Land plants begin to appear in the fossil record as early as the late Ordovician (~450 Ma), show signs of diversification during the Silurian and by the end of the Devonian Period most of the basic features of plants are apparent. During the Carboniferous Period terrestrial biomass became so high as to cause a fall in atmospheric carbon dioxide, triggering the longest period of glaciation of the Phanerozoic, and such a boost to oxygen in the air (to over 30%) that insects, huge by modern standards, were able to thrive and the risk of conflagration was perhaps at its highest in Earth’s history. Yet surprisingly, the first signs of massive forest fires appear in the Devonian when vegetation was nowhere near so widespread and luxuriant as it became in the Carboniferous (Kaiho, K. et al. 2013. A forest fire and soil erosion event during the Late Devonian mass extinction. Palaeogeography, Palaeoclimatology, Palaeoecology, v. 392, p. 272-280). Moreover, Devonian oxygen levels were well below those of the present atmosphere and CO2 was more than 10 times even the post-industrial concentration (387 parts per million in 2013). Such atmospheric chemistry would probably have suppressed burning.
Kunio Kaiho of Tohoku University in Japan and colleagues from Japan, the US and Belgium analysed organic molecules in Belgian marine sediments from the time of the late-Devonian mass extinction (around the Frasnian-Famennian boundary at 372 Ma). A range of compounds produced by hydrocarbon combustion show marked ‘spikes’ at the F-F boundary. The thin bed that marks the extinction boundary also shows sudden increase then decrease in δ13C and total organic carbon, indicative of increase burial of organic material and a likely increase in atmospheric oxygen levels. Another biomarker that is a proxy for soil erosion follows the other biogeochemical markers, perhaps signifying less of a binding effect on soil by plant colonisation: a likely consequence of large widlfires. Unlike the biomarkers, magnetic susceptibility of the boundary sediments is lower than in earlier and later sediments. This is ascribed to a decreased supply of detrital sediment to the Belgian marine Devonian basin, probably as a result of markedly decreased rainfall around the time of the late-Devonian mass extinction. But the magnetic data from 3 metres either side of the boundary also reveal the influence of the 20, 40, 100 and 405 ka Milankovich cycles.
This set of environmentally-related data encourages the authors to suggest a novel, if not entirely plausible, mechanism for mass extinction related to astronomically modulated dry-moist climate changes that repeatedly killed off vegetation so that dry woody matter could accumulate en masse during the Frasnian while atmospheric oxygen levels were too low for combustion. A mass burial of organic carbon at the end of that Age then boosted oxygen levels above the burning threshold to create widespread conflagration once the wood pile was set ablaze. Makes a change from continental flood basalts and extraterrestrial impacts… Yet it was about this time that vertebrates took it upon themselves to avail themselves of the new ecological niche provided by vegetation to haul themselves onto land.
The later part of the Devonian (the Frasnian and Famennian Stages) once marked the second largest marine mass extinction (~375 Ma) of the Phanerozoic Eon: it was one of the ‘Big Five’. For the last decade the drop in marine biodiversity in that interval has come under scrutiny: partly because it may have involved several events; no well-supported extinction mechanism has emerged; and extinctions seem have been concentrated on three animal groups – trilobites, brachiopods and reef corals. Something large did happen, as reef-building corals almost disappeared and coral reefs only returned with the rise of modern (scleractinian) corals in the Mesozoic. While the end of the Devonian still figures widely as having experienced a mass extinction event, more detailed palaeontological work at the genus and species level suggests another possibility.
‘Officially’ a mass extinction event must exceed the background extinction rate throughout the Phanerozoic and be above that in immediately preceding and following stages: statistically significant, that is. They always give rise to a marked reduction in biodiversity, but another mechanism can do that without extinctions suddenly increasing. The rate at which new species emerge can fall below that of species extinctions, when the overall number of living species falls. As far as ecosystems are concerned both processes are equally severe, but the causes may be very different.
