Abstract
Eleven conodont zones are established for the Lower Triassic of the Motianling section in the Nanpanjiang Basin, South China. Detailed size measurements on 2244 P1 conodont elements demonstrate dynamic size variation patterns during the Early Triassic that are calibrated to the Early Triassic timescale and conodont zonation. Correlation of conodont size variation with various environmental indicators shows that these size changes are primarily controlled by rapidly changing and episodic environmental conditions and ecological recovery in the aftermath of the end-Permian mass extinction. Decreasing size trends occur in the early Griesbachian, late Griesbachian and at the Smithian–Spathian boundary. Episodes of size increase (rebounds) occur in the middle Griesbachian, Induan–Olenekian boundary and middle–late Spathian. The observed temporal size variations indicate that conodont animals were significantly affected by rapidly changing environmental conditions (mainly temperature and redox conditions) during the Early Triassic and recurrent size increases (rebounds) occurred during periods of environmental and climatic amelioration (climatic cooling and oxygenation), demonstrating a complicated recovery history of marine ecosystems in the aftermath of the largest Phanerozoic biotic crisis.
Supplementary material: Data for all species are available at https://doi.org/10.6084/m9.figshare.c.4528616
The end-Permian mass extinction wiped out more than 90% of marine species (Sepkoski 1984; Payne & Clapham 2012). Extensive studies indicate that this mass extinction was caused by multiple factors, including massive volcanism and release of gas hydrates triggering global warming (Benton 2003), lethally hot sea surface temperatures (Joachimski et al. 2012; Sun et al. 2012), global anoxia (Wignall & Twitchett 1996; Isozaki 1997), ocean acidification (Hinojosa et al. 2012; Clarkson et al. 2015; Garbelli et al. 2016) and abnormal productivity (Meyer et al. 2011; Grasby et al. 2016).
The hostile environmental conditions that precipitated the mass extinction persisted to the end of the Early Triassic or beginning of the Middle Triassic, leading to a prolonged recovery (Payne et al. 2004; Chen & Benton 2012). Numerous recent studies have shown that the Earth's environment experienced recurrent and dynamic change from the latest Permian to the Middle Triassic (e.g. Sun et al. 2012; Grasby et al. 2013; Tian et al. 2014). Transient faunal rebounds or recovery of some groups or clades coincided with rapid environmental amelioration intervals (e.g. Song et al. 2014; Dai et al. 2018). According to trophic pyramid theory, where lower levels of food chains will probably recover first, followed by ever higher trophic levels (Chen & Benton 2012), interspecific impact (e.g. competition) also played an important role in controlling biotic recovery (Hautmann et al. 2015).
As a biological trait, body size in marine animals is interpreted to be related to both environment and ecological changes. Previous studies have revealed that numerous invertebrate marine groups underwent size reduction during the end-Permian crisis and its aftermath, primarily controlled by environmental deterioration (e.g. Schaal et al. 2016), although animal sizes evolved as a result of both environmental pressures and evolutionary controls (Heim et al. 2015). Conodont elements are tooth-like structures and the only mineralized apparatus preserved in soft-bodied conodont animals, in front of which lie two lobate structures with fibrous muscle tissues (Briggs et al. 1983; Gabbott et al. 1995). Geometrical analysis of Early Triassic oro-pharyngeal elements supports the contention that conodont animals are vertebrates (Goudemand et al. 2011). Engineering approaches previously applied to mammalian dentition showed that conodont elements were adapted to fracture food items despite the fact that conodont animals are small in size and lack jaws (Jones et al. 2012). During the process of feeding, the outer surfaces of denticles are subject to wear and damage, but damage is also recorded within denticles (Purnell & Jones 2012; Shirley et al. 2018). Repair of damage to conodont elements by accretion of lamellar tissue during ontogeny and even regrowth of denticle tips broken off during life indicate that conodont elements were retained permanently through the life of the animal and their size represents the animal size at death (Donoghue & Purnell 1999). Thus, the size of the conodont elements can be taken as a proxy for the conodont animal body size or body weight, which is related to both environment and ecology (Gingerich et al. 1982; Donoghue & Purnell 1999).
Size variations in Permian–Triassic conodont elements have attracted recent attention (Luo et al. 2008; Chen et al. 2013; Schaal et al. 2016; Zhang et al. 2018; Leu et al. in press). However, these studies all focused on the ‘Lilliput effect’ (size reduction) at the Permian–Triassic boundary (Luo et al. 2008), Smithian–Spathian boundary (Chen et al. 2013; Leu et al. in press) and Middle–Late Triassic boundary (Zhang et al. 2018). Here we present a quantitative profile of conodont size variations, including multiple episodes of size increase (size rebound) throughout the Early Triassic, at the previously unstudied Motianling section in the Nanpanjiang Basin, in relation to environmental and ecological stresses and changes that affected conodont animals during this unprecedented period in the aftermath of Earth's most severe mass extinction event.
