AUSTRALIAN FRONTIERS OF SCIENCE, 2003
Canberra, 31 July to 1 August 2003
Mass extinctions
by Dr Annette George
 |
Annette George is a Senior Lecturer in Geology at the University of Western Australia. Previously she held a teaching position at the University of Melbourne following her PhD graduation from Victoria University of Wellington in 1988. Annette's research interests focus on deciphering the tectonic history of sedimentary basins which involves the application and integration of a variety of geological techniques. She coordinates the Sedimentary Research Group at UWA which includes research and postdoctoral research fellows, PhD and honours students. The research also focuses on application of sedimentary techniques to understanding basin-hosted mineral deposits and petroleum reservoirs. Annette has published widely on a variety of sedimentary basins particularly the Devonian reef complexes of northern Australia in recent years. She received the Australian Academy of Science's inaugural Dorothy Hill Award in 2002. |
The topic of mass extinctions has been one that has attracted a considerable amount of research over the last 25 years. Many scientists over these last 25 years have put forward similar, fairly simple kinds of ideas as to why we have had these mass extinction events through Earth history. You can think of popular ones like bolide impacts; and sea level changes has been popular as well. The idea I would like to get across to you today is that, while these mechanisms in themselves are quite valid, what the research has really been showing in the last 5 years or so is that it is a more complicated story and we need to look at the way these mechanisms interact to really establish the causes of mass extinction events.
Click on image for a larger version of figure 1
If we put these extinctions into what we understand of the evolution of our biosphere, in most of the research over many years, there has been a great deal of emphasis on the origin of life (which is currently a hot topic in the Earth sciences) as well as a great deal of interest in how life has evolved. There has been less emphasis on the periods of loss of diversity.
If we consider the rock record, we can see that through a considerable amount of Earth history there have been some very prominent events including the early development of life and the first stromatolites, the start of oxygenation of the oceans, the development of abundant oxygen through a long time of photosynthesis, and, right up at the top of the geological column, at 550 million years ago when there was a tremendous increase in the number of complex and hard-bodied organisms at the Pre-Cambrian-Cambrian boundary. Hence the notion of the 'Cambrian explosion'. If you focus on the green funnel in the diagram (labelled faunal and floral diversity on figure 1), what you can see is that today our modern biosphere is the most diverse it has ever been.
What we want to do, then, is have a look at some of these times when there has been tremendous biotic loss. It is important not just to think of the extinction events as times of great loss. They also provided a tremendous opportunity for the radiation of other kinds of organisms.
What does mass extinction mean? There are two key aspects. The extinction or this biotic loss needs to occur over a geographically wide area, and it also has to occur over a very narrow interval of time, for example 1 to 2 million years or considerably less.
Recognising these mass extinction events relies on some important things. One is that we can actually measure the amount of loss: how many species did we have before, and how many did we have after a particular event? Another is being able to measure the time. One million years in geological time is very short and we are pushing the resolution of many of our dating techniques to be able to separate events that might be 100,000 years long or potentially less.
You will recognise those first two aspects as having a lot of relevance to the way we think about extinctions today. There many arguments about the loss of species during our lifetimes, based simply on the difficulties of measuring the amount of loss and whether we think these losses have occurred over a short period of time or not.
Click on image for a larger version of figure 2
When we look at rocks, the third one becomes important as well. That is to say: how much time is recorded by the rock record? We may have periods of very, very slow sedimentation so that not much material is deposited. An extinction event recorded in these rocks then looks like it happened very quickly. Nonetheless, looking back through the last 550 million years of Earth history we find that there have been times of profound crisis. The top five are highlighted in figure 2. I will focus on three of these, in the Late Devonian, at the Permo-Triassic boundary and at the Cretaceous-Tertiary boundary; Dr Kliti Grice will focus on the Permo-Triassic event.
Click on image for a larger version of figure 3
What are some of the common characteristics of these crises? There has always been a great deal of argument about whether some of these events truly represent 'mass' extinctions, or whether they are not simply the result of ongoing species extinctions. The graphs in figure 3 summarise some of these ideas.
The graph on the left of figure 3 shows you what background extinction rates through geological time are considered to have been. You can see the big five mass extinctions shown by peaks standing up above the envelope of background extinction.
Another important point to keep in mind as well is the nature of extinction summarised on the right of figure 3. Generally with the big crises three end-members can be recognised. The very popular bolide impact theory suggests that extinction would be very catastrophic, something like the middle scenario in figure 3. In fact, most of the evidence from the rock record suggests that extinctions tend to be either fairly stepwise (bottom scenario), or rather graded. This has driven a great deal of argument because it means that some of the mechanisms proposed do not seem to fit very well with the evidence that we have from rocks where the fauna and flora prserved in those rocks actually went extinct well before the supposed extinction horizon.
