The end-Ordovician mass extinction was the first of five during the Phanerozoic, andthe first that involved multicelled organisms. It happened in two distinct phases that roughly coincided with an intense but short-lived glaciation at the South Pole, then situated within what is now the African continent. Unlike the other four, this biotic catastrophe seems unlinked to either a major impact structure or to an episode of flood volcanism.
Artist’s impression of an Ordovician shallow-sea community (credit: drtel)
In 2009 Earth Pages reported the curious occurrence in 470 Ma (Darriwilian Stage) Swedish limestones of a large number of altered chondritic meteorites, possible evidence that there may have been an extraterrestrial influence on extinction rates around that time. In support is evidence that the meteorite swarm coincided with megabreccias or olistostromes at what were then Southern Hemisphere continental margins: possible signs of a series of huge tsunamis. But in fact this odd coincidence occurred at a time when metazoan diversity was truly booming: the only known case of impacts possibly favouring life.
Number One of the Big Five mass extinctions occurred during the late-Ordovician Hirnantian stage (443-445 Ma) and has received much less attention than the later ones. So it is good see the balance being redressed by a review of evidence for it and for possible mechanisms (Harper, D.A.T et al. 2014. End Ordovician extinctions: A coincidence of causes. Gondwana Research, v. 25, p. 1294-1307). The first event of a double-whammy mainly affected free-swimming and planktonic organisms and those of shallow seas; near-surface dwellers such as graptolites and trilobites. The second, about a million years later, hit animals living at all depths in the sea. Between them, the two events removed about 85% of marines species – there were few if any terrestrial animals so this is close to the extinction level that closed the Palaeozoic at around 250 Ma.
No single process can be regarded as the ‘culprit’. However the two events are bracketed by an 80-100 m fall in sea level due to the southern hemisphere glaciation. That may have given rise to changes in ocean oxygen content and in the reduction of sulfur to hydrogen sulfide. Also climate-related may have been changes in the vertical, thermohaline circulation of the oceans, falling temperatures encouraging sinking of surface water to abyssal depths providing more oxygen to support life deep in the water column. Sea-level fall would have reduced the extent of shallow seas too. Those consequences would explain the early demise of shallow water, free swimming animals. Reversal of these trends as glaciation waned may have returned stagnancy and anoxia to deep water, thereby affecting life at all depths. The authors suggest generalized ‘tipping points’ towards which several global processes contributed.
The global glaciations of the Neoproterozoic that reached low latitudes – the so-called ‘Snowball Earth’ events have dominated accounts of ancient glaciations since the start of the 21st century. Yet they are not the oldest examples of large-scale effects of continental ice sheets. Distinctive tillites or diamictites that contain large clasts of diverse, exotic rocks occur in sedimentary sequences of Archaean and Palaeoproterozoic age. The oldest are dated at around 2.9 Ma in the Pongola Supergroup of Swaziland, South Africa and formed at an estimated palaeolatitude of 48°; within the range of the equatorward extent of Pleistocene ice sheets. No evidence has turned up for glaciation of that age in other regions, and therefore for a ‘Snowball Earth’ at that time. The surprise is not the antiquity of the Pongola glaciation but the fact that tillites formed by glaciers are not more common in the early part of geological history. The sun has increased in its warming effect since the Earth formed so that the very absence of glaciations over huge spans of early Precambrian time points strongly towards an early atmosphere far richer in greenhouse gases than it is now.
Evidence for Palaeoproterozoic glaciation is more widespread, important tillites occurring in the Great Lakes region of North America and in the Transvaal and Griqualand regions of South Africa. Those of South Africa formed at a latitude of around 10°, suggesting ‘Snowball’ conditions, and in each region there are multiple tillites in the stratigraphic column. Accurate dating of volcanic ash horizons in the sequences of both areas (Rasmussen, B. et al. 2013. Correlation of Paleoproterozoic glaciations based on U-Pb zircon ages for tuff beds in the Transvaal and Huronian Supergroups. Earth and Planetary Science Letters, v. 382, p. 173-180) has made it possible to correlate three glacial deposits precisely between the two now widely separated areas. The dating also reveals that four glacial events occurred over a period of 200 Ma between 2.45 and 2.22 billion years ago: longer than the duration of the Mesozoic Era of the Phanerozoic and about the same as the time span during which 3 or 4 ‘Snowball’ events plastered the planet with ice in the Cryogenian and Ediacaran Periods of the Neoproterozoic.
Diamictite from the Palaeoproterozoic Gowganda Formation in Ontario Canada (credit: Canadian Sedimentology Research Group)
This episode of the first large-scale glaciations neatly brackets the first appearance of significant amounts of oxygen in the Earth’s atmosphere during the Great Oxidation Event from 2.45 to 2.2 Ga. It is hard to avoid the conclusion that the two were connected as an increase in oxygen in the air must have influenced the concentration of greenhouse gases, especially that of methane, the most powerful of several that delay loss of heat to space by radiation from the surface. Once oxygen production by photosynthetic organisms exceeded a threshold atmospheric methane would very rapidly have been oxidized away to CO2 plus water vapour, leaving excess oxygen in the air to prevent the build-up of methane thereafter as is the case nowadays. But what pushed atmospheric composition beyond that threshold? A key piece of evidence lies in the record of different carbon isotopes in seawater of those times, which emerges from their study in Precambrian limestones.