Geological setting and lithostratigraphy
The well-exposed Motianling section (23°24'34"N, 107°00'18"E), which spans the uppermost Permian to lowermost Middle Triassic, is located on the margin of the Debao isolated carbonate platform in the Nanpanjiang Basin (Fig. 1). The studied section is composed of three formations, the uppermost part of the Heshan Formation, the Luolou Formation and the basal part of the Baifeng Formation, and can be divided into four lithological units (Fig. 2). These units are 4, 21.11, 41.42 and 7.15 m in thickness, respectively. Unit I, the uppermost Heshan Formation, is marked by dark grey thin-bedded bioclastic limestone (Fig. 3a). This unit contains large numbers of fusulinid and non-fusulinid foraminifers, algae, ostracods, bivalves and other marine fauna, which implies an open-marine platform environment. The Luolou Formation is divided into two units, Unit II and Unit III. Unit II comprises thick-bedded microbialites, dolomitized limestone and oolites (Fig. 3a and b). Microbialites in the Motianling section are thrombolytic and consist of a digitate micritic framework enclosed in sparry cement. Because the transition from thrombolytes to oolites can be a function of deepening (Wang et al. 2019), Unit II indicates a relatively restricted platform environment with a slight overall deepening trend upwards. Unit III comprises dark grey and reddish, medium- to thick-bedded, argillaceous limestone (Fig. 3c and d). Together with abundant ammonoid fossils (Fig. 3e), the sedimentary facies of Unit III indicates a low-energy inter-shelf basin. Unit IV, the lowermost Baifeng Formation, consists of claystone or sandstone (Fig. 3f). It should be noted that a possible hiatus is recorded between Unit I and the overlying microbialites (basal Unit II), which is also seen at the Taiping and Dajiang sections in the Nanpanjiang Basin (Jiang et al. 2014; Xiao et al. 2018; Tian et al. 2019).
Palaeogeography of (a) the world and (b) the Nanpanjiang Basin (NB) of the South China Craton during the Early Triassic. (b) shows the study section locations (red) on isolated carbonate platforms (blue): MTL, Motianling. (a) is edited from Metcalfe (2017), and (b) from Lehrmann et al. (2003) and Algeo et al. (2007). It should be noted that the South China Craton has rotated c. 80° clockwise since the Permian–Triassic. EM, East Malaya; WS, West Sumatra; SI, Simao; S, Sibumasu; SQ, South Qiangtang; NQ, North Qiangtang; L, Lhasa; SWB, South West Borneo; WC, Western Cimmerian Continent.
Stratigraphy, sample horizons, conodont ranges and size variation of conodont P1 elements at the Motianling section. A black dot or a star indicates a sampled horizon; 40 numbered stars show the samples chosen for statistical size analysis. For segminate elements, the 25–75% quartiles of length are shown by a box (with specimen number >2). The number beside the box indicates the number of conodont P1 elements measured. U.P., Upper Permian; Ch., Changhsingian; B.Fm, Baifeng Formation; He. Fm., Heshan Formation; M.Tr., Middle Triassic.
Outcrop photographs of the Motianling section. (a) Uppermost Heshan and lowermost Luolou formations with a lithological truncating surface. (b) Microbialites and dolomite limestone of the lower part of the Luolou Formation. (c) Dark grey thin-bedded limestone of the middle Luolou Formation. (d) Grey to reddish thin-bedded limestone of the upper Luolou Formation. (e) Ammonoid (Owenites) bed from the upper Luolou Formation. (f) Uppermost Luolou and lowermost Baifeng formations.
Materials and methods
A total of 97 bulk samples (Fig. 2), each weighting c. 3–5 kg, have been collected and processed by acetic acid dissolution following the standard method described by Jiang et al. (2007). Insoluble residues were sieved using 20 mesh (0.850 mm) and 160 mesh (0.095 mm) sieves. Heavy liquid separation of conodont elements was undertaken by high-density liquid (HST, heteropolytungstate heavy liquid, ∼2.78 g cm−3) to eliminate the lower density minerals. The conodont fossils were carefully picked from prepared sample residues under a stereomicroscope. All photographs were taken using a scanning electronic microscope (Hitachi, SU8010) at the China University of Geosciences (CUG). A total of 2244 P1 conodont elements from 40 samples were obtained, which represent 54 species of the genera Hindeodus, Isarcicella, Clarkina, Sweetospathodus, Neospathodus, Novispathodus, Eurygnathodus, Discretella, Triassospathodus, Neogondolella, Pachycladina, Spathicuspus, Parachirognathodus, Icriospathodus and Cratognathodus.
In segminiplanate conodont elements, the length exhibits positive correlation with the platform area (Purnell 1993), whereas the length of segminate conodont elements has a positive correlation with their lateral area (LA = length × height; Chen et al. 2013). Thus, only the length of the basal part of P1 elements was used as an overall size proxy in this study; conodont height was not used owing to denticles of conodont P1 elements being fragile (e.g. figs 2, 3, 6, 8, 12 and 14 of Chen et al. 2013) and thus rendering height determination, in particular to segminiplanate elements, difficult to acquire. Because denticles of some conodont species might be inclined backward or forward, they may extend beyond or within the carina (e.g. Fig. 4: 19, 21 and 29). In this study, the lengths of conodont P1 elements have been measured from the anterior edge of the carina to the posterior end of the cavity, which is slightly different from the length proxy of Leu et al. (in press), who used the distance from the anterior edge of the carina to the edge of the last denticle (Fig. 4). Based on species richness and individual data, rarefaction curves were also generated by the palaeontological statistics software package (PAST) to evaluate sampling sufficiency (Hammer et al. 2001).
