Different wildfire types promoted two-step terrestrial plant community change across the Triassic-Jurassic transition

Frequent wild ﬁ res associated with emplacement of the Central Atlantic Magmatic Province (CAMP) are thought to have been important drivers of two signi ﬁ cant changes in terrestrial plant communities and diversity during the Triassic-Jurassic Mass Extinction (TJME, ca. 201.51 Ma). However, it remains to be investigated whether these two changes are potentially related to different wild ﬁ re types. To better understand this relationship, we used a new method to reanalyze fossil pollen and spores across the Triassic-Jurassic transition in the Jiyuan Basin from the North China Plate. Results show that two peaks in wild ﬁ re frequency experienced different types of wild ﬁ res, with each linked to signi ﬁ cant changes in plant communities and diversity losses. In the ﬁ rst wild ﬁ re peak, canopy ﬁ res dominated and are accompanied by signi ﬁ cant losses of canopy forming plants, while in the second wild ﬁ re peak, ground cover ﬁ res dominated accompanied by signi ﬁ cant losses of ground cover plants. Changes in atmospheric humidity conditions were an important control on the two different wild ﬁ re peaks. Relatively humid climatic conditions corresponded to the prevalence of canopy ﬁ res and hindered the spread and development of ground cover ﬁ res in wet surface conditions. Conversely, relatively arid climatic conditions corresponded with the prevalence of ground cover ﬁ res in dry surface environments. Our results provide a potential relationship between terrestrial plant communities and wild ﬁ re types, which is important to further understanding of terrestrial environmental and ﬂ oral changes driven by Large Igneous Provinces.

Frequent wildfires across the T-J transition are thought to be the main cause of changes in terrestrial plant communities and diversity at that time (e.g., Song et al., 2020;Zhang et al., 2022).Studies from the western Tethys (including North America, East Greenland, Poland, Denmark, and Sweden) (Jones et al., 2002;Marynowski and Simoneit, 2009;Belcher et al., 2010;Pole et al., 2018;Lindström et al., 2019;Lindström et al., 2021) and eastern Tethys (including South and North China) (Song et al., 2020;Zhang et al., 2022) have demonstrated that the influence of global warming on wildfires has had a significant impact on altering land vegetation and ecosystems, ultimately resulting in soil erosion after the loss of vegetation (van de Schootbrugge et al., 2020;Lindström et al., 2021).Furthermore, previous investigations on inertinite (charcoal) content and palynology in Sweden and Denmark in the western Tethys have shown that changes in plant communities are closely related to wildfire types, with different types of wildfires including ground and canopy fires causing plant communities changes (Petersen and Lindström, 2012).This relationship may also exist in the eastern Tethys region, but further study is required to evaluate this.
In this study, we investigate the ZJ-1 borehole in the Jiyuan Basin on the southern North China Plate (NCP) in the eastern Tethys.We use a new method to reanalyze previously published palynological data from Zhang et al. (2022), and combine them with published fossil charcoal and geochemical evidence (also from Zhang et al., 2022), to reveal the potential relationship between terrestrial floras and wildfire types through the T-J transition.

Geological setting
During the Late Triassic, the NCP was situated around 30-40°N in the eastern Tethys (Greene et al., 2012;Zhang et al., 2022) (Figures 1A, B).Prototype basin reconstructions indicate that in the Late Triassic the Jiyuan Basin was located on the southeast part of the Ordos Basin (Liu et al., 2013) (Figure 1B).Provenance of sediment in the Jiyuan Basin was mainly from the northern Qinling-Dabie Orogenic Belt and the southern NCP located to the south (Yang et al., 2012) (Figures 1A, B).
Figure 1C displays the stratigraphic succession of lithological changes and the position of fossil plants assemblages from the Late Triassic to the Middle Jurassic in the Jiyuan Basin.Previous studies have constrained the geological ages of the Tanzhuang and Anyao formations from the Late Triassic to Early Jurassic based on zircon U-Pb dating, d 13 C org chemostratigraphic patterns, and palynology biostratigraphy (Lu et al., 2021;Zhang et al., 2022) (Figure 1C).

