Charcoal is one of the by-products of fire. It is almost pure carbon, and at first sight may be considered an unlikely source of information about past life on the planet. However, when plants are charred by fires many of their very finest anatomical details are not affected. Just as importantly, charcoal is often subsequently left virtually unaffected by the fossilization process, as it is relatively chemically inert. As long as this brittle material is not crushed, beautiful anatomical detail can be seen microscopically, even in specimens as much as 419 million years old. This incredible preservation is not just restricted to wood but can also affect any other plant organ that was charred, no matter what its size or delicacy. Many of the earliest land plant fossils, often measuring just a few millimeters in height, are known in detail because they were preserved as charcoal. Similarly, much of our knowledge of early angiosperm evolution and diversification during the Cretaceous comes from the charcoalification of reproductive structures, even flowers.
SEM Image of Cretaceous flower from Allon, Georgia, USA preserved as charcoal
Ancient charcoal can tell us both about the nature of ancient fire itself, including the minimum temperature any given fire may have reached, as well as the plants being burned. However, research by Ian Glasspool (Field Museum of Natural History) and Andrew Scott (Royal Holloway University of London) also suggests that ancient charcoal in coal deposits can be used to predict past concentrations of atmospheric oxygen. At present the atmosphere is composed of about 21% oxygen but this level was suspected to have been much higher during the past. These levels are of more than just passing interest; relatively high concentrations of atmospheric oxygen on Earth have strongly impacted the evolution of life with links suggested to the evolution of tetrapods, gigantism in sauropods, placental mammals, insect gigantism, powered flight in both birds and insects and to the Permo-Triassic mass extinction.
Predictions of Phanerozoic atmospheric oxygen concentration based on the work of
Bergman et al., 2004, Berner, 2009 and Glasspool and Scott 2010.
The work by Glasspool and Scott exploits the correlation between atmospheric oxygen concentration and flammability to make predictions about the past based on the abundance of charcoal. At low levels (around 15%) fires are unlikely ever to have propagated, no matter how dry the plants; but at oxygen levels of about 30% even wet plants can burn and fires would become rampant producing large volumes of charcoal. The charcoal data indicates that about 400-390 million years ago, during the Middle Devonian, atmospheric oxygen levels were very low, well below modern levels. However, the concentration then rose rapidly over a 30 million year interval following the growth of the Mid-Late Devonian forests and from the Mississippian remained at levels above 26% until the Mesozoic. Phanerozoic atmospheric oxygen levels peaked at about 29% during the Early Permian and remained relatively high until at least the last 2.8 million years of the Permian. A pronounced collapse in atmospheric oxygen concentration occurred following the Permo-Triassic extinction, but this decline appears to have followed rather than driven the extinction. The Mesozoic appears to have experienced large amplitude cyclical fluctuations in atmospheric oxygen concentration throughout the Triassic and Jurassic, levels then rising throughout the Cretaceous until reaching another peak during about 100 million years ago. From this point onwards oxygen levels declined until they appear to have reached relative stability at about 21-22% over the last 40 million years.