SEM photographs of the key elements of conodonts from the Motianling section and length of the conodont elements. L represents the measurement in this study, L represents the measurement given by Leu et al. (in press). All specimens illustrated in this figure are housed at the State Key Laboratory of Biogeology and Environmental Geology, China University of Geosciences (Wuhan). 1, Isarcicella staeschei Dai et al. 1989: oral view, from sample ZM-006; 2, Hindeodus bicuspidatus Kozur, 2004: lateral view, from sample ZM-005; 3, Isarcicella turgida (Kozur 1975): oral view, from sample ZM-006; 4, Hindeodus praeparvus (Kozur, 1996): lateral view, from sample ZM-005; 5, Isarcicella inflata Perri & Farabegoli, 2003: oral view, from sample ZM-005; 6, Isarcicella lobata Perri & Farabegoli, 2003: oral view, from sample ZM-005; 7, Hindeodus peculiaris (Perri & Farabegoli, 2003): lateral view, from sample ZM-006; 8, Isarcicella isarcica (Huckriede, 1958): oral view, from sample ZM-006; 9, Hindeodus anterodentatus (Dai et al. 1989): lateral view, from sample ZM-005; 10, Hindeodus inflatus Nicoll et al. 2002: oral view, from sample ZM-006; 11, Hindeodus latidentatus (Kozur, 1975): lateral view, from sample ZM-006; 12, Hindeodus postparvus Kozur, 1989: lateral view, from sample ZM-007; 13, Hindeodus sosioensis Kozur, 1996: lateral view, from sample ZM-007; 14, Hindeodus parvus (Kozur & Pjatakova, 1976): lateral view, from sample ZM-002; 15, Neospathodus dineri Sweet, 1970, Morphotype 1: lateral view, from sample ZM-019; 16, Neospathodus dieneri Sweet, 1970, Morphotype 2: lateral view, from sample ZM-019; 17, Neospathodus cristagalli (Huckriede, 1958): lateral view, from sample ZM-019; 18, Neospathodus chii Zhao et al. 2007; lateral view, from sample ZM-019; 19, Neospathodus dieneri Sweet, 1970, Morphotype 3: lateral view, from sample ZM-019; 20, Euryganathodus costatus Staesche, 1964: oral view, from sample ZM-028; 21, Neospathodus pakistanensis Sweet, 1970: lateral view, from sample ZM-028; 22, Euryganathodus hamadai (Koike, 1982): oral view, from sample ZM-028; 23, Neospathodus novaehollandiae McTavish, 1973: oral view, from sample ZM-028; 24, Parachirognathus peculiaris Zhang & Guo, 1991: lateral view, from sample ZM-028; 25, Novispathodus waageni waageni (Sweet, 1970): lateral view, from sample ZM-028; 26, Hindeodus praeparvus (Kozur, 1996): lateral view, from sample ZM-07; 27, Sweetospathodus kummeli (Sweet, 1970): lateral view, from sample ZM-020; 28, Neospathodus tongi Zhao et al. 2007: lateral view, from sample ZM-023; 29, Neospathodus chaohuensis Zhao et al. 2007: lateral view, from sample ZM-023; 30, 31, Novispathodus waageni eowaageni (Zhao et al. 2007): lateral view, from sample ZM-028; 32–34, Novispathodus pingdingshanensis (Sweet, 1970): lateral view, from sample ZM-035; 35(a,b), Icriospathodus collisoni (Solien, 1979): a, oral view; b, lateral view; from sample ZM-037; 36, Triassospathodus brevissimimus (Orchard, 1995): lateral view, from sample ZM-037; 37, Novispathodus abruptus (Orchard, 1995): lateral view, from sample ZM-037; 38, Clarkina planata (Clark, 1959): oral view, from sample ZM-015; 39, Clarkina carinata (Clark, 1959): oral view, from sample ZM-013. Scale bar represents 400 µm.
Results
Conodont zonation
We here establish 11 conodont zones at the Motianling section, with the base of each zone defined by the first occurrence (FO) of a designated taxon. The conodont zones are described as follows and the key fossils are shown in Figure 4. A table of data for each species is given in the supplementary material.
Hindeodus parvus Zone (16.8–23.32 m)
The lower limit of this zone is defined by the FO of Hindeodus parvus and the upper limit is defined by the FO of Isarcicella lobata (Perri & Farabegoli, 2003). Associated taxa include H. preaparvus (Kozur, 1996) and H. typicalis (Sweet, 1970). H. parvus (Kozur & Pjatakova, 1976) first occurs at 16.8 m in the Motianling (MTL) section (Fig. 2). As a cosmopolitan species, the first appearance datum (FAD) of H. parvus marks the base of the Induan Stage in the Meishan Global Stratotype Boundary and Point (GSSP) section of South China and elsewhere (Yin et al. 2001).
Isarcicella lobata Zone (23.31–24.11 m)
The FO of Isarcicella lobata defines the lower limit of this zone and the FO of Isarcicella isarcica (Huckriede, 1958) marks the upper limit. Associated taxa include H. parvus, H. typicalis, H. bicuspidatus Kozur, 2004, H. anterodentatus (Dai et al. 1989), H. eurypyge Nicoll et al., 2002, I. inflata Perri & Farabegoli, 2003 and I. turgida (Kozur 1975). Perri & Farabegoli (2003) first established the I. lobata Zone in the Southern Alps. In the Meishan section, I. lobata was also identified in bed 28, but its FO was later than that of I. staeschei Dai et al. 1989, so the I. staeschei Zone was established after the H. parvus Zone (Jiang et al. 2007). However, Jiang et al. (2011) established an I. lobata Zone in the Shangsi section because of the earlier occurrence of I. lobata. A phylogenic tree of Hindeodus–Isarcicella during the Permian–Triassic transition indicated that I. lobata was more primitive than I. staeschei (Jiang et al. 2010). Recently, Jiang et al. (2014) recognized this zone in microbialites in the Dajiang section, similar to what is here reported for the Motianling section.