Materials and methods
Palynological and charcoal data (Zhang et al., 2022; sampling locations are shown in Figure 2) were re-analyzed using new method (see below) from the Anyao Formation of the ZJ-1 borehole (35.07001°N, 112.47338°E) in the Jiyuan Basin (Figures 1A, B).Variations in spore and pollen compositions through the studied strata were used to reconstruct climatic conditions based on the climatic preferences of the parent plants (e.g., Li et al., 2020;Zhang et al., 2022;Zhang et al., 2023a;Zhang et al., 2023b;Zhang et al., 2023c).To reveal detailed changes in climatic conditions and plant diversity in the study area, we undertook non-metric multidimensional scaling ordination analysis (nMDS) and Simpson (1-D) Diversity Index analysis based on palynological fossils (e.g., Slater et al., 2019;Zhang et al., 2023b).Spore and pollen data were ordinated using nMDS analysis (at two dimensions) with the Bray-Curtis similarity matrix, a useful ordination method for detecting patterns of co-occurrence among taxa as well as ecological gradients (e.g., Zhang et al., 2023b).Simpson (1-D) values were calculated as 1-S (p i ) 2 , where p is the proportional abundance of species and i is an unbiased measure of evenness, which ranges from zero (one taxon dominates the community completely) to one (all taxa have equal abundance) (e.g., Slater et al., 2019).According to the preferred ecological habits of plants, we reconstructed the vegetation types reflected by the palynological fossils, and identified ground cover, understory, and upper canopy plants (e.g., Petersen and Lindström, 2012).Charcoal abundance was used as a proxy for wildfire (e.g., Scott, 2000;Glasspool and Scott, 2010;Glasspool et al., 2015;Lu et al., 2020;Zhang et al., 2023a).To reveal burning temperatures and type of wildfires, we measured inertinite (charcoal) reflectance values on polished blocks using a Leica DM4500P LED reflected light microscope with a × 50 oil immersion objective.Reflectance values were identified for over 100 points in each sample.

Palynological assemblages and plant diversity
Thirty-four spore, thirty-two pollen, and eight algae genera have been identified in the study area (Zhang et al., 2022) (Supplementary Table 1).According to the palynomorph composition, palynomorph abundance variations, and straigraphically constrained cluster analysis (CONISS) of sporepollen data, it can be divided into the six palynological assemblage zones (Zhang et al., 2022) (Figure 2).In these palynological assemblage zones, two significant plant diversity declines occurred (Figure 2).In AZ-II, there was a decrease of around 45% in the proportion of spores and pollen, with the number of genera decreasing from 38 to 21 (Zhang et al., 2022) (Figure 2).Similarly, in AZ-IV, there was a decrease of approximately 44%, with a reduction from 50 to 28 genera (Zhang et al., 2022) (Figure 2).Moreover, AZ-IV experienced a significant rise in the abundance of Cythidites spores and Classopollis pollen, and it marked the first appearance of the Cerbropollenites pollen type (Zhang et al., 2022).

nMDS and Simpson (1-D) values
In this study, re-analyzing the palynological fossils using nMDS show that AZ-II and AZ-IV were intervals with relatively warm climatic conditions (Figure 3).This is consistent with previous results from principal component analysis (PCA) in the Jiyuan Basin (Zhang et al., 2022) (Figure 2).Results from the Simpson (1-D) analysis also identified two obvious declines in plant diversity during AZ-II and AZ-IV (Figure 2).This is consistent with the vegetation reconstruction results of palynological fossils in the studied Jiyuan Basin (Zhang et al., 2022).

Changes in terrestrial plant community
In this study, the ecological habits of palynological fossil genera were divided according previous classifications (see Petersen and Lindström, 2012) to include ground cover, understory, and upper canopy plants (Supplementary Table 2).Changes in the plant community structure in the study area are significant and can be divided into six stages that correspond to the changes in the palynological assemblages (Figure 2).

