Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Tuatara: Volume 16, Issue 1, April 1968

Oxygen Isotope Paleotemperatures from the Tertiary of New Zealand

page 41

Oxygen Isotope Paleotemperatures from the Tertiary of New Zealand


The Study of past climates by the oxygen isotope method has been increasingly used over the past fifteen years. Most work has been carried out in the U.S.A., where the method was first discovered and developed by a team headed by Professor Urey at the University of Chicago. Up till now most effort has been concentrated on the Mesozoic Period and the Pleistocene with the Tertiary period, rather surprisingly, being largely ignored. A notable exception however is the work of Dorman and Gill, who have spent many years using this isotope technique on Tertiary fossil material from Victoria, Australia.

The isotope method has at various times come in for a fair share of criticism and some of this has been quite legitimate. This is especially so, when inadequate attention has been paid to the selection of suitable samples and to good geological correlation. The method has some fundamental weaknesses, but if sufficient care is taken, these may be avoided. To appreciate these difficulties, and to be in a position to critically examine isotope results, it is essential to understand the fundamentals on which the method is based.

The oxygen isotopes which are of interest are oxygen-16 and oxygen-18, usually written O16 and O18. Both are stable isotopes and they react in a similar fashion, e.g. they both combine with hydrogen to form water. The water made from O16 would appear the same as that made from O18, but the latter is more dense because the O18 atoms are heavier than the O16 atoms. All naturally occuring compounds that contain oxygen, contain both isotopes but the O16 predominates. There is approximately one O18 for every 500 O16 atoms. In different compounds there are slightly different ratios of the O18 to the O16 atoms. For instance if calcite is precipitated in equilibrium with water, the ratio O18/O16 in the calcite is slightly more than the O18/O16 in the water i.e. the O18 tends to concentrate in the calcite.

This leads to the two fundamentals on which the method is based. Firstly, the difference between the O18/O16 ratio in the calcite and the O18/O16 ratio in the water is constant for any defined temperature. Secondly, the difference in the ratios depends on the page 42
Fig. 1. Oxygen isotope paleotemperature curve for the Tertiary period of New Zealand. (Followed lines: Planktonic Foraminifera. Broken lines: Macro fossils).

Fig. 1. Oxygen isotope paleotemperature curve for the Tertiary period of New Zealand. (Followed lines: Planktonic Foraminifera. Broken lines: Macro fossils).

temperature at which the calcite is precipitated; as the temperature rises the differences between the O18/O16 of the calcite and the O18/16 of the water decreases.

How are these two factors used to measure the temperature of past seas and hence deduce the pattern of ancient climates. It has been shown experimentally that organisms deposit their shells in equilibrium with the sea water in which they live. By measuring the O18/O16 of waters from all the World's oceans it has been found that the sea has a very constant O18/O16 ratio and it seems likely that this has been so for a very large part of geological time. By using this ocean water as a standard reference and measuring the O18/O16 ratio of a shell, we can determine the temperature of the sea water in which the animal lived. By measuring the O18/O16 of shells from formations of various geological ages, a picture of the temperature versus time can be built up.

The explanation of the fundamentals leads on to the uncertainties that can arise in using this method. To calculate the temperature of the water from the O18/O16 ratio of the shells, we have to assume that the animal was living in the open ocean water that is our reference material. However, if the animal was living in waters which were diluted with river water, an error is introduced into the calculations. This is because river water has a lower O18/O16 ratio than sea water and this will lower the O18/O16 ratio of the calcite in the shell. Because it is impossible to determine precisely what percentage of river water has diluted the sea water, we cannot make any correction for it. If we ignore this contamination and just apply the same formula, we get a temperature which is higher than the real temperature. Depending on the amount of contamination page 43 it could be only a few degrees centigrade, 20°C, or even more. It cannot however be any lower because any contamination makes the result higher.

A very similar effect happens if after deposition some of the shell is dissolved and redeposited by ground waters. This is especially so with shells that were originally aragonite and have subsequently changed to calcite. Here the effect is the same i.e. the result obtained is too high.