Isarcicella isarcica Zone (24.11–25.39 m)
The lower limit of this zone is the FO of I. isarcica and the upper limit of this zone is the FO of H. postparvus Kozur, 1989. H. parvus, H. typicalis, H. anterodentatus, H. eurypyge, H. inflatus Nicoll et al., 2002, H. latidentatus, I. inflata, I. lobata and I. staeschei are also identified in this zone. This zone is widely developed around the world, including South China, Tibet, Pakistan, Kashmir, India, Italy, Iran, and Slovenia (Jiang et al. 2011). In the Nanpanjiang Basin, this zone was also identified in many other sections, such as Dawen (Chen et al. 2009), Dajiang (Jiang et al. 2014) and Heping (Lehrmann et al. 2003).
Hindeodus postparvus Zone (25.39–27.34 m)
This zone begins with the FO of H. postparvus and ends with the FO of H. socioensis Kozur, 1996. Other species include H. parvus, H. praeparvus, H. typicalis, H. anterodentatus and H. latidentatus. In the upper Griesbachian of the Lukač section of Slovenia, the H. postparvus Zone was established to partly correlate with the H. postparvus–H. socioensis Zone (Kolar-Jurkovšek & Jurkovšek 2015), which was established first by Kozur (1998). Recently, Wu et al. (2014) also established an H. postparvus Zone, which begins with the FO of H. postparvus and ends with the disappearance of the Hindeodus genus in Tibet.
Hindeodus socioensis Zone (27.34–29.69 m)
The lower limit of this zone is limited by the FO of Hindeodus socioensis, and the upper limit of this zone is limited by the FO of Clarkina carinata (Clark, 1959). Associated taxa include H. typicalis, H. bicuspidatus, H. anterodentatus, H. postparvus and H. peculiaris (Perri & Farabegoli, 2003). This zone was first established by Kozur (1996) at Sosio Vally in Sicily; H. socioensis was used as an assemblage zone together with H. postparvus (Kozur, 1998). In the Wadi Waist area, Central Oman, Krystyn et al. (2003) also established this zone. Recently, this zone was also reported from South China in the Dajiang (Jiang et al. 2014) and Jiarong (Chen et al. 2015b) sections.
Clarkina carinata–Clarkina planata Zone (29.69–35.94 m)
The FO of Clarkina carinata defines the base of this zone and its upper limit is unclear owing to poor exposure. Associated taxa include C. planata (Clark, 1959), C. kazi (Orchard & Krystyn, 1998), C. taylorae (Orchard et al., 1994), C. krystyni, C. lehrmanni Chen et al., 2009, Hindeodus bicuspidatus, H. anterodentatus, H. postparvus and H. socioensis. In the Meishan section, Zhang et al. (1995) recognized this conodont zone above the Isarcicella isarcica Zone, and they revised it as the C. planata–C. tulongensis Zone in a recently reported study (Zhang et al. 2009). Recently, some researchers established C. krystyni (or C. planata) and Nc. discreta zones from the Jiarong, Xiakou and Ganxi sections of South China, and these were equated to the C. krystyni Zone in Chaohu and C. planata–C. tulongensis Zone in the Meishan section (Zhao et al. 2007, 2013; Chen et al. 2015b; Lyu et al. 2017).
Neospathodus dieneri Zone (35.94–50.96 m)
The lower limit of this zone is marked by the FO of Neospathodus dieneri Sweet, 1970, and the upper limit of this zone is marked by the FO of Novispathodus ex. gr. waageni (Sweet 1970). Associated conodonts include Sweetognathodus kummeli (Sweet 1970), N. tongi Zhao et al. 2007, N. chaohuensis Zhao et al. 2007, N. cristagalli (Huckriede, 1958), and N. chii Zhao et al. 2007. This zone was first established by Sweet (1970) from Pakistan and it has been reported widely around the world. Based on the differences in the configuration of the cusp and penultimate denticle, Zhao et al. (2007) identified three morphotypes within N. dieneri from the Chaohu area, and three subzones were recognized according to these morphotypes. Recently, these subzones were reported from the western Hubei Province of South China (Zhao et al. 2013; Lyu et al. 2017). In this study we establish the N. dieneri Zone for the co-occurrences of these three morphotypes.
Novispathodus ex. gr. waageni Zone (50.96–57.23 m)
The base of this zone is defined by the FO of Novispathodus ex. gr. waageni and the upper limit by the FO of Nv. pingdingshanensis (Zhao et al. 2007). Associated taxa include Eurygnathodus costatus Staesche, 1964, E. hamadai (Koike, 1982), Discretella discreta (Müller, 1956), Neospathodus novaehollandiae McTavish, 1973, N. pakistanensis Sweet 1970, N. cristagalli, N. dieneri, Spathicuspus sp., Pachyladina sp., and Parachirognathus peculiaris Zhang & Guo, 1991. This conodont zone was originally described from the Mianwali Formation, Western Pakistan by Sweet (1970), and was then reassigned to Novispathodus waageni by Orchard (2005). According to the specimens from Chaohu, Nv. ex. gr. waageni was assigned to two subspecies, namely Nv. waageni waageni and Nv. waageni eowaageni, and the Nv. ex. gr. waageni Zone was comparatively subdivided into these two conodont zones (Zhao et al. 2007). In ascending order, the Nv. waageni eowaageni Zone and the Nv. waageni waageni Zone have been reported from South China (Zhao et al. 2013; Chen et al. 2015; Lyu et al. 2017), Western Australia (Metcalfe et al. 2013), and Northern Vietnam (Maekawa et al. 2016). We only established the Nv. ex. gr. waageni Zone here according to the co-occurrences of Nv. waageni waageni and Nv. waagni eowaagni in the Motianling section (Fig. 2).