Inertinite (charcoal) reflectance
In this study, we tested the inertinite reflectance during two peaks in wildfire frequency across the T-J transition in which there are elevated levels of inertinite (Figure 2); results are presented in Supplementary Table 3.During the first wildfire peak, inertinite reflectance values varied between 1.7 and 4.7%Ro, and most (65%) values were between 3.8 and 4.3%Ro.During the second wildfire peak, inertinite reflectance values varied between 0.7 and 3.1%Ro, and most (62%) were between 1.1 and 1.7%Ro.

Changes in terrestrial plant communities and wildfire styles
Recent studies based on chemostratigraphy (d 13 C org patterns) and palynological biostratigraphy have constrained the geological age of the target strata in the studied Jiyuan Basin and indicate that the T-J boundary is located in the upper part of Anyao Formation; this position roughly corresponds to palynological sample #JY-6 in the JZ-1 core (Zhang et al., 2022) and is consistent with our nMDS analysis (Figure 3).In this study, nMDS analysis showed that there were significant differences between Late Triassic and Early Jurassic plant types accompanied by a significant decline in plant diversity inferred from the Simpson (1-D) value (Figure 2).Similar phenomena have been observed in the United Kingdom (Bonis et al., 2010), Germany (van de Schootbrugge et al., 2009;Lindström et al., 2019;Lindström et al., 2021), Denmark (Lindström et al., 2019;Lindström et al., 2021), Austria (Bonis et al., 2009), North America (Olsen et al., 2002), Australia (De Jersey and McKellar, 2013), and New Zealand (De Jersey and McKellar, 2013) and appears to represent a global signal (Wignall and Atkinson, 2020).
Changes in terrestrial plant types and diversity across the T-J transition were greatly influenced by the occurrence of frequent wildfires (Song et al., 2020;Zhang et al., 2022).In the study area, there were two significant peaks in the abundance of fusinite content and charcoal fossils across the T-J transition, indicating that there were two intervals with more frequent wildfires (Zhang et al., 2022).Frequent wildfires seem to be a global phenomenon across the T-J transition, with similar records recorded in North America, East Greenland, Poland, Denmark, Sweden, and South and North China (Jones et al., 2002;Belcher and McElwain, 2008;Marynowski and Simoneit, 2009;Pole et al., 2018;Lindström et al., 2019;Song et al., 2020;Lindström et al., 2021;Zhang et al., 2022).The two wildfire peaks in the study area were accompanied by significant changes in plant types and diversity, suggesting that frequent wildfires are the primary cause of significant changes in terrestrial plant ecosystems (Zhang et al., 2022).In this study, ground cover plants flourished during the first wildfire peak across AZ-II and AZ-III, while the abundance of understory and upper canopy plants decreased (Figure 2).During the second wildfire peak across AZ-III and AZ-IV, upper canopy plants flourished and ground cover plant significantly decreased (Figure 2).
Different types of wildfires may be responsible for changes in different plant types across the T-J transition.Previous studies have shown that ground, surface, and canopy wildfires are characterized by significant differences in burning temperature (Jones, 1997;Scott, 2000).Ground fires have a maximum burning temperature Non-metric multidimensional scaling (nMDS) plot of spore-pollen data from the studied borehole in the Jiyuan Basin.Insert box indicates inferred climatic changes through this interval based on spore-pollen fossils and sedimentary features (Zhang et al., 2022).
of 400°C during ground litter burns (Scott and Jones, 1994;Scott, 2000), while surface fires burn litter, herbs, and shrubs and attain a maximum temperature of 600°C (Scott and Jones, 1994;Scott, 2000).However, canopy fires burn larger trees and shrubs and can reach higher maximum temperatures of 800°C (Scott and Jones, 1994;Scott, 2000).Extensive experimental data in previous studies have shown that changes in inertinite reflectance values can reflect burning temperatures and type of wildfires (e.g., Jones et al., 1991;Scott, 2000;Scott and Glasspool, 2007;Petersen and Lindström, 2012;Xu et al., 2022).The burning temperature, denoted as T, can be represented by the linear regression equation T = 184.10+ 117.76%Ro (r 2 = 0.91) (Jones, 1997), where %Ro corresponds to the measurable inertinite reflectance.In this study, changes in inertinite reflectance values provide evidence for wildfire types across the T-J transition.According to the above equation, burning temperature of the two wildfire peaks in the study area vary from 631.59-690.47°Cand 313.64-384.29°Crespectively.The first wildfire peak had higher burn temperatures and was likely dominated by canopy type wildfire, while the second wildfire peak had lower temperatures and was dominated by ground cover type wildfire (Figure 2).We consider that in the first wildfire peak canopy fires promoted significant change in plant community composition marked by the disappearance of canopy plants, whereas in the second peak ground cover fires promoted significant changes in plant community composition marked by the disappearance of surface plants.Similar changes in terrestrial wildfire patterns have been observed in Sweden and Denmark (Petersen and Lindström, 2012).

