Paleontology 137
J Bret Bennington

Paleoclimatology

Fossils can be used in a variety of ways to gain insights into the climate conditions of the past. Obviously, climate has a major influence on the distrubution and form of organisms. If we understand these influences and their effects on modern floras and faunas, and if we can confidently assume that fossil floras and faunas responded in the same ways to the same climatic effects (a big if) then we can use fossils to reconstruct past climates.

Generally, because of the fact that paleoclimate inferences based on fossils usually rely on some assumption of comparability of modern organisms to past organisms, the farther back in time one tries to make paleoclimate inferences the less precise or reliable they become. Thus, an abundance of detailed fossil-based paleoclimate data have been amassed for the Quaternary, but for the Paleozoic the conclusions are fewer, more general, and often less reliable.

Paleotemperature Estimates

Direct estimates of the temperature of seawater can be obtained from certain calcareous fossils using the ratio O18:O16 preserved in the calcite of the shell.

This is made possible by the fact that O18 is preferentially removed from the water and incorporated into the calcite during the crystallization process. This effect diminishes as temperature increases so that, all other things being equal, calcite precipitated at colder temperatures will have more O18 than calcite precipitated at warmer temperatures.

A sample of shell material can be analyzed in a mass spectrometer to determine the ratio of O18 to O16 and then compared to some arbitrary standard to get a relative estimate of the difference in temperature of the waters in which the shell formed.

For calcite O18 measurements the standard that is used is the calcite in the phragmacone of a belemnite from the Late Cretaceous Pee Dee Formation of South Carolina (PDB).

Deviations from the standard value are calculated as as follows:

Now, if the ratio of O18 to O16 in seawater was constant through time, then could be related to seawater temperature in a straightforward manner. Unfortunately, this ratio is not constant.

The effect of glaciations on

The ratio of O18 to O16 in seawater changes through time in response to glaciations. O16 evaporates more readily than O18 so that freshwater is depleted in O18. In general this would not be an issue because freshwater is continually recycled into the oceans, so that soon after being removed the missing excess O16 is replaced.

However, if the O16 enriched freshwater is locked up for long periods of time as glacial ice in polar ice caps, then the ratio of O18 to O16 will increase in the oceans and therefore in the calcite shells of ocean organisms.

Of course, glaciations have the effect of cooling the ocean waters, which also causes an increase in the ratio of O18 to O16 in calcite shells.

So, the question becomes, if you measure an increase in the ratio of O18 to O16, how much of that increase is due to temperature and how much is due to global enrichment of O18 in the seawater due to glaciation?

Overhead: Oxygen isotope trends for the past 700,000 years

Indeed, the original trends in plotted were at first assumed to represent fluctuations in ocean water temperature, but later it was realized that they were really an indication of global ice volume showing the waxing and waning of polar ice sheets. Important evidence corraborating the existence of a large number of Pliocene and Pleistocene glacial episodes, but not quite what the researchers were after!

To obtain a true temperature estimate, one has to subtract the due to glacial ice buildup from the total measured. due to ice is the measured directly from seawater and compared to Standard Mean Ocean Water (SMOW). Unfortunately, it is not possible to measure this quantity in the past, it can only be estimated.

Assuming that seawater can be measured, then empirical studies show that temperature can be estimated using the following equation:

Now, how can one get around the problem of not knowing seawater for times in the past?

This is difficult.

a. Assume that there is no substantial ice buildup during stretches of the past when other geological evidence supports this. Use theoretical estimates of seawater for non-ice times. (example from study of Paleozoic Brachiopods).

b. Assume no substantial changes in ice volume and bracket the estimates between ice-free values and values based on modern (moderate ice volume) seawater. (K-Tertiary surface water study)

c. Compare surface water values to deep sea values. Deviations in deep sea water temperature from surface water temperature indicates onset of glaciation as cold water flows from poles to equator.

There are other factors that complicate the relationship between temperature and the ratio of O18 to O16 in shell material.

Vital effect: some taxa show different degrees of fractionation of the isotopes from inorganic precipitation. The equation presented above has been shown to be valid for brachiopods, planktonic forams, bivalves, and brachiopods.

