater temperatures

began a long-term warming over the next 18 million years (33–

15 Ma). The coldest deep-water temperatures of 3



C were

recorded at 33 Ma, while temperatures reached 9



Cat25

and 15 Ma (Figure C34b).

There is a gap in the planktonic foraminiferal d

O record

for the latest Eocene that hampers our assessment of the tropi-

cal response during the Eocene/Oligocene climate event. How-

ever, it is clear from the data that do exist that the planktonic

response across the Eocene/Oligocene boundary differed

from the benthic response. The planktonic foraminiferal d

values for the early Oligocene are similar to late Eocene val-

ues, whereas there was a 1.5% increase in benthic values.

Planktonic foraminiferal records from other regions that span

the Eocene/Oligocene boundary indicate that the surface water

O increase was on the order of 0.5%. This change is

approximately equal to the effect of the modern Antarctic ice

sheet. Combined with the physical evidence, it seems probable

that the planktonic foraminiferal d

O increase at the Eocene/

Oligocene boundary recorded the ice volume influence with lit-

tle temperature effect. Therefore, tropical surface temperatures

remained around 22



C while the deep ocean cooled during this

O shift. Following the boundary event, planktonic foramini-

fera d

O record during the Oligocene and early Miocene mir-

rored the benthic record in many respects. For much of the

Oligocene and early Miocene, the absolute values were close

to –0.5%, which translates into a temperature estimate of



C(Figure C34b). By 15 Ma, tropical surface waters had

warmed to 26



The middle Miocene d

O shift represents an increase of

1.5% in the benthic record between 15 and 13 Ma. This transi-

tion is composed of two sharp increases around 14 and 13 Ma

(Figure C35b). These d

O steps occurred in less than 200

thousand years with each showing an increase of 1%

followed by a small decrease. During these two shifts, deep-

waters cooled from 9–5



C. The planktonic foraminiferal d

record from 15 to 13 Ma shows two increases as recorded

in the benthic foraminiferal record (Figure C34). However, it

does not show the large permanent shift recorded by the

benthic foraminifera, indicating a small cooling from 26 to



C. From 13 to 3 Ma, the deep ocean cooled slightly from

5to3



C while the surface waters warmed from 24 to 26



(Figure C34b).

The last of the large d

O steps in the Cenozoic was re-

corded during the late Pliocene from 3.2 to 2.6 Ma. This “step”

is better characterized as a series of d

O cycles with increasing

amplitudes and values over this interval (Figure C35c). The

cycles have been subsequently determined to be 40 kyr obli-

quity cycles related to variations in the solar radiation received

at high latitudes. This interval ushered in the growth of large-

scale Northern Hemisphere ice sheets that have since domi-

nated Earth’s climate. At 2.6 Ma, the first ice-rafted detritus

was deposited in the open North Atlantic and was coeval with

the d

O maximum. Prior to 2.6 Ma, IRD was confined to the

marginal basins to the north: Greenland and Iceland’s contin-

ental margins. Subsequent d

O maxima were associated with

ice-rafted detritus. Between 2.6 and 1 Ma, large Northern

Hemisphere ice sheets waxed and waned on a 40 kyr beat.

Some data indicate that smaller ice sheets may have existed on the inland

parts of Antarctica during the late Eocene. However, these were not large

enough to deposit ice rafted detritus in the ocean. Therefore, their effect on

the d

O values of the ocean was probably less than 0.3%.

CENOZOIC CLIMATE CHANGE 151

Beginning around 1 Ma, the ice sheets increased in size and

switched to a 100 kyr beat (Figure C36). During this interval,

deep-water temperatures remained similar to those in the mod-

ern ocean (0–3



C).

The planktonic foraminiferal d

O response during the late

Pliocene event shows the cyclic behavior, but not the overall

increase recorded by the benthic foraminifera. As with the mid-

dle Miocene d

O shift, the late Pliocene increase represents the

cyclic build up of ice sheets accompanied by deep-water cool-

ing. The tropical surface water temperatures, however, varied

between 26 and 28



Pleistocene ox ygen isotope variations

The first systematic downcore examination of the marine stable

isotope record was made by Cesaré Emiliani during the 1950s

on d

calcite

records generated from planktonic foraminifera

in Caribbean deep-sea cores. Emiliani recognized the cyclic

pattern of low and high d

calcite

values and concluded that

these represented glacial-interglacial intervals. Emiliani identified

the seven most recent climate cycles and estimated that they

spanned the last 280,000 years.

