Late Ordovician mass extinction caused by global warming or cooling?

Renqiang Liao· Weidong Sun,3

Abstract The Late Ordovician mass extinction (LOME)was the first global extinction with the destruction of 85%of marine species. However, the cause of LOME is still controversial. Most studies attribute it to large-scale volcanism caused by global cooling or warming. Through analyzing the driving difference between global cooling and warming on large-scale magmatism,the perspective is intended to evoke a hot discussion on the cause of LOME.Did global cooling or warming trigger the LOME?

Keywords Late Ordovician · Mass extinction ·Volcanism · Glaciation

There have been ‘Big Five’extinctions on Earth since the Phanerozoic Era. The Late Ordovician mass extinction(LOME) was the first global mass extinction, which ranks second among the five mass extinctions, with the destruction of 85 % of marine species (e.g., Sheehan 2001). Previous studies suggested that the majority of extinction events are temporally related to large igneous provinces(e.g., Courtillot and Olson 2007). However, the cause of LOME is still controversial, with models ranging from glaciation (Ling et al. 2019; Trotter et al. 2008),environmental perturbations (Algeo et al. 2016), oceanic redox changes (Hammarlund et al. 2012), large-scale volcanism (Grasby et al. 2019; Jones et al. 2017), or multifactors co-evolution (Harper et al. 2014).

Recently, Bond and Grasby (2020) analyzed Hg contents and other biogeochemical indicators (e.g., Mo, U,TOC, Zn) of a set of Ordovician-Silurian deep-water sediments exposed at Dob’s Linn (Scotland). Based on these data,they suggested that LOME was driven by global warming caused by large-scale volcanism. This challenges previous models that LOME was the result of global cooling and fast glaciation based on oxygen isotopes (e.g.,Trotter et al. 2008) and high precision U-Pb dating results(Ling et al. 2019).

The use of mercury (Hg) as an indicator of ancient volcanoes in sediments has arisen in the past decade,which was taken as key evidence for solving this issue (Grasby et al. 2019). Before human activities, Mercury is mainly released from volcanoes or hydrothermal fluids (Pyle and Mather 2003). Hg has a long residence time of ～0.5 to 1 year in the atmosphere,allowing long-distance transport,dispersal, and mixing (Selin 2009). Meanwhile, Hg has a relatively short residence time in the ocean (～102to 103year), such that the signal can be recorded by marine sediments (Gill and Fitzgerald 1988).

Many factors cause the extinction of marine organisms,such as the import of toxic substances,ocean acidification,and ocean anoxia (Grasby et al. 2019). Among them, the input of toxic Hg is particularly eye-catching. The Hg content in the crust and mantle is low (＜10 ppb) (Canil et al.2015;Rudnick and Gao 2014).When there is organic matter or sulfide in the geological body, the Hg content increases sharply, which can reach hundreds or even thousands of ppb (Grasby et al. 2019).

The Hg content in the Hirnantian strata where the LOME event occurred was abnormally high(～ 500 × background) (Jones et al. 2017; Shen et al.2019). Most scholars suggested that the high Hg contents are originated from the Late Ordovician large-scale volcanic activities(Gong et al.2017;Grasby et al.2019;Jones et al. 2017). However, based on geochemical analyses of three Ordovician-Silurian classic sections(Yanzhi Section,Jiaoye Section, and Qiliao Section) in South China, Shen et al.(2019) attributed the Hg enrichments of the Ordovician-Silurian section in South China to the presence of Hgrich sulfide in the strata, which does not require volcanic origin.This finding challenges the hypothesis of a volcanic driver for the LOME. Nevertheless, ppb to ppm levels of Hg are far too low to cause mass extinction.

Fig. 1 Schematic illustration showing the relationship between the strength of volcanic activities and glacier variabilities.a Deglaciation leads to the unloading and decompression of land glaciers, promoting the eruption of land volcanoes; b Glaciation reduces sea level, which induces decompression and rising of oceanic plate,promoting the melting at the mid-ocean ridges.Meanwhile,the mid-ocean ridge rises relative to the old oceanic crust, which results in an increase of the inclination angle of oceanic plate and then acceleration ofthe subduction rate of oceanic plate, enhancing the island arc volcanic activities (modified after Sun and Liao 2020)

In addition to the input of toxic substances, ocean acidification and anoxia can also lead to the extinction of organisms. Both of them are related to large-scale volcanism (Grasby et al. 2019). Bond and Grasby (2020)proposed that LOME is due to the release of a large amount of greenhouse gases (e.g., CO2and CH4) from volcanism,which caused global warming (i.e., LOME pulse1 is located at the end of global warming),resulting in deglaciation,expanding the pre-existing oxygen minimum zone (OMZ),and triggering a biological extinction event. Previous studies suggested that global warming will lead to the unloading and decompression of land glaciers, promoting the eruption of land volcanoes (Huybers and Langmuir 2009), which further releases greenhouse gases (Fig. 1a).Meanwhile, a volcanic eruption brings a large amount of nutrients and promotes biological prosperity (e.g., the Great Ordovician Biodiversification Event before LOME;Grasby et al. 2019), which in turn brings organic matter increases. The decomposing of organic matters consumes more oxygen in the deep sea, and further expands OMZ(Grasby et al. 2019). All these together form a positive feedback mechanism that causes the rapid death of organisms. However, this scenario is not supported by oxygen isotopes and other observations, which suggest cooling, instead of global warming (e.g., Ling et al. 2019;Trotter et al. 2008).

In fact,large scale glaciation during global cooling may also trigger large-scale volcanism (Crowley et al. 2015).Because sea level decreases during the formation of glaciers, which induces decompression and rising of the oceanic plate, promoting the melting of the asthenosphere and upper mantle(Boulahanis et al.2020).Meanwhile,the mid-ocean ridge rises relative to the old oceanic crust,which consequently results in an increase of the inclination angle of the oceanic plate and then acceleration of the subduction rate of the oceanic plate,enhancing the volatilerich island arc volcanic activities(Fig. 1b)(Sun 2019).The distribution of multiple layers of K-bentonite characterized by continental marginal arc in south China is considered to be the product of volcanism during global cooling (Ling et al. 2019).

The dropping seawater87Sr/86Sr ratios around the Middle-Late Ordovician was attributed to high volcanic activities (Young et al. 2009) or an unknown super plume event (Lefebvre et al. 2010), that might have been related to a trap emplacement of basaltic rocks on a continent that now has been eroded. Large mass-independent sulfur isotope fractionation at the LOME suggests ‘stratospheric volcanic eruption’, which presumably resulted in environmental deterioration (Hu et al. 2020).

The drastic and rapid changes in the ecological environment during geological time have caused mass extinctions. Related scientific questions are worthy of further studying: Are there large-scale volcanic eruptions at the end of the Ordovician?What are the controlling factors that trigger large scale volcanic eruptions, global warming, or global cooling? How did different ecosystems respond to global change? These issues involve the interaction between the Earth’s interior and the surficial environment and help to understand the formation of a livable planet.