CAMP as a potential driver of the Triassic-Jurassic transition changes
Recent studies have shown that CAMP emplacement was an important driver of global environment, climate, and floral change across the T-J transition (e.g., Hesselbo et al., 2002;Ruhl and Kürschner, 2011;Greene et al., 2012;Percival et al., 2017;Lindström et al., 2019;Ruhl et al., 2020;Shen et al., 2022;Zhang et al., 2022).CAMP released massive amounts of isotopically light CO 2 into the reservoir of the exogenic carbon cycle and drove major global warming (Ruhl and Kürschner, 2011;Ruhl et al., 2020;Zhang et al., 2022).Warming climatic conditions at this time are supported by oxygen isotopes from Lavernock Point (United Kingdom) in the northwest Tethys (Korte et al., 2009), pCO 2 increases in pedogenic carbonates from the Newark Basin and the stomatal index (plant fossil cuticles) from Greenland, Sweden, and Northern Ireland (McElwain et al., 1999;Bonis et al., 2010;Schaller et al., 2011;Steinthorsdottir et al., 2011;Schaller et al., 2012;Lindström et al., 2021), and PCA analysis and sedimentological evidence (increased calcareous nodules) in North China (Zhang et al., 2022).
In the context of global warming climatic conditions, in higher temperatures lightning becomes more frequent and wildfires can be more destructive (Glasspool et al., 2015;Miller and Baranyi, 2021).As lightning is the main cause of wildfire, higher temperatures associated with CAMP volcanism would be expected to have increased the frequency and destructiveness of wildfire events (e.g., van de Schootbrugge et al., 2008;Belcher et al., 2010;Glasspool et al., 2015;Lu et al., 2020;Lu et al., 2022;Zhang et al., 2022;Zhang et al., 2023a;Zhang et al., 2023c).Wildfires are influenced by a variety of factors, including the supply of combustible material, oxygen concentration required for combustion, a conducive climate with low moisture, and an ignition mechanism (Scott, 2000;Glasspool et al., 2015;Lu et al., 2020;Zhang et al., 2022).During the T-J transition, the abundant terrestrial vegetation in the Jiyuan Basin provided ample fuel for wildfires (Zhang et al., 2022).Widespread wildfires across the globe provide evidence for the presence of atmospheric oxygen concentrations suitable for wildfires during this time (see Section 5.1).Global warming climatic conditions during the T-J transition also promoted frequent lightning activity (the primary ignition mechanism) (Scott, 2000;Glasspool et al., 2015;Miller et al., 2017;Lu et al., 2020;Zhang et al., 2022;Zhang et al., 2023a).However, the two periods of warming climatic conditions across the T-J transition that saw significant changes in plant communities and diversity in the study area were accompanied by different types of wildfires (see Section 5.1) (Figure 2).Different atmospheric humidity conditions in the study area under global warming may be the main reason for the different types of wildfires.
In the study area, changes in vegetation types, PCA analysis results, and the widespread occurrence of calcareous nodules indicate that there are differences in atmospheric humidity conditions during the two pulses of plant diversity collapse across the T-J transition (Zhang et al., 2022) (Figure 2).During the first pulse of plant diversity collapse and its subsequent recovery phase, relatively humid climatic conditions prevailed in the study area indicated by an increase in hygrophytes (Zhang et al., 2022).Similar phenomena are also seen in the western Tethys (including Hungary, Great Britain, Denmark, Greenland, and Canada) and the southern Pangaea supercontinent (New Zealand and Australia) (van de Schootbrugge et al., 2009;Lindström, 2016).During the second pulse of plant diversity collapse, relatively arid climatic conditions prevailed in the study area inferred from increases in xerophytes and calcareous nodules (Zhang et al., 2022).Similar climatic conditions have been observed in South China (Huang, 2001;Zhou et al., 2021).In this study, during the first decline in plant diversity (TJME-I) that accompanied by the prevalence of canopy fires, relatively humid conditions caused an increase in the humidity of near-surface plants (including herbaceous plants and shrubs) and plant litter (fuel), which was detrimental to wildfires burning near-surface plants and development of ground cover fires (Figure 4).During the second plant diversity decline (TJME-II), the relatively arid conditions during TJME-II also promoted the decrease in humidity of nearsurface plants and plant litter, which was conducive to their burning by ground cover fire, leading to the rapid spread and development of ground cover fires (Figure 4).Therefore, we consider that the different atmospheric humidity conditions in the context of global warming are the main cause of the different types of wildfires in the study area across the T-J transition.Furthermore, we also consider that the combined effect of the variable eruption intensity of the bimodal CAMP, and the southward migration of the tropical pressure convergence zone, may have played a role in promoting different wildfires types through driving climate change (Song et al., 2020;Zhang et al., 2022).

