The Triassic was a crucial period for shaping the modern world, in terms of the evolution of both the biosphere and the geosphere. While successful faunas and floras spread over Pangea, in the Panthalassa and Tethys oceans, the Triassic biosphere was devastated by two extreme extinctions at its dawn (ca. 252 Ma) and its end (ca. 201 Ma) and by a combined extinction and significant radiation episode at ca. 232 Ma, during the Carnian (e.g., Tanner et al., 2004; Song et al., 2013; Bernardi et al., 2018; Dal Corso et al., 2020, 2022; Wignall and Atkinson 2020). The three main biocrises coincided with the emplacement of three Large Igneous Provinces (LIPs), namely, the Siberian Traps, the Central Atlantic Magmatic Province (CAMP), and the Wrangellia (Figure 1). In this article, we review the main features of these LIPs, with a focus on their timing, emplacement styles, and CO₂ budgets. This is followed by an in-depth discussion of their critical roles in the Triassic environmental crises.

High-precision geochronology proves that the main activity of the Siberian Traps and the CAMP (main pulses at 252.3–251.4 Ma and 201.6–201.0 Ma, respectively; for the Siberian Traps: Kamo et al., 2003; Burgess and Bowring, 2015; Burgess et al., 2017; Augland et al., 2019; for the CAMP: Blackburn et al., 2013; Davies et al., 2017, 2021; Marzoli et al., 2019) spanned the end-Permian and end-Triassic extinction intervals, respectively (251.94–251.88 Ma and 201.51–201.36 Ma; Schoene et al., 2010; Burgess et al., 2014; Wotzlaw et al., 2014). The pulsed nature of these LIPs was inferred from magnetostratigraphy and biostratigraphy (e.g., Knight et al., 2004; Panfili et al., 2019; Latyshev et al., 2020), yielding eruption rates possibly two orders of magnitude greater than the most voluminous present-day eruptions, such as those at Hawai’i (e.g., Patrick et al., 2020). The Siberian Traps and the CAMP are among the most voluminous known LIPs (both ca. 3 ± 1 million km³; Saunders and Reichow, 2009; Marzoli et al., 2018), although a precise estimate of the original and preserved magma volume remains challenging. Large parts of the Siberian Traps are intrusive, subaerial, and huge volumes of volcanics were likely removed by erosion. The best-studied part of the province is on the Siberian Craton, at a paleolatitude of ca. 50–60°N, over an area of ca. 2.5 million km² (Figure 1A). The CAMP was emplaced over 10 million km² straddling the equator, and has been similarly partly eroded away, as demonstrated by the sparsely preserved lava piles (Figure 1C).

The first emplacement stage of the Siberian Traps started shortly before 252.24 ± 0.12 Ma with initial pyroclastic eruptions followed by effusive activity emplaced in ca 0.30 Ma. From 251.91 ± 0.07 Ma, the extrusive activity waned and shifted to dominantly intrusive for ca. 0.40 Ma, building the vast network of sills in the Tunguska Basin. These tholeiitic magmas were produced by shallower and extensive melting of a mixed peridotitic-pyroxenitic mantle (Sobolev et al., 2011; Callegaro et al., 2021). Renewed extrusive activity is recognized from 251.48 ± 0.09 Ma, marking the beginning of the third phase of the Siberian Traps. During this phase, tholeiitic extrusive and intrusive activities were accompanied by mafic to felsic alkaline intrusions in Maymecha-Kotuy and Taimyr areas (Elkins-Tanton et al., 2007; Sobolev et al., 2009; Augland et al., 2019). Alkaline melts were produced from a deep (5.5 GPa) and volatile-rich (carbonated) mantle source (Elkins-Tanton et al., 2007). The youngest sill dated in Tunguska (251.35 ± 0.09 Ma; Burgess and Bowring, 2015), and the Guli carbonatitic complex in the Maymecha-Kotuy region (250.20 ± 0.30 Ma; Kamo et al., 2003) represent the youngest known Siberian Traps products.

The beginning of the second (intrusive) phase is considered as the deadly subinterval of the Siberian Traps (Burgess et al., 2017), since it coincides with the beginning of the extinction horizon at Meishan (251.94 ± 0.04 Ma; Burgess et al., 2014), Dongpan and Penglaitan (251.94 ± 0.03 Ma and 251.98 ± 0.03 Ma; Baresel et al., 2017), China. Even if these tholeiitic magmas were probably volatile poor (Sibik et al., 2015), the interaction between the sills and the sedimentary host-rocks (i.e., evaporites, carbonates, marlstones, and coal) likely produced large amounts of thermogenic carbon, sulfur and halocarbon species (Svensen et al., 2018; Callegaro et al., 2021; Sibik et al., 2021).

