Three sections were analyzed in this study, namely, the Jiaodingshan section, the Sanquan section, and the Hongyan section, which are widely spread all over the Upper Yangtze Platform. The Jiaodingshan section spreads into the Malie, Hanyuan, Yaan, and Sichuan provinces and is located 67 km northeast of the Hanyuan government office (102.87N, 29.4E) (Figure 1B). The Jiaodingshan is the eastern remnant of the Daxiangling Mountain Range, which shows a north-south trend where the Middle and Upper Ordovician were well exposed (Figures 2E, F), with the thickness of the study interval being 39.8 m. The Sanquan section is on the west of the riverside, northwest of Sanquan Town, Nanchuan District, Chongqing (29.14°N, 107.19°E). The study interval, which is 23.0 m, is a section discovered with three continuous and well-exposed Ordovician series (Figures 2B, C, D, G, H). The Hongyan section is a section situated in the south of Xikou Town, Huaying, Sichuan Province (30.16°N, 106.69°E), with the thickness of the study interval being 26.4 m (Figure 2A) (Ma et al., 2022).

The Upper Ordovician developed well in the Jiaodingshan section. It developed the Qiaojia, Pagoda, and Linxiang formations in ascending order (Table 1). The Lower Qiaojia Formation is made of purple-red mudstone that totals 7.8 m in thickness with 2.2 m of limestone developed at the top. The Pagoda Formation totals a thickness of 33.5 m including, from bottom to top, 8 m of purple-red limestone, 14.5 m of dark grey limestone, and 11 m of grey limestone. The Linxiang Formation is mainly formed of 3 m of dark grey mudstone and 2.5 m of grey limestone in ascending order. The partially covered black graptolite mudstone of the Longmaxi Formation can be seen on the upper outcrop.

In the Sanquan section, the Guniutan Formation (Table 1) is mainly composed of bioclastic limestone with a thickness of 17.3 m, with the Baltoniodus variabilis conodont zone being recognized on the top of this formation. The middle and lower Pagoda Formation is characterized by reddish and medium thick-bedded limestone, with grey mudstone about 30 cm thick at the bottom. The middle and upper Pagoda Formation is characterized by grey medium-thick bedded limestone with a thickness of 23.5 m, which is in parallel unconformity contact (paraconformity) with the underlying Guniutan Formation. It also yielded conodont fauna, and Protopanderodus insculptus and Hamarodus brevirameus zones were established in the Pagoda Formation. The Linxiang Formation is characterized by grey micritic limestone and nodular limestone, which is about 4 m thick. The Wufeng Formation (Table 1) is covered with black carbonaceous shale, calcareous mudstone, and marl, about 10 m thick, and produces a large number of graptolites.

In the Hongyan section, the Miaopo Formation is constituted by grey limestone (9.5 m) with a ‘horse hoof crack’ texture, and the Pagoda Formation presents grey limestone (22.5 m). The Miaopo Formation is defined according to the occurrence of the conodont fauna, Yangtzeplacognathus jianyeensis. The Y. jianyeensis zone was widely established in the Ordovician Miaopo Formation of South China. The Miaopo Formation was also established in the Chenjiahe section of the Hubei province and is located near the Global Stratotype Section and Point for the Middle Ordovician Series.

The limestone of the Pagoda Formation is a large-scale and widely distributed strata in southern China. It has high consistency in the lithology in study sections, consisting of grey to light purplish red, middle-thick-bedded limestones with polygonal reticulate structure or ‘horse hoof crack’ texture that is especially apparent on bedding planes. Many scholars have studied its genesis, biological fossils, and duration (Sheng and Ji, 1984; An et al., 1985; Sheng and Ji, 1986; Mu et al., 1987; Ma et al., 2019b) and the results have significant implications for fossil fuels generation and paleoenvironment of the Yangtze Platform.

In this study and search for conodonts, 28 samples were collected from the Qiaojia to the lower Linxiang Formation in the Jiaodingshan section. In total, 4, 21, and 3 samples were collected in the Qiaojia Formation, Pagoda Formation, and Linxiang Formation, respectively, with a weight of 5–7 kg per sample (Figure 3). The collected samples were broken into 2–3 cm³ small pieces and digested in 10% acetic acid, and the insoluble substances at the bottom were separated every 2–3 days. The samples were screened to remove the residual acid and impurities with 20 mesh and 160 mesh sample sifters, respectively. After drying the screened samples, the conodonts were further separated by LST liquid with a density of 2.80 g/mL. At last, abundant conodonts were obtained under binocular stereography.