To apply the paleotemperature

equation to these records, Emiliani estimated that ice sheet

induced ocean d

water

variability was relatively small, 0.3%.

Therefore, most of the d

calcite

variability between glacial

and interglacial intervals represented temperature changes of

5–10



C. Emiliani divided the d

calcite

record into warm

stages (designated with odd numbers counting down from the

Holocene) and cold stages (even numbers). Hence, “Isotope

Stage 1” refers to the present interglacial interval and “Isotope

Figure C35 High-resolution d

O records representing the Eocene /Oligocene boundary (a), middle Miocene (b), and late Pliocene (c) d

O shifts.

Current age estimates indicate that the duration of the cycles is

approximately 525,000 years.

As shown above, the maximum glacial – interglacial ice sheet signal was

closer to 1.2%.

152 CENOZOIC CLIMATE CHANGE

Stage 2 ” refers to the LGM (Figure C36). During the 1960s and

1970s, many argued that most of the glacial to interglacial dif-

ference in d

calcite

values resulted from ice volume changes.

Nicholas Shackleton of Cambridge University made the key

observation that benthic foraminiferal d

O values show a gla-

cial to interglacial difference of 1.8% (which includes both

ice volume and temperature effects). If the ice volume contribu-

tion were only 0.3%, as argued by Emiliani, then the deep

ocean temperatures would have been between 6 and 7



colder than the present temperatures of 0 to 3



C. Seawater

freezes at –1.8



C, precluding Emiliani’s “low” ice volume esti-

mate. By the early 1970s, numerous d

O records had been

generated and showed a cyclic variation through the Pleisto-

cene and into the late Pliocene. One hundred oxygen isotope

stages, representing 50 glacial/interglacial cycles, have been

identified for the interval since 2.6 Myr ago (Figure C36).

Figure C36 Planktonic and benthic foraminiferal d

O records for the last 3 million years. Note the high frequency signals in the records. For the

interval between 3 and 1 Ma, a 40 kyr cycle dominates the records. After 1 Ma, the beat changes to a 100 kyr cycle and the amplitudes increase.

CENOZOIC CLIMATE CHANGE 153

Mechanisms for climate change

Most climate change hypotheses for the Cenozoic focus on

either oceanic heat transport and/or greenhouse gas concentra-

tions. Each mechanism produces different responses in the

equatorial-to-pole and surface-to-deep temperature gradients.

An increase in the meridional heat transport generally cools

the tropics and warms the poles. If poleward transport of heat

decreases, then the tropics will warm and the poles will cool.

Variations in greenhouse gas concentrations should produce

similar changes in both the tropical and polar regions.

Tropical surface water and deep ocean records co-varied for

the first part of the Cenozoic. The warming and subsequent

cooling between 65 and 34 Ma are most often ascribed to chan-

ging greenhouse gas concentrations. The interval of warming

that began around 58 Ma and peaked at 52 Ma coincided with

the release of large amount of CO

into the atmosphere as a

consequence of tectonic processes. The eruption of the Thulean

basalts in the northeastern Atlantic Ocean began during the

Paleocene and peaked around 54 Ma. It is also recognized that

there was a large-scale reorganization of the mid-ocean ridge

hydrothermal system beginning during the late Paleocene and

extending into the Eocene. Both tectonic processes accelerate

mantle degassing, which raises atmospheric levels of CO

Recently, evidence for another potentially large CO

reservoir

was found along the eastern continental margin of North

America. Methane hydrates frozen within the sediments appear

to have been released catastrophically at least once and possi-

bly multiple times during the latest Paleocene and early Eocene

(58–52 Ma). One or all of these sources could have contrib-

uted to the build-up of greenhouse gases in the atmosphere

between 58 and 52 Ma.

Following the thermal maximum, the long-term cooling in

both the surface and deep waters implies that greenhouse gas

concentrations slowly decreased. Proxies for estimating pCO

concentrations (d

C fractionation within organic carbon and

boron isotopes) are still being developed and refined. However,

preliminary indications suggest that atmospheric pCO

levels

were high (>1,000 ppm) during the early Eocene, dropped to

400–500 ppm during the middle to late Eocene, and reached

late Pleistocene concentrations (200–300 ppm) by the early

Oligocene. It is generally thought that accelerated weathering

of continents lowered atmospheric p CO

levels. The tectonic

agent responsible was probably the uplift of the Himalayas

and Tibetan Plateau.