Most people are familiar with the Cretaceous-Tertiary extinction (Cretaceous is abbreviated to K, hence the 'KT boundary'.) In the Late Cretaceous the landscapes had large, tall but weak kinds of trees, and reptiles including the dinosaurs. From the evidence that has been collected from Cretaceous and Tertiary rocks, the ideas that have been put forward are as follows: important environmental changes, the asteroid impact hypothesis and a second hypothesis that has gone alongside the asteroid impact idea, is that of voluminous volcanic outpourings.
Of all of the extinction events, the one at the KT boundary has always been the considered the most straightforward, and there has been a great deal of interest and emphasis placed on the bolide impact. The volcanic hypothesis involves so much lava being poured out that we could cover the Earth in about a metre of basalt – this means considerably larger kinds of eruptions than those that we see at present day. By comparison, the recent eruption of Mount Pinatubo, which put a great deal of ash into the atmosphere, decreased global mean temperature by about half a degree. If you think about that kind of effect on the considerably larger scale of the Late Cretaceous eruptions, then you can imagine the same kinds of atmospheric conditions that the impact hypothesis people have proposed – lots of dust, a much cooler climate killing off plant material, the plant eaters are in trouble, and the meat eaters are then in trouble as well.
Even so, the KT extinction attracts a lot of controversy, even though this event has benefited from the ability of scientists to date the actual event (65 million years) and events around that time. The basalts that were erupted, and have been well studied in India, show that the bolide impact occurred well after volcanism had started. So the stepwise reduction in the biota that is associated with the Late Cretaceous seems to fit the suggestion of there being multiple causes. Environmental changes, enhanced by the effects of extreme volcanism was the key mechanism, with the asteroid impact effectively being the 'last straw'.
Click on image for a larger version of figure 4
Interpretations begin to get a bit more complicated when we start to look at some of these other mass extinction events. The Permian-Triassic boundary has always been considered the biggest one. Figure 4 gives you an idea of the scale of biotic loss that has been suggested. The more conservative estimates are around 75 per cent of species lost; more dramatic estimates are up around 96 per cent. (The Permo-Triassic boundary is marked by the red line and it is relatively easy to compare the biota either side of the boundary.)
It is also noticeable that before the Permo-Triassic event, there were other events going on that were also acting to decimate some of these fossil groups. The Permian was a time of very dramatic reef growth, but the reef builders suffered considerably prior to the Permo-Triassic boundary.
Click on image for a larger version of figure 5
If we consider what the world looked like at the end of the Permian, reconstructions generally show the continents of the world lumped together as a large supercontinent called Pangaea, with large oceans (Panthalassa and the Paleotethys). The edges of the supercontinent featured quite wide continental shelves (figure 5).
There is a variety of mechanisms proposed for the PT extinction. There has been considerable emphasis on changes in sea level. For a long time researchers supported the idea of major sea-level falls which would expose large areas that were previously flourishing shallow marine environments, or, alternatively, that sea level rises would create much deeper conditions in these shallow marine environments. Global warming is an important mechanism – a lot of the data that have been used to support global warming come from the loss of cool water and cool temperate organisms, for example Glossopteris forests in the higher latitudes.
Anoxia has been an important mechanism that has come up time and time again, particularly for the Permo-Triassic extinction, but it has also been one that has been interpreted in several of the others. The anoxia was probably driven by collapse of the temperature gradient (from the poles to the Equator) as the higher latitudes warmed, creating ocean circulation which was very, very sluggish and a lack of oxygenated water through much of the middle latitudes.
Click on image for a larger version of figure 6
Let us now consider an older event, the Late Devonian: this was a time of prolific reef growth. Figure 6 gives you some idea of the kinds of organisms that built and lived around the Devonian reefs, including types of calcareous types sponges, called stromatoporoids, with various growth forms, stromatolites and molluscs. There would have been fish and all sorts of other organisms producing a diversity that we associate with our modern-day reefs.
The mechanisms that have been proposed for the extinction of these reef complexes include those we have already considered for the other extinctions. Asteroid impact was a very popular theory. However, no evidence has ever been found that fits the actual rock boundary associated with the Late Devonian extinctions. For the KT boundary, the iridium anomaly and various other features have been widely recognised; in the Devonian rocks, no such iridium anomaly at the actual boundary associated with the extinction has been found.
Quite a large list of possible mechanisms has been proposed. Warming oceans was based on the proposal that, as the water got warmer, the organisms were no longer in their habitable 'comfort' zone. However, there is very little data for this, and any warming would have to have involved very high temperates in the the tropical latitudes for this to have affected the reef-building organisms.
Click on image for a larger version of figure 7
The work we have been doing in the Canning Basin over a number of years is is based on understanding the very well-exposed Devonian reef complexes (figure 7) with nice big platforms, that have been called pinnacle reefs, and deeper water areas. The Canning Basin is located in northwestern Australia. The reefs grew along that northwestern edge of the Canning Basin, when that area was a sea, so they snuggled up against what was the elevated landmass of the Kimberley Block. At that time, northern Australia was sitting at quite a low latitude which enabled the prolific growth of reef complexes, which, in the Canning Basin, are about 350 kilometres long. In our research we want to understand how reefs grow, and more recently we have built reef models that allow us to really try and determine the extinction mechanisms in a systematic way from the rock record.