After the end of the Archaean Eon at 2.5 Ga the proportion of marine 13C to 12C increased dramatically. Its accepted measure (δ13C) changed rapidly from the near-zero values that had previously characterised the Archaean to more than 10; an inflated value that lingered for much of the half-billion years that spanned the Great Oxidation Event and the Palaeoproterozoic glaciations (Martin, A.P et al. 2013. A review of temporal constraints for the Palaeoproterozoic large, positive carbonate carbon isotope excursion (the Lomagundi–Jatuli Event). Earth-Science Reviews, v. 127, p. 242-261). Later times saw δ13C return to hovering between slightly negative and slightly positive values either side of zero until the Neoproterozoic when once more ‘spikes’ affected the C-isotope record during the period of the better known ‘Snowball’ events. What lay behind this very broad carbon-isotope anomaly?
To increase 13C at the expense of 12C requires to removal from seawater of very large amounts of the lighter isotope. The only likely mechanism is the prolonged and permanent burial of masses of organic material, the only substances that selectively take up 12C. In turn, that implies a huge increase in biological productivity and its efficient burial without being oxidised to CO2 plus water. There are three possibilities: oxygen was absent from the ocean floor; sedimentation was too fast for oxidising bacteria to keep pace or such bacteria did not evolve until the end of the Lomagundi–Jatuli Event. It seems likely that such a dramatic change in the biosphere may have marked some fundamental shift in biological evolution not long after the close of the Archaean. Whichever, the biosphere somehow increased its capacity to generate oxygen. Since oxygen is anathema to many kinds of anaerobic bacteria and archaea, probably the only kinds of organism at the outset of these events, it is possible to imagine continual extinctions yet to maintain high biological productivity new organisms may have emerged to replace those that vanished. By 2.0 Ma, the first putative eukaryote cells (those with nuclei and a variety of organelles) had appeared.
More on ‘icehouse’ and ‘greenhouse’ Earths
Geologists realized long ago that volcanic activity can have a profound effect on local and global climate. For instance, individual large explosive eruptions can punch large amounts of ash and sulfate aerosols into the stratosphere where they act to reflect solar radiation back to space, thereby cooling the planet. The 1991 eruption of Mt Pinatubo in the Philippines ejected 17 million tones of SO2; so much that the amount of sunlight reaching the Northern Hemisphere fell by around 10% and mean global temperature fell by almost 0.5 °C over the next 2 years. On the other hand, increased volcanic emissions of CO2 over geologically long periods of time are thought to explain some episodes of greenhouse conditions in the geological past.
Ash plume of Mount Pinatubo during its 1991 eruption. (credit: Wikipedia)
The converse effect of climate change on volcanism has, however, only been hinted at. One means of investigating a possible link is through the records of volcanic ash in sea-floor sediment cores in relation to cyclical climate change during the last million years. Data relating to the varying frequency volcanic activity in the circum Pacific ‘Ring of Fire’ has been analysed by German and US geoscientists (Kutterolf, S. et al. 2013. A detection of Milankovich frequencies in global volcanic activity. Geology, v. 41, p. 227-230) to reveal a link with the 41 ka periodicity of astronomical climate forcing due to changes in the tilt of the Earth’s axis of rotation. This matches well with the frequency spectrum displayed by changes in oxygen isotopes from marine cores that record the waxing and waning of continental ice sheets and consequent falls and rises in sea level. Yet there is no sign of links to the orbital eccentricity (~400 and ~100 ka) and axial precession (~22 ka) components of Milankovitch climatic forcing. An interesting detail is that the peak of volcanism lags that of tilt-modulated insolation by about 4 ka.
At first sight an odd coincidence, but both glaciation and changing sea levels involve shifting the way in which the lithosphere is loaded from above. With magnitudes of the orders of kilometres and hundreds of metres respectively glacial and eustatic changes would certainly affect the gravitational field. In turn, changes in the field and the load would result in stress changes below the surface that conceivably might encourage subvolcanic chambers to expel or accumulate magma. Kutterolf and colleagues model the stress from combined glacial and marine loading and unloading for a variety of volcanic provinces in the ‘Ring of Fire’ and are able to show nicely how the frequency of actual eruptions fits changing rates of deep-crustal stress from their model. Eruptions bunch together when stress changes rapidly, as in the onset of the last glacial maximum and deglaciations, and also during stadial-interstadial phases.
Whether or not there may be a link between climate change and plate tectonics, and therefore seismicity, is probably unlikely to be resolved simply because records do not exist for earthquakes before the historic period. As far as I can tell, establishing a link is possible only for volcanism close to coast lines, i.e. in island arcs and continental margins, and related to subduction processes, because the relative changes in stress during rapid marine transgressions and recessions would be large.. Deep within continents there may have been effects on volcanism related to local and regional ice-sheet loading. In the ocean basins, however, there remains a possibility of influences on the activity of ocean-island volcanoes, though whether or not that can be detected is unclear. Some, like Kilauea in Hawaii and La Palma in the Canary Islands, are prone to flank collapse and consequent tsunamis that could be influenced by much the same process. Another candidate for a climate-linked, potentially catastrophic process is that of destabilisation of marine sediments on the continental edge, as in the Storegga Slide off Norway whose last collapse and associated tsunami around 8 thousand years ago took place during the last major rise in sea level during deglaciation. The climatic stability of the Holocene probably damps down any rise in geo-risk with a link to rapid climate change, which anthropogenic changes are likely to be on a scale dwarfed by those during ice ages.