Novispathodus pingdingshanensis Zone (57.23–58.23 m)
The FO of Novispathodus pingdingshanensis defines the lower limit of this zone and the FO of Icriospathodus collinsoni (Solien 1979) defines the upper limit. Associated taxa include Discretella discreta, Cornudina sp., Nv. waagni waageni, Nv. waageni eowaageni, Spathicuspus spathi (Sweet, 1970), Neospathodus dieneri Morphotype 1, N. dieneri Morphotype 2, N. dieneri Morphotype 3 and Cratognathodus sp. Zhao et al. (2007) originally described this species from the Chaohu area, and recently Goudemand et al. (2012) and Komatsu et al. (2016) revised the diagnosis of Nv. pingdingshanensis. So far, Nv. pingdingshanensis has been reported widely from South China (Zhao et al. 2007; Ji et al. 2011; Liang et al. 2011; Goudemand et al. 2012), northeastern Vietnam (Maekawa et al. 2015; Komatsu et al. 2016) and Japan (Maekawa et al. 2018).
Icriospathodus collinsoni Zone (58.23–62.73 m)
This zone extends from the FO of Icriospathodus collinsoni to the FO of Triassospathodus homeri (Bender, 1970). Other co-occurring taxa include Discretella discreta, Spathicuspus sp., Cratognathodus sp., Triassospathodus brevissimus (Orchard, 1995), Neospathodus sp., N. triangularis (Bender, 1970), and Novispathodus abruptus (Orchard, 1995). This zone has been recognized globally as follows: South China (Zhao et al. 2007; Ji et al. 2011; Chen et al. 2013, 2015b; Sun et al. 2015), western USA (Solien 1979), Vietnam (Komatsu et al. 2016) and Japan (Maekawa et al. 2018). According to these reports, the range of this zone is limited to the early–middle Spathian.
Triassospathodus homeri Zone (∼62.73 m)
This zone is indicated by the range of Triassospathodus homeri. The upper limit of this zone is unclear because no other key conodonts are found. Associated taxa are Tr. symmertricus (Orchard, 1995), Spathicuspus sp., Neogondolella sp., Neospathodus triangularis and Novispathodus abruptus. This zone has been recognized worldwide as follows: South China (Zhao et al. 2007; Chen et al. 2015b), Pakistan (Sweet 1970), Spiti (Goel 1977) and North America (Orchard 2008). According to these reports, this zone belongs to the middle–late Spathian.
Stage and sub-stage boundaries of the studied section
As no Permian conodont has been found at the Motianling section, the position of the Permian–Triassic boundary is not precisely located yet. Studies from the Yangtze platform and different isolated platforms in the Nanpanjiang Basin indicate that the FO of H. parvus occurs at the bottom of the microbialites (Jiang et al. 2014; Wang et al. 2016; Chen et al. 2018b). In addition, recent work from the Taiping section also shows that the Permian–Triassic boundary is located at the top of the Heshan Formation (Xiao et al. 2018). Based on this stratigraphic correlation scenario, we infer the Permian–Triassic boundary in the MTL section to be located at the base of the microbialites (Fig. 2).
The Griesbachian–Dienerian boundary has been traditionally defined by the FO of Sweetospathodus kummeli (e.g. Zhao et al. 2013), but some studies in South China have defined this boundary by the FO of Neospathodus dieneri (Liang et al. 2016; Lyu et al. 2017). Here, we define this boundary by the FO of Neospathodus dieneri because of its earlier occurrence than Sweetospathodus kummeli in this section. As stated above, Novispathodus ex. gr. waageni has a worldwide distribution and the FO has been widely used as the marker of the Dienerian–Smithian boundary (e.g. Lyu et al. 2017), thus the Induan–Olenekian boundary in the Motianling section is also marked by the base of this zone. For the Smithian–Spathian boundary, the FO of Novispathodus pingdingshanensis has been widely used to recognize this boundary in South China (Tong et al. 2004; Liang et al. 2011; Chen et al. 2013; Zhao et al. 2013) and we here follow this. In this paper we recognized the Induan–Olenekian boundary and Smithian–Spathian boundary by the FOD of Nv. ex. gr. waageni and Nv. pingdingshanensis respectively. Thus, the Griesbachian–Dienerian boundary, Induan–Olenekian boundary and Smithian–Spathian boundary are respectively located at 35.94 m, 50.96 m and 57.23 m of the studied section (Fig. 2).
Conodont distribution and size variation
The discovered conodont elements are mainly distributed in Unit III, but some Hindeodus and Isarcicella species are recorded in the upper part of Unit II. Segminiplanate P1 elements are from the lowermost, middle and uppermost Unit III (samples 13, 15, 17 and 28, and only one specimen in sample 39; Fig. 5). From samples 1 to 10, the mean value of conodont element size decreases from 462 to 295 µm. This trend ends with a transient size increasing interval from samples 10 to 15. Also, the segminiplanate elements (Clarkina) decrease persistently from 658 to 567, then to 363 µm, until their disappearance in sample 17. From samples 16 to 26, no distinct trend has been observed. Sizes increase again from samples 27 to 33. Following this size increase, a size decrease has been shown from samples 33 to 37.
