oxygen isotope records

oxygen isotope records Oxygen stable isotope records provide a vital method in understanding the Earth's earlier climates. Major long-term variations in ¹⁶O/¹⁸O, the ratio of the isotopes oxygen-16 and oxygen-18, in sea water are controlled by variations in ice-sheet volume. They reflect past sea levels, ice volumes, and marine temperatures and can be directly correlated to a range of parameters, including recorded levels of atmospheric greenhouse gases. In the 1940s, Harold Urey at Chicago University argued that although supposedly identical in chemical composition, the stable isotopes of an element do not behave similarly in all chemical processes. When water evaporates more of the lighter ¹⁶O isotope is removed by the vapour, enriching the remaining water in ¹⁸O. Also, water molecules containing the heavier isotope tend to condense and fall as precipitation more readily than molecules containing the lighter isotope. This results in fresh water having a higher ¹⁶O/¹⁸O ratio than sea water. As ice-sheet volume increases, the amount of ¹⁶O removed from the sea increases, thus enriching sea water in ¹⁸O.

Urey suggested that any oxygen-bearing substance precipitated from water, for example organic or inorganic carbonate, should have an identical isotopic ratio. Changes in the isotopic composition should be monitored directly by any contemporary precipitates, which may be preserved for example as fossils. Cesare Emiliani in the early 1950s produced the first complete oxygen isotope record from shells of foraminifera.

Oxygen isotope records have been most commonly developed from the carbonate precipitated by planktonic and benthic foraminifera, corals, and mollusca. Ice cores and oxygen trapped in bubbles within ice are also increasingly being analysed for oxygen isotopes.

Analytical results are expressed in relation to the ¹⁶O/¹⁸O ratio of a standard, given as δ¹⁸O. A common standard in the past was based on a carbonate from the rostrum of a belemnite from the Cretaceous Pee Dee Formation, South Carolina (PDB). Today, however, Standard Mean Ocean Water (SMOW) is used for it is more widely available and gives more reproducible results. A correction factor can be applied to convert PDB results to SMOW. Samples are processed using a mass spectrometer. The sample is converted into CO₂ and is analysed by comparing its ¹⁶O/¹⁸O ratio to that of a standard gas, which is in equilibration with SMOW.

It is assumed that the organisms which deposited the calcite in their shells did so at equilibrium or constant disequilibrium through time. This may not always be true, for certain species are known to vary their habit during their life cycle and also as a result of evolutionary changes. For example, certain species of benthic foraminifera may be either epifaunal or infaunal, respectively recording ambient benthic-water or pore-water changes. Planktonic foraminifera are also known to migrate through a range of water depths. The ratio of δ¹⁸O between ambient water compared to carbonate precipitates can also be altered by temperature, salinity, and biological influences such as respiration. Measurements from foraminifera cultured in water of a known isotopic value are used to develop our understanding of their degree of fractionation.

Sea level

Global sea level is affected, among other things, by the amount of water contained within continental ice sheets; this in turn is recorded in the δ¹⁸O of marine organisms. It is therefore possible to use the δ¹⁸O record to record former sea-level changes. A δ¹⁸O increase of 1 ‰ (one per thousand) represents a drop in sea level of about 10 m. However, it is impossible to apply a simple linear relation between δ¹⁸O and sea level. Estimates of past sea level based on marine δ¹⁸O values show that discrepancies can arise because of the effect of sea temperature variations and the form in which the ice exists. The formation of sea ice will remove ¹⁶O, but will have no effect on the sea level.

During the Miocene epoch a series of sea-level falls related to ice-sheet growth are recorded in the marine δ¹⁸O record. Records for the past 16.5 million years have a series of step-like increases in the δ¹⁸O values, because of the transition from a relatively unglaciated world to one similar to that of today.

Temperature

Temperature is also known to influence the δ¹⁸ O record; becoming more negative with warmer temperatures. Using the δ¹⁸O of planktonic foraminifera as a record of sea surface temperature (SST) is problematic because (1) these foraminifera respond to environmental changes and vary in their habitat, and (2) temperature changes are often associated with salinity changes, which also have a major effect on the δ¹⁸O record. For benthic records, complications also arise because of variations in the original surface sea water δ¹⁸O during deep water formation, caused by changes in evaporation rates and freshwater input.

Corals

Corals provide the δ¹⁸O records with highest resolution. Their growth rings are deposited monthly, and by counting back from those presently being formed it is possible to produce a monthly record over many decades. Corals in the western Galápagos Islands show a close correlation between δ¹⁸O and SST since 1965, when SST records for this region start. Using this information it is possible to estimate earlier SST values from the coral δ¹⁸O record. El Niño events, characterized by periods of very high SST in the eastern Pacific are clearly visible in the coral δ¹⁸O record. Also an 11-year cycle is seen in the records of δ¹⁸O and the growth rate of corals, supporting the idea that climate variability in the tropics is driven by solar cycles.