Diagenesis: Recrystallization and reprecipitation can change isotopic ratios in shell material. Samples used must be as pristine as possible and unaffected by significant diagenesis.

Leaf Margins

Most attempts to infer temperature trends over time are less quantitative than isotopic methods. However, a general sense of the climate conditions can often be obtained by comparisons with modern conditions. This is especially true with plants.

Because most plants are sensitive to changes in climate they can provide very useful indications of changes in terrestrial climates.

For example, flowering plants show a very interesting relationship between leaf shape and climate. Plants with non-entire leaf margins (toothed, indented, or lobed outlines) are common in cool climates, but plants with entire leaf margins (smooth outlines) are common in tropical climates. Also useful is the pattern of venation seen in leaves. Tropical plants tend to have pinnate (feather-like) venation and temperate plants tend to have palmate (radiating) venation. Finally, compound leaves are more common in the tropics than they are in temperate climates. These characteristics are found across a variety of tree and shrub groups, suggesting that they are functional adaptations to living in certain climates, rather than peculiarities of particular plant families.

Overhead: Percentages of entire margined leaves in climate zones

For whole floras, the ratio of non-entire to entire margin species changes with latitude and climate, as do the ratios of compound to non-compound leaf species. By estimating such percentages in fossil floras, a semi-quantitative sense of climate change through time can be arrived at.

Tree Rings

The woody tissue in trees and other plants grows outward from the center of the stem in concentric layers.

If growth is interrupted or slowed during the year for any reason, the cells in the current layer of wood become bunched together forming a growth ring. Generally growth rings are a good indicator of seasonality, which itself is usually associated with either temperate climates or high latititude locations where sunlight is reduced for part of the year.

Fossil wood is known from the Upper Devonian through to the present. Some of the most famous Devonian fossil wood comes from the 'Gilboa Forest', an Upper Devonian petrified forest in the Catskills that now lies beneath a reservoir.

Devonian fossil wood agrees very consistently with paleolatitudes inferred from paleomagnetic data. New York specimens have very weakly defined growth rings suggesting that New York was indeed on the margin of the Devonian tropics (paleolat. 17° S). Russian specimens from the Donetz Basin (paleolat. 3°) have no growth rings, corraborating the tropical position of the Donetz Basin in the Devonian.

Carboniferous fossil wood from throughout Europe and America is consistantly lacking in growth rings confirming the tropical, coal forest environment inferred for these landmasses.

Permian wood is without significant growth rings in Europe and North America. However, fossil wood from Gondwanaland, inferred to have been at this time at high southern latitudes, shows well-marked growth rings.

In the Mesozoic and the early Tertiary there is perhaps the greatest latitudinal spread of trees of any time in Earth history. In the Lower Cretaceous fossil wood shows growth ring patterns that coincide predictably with paleolatitude.

Of great interest is the observation that trees growing at extremely high latitudes (inside of the arctic and antarctic circles) show not only growth rings (as would be expected - at high latitudes one would expect daylength to produce a marked growing season) but thick growth rings that indicate rapid growth for at least part of the year. This suggests that the high latitude climate at this time was much more temperate than it is at the present time, and that, by extension, global climates were much warmer overall from pole to pole than they are today.

Carboniferous paleoclimate

Studies of the paleoflora of the Carboniferous have been used to infer climate conditions, particularly with respect to climatic wetness.

For example, a study by Winston (1990) used a combination of Lycopod abundance and coal seam thickness and quality to reconstruct trends in climatic wetness through the Pennsylvanian in the Appalachian basin. Lycopods in the carboniferous were large trees that formed one of the major components of coal swamp vegetation. The morphology of lycopod roots show them to be adapted to grow in standing water and previous studies have argued that lycopod abundance increases with increasing 'ever-wet' conditions. Seasonality of rainfall or lessening of annual rainfall leads to a decrease in lycopod abundance.

Furthermore, coal seams in general, being a product of standing freshwater environments, require extended intervals of wet climate to attain minable thicknesses. Extended dryness or seasonal dryness leads to conditions that oxidize and degrade the peat, preventing significant accumulations from forming.