In order to carry out the carbon and oxygen isotope stratigraphic analysis, 61 samples were obtained from the Qiaojia to Linxiang Formation of the Jiaodingshan section. In all, 42 samples were obtained from the Sanquan section and 53 samples were obtained from the Hongyan section. The samples were carefully selected in the Palaeontology Laboratory of the Research Institute of the Petroleum Exploration and Development, PetroChina, and ground into 200 g of powder using a micro-drill, avoiding the samples from becoming polluted. The samples were first ground before a mm-sized hole was drilled. Then the samples were processed in the Isotope Laboratory of the Nanjing Institute of Geology and Palaeontology, CSA. The powder was reacted with 100% anhydrous phosphoric acid for 24 h at 22°C± 1°C and 50% RH± 5% in humidity to produce CO₂. The samples were then tested on the Thermo Scientific MAT 253 stable isotope mass spectrometer. The standard isotope sample is GBW-04405 (National Standard) (δ¹³C= 0.57, δ¹⁸O= −8.49). The analysis results of δ¹³C and δ¹⁸O (both Vienna Pee Dee Belemnite values) were obtained with 0.03 and 0.14 standard deviations, respectively (Table 2).

Three conodont zones were established in ascending order based on the conodont fauna collected from the Jiaodingshan and Hongyan sections, respectively.

The δ¹³C isotope samples analyzed were collected in the Qiaojia, Pagoda, and Linxiang Formations. According to the characteristics of lithology and isotopic value, the Upper Ordovician in the Jiaodingshan section was divided into seven isotopic stages from the bottom to the top, respectively, named A1 to A7 (Figure 6). The A1 and A2 could be correlated with the Upper Qiaojia Formation and the bottom of the Pagoda Formation. A3–A6 and lower A7 correlated with the Pagoda Formation. The middle and upper A7 stages were defined in the Lower Linxiang Formation.

In order to ensure the validity of carbon and oxygen isotope data, fresh rocks without veins and fractures were selected for sample collection in the rock outcrop, and the structure of late striking, recrystallization, and pore fracture were avoided as far as possible. The fresh rock samples were ground indoors, and the quality was guaranteed by avoiding the calcite vein. First, the validity of the carbon isotope data was analyzed based on oxygen isotope data. The accepted range of the effective δ¹⁸O was greater than −10‰ (Kaufman and Knoll, 1995; Zhu et al., 2004; Li et al., 2009), and the results demonstrated that almost all the δ¹⁸O values obtained in the Jiaodingshan section were greater than −10‰. However, a δ¹⁸O value of less than −10%∘ carbon isotope value is not fully determined to be invalid data because of the following reasons. Small amounts of carbon are exposed to carbonate rocks based on the low carbon concentration of diagenetic pore fluid. The atmospheric freshwater or hydrothermal water, which promotes the oxygen isotope rebalancing, sometimes is not strong enough to change the value of the carbon isotope, so the carbon isotope is less easily affected by the diagenesis compared with the oxygen isotope (Anderson and Arthur, 1983; Glumac and Spivak-Birndord, 2002). Hence, in the actual study, even if the value of δ¹⁸O is less than −10‰, the value of δ¹³C may still be valid (Dilliard, 2007; Wotte et al., 2007) ¹ . Therefore, although the negative value of δ¹⁸O of a sample could be used as a good index to determine whether the value of δ¹⁸O is affected by the diagenetic, how it effectively indicates the change of carbon isotope composition needs to be further studied and should be used in combination with other methods (Dilliard, 2007; Wotte et al., 2007) ¹ .