Conclusions
1. Through the Triassic-Jurassic transition in the Jiyuan Basin, two intervals with peaks in wildfire frequency are identified.Inertinite reflectance values, serving as a proxy for burn temperature, records different types of wildfires for the two peaks.In the first peak, wildfires had high burn temperatures indicative of canopy fires and were accompanied by the loss of canopy plants, while in the second peak, wildfires had lower burn temperatures characteristic of ground cover fires and were accompanied by the loss of ground cover plants.The different types of wildfire resulted in significant but different changes in plant community composition, and each included a significant loss in plant diversity.Wildfire was important through the T-J transition in the Jiyuan Basin in shaping plant community composition and was a major factor contribuing to the Triassic-Jurassic boundary terrestrial diversity loss and mass extinction.
2. Changes in atmospheric humidity conditions were a significant driver of the two different wildfire peaks.In the study area, atmospheric humidity was significantly different during the two intervals of plant community change.Relatively humidity climatic conditions corresponded first to the prevelance of the canopy fires with very humid conditions and surface environments hindering the development and spread of ground cover fires.Relatively arid climatic conditions corresponded with the second peak in wildfire with the prevalence of ground cover fires, with arid conditions promoting the development and spread of ground cover fires.

1
FIGURE 1 Location and geological context for the study area.(A), Palaeogeographic reconstruction for the Late Triassic (~202 Ma) showing location of the NCP and approximate extent of CAMP (revised after Greene et al., 2012); (B) Simplified tectonic map of the southern NCP during the Late Triassic showing the location of study area (modified from Liu et al., 2013); (C) Stratigraphic framework of the Tanzhuang Formation to the Lower Yangshuzhuang (YSZ) Formation from the Jiyuan Basin (modified from Yang et al., 2012).CAMP, Central Atlantic Magmatic Province; NCP, North China Plate; SCP, South China Plate; QDOB, Qinling-Dabie Orogenic Belt; S-NCP, southern North China Plate; SCP, South China Plate; Fm., Formation; m & s, coal, mudstone and silty mudstone; s., sandstone; c, conglomerate; Dep., Depositional environment; C.-P., Coniopteris-Phoenicopsis.

FIGURE 2
FIGURE 2 Results of d 13 C org values, Hg concentrations, Hg/TOC ratios, kerogen macerals, floral composition, Simpson (1-D) diversity values, generic richness (number of different genera), and principal components analysis (PCA) from the studied borehole in the Jiyuan Basin.d 13 C org values, Hg concentrations, Hg/TOC ratios, kerogen macerals, generic richness, and PCA data are from Zhang et al. (2022).