Size analysis (in log µm) with the sample numbers as sampling bin. Filled diamonds, segminiplanate conodont elements; open diamonds, segminate conodont elements.
From sample 27 to sample 28, the most distinct conodont size increasing trend of the Motianling section is recorded, with the mean length of segminate conodont elements varying from 292 to 473 µm (Fig. 5). Conodont elements with size between 2.40 and 2.94 log µm occur most frequently and form 91% of segminate elements in sample 28 (Fig. 6c), whereas conodont elements with size between 2.40 and 2.89 log µm occur most frequently and form 96% of segminiplanate elements in sample 28 (Fig. 6d). Meanwhile, in sample 28, segminiplanate element mean size is 2.66 log µm whereas segminate element mean size is 2.64 log µm. In contrast, conodont elements with size between 2.2 and 2.69 log µm dominate sample 27, forming 97% of all conodont elements (Fig. 6a), whereas conodont elements with size between 2.4 and 2.99 log µm contribute 97% of all conodont elements in sample 28 (Fig. 6b). Likewise, frequency of the conodont sizes of these two samples show wider size distribution range in sample 28 than in other samples (Fig. 6a and b).
Rarefaction curves of conodont elements based on sample numbers. Note that only well-sampled samples are indicated by their numbers. (a) Samples from Grisbachian; (b) samples from Dienerian; (c) samples from Smithian; (d) samples from Spathian.
The rarefaction curves for sampled beds are shown in Figure 7; the results show that only 21 samples are effectively robust. To minimize the sampling bias, the size variation profile has been established with conodont zones as the sampling bin. The rarefaction curves (Fig. 8) show that nine of 11 samples are efficient in sample size.
Size distribution histograms of conodont P1 elements near the Induan–Olenekian boundary (in log μm). n, number of specimens.
Size analysis with conodont zones as sampling bins. Age data are from Ovtcharova et al. (2006, 2015), Galfetti et al. (2007) and Burgess et al. (2014). It should be noted that for numerical timescale ordinal axis is non-linear.
Taking conodont zones as the sampling bins, the mean sizes decrease from the H. parvus Zone to the H. socioensis Zone (Fig. 9). Then, a positive excursion occurs in the C. carinata–C. planata Zone, but this trend is quickly reversed with a smaller mean size of conodonts in the N. dieneri Zone (Fig. 9). From the C. carinata–C. planata Zone to Nv. pingdingshanensis Zone, no mean size variation can be identified (Fig. 9). Following this, mean sizes increase from the Nv. pingdingshanensis Zone to the Tr. homeri Zone (Fig. 9). It is worth noting that segminiplanate conodont elements occur in the C. carinata–C. planata Zone, Nv. ex. gr. waageni Zone and Tr. homeri Zone, with larger size than the segminate conodont elements from the same beds (Fig. 5).
Rarefaction curves of conodont elements with conodont zones as sampling bins of these four substages. 1, Hindeodus parvus Zone; 2, Isarcicella lobata Zone; 3, Isarcicella isarcica Zone; 4, Hindeodus postparvus Zone; 5, Hindeodus sosioensis Zone; 6, Clarkina carinata -Clarkina planata Zone; 7, Neospathodus dieneri Zone; 8, Novispathodus waageni Zone; 9, Novispathodus pingdingshanensis Zone; 10, Icriospathodus collinsoni Zone; 11, Triassospathodus homeri Zone.
Taking substages as the sampling bins, conodont elements from the Smithian show the widest size distribution range. Conodont elements with size between 2.30 and 2.89 log µm contribute 99% of the conodont elements in the Griesbachian (Fig. 10a), whereas 99% of Dienerian conodonts are 2.20–2.79 log µm in size (Fig. 10b). In the Smithian, conodont elements of 2.20–3.09 log µm size form 99% of the specimens (Fig. 10c), whereas conodont elements sized 2.25–2.94 log µm represent 99% of the Spathian conodont elements (Fig. 10d).
Size distribution histograms of conodont P1 elements of each substages (in log µm). n, number of specimens. Dienerian (All conodonts) n = 521.
Discussion
Conodont element size variations during the Early Triassic
The size of conodonts fluctuates significantly throughout the Early Triassic (Fig. 11). In the early Griesbachian, conodont elements exhibit a progressive decreasing trend in size, ending with a transient size increasing trend in the middle Griesbachian, accompanied by the occurrence of segminiplanate elements (Fig. 9). No distinct trend or fluctuating trend can be identified during the Dienerian. The most distinct trend in increasing size is recorded across the Induan–Olenekian boundary, suggested by size variation profiles within all analysed sampling bins, including field samples, conodont zones and substages (Figs 2, 5 and 9). A decreasing trend in conodont size near the Smithian–Spathian boundary also is also evident.
Correlation of conodont, brachiopod and foraminifera diversity and size with key indicators of environmental change from the latest Permian to Middle Triassic. It should be noted that for numerical timescale ordinal axis is non-linear. Isotopic ages are U–Pb zircon ages from Ovtcharova et al. (2006, 2015), Galfetti et al. (2007) and Burgess et al. (2014). Carbon isotope curve simplified from Payne et al. (2004). Conodont oxygen isotopic curve modified from Sun et al. (2012). Redox conditions of the Nanpanjiang Basin are based on pyrite framboids (Tian et al. 2014). Unusual facies from Pruss et al. (2006) and Chen & Benton (2012). Foraminiferal diversity and size data from (Song et al. 2011a, b). Brachiopod diversity and size data from Chen et al. (2015a, 2018a). Conodont diversity data from Stanley (2009). Conodont size data from this study. M. Tr., Middle Triassic; U. P., Upper Permian; Lop., Lopingian; Chang., Changhsingian; Diver., diversity; C. S. P., carbonate seafloor precipitates; F. P. C., flat pebble conglomerates; W. S., wrinkle structures.
