Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Tuatara: Volume 16, Issue 1, April 1968

Oxygen Isotope Palaeotemperature Determinations from Victoria, Australia

Oxygen Isotope Palaeotemperature Determinations from Victoria, Australia


Temperature is the most fundamental factor in climate; from the present point of view it is the earth functioning as a heat engine.

Just as in isotopic dating the main problem is to know what exactly is being dated, so in isotopic thermometry, the main problem is to know just what is being measured. To what extent is growth temperature being measured, and to what extent subsequent ionic exchange? Even where the growth temperatures of fossil organisms have been accurately determined, what is the relationship of such ambient temperatures to the climatic mean temperature of the time, which is the desideratum for a palaeotemperature curve?

Large numbers of assays are required to provide some kind of statistical mean, especially where shelf faunas (as against deep sea faunas) are used. For the Victorian assays, only shelf faunas were available. The use of pelagic foraminifera included with such faunas were investigated, but sufficient well - page 57 preserved forms in sufficient places were not available to provide a Cainozoic paleotemperature curve. Chemically stable molluscan shells from the continental shelf having a long range in time were selected. A large number of assays was carried out, most of which have been published (Dorman and Gill 1959a, b, Dorman 1966, Gill and Dorman 1967).


The aim of oxygen isotope palaeotemperature projects is to establish from the assay of fossils a curve of mean temperature that represents variation in temperature against time. Before this method was available, biologic data were the chief source of temperature estimates. These could only be generalised. The isotope method provided a check from another discipline, and also a more accurate way of determining past temperatures where suitable samples for assay were available. Although even a rough approximation of temperature is useful, the isotope method seeks to refine its procedures so as to give a reliable account of the temperatures of the past. If the parameter of temperature can be defined, many inferences can be safely made from the data. However, the chief difficulty is to cope with the many variables which can cause deviation from the actual temperature. Some of these are:

1. The assay variable.

More accurate mass spectrometers built for this special purpose have done much to overcome this variable, making possible determinations within + or - 1°C, other factors being correct. However, there is still the assumption that the proportion of O18 to O16 has not changed significantly through the span of geologic time concerned.

2. The latitude variable.

Many past attempts at palaeotemperature estimation have failed because the data were collected over too wide a range of latitude. The assays in Victoria have been restricted to fossils from one degree of latitude in order to eliminate this variable.

3. The biology variable.

Genera and species of marine organisms vary in their times and rates of physiologic activity, i.e. the times of acceptance of the isotopes whereby the record of temperature is retained. Some organisms are active during the heat of the day (for example), and some during the cool of the night. The physiologic characteristics of a fossil cannot be determined as a rule, although a probability can be established from its living relatives. In this page 58 connection, the assessments for recent fossils will be more accurate than those for ancient fossils, and those for living forms more accurate than those for extinct forms.

This variable applies particularly to intertidal or shallow water marine organisms. Pelagic or deeper shelf taxa are therefore to be preferred. However, even the most suitable pelagic fossils may record different temperatures depending on where the ocean currents take them.

4. The ecologic variable.

Some organisms live at the surface of the sea and so record marine surface temperatures. Others are benthic and so record ocean bottom temperatures. The latter have a numerically lower range, and are slower to adjust to climatic change. This variable can be countered to a certain extent by using the same genus (or the same evolutionary series of species) throughout the geologic periods studied. Better still is to assay a benthonic series, a pelagic series, and a nectonic series, and compare the results. As far as possible, this was done by Dorman and Gill (1959a).

5. The chemical exchange variable.

As any chemical exchange between the minerals of the assayed fossils and their environment, and any reconstitution (e.g. change from aragonite to calcite) alters the proportion of the isotopes, serious error may result from this variable. Calcitic fossils, being chemically and physically more stable, are therefore to be preferred, and well-preserved specimens chosen. The removal of exterior surfaces such as may have suffered exchange is also a prophylactic against this type of error. It may be assumed in all cases that some exchange has occurred, and the question is whether the amount has been significant.

6. The palaeogeography variable.

In Victoria within the degree of latitude from which fossils were selected for O18/O16 assay, there are three sedimentary basins — the Gippsland Basin in the east, the Port Phillip Sunkland centrally situated, and the Mt. Gambier Sunkland in the west. The first two basins are separated by the South Gippsland horst, and the second two by the Otway horst. Significant palaeotemperature differences were noted between these basins, and these may well be connected with their paleogeography — the degree to which (at a given time) they were open to the ocean, the nature of the water circulation within them, and other factors that would affect the mean temperature of a given site.

7. The chronometry variable.

In order to plot temperature against time, some satisfactory method is needed of dating the fossils whose isotope composition is page 59 measured. As such knowledge is limited, this constitutes another variable. It can be countered to a certain extent by taking fossils from a superposed series, if such be available.