The second method to ensure the validity of carbon isotope data is the scatter plot of the δ¹³C and δ¹⁸O (Table 2). There is no linear correlation between the two data series, showing a dispersed distribution. This indicates that the strata were not deposited under the influence of diagenetic transformation (Bathurst, 1972; Brand and Veizer, 1980; Brasier et al., 1990; Brasier et al., 1996; Dilliard, 2007). All in all, combining these two methods, the δ¹³C and δ¹⁸O values in the study sections were considered to be reliable (Table 2). The isotopic results in the Sanquan and Hongyan sections are quoted from Ma et al. (2019c), Ma et al. (2022).

In the isotopic seven stages, there were three distinct positive excursions of δ¹³C values in the range of 2.0‰–3.0‰ in A2, A4, and A6, with the first lower peak occurring at the first appearance of conodont H. brevirameus. This positive excursion of δ¹³C isotopes was found not only in the Jiaodingshan area but also in the Xiaogangwan (Qixian, Chongqing) and Sanquan (Nanchuan, Chongqing) areas, both located in the eastern Upper Yangtze Platform, the Liangcun area, located in the southern part of the Upper Yangtze Platform, the Hongyan area, and even in the Dawangou section, which is the Late Ordovician GSSP (Global boundary Stratotype Section and Point), in the Kelpin area and is located in the western margin of the Tarim Basin (Hennissen et al., 2016).

Some authors interpreted this event as a global carbon isotope excursion event, Guttenberg δ¹³C Excursion (GICE) (Bergström et al., 2009). However, we compared it with the strata developing conodont fossils of North America (Bergström and Ferretti, 2017; MacDonald et al., 2017) and defined this positive excursion event as PPCE (Pagoda Positive δ¹³C Excursion), which occurred in a higher layer than the layer occurring in the GICE event. It is believed that it is not the GICE event, but a new carbon isotope positive event, PPCE, represented at least regionally. PPCE developed in the Middle-Late Katian and Late Ordovician, while the Guttenberg δ¹³C Excursion (GICE) defined in the Baltoscandia and America, extended from the Late Sandbian to the Early Katian. Moreover, it was demonstrated that the PPCE (3.68–4.37 myr) and GICE (2.35 myr) events have different durations based on cyclostratigraphic results by Ma et al. (2022). The unconformity most likely results in different durations of the PPCE. The main difference of the PPCE in different sections is one of the three secondary positive excursions (P1) occurring in the early period of the event. The P1 recognized in the Pagoda Formation is less completed in the Hongyan section than in the Sanquan section, which illustrates that the Pagoda Formation is in parallel unconformity contact with the underlying strata and the stratigraphic gap underlying the Pagoda Formation in the Sanquan section is larger than that in the Hongyan section. The comparison results and the duration time have been published in another study and will not be discussed in-depth (Ma et al., 2022).

It has been found that there are several regions in the Upper Yangtze Platform with very similar carbon isotope characteristics and similar sedimentary backgrounds, all of which are found in the Pagoda Formation. The Pagoda Formation is the most widely distributed and easily recognized early Paleozoic strata in the Yangtze Platform. The lithologic characteristic is grey or light purplish argillaceous limestone, with a “horse hoof crack” texture. A large number of conodont faunas were found there. The study of carbon isotope characteristics and conodont zones in different areas of the Yangtze Platform is instructive for studying the development of the Pagoda Formation and estimating the missing amount of the underlying strata. The Xiaogangwan section (Qixian, Chongqing province) and Sanquan section (Nanchuan, Chongqing province), located in the eastern Upper Yangtze Platform, the Liangcun section (Xishui, Guizhou province), located in the southern portion, and the Jiaodingshan section, located in the western portion, were chosen to discuss the hiatus underlying the Pagoda Formation.

The characteristics of the δ¹³C isotope curve in the strata of the studied Jiaodingshan section could be correlated with the Xiaogangwan section and the adjacent Sanquan section (Fan et al., 2015b; Ma et al., 2019a) (Figure 6). Comparing the two profiles, the Middle and Upper Ordovician carbon isotopic assemblages in the Jiaodingshan section were composed of three positive secondary excursions, with peak values ranging from 2.0‰ to 3.0‰. All of them began to develop at the bottom of the Pagoda Formation and extended upward to the middle and upper Pagoda Formation. The positive excursion events all began at the first appearance of the conodont H. brevirameus, and the largest peak in the Jiaodingshan section belongs to the upper Pagoda Formation, while the largest peaks in the Xiaogangwan and Sanquan sections belong to the lower Pagoda Formation. Regarding the δ¹³C isotope sequence at the bottom of the Pagoda Formation, it was observed that the development of the δ¹³C isotope sequence in the Sanquan section was different from that in the Jiaodingshan and Xiaogangwan sections. This may be due to the possibility that the real FAD of the H. brevirameus in the Xiaogangwan and Jiaodingshan section was lower than the level of the P1 peak, but the earlier occurrences of H. brevirameus were not identified in our samples. Another possibility is that the P1 peak in the Sanquan section was misidentified and should be in the lower layer.