The deepwater temperature cooling across the Eocene/

Oligocene boundary (Figure C34b) was not accompanied by

tropical cooling, and resulted from the first step in the thermal

isolation of Antarctica. In the modern ocean, the Antarctic

Circumpolar Current is a vigorous surface-to-bottom current

that provides an effective barrier to southward-flowing warm

surface waters. The development of this current during the

Cenozoic hinged on the deepening of the Tasman Rise and

the opening of the Drake Passage. Recent drilling indicates that

marine connections developed across the Tasman Rise at or

near the Eocene/Oligocene boundary (33.5 Ma). Tectonic con-

straints on the separation of the Drake Passage are less precise.

Estimates range from 35 to 22 Ma for the opening of this gate-

way. The uncertainty lies in the tectonic complexity of the

region, and the question of what constitutes an effective open-

ing for water to flow through. The climatic consequence of

creating a circum-polar flow was to thermally isolate Antarctica

and promote the growth of the Antarctic ice sheet. As noted

above, the first large ice sheet developed beginning at the

Eocene/Oligocene boundary.

The most notable divergence in the d

O records began during

the middle Miocene (15 Ma). For the first time during the Cen-

ozoic, the tropical surface and deep waters recorded a clear diver-

gence in d

O values, a trend that increased in magnitude and

reached a maximum in the modern ocean. Any poleward

transport of heat appears to have been effectively severed from

Antarctica by 15 Ma, promoting further cooling. On the

other hand, the tropics have been warming over the past

15 Ma. A combination of different factors fueled this warming.

First, less heat was being transported out of the low- and

mid-latitude regions to the high southern latitudes. Second, the

opening of the Southern Hemisphere gateways that promoted

the formation of the circumpolar circulation led to the destruction

of the circum-equatorial circulation. The effects of the closure of

the Tethys Ocean (predecessor to the Mediterranean) during the

Miocene, shoaling of the Panamanian Isthmus (4.5–2.6 Ma),

and constriction in the Indonesian Passage (3 Ma to present)

allowed the east-to-west flowing surface waters in the tropics

to “pile” up and absorb more solar radiation. A consequence of

the equatorial warming and high-latitude cooling was an increase

in the equator-to-pole temperature gradient. As the gradient

increased, winds increased, promoting the organization of the

surface-ocean circulation patterns that persist today.

Some caveats

A concern in generating marine isotope records is that the isotopic

analyses should be made on the same species. This is important

because d

calcite

values can vary between different species. Co-

existing taxa of benthic foraminifera record d

O values that can

differ by as much as 1%. In planktonic foraminifera, variations

between species can be as great as 1.5%. For both the planktonic

and benthic foraminifera, inter-specific differences are as large as

the glacial-interglacial signal. These inter-specific d

calcite

varia-

tions are often ascribed to a vital effect or kinetic fractionation of

the oxygen isotopes within the organism. However, some of the

difference in the planktonic taxa results from different seasonal

or depth habitats and therefore provides important information

about properties in the upper part of the watercolumn. It is interest-

ing to note that the first d

O syntheses were based on mixed spe-

cies analyses and yet basic features captured in these curves still

persist today. This attests to the robustness of these records and

method for reconstructing climate changes in the ocean.

The high-frequency signal that dominates the late Pliocene to

Pleistocene records is also present in the Miocene and Oligocene

intervals. The cloud of points about the mean shown in

Figure C34a reflects records that were sampled at a resolution

sufficient to document the high frequency signal. For the interval

between 35 and 1 Ma, the benthic foraminiferal d

O record has a

40 thousand year frequency superimposed on the long-term

means that are represented by the smoothed line. The origin of

the 40 thousand year cycles lies in variations in the tilt of the

Earth’s axis (obliquity cycles) that influenced the amount of solar

radiation received at high latitudes. This insolation signal was

transmitted to the deep ocean because the high latitudes were

the source regions for deep waters during much if not all of the

Cenozoic. The record prior to 35 Ma is unclear with regard to

the presence or absence of 40 thousand year cycles.

Finally, relatively high planktonic foraminiferal d

Ovaluesin

the late Eocene to early Miocene indicate cooler than expected

tropical temperatures. Recent work by Pearson and Palmer

154 CENOZOIC CLIMATE CHANGE