The end of the Triassic witnessed the emplacement of the CAMP mainly between 201.64 ± 0.03 Ma and 201.36 ± 0.02 Ma (Blackburn et al., 2013; Davies et al., 2017, 2021; Marzoli et al., 2019), coincident with the main end-Triassic extinction interval (201.51 ± 0.15 to 201.36 ± 0.15 Ma; Schoene et al., 2010; Wotzlaw et al., 2014). Basaltic magmas erupted as short-lived pulses (Knight et al., 2004; Marzoli et al., 2019) during this magmatic phase have been shown to be rich in CO₂ and SO₂ (Callegaro et al., 2014a; Capriolo et al., 2020) and possibly triggered the dramatic increase in CO₂, the rise of temperature and the climatic perturbations heralding the end of the Triassic (Landwehrs et al., 2020; Lindström et al., 2021; Capriolo et al., 2022). Later erupted magmas, for example, in North America and Africa (Holyoke and Recurrent groups; Blackburn et al., 2013; Marzoli et al., 2018, 2019) appear to be less voluminous and possibly had limited environmental effects.

Similar to the Siberian Traps, CAMP magmas are mainly preserved as large sills intruding Paleozoic terrestrial sediments in South America and North-western Africa, and Triassic terrestrial sediments in North America and Europe (Marzoli et al., 2018 and references therein). The CAMP basalts mobilized large amounts of organic carbon and possibly halogens from evaporitic deposits (Heimdal et al., 2018, 2019; Capriolo et al., 2021). Available geochronological data indicate that some CAMP intrusions are older than the preserved (and dated) erupted basalts (Davies et al., 2017). However, the intrusion of sills in Amazonia and north-western Africa seems to be relatively prolonged from ca. 201.53 ± 0.07 Ma to ca. 201.35 ± 0.03 Ma (Davies et al., 2017, 2021), i.e., spanning the entire end-Triassic extinction interval (Schoene et al., 2010). However, it is still unclear whether CAMP intrusions occurred in pulses, thus emitting thermogenic volatiles at high rates, and playing a catastrophic role in the end-Triassic crisis.

Unlike the Siberian Traps, CAMP magmas are relatively uniform in composition, mainly being tholeiitic basalts. Whether such relatively high degree mantle melts originated from the deep or shallow mantle is still disputed (e.g., Ruiz-Martinez et al., 2012; Tegner et al., 2019; Boscaini et al., 2022). CAMP basalts lack geochemical signatures of a mantle plume originating from the deep mantle but show a subduction signature, which is best interpreted as reflecting the presence of recycled subducted continental rocks (sediments, mainly; Callegaro et al., 2013; Merle et al., 2014; Shellnutt et al., 2018; Marzoli et al., 2019). The absence of alkaline mafic rocks or carbonatites in the CAMP calls against a significant contribution from particularly CO₂-rich metasomatized mantle portions. In contrast to the Siberian Traps, evidence for explosive volcanic activity is rare for the CAMP, whose lavas are mostly compound pahoehoe flows (e.g., El Hacimi et al., 2011) and therefore occurred as fissure eruptions like the historical Laki eruption in Iceland.

Eruptions of the Siberian Traps, the CAMP, and the Wrangellia LIPs deeply reshaped the Triassic world. In particular, the Siberian Traps and the CAMP emplaced in continental settings as pulsed magmatic events, and both bear strong evidence of interaction between magmas and sedimentary country rocks. A clear difference between these LIPs is shown by the widespread explosive volcanism and abundant alkaline magmatism marking the early and final phases of the Siberian Traps, but unknown for the CAMP. On the other hand, the Wrangellia presents substantial differences as it was emplaced in an oceanic setting, and partially consists of subaqueous lava flows. Due to the lack of high-precision geochronology it is also impossible to clarify whether its magmatic activity was pulsed or continuous.

