Despite the downgrading of the raw data for mass extinctions by their stratigraphic weighting (see The “Big Five” become the “Big Three”? Earth Pages, January 2002), which cast doubt on the magnitude of the Triassic-Jurassic boundary event, the Tr-J draws attention because something out of the ordinary did happen then. In terms of changes in fauna and flora, the boundary is globally recognisable. The giant Manicouagan crater in Quebec formed almost at the boundary, and the Central Atlantic Magmatic Province (CAMP) had about the right timing too – 202 to 198 Ma ago. The Camp was the magmatic expression of the first substatially break up of Pangaea. Carbon isotopes in sediments reflect changes in the inorganic and organic parts of the C-cycle, and it seems that a significant excursion from the norm characterizes the change from Triassic to Jurassic biota at four important sections in western Canada, Hungary, Britain and East Greenland (Hesselbo, S.P. et al., 2002. Terrestrial and marine extinction at the Triassic-Jurassic boundary synchronized with major carbon-cycle perturbation: A link to initiation of massive volcanism? Geology, v. 30, p. 251-254).
Around the palaeontologically defined boundary, each section records a sharp decrease in the proportion of heavy 13C to light 12C (d13C), followed by a protracted period dominated by isotopically light carbon in the earliest Jurassic. Such a shift must indicate some kind of global change in the C-cycle. One explanation is an increase in the amount of CO2 expelled from the mantle by major volcanic activity, which tallies with the CAMP. If that was the reason, then d13C would be a proxy for all manner of other effects of volcanic activity – acid rain, atmospheric dust as well as volcanic enhancement of the “greenhouse effect”. However, as several long-running sagas have shown, isotopically light carbon recorded in sediments can also be explained by methane escape from deep sediments, and even by a massive reduction in biological activity that would otherwise sequester light carbon through metabolism. More confusion arises from the carbon that is isotopically analysed. That in carbonate sediments records the isotopic composition of carbon dissolved in seawater, while that found as carbon in preserved organic matter tells a different story. The Tr-J data are from bulk organic carbon in reduced sediments, and relates to the reservoir of carbon on which cell metabolism has drawn – CO2 dissolved in seawater, and that in the air for marine and terrestrial life respectively, both of which are in equilibrium. It looks like the excursion stemmed from an increase in volcanic emissions to the atmosphere, and the CAMP. However, the study by Stephen Hesselbo and colleagues from the Universities of Oxford and Copenhagen, and the Geological Survey of Denmark and Greenland, reveals a glitch. The appearance of the ammonite genus Psiloceras has generally been taken to mark the start of the Jurassic. In Canada, such fossils occur just after the first major shift in carbon isotopes, whereas in Britain Psiloceras appears considerably later. One takes ones’ choice: either ammonite species appeared synchronously everywhere (assuming that they were pelagic ocean-crossers like modern Nautilus), or signs of change in the C-cycle are truly global. Because of very rapid mixing of air and water, on geological timescales, the latter is like to be true and faunal zones are not so reliable as stratigraphers believed over the last century. It seems that as well as tying whatever happened at the Tr-J boundary to massive volcanism, the study marks a turn from traditional palaeontology to geochemical markers as the “golden spikes” in stratigraphy.
There are other means of linking changes in the pace of volcanism and the surface environment, that emerge by careful choice of geochemical proxy data. A long used, but imprecise approach depends on the much slower rate at which radiogenic 87Sr builds up in the mantle than in the continental crust, because of the much lower rubidium content of the mantle. Oceanic strontium isotopes, measured for the past by analysing marine carbonates, reflect the derivation of strontium by erosion and weathering of continental crust, and addition of strontium to sea water by its hydrothermal interaction with newly emerged volcanic rocks at constructive margins or on oceanic plateaux (submarine flood basalts). Since the 87Sr/86Sr ratio of seawater is a commonly used proxy for varying rates of continental weathering, identifying signs of massive increases in submarine volcanism is only possible when carbonates reveal extremely low values of the ratio. The mantle is enormously richer than continental crust in elements likely to have entered the core preferentially, such as gold and especially the platinum group elements. Partial melting allows unusually high amounts of such elements to enter basalt magmas, and the lavas and ejecta that enter the surface environment. Potentially, the environmental abundance of such normally exceeding rare elements is a reliable proxy for major magmatic events. Osmium isotopes are particularly promising, even in tiny concentrations, because two (187Os and 188Os) are daughters of the decay of unstable rhenium isotopes. The details of their use is complex. As well as the exceedingly light carbon isotopes in sediments at and around the Tr-J boundary, marine shales also reveal a sudden increase in osmium abundance, the 187Os/188Os ratio and the abundance of rhenium (Cohen, A.S. and Coe, A.L. 2002. New geochemical evidence for the onset of volcanism in the Central Atlantic Province and environmental change at the Triassic-Jurassic boundary. Geology, v. 30, p. 267-270). The most likely source for globally distributed anomalies of this kind are the volcanic rocks associated with the CAMP, and their reaction with both rainfall and hydrothermal fluids. However, the data do not rule out an extraterrestrial influence by a major impact. That the Tr-J boundary is associated with both the CAMP and the Manicougan impact seems likely to vex geochemists hoping to tie things down to any single trigger.