The δ¹³C isotope sequence in the Jiaodingshan section could be correlated with that in the Liangcun section (Fan et al., 2015b), with three secondary excursions and similar curve evolution. The first excursion was located in the first appearance of the conodont H. brevirameus, and the secondary excursion was located in the first appearance of the conodont P. insculptus, which is highly consistent within the two sections. The difference is that the range of the peak value was more pronounced than that of the Liangcun section, with the secondary peak value of carbon isotope in the Liangcun section in the range of 1.8‰–2.2‰. The cyclostratigraphic results indicated the duration of the PPCE in the Liangcun and Sanquan sections. The duration of the PPCE in the Sanquan section was 4.37 myr and that in the Liangcun section was 3.75 myr (Ma et al., 2022).

Comparing the δ¹³C isotope sequence in the Hongyan section with the sequence in the Jiaodingshan section, it was observed that three secondary excursions were identified in both sections with similar shape curves, and the largest peaks were both in the upper strata (Figure 6). The secondary peak value of the Jiaodingshan section ranged from 2.0‰ to 2.8‰, and the value of the Hongyan section ranged from 1.5‰ to 2.4‰. There is only one conodont zone, H. brevirameus zone, that occurs in the Pagoda Formation, and no conodonts were found in the upper formation. The cyclostratigraphic result in the Hongyan section illustrated that the excursion event lasted approximately 3.68 myr (Ma et al., 2022).

To conclude, the carbon isotopic excursion identified in the Pagoda Formation was defined as the Pagoda Positive δ¹³C Excursion (PPCE) with three secondary excursions, ranging from 2.0‰ to 3.0‰, which started developing in the same layer with the first appearance of the conodont H. brevirameus. The different features of the earliest secondary excursion P1 in the study sections led to the incompleteness of the PPCE. This may be due to the stratigraphic record underlying the Pagoda Formation.

The global and regional tectonic movements intensified since the Ordovician, resulting in the gradual disintegration of the Gondwana supercontinent. The marginal Gondwana, such as South China, migrated away from Gondwana to the middle and low latitudes, forming a natural geographically isolated region, which stimulated the evolution of various species. It is speculated that the South China plate of the Late Ordovician, namely, the Late Ordovician Pagoda Formation of the Upper Yangtze Platform, was affected by tectonic movement during its depositional period, resulting in stratigraphic discontinuity and stratigraphic chronological discontinuity between the Pagoda Formation and the underlying formation. The mass generation of methane hydrate is believed to be richer in ¹²C, and when the global climate cooled, the reserves of methane hydrate in the ocean increased by storing the organic ¹²C in the methane hydrates or the frozen grounds (e.g., permafrost), leading to an increase in the δ¹³C of atmosphere and oceans. This may be related to the cold paleoenvironment of the Upper Yangtze Platform, that is, the South China plate in the Late Ordovician, which was subjected to the Ice Age of the Late Ordovician. Furthermore, the Ordovician radiation caused the organisms to flourish, a process that lasted throughout the whole period of the Ordovician, and the peak of biodiversity before the mass extinction pushed the Ordovician radiation to its climax at the end of the Ordovician (Sepkoski, 1979; Miller and Mao, 1995; Zhan et al., 2013). The statistical results of various marine species showed that two of the three climax acts occurred in the Late Sandbian and Late Katian ages, dominated by brachiopods and other filter feeders, which were in the same period as that examined by this study. During this period, the initial productivity increased, and a large amount of ¹²C-rich organic carbon was buried relative to inorganic carbon (carbonate), resulting in a positive excursion in the δ¹³C value of carbonate.