Click on image for a larger version of figure 8
On a schematic model of a carbonate reef (figure 8), there is the actual reef margin and the area behind, which is very shallow, where sediments are produced by the organisms living there. Much of that sediment is transported down the front, down the slope, and out into the basin proper. So reefs have a wide range of dynamic processes going on – different sorts of sediments being deposited and a great range of organisms living in those different environments.
The Canning Basin reefs have had a long history of study, with some fairly important models proposed at various times. Playford recognised a fall at the level that is associated with the extinction event, although he did not think that sea level fall was really responsible for the extinction. More recently, the work of Geoscience Australia geologists recognised several fluctuations in relative sea level, and some of the work that has been going on around the globe in these sections is starting to recognise that it is really the notion of fluctuations in sea level rather than a single fall or a single rise that was potentially creating environmental stresses in these ecosystems.
Click on image for a larger version of figure 9
The Late Devonian extinction is recorded by having earlier reef growth – and the phase of time is here referred to as the Frasnian stage – with the stromatoporoids (shown with a platy habit in figure 9) encrusted by cyanobacteria (white speckling material on figure 9), forming a very rigid framework. The Famennian stage records the second phase of reef growth, so the Frasnian-Famennian boundary is considered the major Late Devonian extinction event. Thus the Famennian reefal rocks, as shown in figure 9, are made up of the calcified cyanobacteria and sediment only.
Click on image for a larger version of figure 10
Finding the Frasnian-Famennian boundary has always been afforded 'Holy Grail' status in the Canning Basin. Recently, we found a surface where we had Frasnian microbial limestones – very pale-coloured on figure 10 – which have been planed off and overlain by these brown limestones of Famennian age. Fragments of the underlying limestones are included. Up close (bottom left of figure 10) the Frasnian microbial fabrics are clearly truncated by erosion. The geometry of the rock layers in this area suggests that erosion of the Frasnian reef occurred as faults uplifted the area creating a relative fall in sea level which exposed the reef complex.
Click on image for a larger version of figure 11
What is really critical for looking at mass extinction events is a strong stratigraphic or rock-based framework in which to look at where the evidence occurs. Here (in Dingo Gap and Barker River, figure 11) is the Frasnian-Famennian boundary. The key thing that we noticed in analysing our data was that the stromatoporoid debris is markedly absent well before we get to the end of the Frasnian. When we look at blocks of reef material, (shown as yellow blobs on figure 11), they are not stromatoporoid reef fabrics; they are the microbial limestones – which means that the reef complexes were being built by cyanoabacteria and microbes well before the end of the Frasnian stage, that is, prior to the major extinction phase.
Our data show relative sea level fluctuations. We have no evidence for anoxia which has been widely proposed in other Late Devonian sections elsewhere. At Dingo Gap (shown in cross-section on figure 11) we see only oxygen-related types of rocks. When we look at the subsurface (eg, Barker River cross-section on Figure 11) we also don't find any evidence for rocks deposited under anoxic conditions. This is a bit of an issue, because the Canning Basin rocks are not really supporting what has become the favoured view elsewhere.
Click on image for a larger version of figure 12
What we aimed to do in our work is generate more dynamic models for reef growth (figure 15). We represent the changes in sea level by these simple sinusoidal curves. When sea level is low, lots of siliciclastic sediment being eroded from the Kimberley mountains is coming out into the reef system generating a lot of environmental stress, either by physically smothering the organisms or bringing out too many nutrients through the run-off, or by making far more turbid conditions than most of these reef organisms can tolerate.
Hallam and Wignall suggested that a relative fall at the Frasnian-Famennian boundary was not very sensible because there was not much evidence worldwide for low sea levels – in fact they suggest that it was a time of global highstand. Yes, there is a fall at the Frasnian-Famennian boundary in the Canning Basin, but it is very much driven by the tectonic activity in the basin (as it opened and stretched) rather than related to any global falls.
Click on image for a larger version of figure 13
Where are we at in terms of understanding the mechanisms responsible for the Late Devonian extinction? From our discussion we can suggest that some are much less likely than others, but there are several that remain as highly viable types of mechanisms to have been operating (figure 13).
How do we think we might solve some of these problems? I think we are starting to put together enough information to know more about the history of sea level change. We still very much lack understanding of the palaeoecology of some of the organisms, particularly things like the stromatolites and the microbial communities., and in the case of the Canning Basin, have good stable isotopic datasets with which to investigate potential environmental changes. The big answers will come from understanding more about microbes, which is also true of several areas in Earth science. So we are hoping that with our good outcrop and subsurface stratigraphic sections and our improving understanding of how the reefs evolved through time, in terms of the rocks and organisms that helped form them, we should be able to generate some more meaningful data to solve this problematic extinction phase.