From the initial oxygen isotope assays, palaeotemperature graphs were constructed (Dorman and Gill 1959a, b). Recently Dorman (1966) has published new graphs based on a more extensive series of assays and a critical re-assessment of the results. A review of work so far has been prepared (Gill and Dorman, 1967). It is instructive to compare and contrast the two sets of graphs.

1. The Ostrea Series.

This genus was chosen as an example of a bottom-living mollusc with a relatively stable calcitic shell, and a long geologic history. It has been noted that a shallow water form such as Ostrea manubriata is more variable in isotope temperature than species belonging to deep waters. Figure 1 shows the two oyster graphs.

Fig. 1: Oxygen isotope palaeotemperature curves for two series of Tertiary mollusc shells from Victoria, Australia.

Fig. 1: Oxygen isotope palaeotemperature curves for two series of Tertiary mollusc shells from Victoria, Australia.

The first was a generalised curve indicating lower temperatures in the lower Cainozoic, higher temperatures in the Middle Cainozoic, and lower temperatures again in the Upper Cainozoic. The same idea is conveyed in the new graph but (a) the peak is transferred. This was expected from the other palaeotemperature graphs (Pecten series of Glycymeris series) published in the original paper, (b) The peak is higher, (c) There is evidence of temperature drops at the Eocene/Oligocene boundary, and in the Upper Miocene.

page 60

The new graph, being based on a much larger series, points up the reality of variables 3 to 7 above.

2. The Pecten - Chlamys Series.

Again the same general idea of a mid-Tertiary period of relatively high temperatures is indicated, but the peak is higher.

A result of the larger series of determinations is that the two graphs have become much more alike. Indeed, there is no major difference. The resultant curve fits well the biological evidence. (Gill 1961 a, b). As shells are used that record the rather variable conditions of marine waters over the continental shelf, it is probably significant that with greater numbers of assays, the two curves have approached each other. When this work was commenced, only deep sea faunas had been used for oxygen isotope palaeotemperature measurements. The present work indicates the feasibility of using shelf-faunas, provided a sufficient number of determinations is made.

Further progress could probably be achieved by studying the depth at which the organisms lived and a series of molluscs of the same genus selected from deposits emplaced at the same depth.


Dr. N. deB. Hornibrook. Could you tell us where the peaks come in the Australian stages?

Mr. I. Devereux. The figures supplied by Mr. Gill do not give any stages but I will refer to their lastest published results. The dip is in the Upper Johannian Stage, below the Janjukian Stage.

Dr. N. deB. Hornibrook. That would correspond to our Ak to Ar, certainly Upper Eocene.

Mr. I. Devereux. The high peak is in the Batesfordian and the following dip is in the Mitchellian.

Dr. N. deB. Hornibrook. The Batesfordian corresponds to the Altonian and the Mitchellian our Tongaporutuan to Kapitean.

Professor P. Vella. With the correlation the Australian and New Zealand curves are remarkably similar.

Dr. I. Speden. I would like to comment on the choice of oysters as sample material. They have drawbacks which have made them unpopular for use in this work. Firstly, they live in shallow water environments which often experience salinity fluctuations, and secondly their shells have a foliated structure which would seem susceptible to group water exchange. They also have soft patches in the shell layers and these are permeable and are often recrystallised. For these reasons people have been rather sceptical about using oysters and yet here we seem to have a very comprehensive graph.

Mr. I. Devereux. Th problem of fresh-water contamination of the sea-water can probably be overcome by widespread sampling. It seems that the shells that are originally, deposited as calcite are page 61 more stable than those deposited as aragonite. In assemblages of shells I have examined, in the Wairarapa, the aragonite ones are quite useless but the calcitic ones have remained unaltered. However from very old samples, say Mesozic, the oysters have all exchanged with ground waters and only belemnite guards seem satisfactory.


Dorman, F. H., 1966. Australia Tertiary paleotemperatures. J. Geol. 74:49-61.

—— and Gill, E. D., 1959. Oxygen isotope palaeotemperature measurement on Australian fossils. Proc. Roy. Soc. Vict. 71:73-98.

Gill E. D., 1961a. The climates of Gondwanaland in Cainozoic times. Chapter XIV of Descriptive Paloeoclimatology, Ed. Nairn (Interscience), pp. 332-353.

——, 1961b. Cainozoic climates of Australia Ann. New York. Acad. Sci. 95 (1): 461-464.

——, and Dorman, F. H., 1967. Review of oxygen isotope palaeotemperature measurements in Victoria, Australia. In press.

* Presented by Mr. I. Devereux.