Reviewing detailed records of Devonian species of two genera of brachiopods and one bivalve genus (50 species in all) in five North American stratigraphic sequences, Alycia Stigall of Ohio University, USA noted apparently significant variations in the local assemblages (Stigall, A. L. 2012. Speciation collapse and invasive species dynamics during the Late Devonian ‘Mass Extinction’. GSA Today, v. 22(1), p. 4-9). Speciation overall fell in the Frasnian and the preceding Givetian, while rate of extinction barely changed. For the three studied genera ,speciation reached low values in the Frasnian and Famennian, but that was accompanied by an equally large fall in extinctions. In this narrow sample we seem to be seeing not an extinction crisis but one of biodiversity. Why?
The Late Devonian saw repeated ups and downs in sea level, which repeatedly connected and disconnected shallow- to moderate-depth marine basins. The fossil record shows repeated cases of species from one basin colonising another, each invasion accompanying rapid marine transgression.. One means whereby species arise is through prolonged isolation of separate populations of the ancestral species through independent genetic drift and mutation. The episodic connection of basins may have prevented such allopatric speciation. Interestingly, the invading species were dominantly animals with a broad tolerance for environmental conditions.
Whether this mechanism applied to all three main animal groups whose diversity plummeted in Late Devonian times remains to be seen, and it begs the question ‘why didn’t it happen among other animal groups that were less affected by whatever the events were?’ One of the problems associated with decreasing biodiversity in modern marine (and terrestrial) settings is growth in the numbers of invasive species, so the work on 375 Ma fossils might help understand and mitigate current ecological issues. The only difference is that for many of the hyper-successful invader species the means of invasion has been provided by human activities. brachiopod brachioopod
Plot the ages of major extinctions against those of flood basalt events and you will get a straight line graph for six co-occurrences since 250 Ma, with very little error. Although the exact mechanism for mass death of species and families is argued over interminably, for those six, flood basalt events have to be deeply implicated. There again, every geologist and their aunties dispute the mechanisms behind monster basalt effusions that bury whole landscapes beneath flow after flow and create very distinctive landforms. When they are eroded they form regularly stepped mountain sides, hence their formerly popular name trap basalts, after the Swedish word trappa meaning staircase. There is a hint of cyclicity in their age distribution. But most important of all, no-one has witnessed these vast, pulsating events, the last having mantled the surroundings of the Columbia and Snake River catchments in the US states of Oregon and Washington between 14-17 Ma ago in the Middle Miocene. Some mark episodes of continental break-up, such as those flanking the Central Atlantic at the time of the end-Triassic (~200 Ma) mass extinction, while others are associated with hot spots, such as the Deccan Traps of western India erupted between 60-68 Ma as India drifted over the Reunion hot-spot and those of the Ethiopian highlands (30 Ma) associated with the Afar hot spot.
A common geochemical feature is beginning to emerge concerning the mantle from which the basalts were partially melted. Six sets of flood basalts exhibit the same trace-element and isotopic (Nd, Pb, Hf and He) characteristics, which suggest that their source had been little effected by previous extraction of crust-forming magmas; it is primitive and may be a relic of the original mantle formed at about 4500 Ma shortly after the catastrophic collision between the early Earth and a wandering Mars-sized planet that flung off the Moon (Jackson, M.G. & Carlson, R.W. 2011. An ancient recipe for flood basalt genesis. Nature, online (27 July 2011)doi:10.1038/nature10326). Although undepleted, the chemistry of the mantle source, worked out by back-calculation from that of the flood basalts, is not the same as the once-postulated original accretion of carbonaceous chondrite meteorites: conceivably a result of the chemical reworking when the Moon formed and the remaining Earth was probably molten from top to centre. The important feature is that the recast chemistry is rich in heat-producing elements compared with the source of ‘common-or-garden’ basalts that continually contribute to the ocean floors and island arcs. Wherever the relic mantle is, it is capable of heating itself, over and above the heating from the core and surrounding mantle, and thus likely to generate thermal and material plumes rising through the mantle.
Preceding the work of Jackson and Carlson, another group discovered that when flood basalt events since the Carboniferous are restored to their former geographic positions at the time they were erupted, they cluster above what are now two patches of more ductile mantle close to the cure-mantle boundary (Torsvik, T.H. et al. 2010. Diamonds sampled by plumes from the core–mantle boundary. Nature, v. 466, p. 352–355). If that is the source of basalt flood-forming plumes, then it is still there and, aside from giant impacts with extra-terrestrial projectiles, the most catastrophic upheavals of the Earth system inevitably will continue, perhaps in the next few million years.