A recurrent aspect of the Triassic period is that multiple NCIEs were reported worldwide for the end-Permian (e.g., Wu et al., 2021) and end-Triassic mass extinction intervals (Ruhl et al., 2020), as well as for the CPE (Sun et al., 2019; Dal Corso et al., 2018, 2020). NCIEs would testify for the injection of large quantities of isotopically depleted CO₂ into the atmosphere and oceans. Likely, CO₂ (and other greenhouse and poisonous gases) would have been released either from the erupted magmas (Capriolo et al., 2020) or from devolatilization of the intruded host-rocks (Svensen et al., 2009; Capriolo et al., 2021), implying that LIPs-related magmatism was the main driver for catastrophic environmental changes of the Triassic. Since the magnitude of these global crises is also related to the rates at which volatiles were generated, quantifying the process of magma degassing is crucial. Direct estimates of released CO₂ were obtained from fresh, non-recrystallized melt inclusions (e.g., Sobolev et al., 2009; Black et al., 2014; Sibik et al., 2015; Capriolo et al., 2020). However, these are few (and potentially not entirely representative) for the Siberian Traps and the CAMP, and completely lacking for the Wrangellia. We therefore adopted here the approach of Rosenthal et al. (2015) to estimate the initial CO₂ contents of basaltic magmas from their incompatible trace element compositions. In fact, Nb and CO₂ behave similarly during partial mantle melting, both strongly partitioning into the melt phase. However, at shallow depth in the crust, CO₂ is degassed as magma ascends towards the surface (Capriolo et al., 2020), while Nb remains in the melt. We applied this proxy to whole-rock compositions of the three Triassic LIPs (Figure 2) filtered for basalts with MgO >7 wt% to consider only those magmas closer to the composition of mantle-derived primitive melts, as suggested by Hernandez Nava et al. (2021).

CO₂ concentrations estimated for magmas of the Siberian Traps are significantly variable within the three magmatic phases previously described (Figure 2). High-Ti lava flows of the Lower Unit and the low-Ti magmas of the Middle-Transitional Unit yielded the highest calculated CO₂ contents (ca. 0.1–1.2 and 0.05–0.7 wt%, respectively; Figure 2). In contrast, lavas of the Upper Unit and correlated intrusions, which characterize the second and more voluminous Siberian main phase, yielded lower CO₂ contents (ca. 0.1–0.3 wt%; Figure 2). Within CAMP, the earliest erupted basalts from north-western Africa are relatively enriched in CO₂ (0.4–0.7 wt%), while the slightly later and ubiquitous low-Ti magmas of the Prevalent group (Callegaro et al., 2013; Merle et al., 2014; Marzoli et al., 2018, 2019) yielded slightly lower CO₂ concentrations (ca. 0.1 to 0.5 wt%; Figure 2). These results would imply that, for both the Siberian Traps and the CAMP, the highest concentrations of mantle-derived CO₂ were produced from magmas of the initial phases. Interestingly, the end-Permian extinction coincides with the initial emplacement of the volatile-poor sills of the Siberian Traps, while the end-Triassic extinction coincides with the effusive activity of the CAMP, whose lavas are relatively CO₂-rich (Capriolo et al., 2020). The fact that variable amounts of mantle-derived CO₂ from the magmas produced similar environmental consequences suggests that additional isotopically depleted CO₂ was likely released during the intrusive Siberian Traps activity. In this case, a significant part of the emitted CO₂ (and possibly other gases, e.g., CH₄, SO₂ and halocarbons) may have been produced thermogenically from the intruded host rocks (e.g., Svensen et al., 2009, 2018; Capriolo et al., 2021) highlighting the crucial role of sedimentary basins as additional suppliers of climate-modifying gases. Noteworthy, the early terrestrial onset of the end-Permian extinction seems to suggest that mantle-derived halogens were key environmental impactors during the first phase of the Siberian Traps (Sobolev et al., 2011; Dal Corso et al., 2022).

Geochemical data for the Wrangellia are relatively scarce. However, the high-Ti basalts have the highest Nb and thus possibly the highest CO₂ contents (ca. 0.3–0.5 wt%), while the low-Ti basalts were relatively Nb and thus CO₂-depleted (≤0.3 wt%). This difference may imply that the overall production of CO₂ increased when magmas were predominantly sourced from the mantle plume source, i.e., with minor or no contribution from the lithospheric mantle.

To summarize, in this study, we reviewed the main features of the Siberian Traps, the Wrangellia and the CAMP LIPs, which impacted the Triassic world. High-precision geochronology strongly links the Siberian Traps and the CAMP to the end-Permian and the end-Triassic mass extinctions, respectively. A similar cause-and-effect scenario between the Wrangellia and the CPE seems plausible based on paleontological, biostratigraphic and geochemical studies. However, further high-precision dating on both lava flows and intrusions is necessary to constrain the onset, evolution, and cessation of this LIP, as well as its emplacement mechanisms.

Using previously published data, we estimated maximum CO₂ contents for magmas of the three Triassic LIPs. Even if these estimates are based on proxy data, our results highlight that the mantle-related CO₂ budgets of the more primitive basalts seem to vary during the overall life cycle of each LIP, suggesting that different mantle sources or melting regimes came into play to produce these exceptionally voluminous volcanic events. These observations stress the fact that each LIP is a unique case-study and building paradigms that apply to all LIPs when discussing their relationship to concurrent Earth crises is not straightforward.