Previous studies of low-latitude faunas showed that conodont elements showed size reductions during the Permian–Triassic and Smithian–Spathian boundary crises (Luo et al. 2008; Chen et al. 2013). Recently, research on conodont element size variation from high-latitude peri-Gondwana faunas also showed dynamic changes near the Smithian–Spathian boundary, including both increasing trends in segminate elements and decreasing trends in segminiplanate elements (Leu et al. in press). In this study, not only size reduction at the Smithian–Spathian boundary has been recognized, but also several other minor size reductions in the early Griesbachian and Middle Smithian (Fig. 11). Meanwhile, there are also at least two size increasing intervals at the Motianling section: in the middle Griesbachian and across the Induan–Olenekian boundary, both associated with the re-emergence of segminiplanate elements (Clarkina and Eurygnathodus) (Fig. 2). In addition, the middle–late Spathian displays a long-term size increasing trend, and it is noticeable that this was interrupted by sample 39, which contains only one segminiplanate element (Fig. 5).
Size reduction across the Permian–Triassic boundary resulted from at least two different modes of size change: extinction of larger species and evolution of surviving lineages toward smaller sizes (Twitchett 2007; Schaal et al. 2016). Histograms of the conodont elements have shown that samples with larger mean sizes contain wider size distributions and a lower proportion of smaller individuals (Luo et al. 2006, 2008; Leu et al. in press), and our study shows the same mechanisms in the late Griesbachian and at the Induan–Olenekian boundary (Figs 6 and 10). Thus, we conclude here that conodont element size increases during the Early Triassic resulted from occurrences of larger species and evolution of extant lineages toward larger sizes.
Temporal conodont size variation as biotic response to environmental changes
Over the long term, marine animals experienced substantial size increase since the Cambrian, following Cope's rule (Heim et al. 2015). It has been long speculated that organisms are prone to increase in size over evolutionary time (Cope 1885, 1896), and these large-bodied animals played a critical role in ecosystem function (Estes et al. 2011). The loss of large taxa may have contributed to the multimillion-year delays in ecosystem recovery following catastrophes (Payne et al. 2016). Many studies have revealed that invertebrate or vertebrate marine clades experienced the ‘Lilliput effect’, a term coined by Urbanek (1993), who described the pattern of size change through extinction events: fossil organisms in the immediate aftermath of such events are typically much smaller than during pre-extinction times; for example, in the cases of the Permian–Triassic boundary and the Smithian–Spathian boundary. During these crises, high temperature (Sun et al. 2012), anoxia (Grasby et al. 2013), intensified continental weathering (Song et al. 2015) and reduced nutrient delivery to the photic zone (Grasby et al. 2016) are the main factors contributing to the Lilliput effect (Twitchett 2007; Mutter & Neuman 2009; Chen et al. 2013; Schaal et al. 2016).
The Lilliput effect at the Smithian–Spathian boundary was first reported by Chen et al. (2013). Their report on conodont elements from the Jiarong section described a significant size decrease before the Smithian–Spathian boundary (defined by the base of Nv. pingdingshanensis Zone). The results for the Motianling section confirm this decrease with a significant size drop below the Nv. pingdingshanensis Zone. An equivocal issue that requires resolution is that Chen et al. (2013) hypothesized that lethally hot seawater temperature affected the size of conodonts, whereas Goudemand et al. (in press) correlated the size decrease with a cooling near the Smithian–Spathian boundary (from Scythogondolella milleri Zone to Nv. pingdingshanensis Zone to Borinella n. sp. C Zone) by synthesizing previously reported data from South China and Pakistan and from measurements on seven new samples from Pakistan. Apparent inconsistencies are caused by the different definition of the Smithian–Spathian boundary: Goudemand et al. (in press) defined the Smithian–Spathian boundary by the FO of a new conodont species, Borinella n. sp. C, rather than Novispasthodus pingdingshanensis as used by Chen et al. (2013). Regardless of the disagreement on the Smithian–Spathian boundary, a negative O isotope excursion below the Nv. pingdingshanensis Zone can also be observed in the study by Goudemand et al. (in press), implying a warming event.
The size of conodont elements increased transiently during the middle Griesbachian (from the H. socioensis Zone to the C. carinata–C. planata Zone), although this trend ended with a rapid decrease (Figs 9 and 11), corresponding to a brief oxic episode recorded at an equivalent horizon (Twitchett et al. 2004; Dai et al. 2018). The conodont size increase at the Induan–Olenekian boundary of the Motianling section is coupled with a short-lived cooling of sea surface temperatures and a positive carbon isotope excursion (Fig. 11; Sun et al. 2012). Studies of pyrite framboids also indicated potential oxic ocean water conditions during this interval (Tian et al. 2014). Early Smithian ecosystem recovery is also suggested by a short absence of unusual sediments (Tian et al. 2015), reoccurrence of metazoan reef–biofilm (Brayard et al. 2011), primary size increase of gastropods (Brayard et al. 2010), recovery of foraminifers and initial size increase of foraminifers (Song et al. 2011a, b).
