Mechanisms of global climate change during the five major mass extinctions

Abstract

Since the emergence of diverse animal phyla around 500 million years ago, five major mass extinction events have occurred, each coinciding with abnormal climate changes. We analyzed sedimentary organic molecules from the first and least understood extinction event at the end of the Ordovician period. We divided all five major extinctions into two phases each, totaling ten events, and examined the relationship between climate shifts and the “coronene index”—an indicator of heating temperatures in sedimentary rocks caused by volcanic activity or meteorite impacts. As a result, we found that four of the five extinctions began with global cooling and ended with warming, while one started with an unknown anomaly and also ended with warming. During the initial extinction phases, two events showed low-temperature heating, two high-temperature, and one moderate-temperature. All subsequent warming phases showed moderate-temperature heating. These findings suggest that large-scale volcanic eruptions and meteorite impacts heated sulfides, sulfates, and hydrocarbons at varying temperatures, releasing SO2 or soot into the stratosphere, blocking sunlight, and triggering global cooling and extinction. This was followed by moderate heating of hydrocarbons and carbonates, increasing CO2 emissions and driving long-term global warming, leading to secondary extinction events.

Introduction

Marine animal diversity at the genus level declined rapidly during five major mass extinction events in the Phanerozoic Eon1,2. The last three of these events also significantly affected terrestrial tetrapods3. Extensive research has examined the relationship between these mass extinctions and climate change4,5,6,7,8,9. While volcanic activity is considered the primary driver of most extinctions, the Cretaceous-Paleogene (K-Pg) extinction is widely attributed to an asteroid impact10,11,12. Global warming played a dominant role in the end-Permian (end-P) extinction, whereas cooling followed by warming characterized the Late Ordovician (LO), Late Devonian Frasnian-Famennian (F-F) boundary, end-Triassic (end-T), and K-Pg extinctions4,5,6,7,8,9,13 (Table 1).

The coronene index serves as a proxy for the heating temperature of sedimentary organic matter and is derived from the ratio of the 7-ring polycyclic aromatic hydrocarbon (PAH) coronene to the sum of 5-, 6-, and 7-ring PAHs, including benzo(e)pyrene, benzo(ghi)perylene, and coronene12,14,15,16,17. Coronene requires significantly higher formation temperatures than smaller PAHs18. Kaiho (2024)12 demonstrated a strong correlation between global surface temperature anomalies and coronene index values, linking them to heating events associated with volcanic eruptions and asteroid impacts.

Under normal conditions, the coronene index remains around 0.114,15,16,19. Diagenetic processes do not significantly generate coronene due to its low abundance, and Proterozoic-Paleozoic strata are dominated by smaller PAHs14,15,20. Wildfires, particularly those in grasslands, burn at relatively low temperatures and do not produce significant amounts of coronene. Marine sediment samples containing wildfire-derived PAHs consistently exhibit low coronene index values, challenging the idea that coronene enrichment results solely from wildfires. Volcanic activity, particularly plume volcanism, can generate coronene due to higher mantle temperatures, resulting in medium coronene index values. In contrast, large-scale continental rift volcanism produces low coronene index values16.

The coronene index distinguishes geological events based on heating intensity. Asteroid impacts and high-temperature plume volcanism result in coronene index values ranging from 0.7 to 1.0, while large plume volcanism produces values between 0.2 and 0.7. Background conditions and other volcanic activity yield values between 0 and 0.212,16. Laboratory experiments indicate that different heating processes have distinct climatic effects. Low-temperature sulfide heating or high-temperature sulfate heating primarily releases sulfur dioxide (SO2), driving global cooling. Intermediate-temperature heating of hydrocarbons and carbonates primarily emits carbon dioxide (CO2), leading to global warming. Medium- to high-temperature shock heating from meteorite impacts generates soot and sulfur trioxide (SO3), further contributing to global cooling12,16.

Kaiho (2024)12 concluded that mass extinctions were triggered by sedimentary rock heating in Large Igneous Provinces (LIPs) and by high-temperature heating of impact target rocks, such as those at the Chicxulub impact site. Low-temperature heating released SO2 from sulfides without significant CO2 emissions, leading to global cooling. Intermediate-temperature heating of hydrocarbons and carbonates released substantial CO2, contributing to global warming. High-temperature sulfate heating resulted in massive SO2 emissions, also driving global cooling.

Despite these findings, coronene index data are missing for key extinction phases, particularly the Late Ordovician Mass Extinction (LOME) and delayed extinction events that followed the primary extinction phases in all five major mass extinctions. This study aims to address these gaps by identifying two distinct extinction events within each of the five major mass extinctions and analyzing coronene index values for ten extinction phases, encompassing both primary and delayed events. To accomplish this, I conducted a new PAH analysis of the first and second LOME using sediment samples from the Wangjiawan site, located on the outer shelf of the South China Craton during the Ordovician–Silurian transition (see Methods for details). I then reviewed temperature anomalies from the literature to further investigate the role of heating temperatures in driving climate change during these mass extinction events.

Results

Identification of two events in each mass extinction event

Each of the five major mass extinctions consisted of two distinct extinction pulses. The LOME experienced two pulses of biodiversity loss4,21,22. The Frasnian-Famennian Mass Extinction (FFME) was followed by a minor biotic decline near the top of the mid-triangularis zone of conodonts6 (Table 1). The End-Permian Extinction (EPE) was accompanied by a second extinction pulse in the earliest Triassic (eTE) within the Permian-Triassic mass extinction (PTME)23. The End-Triassic Mass Extinction (ETME) consisted of primary extinction pulse (Events 1 and 2), followed by Event 3 characterized by low biodiversity (Event 3)16,24. The Cretaceous-Paleogene Mass Extinction (KPgME) initially affected surface marine organisms, followed by a minor delayed extinction event among deep-water species in the P0 zone of planktonic foraminifera19 (Table 1). The time intervals between these paired extinction events are estimated to be 0.5, 0.3, 0.06, 0.2, and ~ 0.1 million years for LOME, FFME, PTME, ETME, and KPgME, respectively, based on the time scales of Gradstein et al.25 and Burgess et al.26 (Table 1).