Ice cores

Ice cores drilled from the ice sheets of Greenland and Antarctica provide δ¹⁸O records back to 250 000 years ago. Ice core δ¹⁸O records are developed from (1) ice crystals, which produce a record directly comparable to atmospheric dust levels and composition, and (2) O₂ trapped in air bubbles, which gives a contemporaneous record of atmospheric gaseous changes; for example CO₂ and methane. At present the atmospheric δ¹⁸O (δ¹⁸O_atm) is + 23.5 ‰ relative to SMOW. This difference is known as the Dole effect and is primarily the result of biological isotope fractionation during photosynthesis (the main method of O₂ production), respiration, and hydrological fractionation during evaporation and precipitation. As δ¹⁸O_atm and oceanic δ¹⁸O (δ¹⁸_ocean), developed from foraminifera, vary at the same time and to the same magnitude during the late Quaternary, it appears that the Dole effect has remained reasonably constant during this period. Therefore, δ¹⁸O_atm can be used as a proxy record for ice volume records in a similar manner to that for δ¹⁸O_ocean. This is very useful, for it provides comparable records of atmospheric CO₂ and ice volume. It appears that atmospheric CO₂ began to increase at least 4000 years before an initial decrease in continental ice volume.

Problems with the ice core δ¹⁸O record include (1) the presence of ice formed elsewhere; (2) changes in temperature between times of ice deposition, commonly due to the elevation of the ice sheet surface as a result of thickening of the sheet; (3) changes in atmospheric circulation, as ‘older’ clouds lead to precipitation enriched in ¹⁶O because of the earlier preferential condensation of ¹⁸O; (4) long-term ice sheet deformation.

A section of the Summit Greenland GRIP ice-core δ¹⁸O record is given in Fig. 1. The section is dominated by rectangular climate cycles averaging a few thousand years (ka), termed Dansgaard–Oeschger (D–O) cycles. Packets of D–O cycles characterized by progressively less negative δ¹⁸O measurements are termed Bond cycles and last up 8 ka. They are marked by a Henrich event (H) at their start and end. These cycles reflect the gradual build-up of the ice sheets and their rapid melting (Henrich events). The colder Younger Dryas is clearly marked by more negative δ¹⁸O values and the warmer Holocene by more positive values at the top of the diagram.

Stratigraphy

As the ice volume increases, the δ¹⁸O_ocean becomes more positive, and vice versa. This produces a cycle of variations, contemporaneous around the world, recorded down cores within calcite shells. From this a chronostratigraphy was developed; the more positive glacial stages have even numbers and the negative interglacial stages have odd numbers. The stages start in the present interglacial, Stage 1. Similarly the last interglacial period refers to isotope Stage 5. Substages within major stages are defined by letters (or numbers), the lowest at the termination of the stage; for example, Substage 5e is a more negative (warmer) period at the start of interglacial Stage 5.

A composite record based on a large number of δ¹⁸O planktonic foraminifera records has been developed by Prell and others. The ends of stages are characterized by rapid terminations with large changes in the δ¹⁸O values. For example, the termination of the last glacial period, Stage 2, occurred over less than 10 000 years with a change in δ¹⁸O values from + 1.8 ‰ to −2.0 ‰ (for the present interglacial). The past 4.5 million years contains over 152 stages. New records are compared with this standard to deduce the age of the core. Because of gaps in the sedimentary record and varying degrees of intensity between sites, not all stages or complete stages are recorded in all cores.

The δ¹⁸O record peaks every 100 000 years, with superimposed, smaller cycles of roughly 23 000 and 41 000 years. These cycles strengthen the argument for astronomically driven climate cycles. The longer cycles correlate with the Earth's eccentricity cycle; the shorter cycles relate to the Earth's precessional and obliquity cycles. Age models based on counting the obliquity cycles in the δ¹⁸O record of planktonic foraminifera commonly yield ages and average sedimentation rates higher in resolution than those from fossil data.

Isotopic data support earlier climate change theories that the importance of different astronomical cycles varies with latitude. For example, the 41 ka (41 000-year) cycle is more pronounced in cores from high latitudes, while cores from middle to low latitudes have a stronger 23 ka cycle. Cycles also vary over time; for example the 100 ka cycle plays a dominant role only during the past 650 ka years.

Stephen King

Bibliography

Prell , et al. (1986) Graphic correlation of oxygen isotope stratigraphy: application to the late Quaternary. Paleoceanography, 1, 137–62.
Sowers, T., and Bender,, and M. (1991) The δ¹⁸O of atmospheric O₂ from air inclusions in the Vostok ice core: timing of CO₂ and ice volume changes during the penultimate deglaciation. Paleoceanography, 6, 679–96.