Similarly, the subsequent size increase of conodonts in the middle–late Spathian discovered in the Motianling section could also be attributed to environmental amelioration and ecological recovery. In South China, a habitable marine environment in the mid-Spathian is indicated by numbers of well-preserved and multi-clade marine reptile specimens at Chaohu (Fu et al. 2016). A widely distributed marine red bed also indicates the development of oxic watermass after the Smithian–Spathian boundary crisis (Sun et al. 2015; Song et al. 2017).
Conodont ecological rebounds to the Early Triassic ecosystem recovery
Previous studies show that segminiplanate conodonts, represented by Clarkina species, live in deeper and cooler water than segminate conodonts (such as Hindeodus species) (Lai et al. 2001; Joachimski et al. 2012). After the Permian–Triassic mass extinction, the dominance of Clarkina was replaced by the dominance of Hindoedus in South China, owing to the warming of the sea surface temperature (Nicoll et al. 2002; Jiang et al. 2007; Joachimski et al. 2012). The re-occurrences of segminiplanate elements and their co-occurrences with the segminate elements during the late Griesbachian, the earlist Smithian (Olenekian) and the late Spathian in Unit III could imply transient cooling events (Figs 5 and 11). However, the hypothesized Smithian–Spathian boundary cooling event (Goudemand et al. in press) cannot be proved by the absence of segminiplanate elements herein.
The pattern and the mechanism of post mass extinction biotic recovery has attracted much attention and discussion. Conventional studies have proposed that the Early Triassic ecosystem recovery was a long-term delayed resilience process (Payne et al. 2004). Recent studies, however, have indicated that this process was also coupled with several sub-extinctions, including both benthic and nektonic invertebrates (Stanley 2009; Payne et al. 2011). Recent high-resolution biodiversity studies show the existence of many transient biotic rebounds during the Early Triassic. For example, the late Griesbachian was marked by several temporal diversifications (Dai et al. 2018) and the early Smithian witnessed another rebound (or onset of recovery) of ecosystem or some organisms as the environment ameliorated (Song et al. 2014). Along with a high-resolution foraminiferal size study (Song et al. 2011a), our high-resolution conodont size study suggests that the size of marine invertebrates can also rebound rapidly with transient environmental changes. Thus, both diversity and size studies imply multi-phased ecosystem perturbations during the Early Triassic rather than the delayed recovery and long-term persistent depression scenario.
Environmental disturbance and ecological change during extinction perturbs the dynamic equilibrium of the Earth system, and biotic forcing is the ultimate driver of ecosystem evolution (Hull 2015). After the Permian–Triassic mass crisis, a newly established stable and complex ecosystem was built in a stepwise fashion and slowly from low to high trophic levels (Chen & Benton. 2012), during which the function of interspecific competition was emphasized in the trophic pyramid during the Early Triassic (Hautmann et al. 2015). Although there were several fast rebounds among different marine clades during the intervals when the environment was ameliorated, recovery of the whole ecosystem relied on the full stabilization of the environment and the recovery of different trophic levels represented by different clades (Payne et al. 2004; Chen & Benton 2012; Song et al. 2018).
Conclusions
Eleven conodont zones are recognized in the Lower Triassic of the Motianling Section of the isolated Debao carbonate platform in the Nanpanjiang Basin: Hindeodus parvus Zone, Isarcicella lobata Zone, Isarcicella isarcica Zone, Hindeodus postparvus Zone, Hindeodus sosioensis Zone, Clarkina carinata–Clarkina planata Zone, Neospathodus dieneri Zone, Novispathodus waageni Zone, Novispathodus pingdingshanensis Zone, Icriospathodus collinsoni Zone and Triassospathodus homeri Zone.
The Griesbachian–Dienerian boundary is placed at 35.94 m based on the FO of N. dieneri. The Induan–Olenekian boundary is placed at 50.96 m and recognized by the FO of Nv. waageni, and the Smithian–Spathian boundary is placed at 57.23 m based on the FO of Nv. pingdingshanensis.
Quantitative analyses of measurements on conodont elements reveal that their size showed episodic changes during the Early Triassic. The early Griesbachian, late Grisbachian and Smithian–Spathian boundary were all accompanied by size reductions of conodont elements. Otherwise, conodont size increased during the middle Griesbachian, the Induan–Olenekian boundary and the middle–late Spathian. The occurrences of larger elements contribute to the size increases of the conodonts, and provide evidence for the rapid rebound of conodont animal resilience with Early Triassic environmental ameliorations. Our high-resolution study indicates that marine vertebrates (conodont animals) could rebound to environmental changes as rapidly as invertebrates.
Acknowledgements
We thank reviewers P. Plasencia and D. Bond for their insightful and helpful reviews, which greatly improved the paper. We thank Ting Song and Ying Cao for their help in the field, and Xincheng Qiu and Dongdong Li for their assistance in the processing of conodonts. We are also grateful to Haishui Jiang for his comments on identification of conodonts.
Funding
This study was supported by the National Natural Science Foundation of China (grant numbers 41530104, 41602024, 41872033 and 41661134047), the international exchange programme funded by the postgraduate school of CUG (to K.W.) and Fundamental Research Funds for the Central Universities (CUGCJ1816) to China University of Geosciences (Wuhan). I.M. acknowledges support from the Australian Research Council.
Scientific editing by Graham Shields
- © 2019 The Author(s). Published by The Geological Society of London. All rights reserved
References
Recurrent biotic rebounds during the Early Triassic: biostratigraphy and temporal size variation of conodonts from the Nanpanjiang Basin, South China