Therefore great care must be taken to select shells that have not lived in estuarine conditions, and samples must be carefully examined for any sign of recrystalisation. Even with these precautions it is possible that fresh water contamination did occur but has not been recognised. Therefore shells from one locality may give higher results than shells from another locality of the same age. If this occurs the lower results are the most likely to be correct.

How is the O18/O16 ratio of the shells measured? The shell is ground to a powder and reacted with acid in an evacuated glass vessel. The reaction produces carbon dioxide which is put into a machine called a mass spectrometer and this measures the O18/O16 ratio. Although the mass spectrometer is a very sophisticated machine, the overall technique is very simple and can easily be learnt in a few days.


Over the last two years I have collected a number of samples from various parts of New Zealand representing various Tertiary stages. Samples have also been provided by various people, especially the New Zealand Geological Survey, at Lower Hutt.

Measurements have been made on planktonic foraminifera and macro fossils from shallow water deposits. The results are summarised in the Figure which is “normalised’ to the latitude of Wellington. Major points of interest are the considerable drop just prior to the Eocene/Oligocene boundary (Upper Runangan Stage) and the rise to the Altonian Stage and subsequent fall, almost steadily, towards the Pleistocene.


During the course of this work a great number of people have assisted in various ways, and I thank all those who have helped me in this study. Special thanks to Professor Vella and Mr. Beu of Victoria University, Geology Department, and Mr. Hornibrook and Mr. Maxwell of the geological survey, Lower Hutt for providing many samples, stratigraphic information and general assistance. Thanks to Dr. Hulston of the Institute of Nuclear Sciences for assistance with the mass spectrometry

page 44


Mr. J. Grant-Mackie. What size of sample is required for one determination.

Mr. I. Devereux. I use about 5 mg. of Foraminifera which is several hundred tests. With macro-fossils, usually 8-10 mg.

Dr. N. deB. Hornibrook. How do you tell if a sample is exchanged or altered?

Mr. I. Devereux. Firstly I would examine the material containing the fossils. It should be clean, not stained or brown, denoting percolating ground waters. The fossils should be carefully examined microscopically. They should be white or light grey, opaque not glassy, they should be hard not chalky. When ground they should give a clean white powder.

Mr. V. Neale. Would samples from near Antarctica be likely to be affected by ice water and ice-bergs melting?

Mr. I. Devereux. Yes, almost certainly.

Dr. P. Webb. Have you tried any measurements on living plankton?

Mr. Idevereux No. I have not been able to get any suitable samples, but other people have done it. I have measured some Recent molluscs.

Dr. N. deB. Horinbrook Have your results been corrected for surface temperatures?

Mr. I. Devereux. The results shown are the temperatures measured, normalised to Wellington. If the forams, are recording a temperature a little below the true surface temperature, which is likely, then the results may have to be lifted 1 or 2°C.

Professor A. Wilson. What error would you place on the results?

Mr. Idevereux The analytical error is only 0.5°C but because of the various uncertainties I think 2°C would be a more realistic figure.

Mr. K. Lewisi. I have read a paper by Dr. Wiseman of the British Museum in which he said that he got different results depending on how he prepared his sample. Could you comment on this?

Mr. I. Devereux. Some shells, especially living ones contain organic material which can contaminate the sample and so must be removed. This can be done by heating under vacuum or soaking in Janola.

Mr. J. Grant-Mackie. I am wondering if your low temperature in the Whaingaroan is supported by the presence there of fossil penguins.

Dr. C. A. Fleming. I don't think we know the climatic requirements for fossil penguins.

page 45

Selected Bibliography

Bowen, R., 1966. “Paleotemperature Analysis’, Elsevier, Amsterdam.

Devereux, I., 1967. Oxygen Isotope Temperature Measurements on New Zealand Tertiary Fossils. N.Z. Jour. Science, (In press).

Dorman, F. H., 1966. Australian Tertiary Paleotemperatures. J. Geol. 74.49-61.