Winston's study of Appalachian Basin coals in Virginia and West Virginia shows an interval of moderate drying or seasonality in the mid-Middle Pennsylvanian and an overall drying trend in the Upper Pennsylvanian.

Paleoclimate studies based on the distributions of organisms.

Most studies that apply paleontological data to inferring ancient climate do so by mapping the geographic distributions of organisms that have some understood climate significance.

Two potential problems exist with these studies that must be examined with a critical eye before the results of such studies can be accepted:

1. When mapping the distribution of fossils, one is often interested in areas where a particular fossil group is absent. However, before the absence of fossils can be considered meaningful, one has to show that the organism in question could have been found as a fossil but was not because it was not present in the area. To cite an obvious example: the absence of Upper Jurassic dinosaur fossils from the eastern United States is not considered significant because there are no Late Jurassic age rocks in this region. If, however, there were late Jurassic rocks in eastern North America, and if these rocks contained non-dinosaur vertebrate fossils, but no dinosaurs, then the absence of Late Jurassic dinosaurs would need to be explained.

Generally, before the absence of a type of fossil can be considered significant, some control group must be shown to be present. The control group is ideally a type of organism that has similar fossilization potential to the organism of interest, but differs in the characteristic of interest.

2. The climate significance inferred for a fossil organism, usually obtained using data obtained by studying living organisms, must be reasonable.

Mapping Tertiary climate using crocodilians

For example, in his study of continental paleoclimate in North America, Markwick (1994) uses crocodilians as his climate indicating organism. Living crocodilians only exist in tropical to subtropical environments. There are no living exceptions to this rule. Furthermore, based on what we know of the physiology of crocodilians (cold blooded - exothermic) there appear to be basic physiological reasons why crocodilians cannot survive in temperate climates. To be conservative in his climate reconstruction, Markwick uses the modern climate distribution of the most cool temperature tolerant crocodilian species, Alligator mississippiensis.

For his control group, Markwick uses the presence of any other terrestrial vertebrates. The assumption here is that if vertebrates are fossilized then the fossilization site will contain crocodilian fossils if the living animals were present. This is also probably reasonable for a number of reasons:

1. Vertebrate fossils are generally associated with freshwater environments - rivers, lakes, etc. which are also the normal habitats of crocodilians.

2. Crocodilian bones and scutes are very distinctive and often abundant when present - i.e. they are less difficult to fossilize and recover than most other vertebrate fossils.

So, it appears that not only should the absence of crocodilian fossils from a vertebrate site be considered to signify the absence of crocodilians from that site, but that the absence of crocodilians does indicate the absence of subtropical to tropical climate conditions.

From the results of his study, Markwick draws the following conclusions:

Eocene - conditions were warm and 'equible' throughout much of the interior of North America.

Late Oligocene - crocodilians are restricted to maritime regions - suggests that interior conditions were cool and or highly seasonal.

Miocene - crocodilians return to the continental interior, presumably because of warmer, less seasonal climate conditions.

Pleistocene - Holocene - crocodilians approach their present day distribution.

Global Ocean temperatures estimated using marine fossil distributions

Many studies have been done using the distributions of marine fossil organisms to infer relative climate conditions and water temperature.

For example, the modern reef-forming corals (hermatypic scleractinians) live symbiotically with zooxanthellae (a type of marine algae related to dinoflagellates) and are entirely tropical in there distribution. Modern species are unable to tolerate temperatures below 16° C and their maximum abundance and diversity occurs in waters that are between 25° and 29° C.

Some corals do live in cooler waters worldwide, but these appear unable to build substantial reef structures, in part because they lack zooxanthellae and in part because it is more difficult to precipitate CaCO3 in cool water. Thus, in the post Triassic, the presence of massive, carbonate reefs is considered a reliable indication of tropical waters.

In the Paleozoic, prior to the evolution of modern corals, their were two groups of Paleozoic corals, the tabulates and the rugosans. It is not known whether or not these were hermatypic, but it seems reasonable to conclude that they only formed massive, high diversity reefs in warm tropical waters. Indeed, the mapped distributions of Paleozoic reefs generally concurs with the positions of continents in tropical latitudes based on other lines of evidence such as paleomagnetic data.