1. Introduction
Extensive recent research, especially in northern Europe and North America, has confirmed the usefulness of δ13C chemostratigraphy for both local and long-range correlations of Lower Palaeozoic rocks. Seven significant positive δ13C excursions have been recognized (Fig. 1) in the Middle and Upper Ordovician (Darriwilian through Hirnantian Global Stages), and recent investigations (Bergström et al. Reference Bergström, Young, Schmitz and Saltzman2007b) suggest that at least six of these excursions can be traced across the Atlantic from North America to Baltoscandia. The currently best known, and most investigated, among these excursions are the early Katian Guttenberg excursion (GICE) and the latest Ordovician Hirnantian excursion (HICE). Both of these excursions have been documented from a considerable number of localities in North America (see, e.g. Finney et al. Reference Finney, Berry, Cooper, Ripperdan, Sweet, Jacobson, Soufiane, Achab and Noble1999; Brenchley et al. Reference Brenchley, Marshall, Hints and Nõlvak1997, Reference Brenchley, Carden, Hints, Kaljo, Marshall, Martma, Meidla and Nõlvak2003; Saltzman et al. Reference Saltzman, Bergström, Huff, Kolata, Albanesi, Beresi and Peralta2003; Melchin, Holmden & Williams, Reference Melchin, Holmden, Williams, Albanesi, Beresi and Peralta2003; Saltzman & Young, Reference Saltzman and Young2005; Ludvigson et al. Reference Ludvigson, Witzke, González, Carpenter, Schneider and Hasiuk2004; Young, Saltzman & Bergström, Reference Young, Saltzman and Bergström2005: Bergström, Saltzman & Schmitz, Reference Bergström, Saltzman and Schmitz2006; Bergström et al. in press; Barta et al. Reference Barta, Bergström, Saltzman and Schmitz2007) and northern Europe (see, e.g. Marshall & Middleton, Reference Marshall and Middleton1990; Middleton, Marshall & Brenchley, Reference Middleton, Marshall and Brenchley1991; Brenchley et al. Reference Brenchley, Marshall, Carden, Robertson, Long, Meidla, Hints and Anderson1994, Reference Brenchley, Marshall, Hints and Nõlvak1997; Ainsaar, Meidla & Martma, Reference Ainsaar, Meidla and Martma1999, Reference Ainsaar, Meidla and Martma2004; Kaljo et al. Reference Kaljo, Hints, Martma, Nõlvak and Oraspõld2004; Kaljo, Martma & Saadre, Reference Kaljo, Martma and Saadre2007; Schmitz & Bergström, Reference Schmitz and Bergström2007). The HICE has also been recorded in China (Wang et al. Reference Wang, Orth, Attrep, Chatterton, Wang and Li1993; Wang, Chatterton & Wang, Reference Wang, Chatterton and Wang1997; Chen et al. Reference Chen, Rong, Fan, Zhan, Mitchell, Harper, Melchin, Peng, Finney and Wang2006a), and in the Subpolar Ural region of Russia (Männik et al. Reference Männik, Beznosova, Majdl, Martma, Hints and Ainsaar2004), but we are not aware of any previous record of the GICE from any locality within the vast Asian continent.
Recently recognized positive δ13C excursions in the Upper Ordovician (Katian and Hirnantian Global Stages) and the upper Middle Ordovician (Darriwilian Global Stage). Note the stratigraphic position of the Guttenberg excursion (GICE) in the lower Katian Stage. The figure is modified from Bergström et al. (Reference Bergström, Chen, Gutttiérrez-Marco and Dronov2008).
For several years we have carried out chemostratigraphic and biostratigraphic studies of the Pagoda, Linhsiang and Yenwashan formations in the Yangtze Platform and Chiangnan (Jiangnan) belt in southeastern China (Fig. 2). Our principal study localities are: the stratigraphically upper continuation of the Huangnitang GSSP section, Zhejiang (Chen et al. Reference Chen, Zhang, Bergström and Xu2006b), which represents the Chiangnan (Jiangnan) slope belt; the natural outcrop along the path on the hillside above Xiaotan village, Hexian, and outcrops at Beigongli, Jingxian, both sites being located in Anhui near the eastern margin of the Yangtze Platform; the inactive quarry on the hill slope above the Puxihe River near the ends of prominent bridges across a tributary of the Puxihe River about 29 km N of Yichang, Hubei; and an inactive quarry at Baiguowan, Donggongsi, near Zunyi, Guizou. Both the latter sites are situated in a more central portion of the Yangtze Platform. For more detailed locations of these sites, see the legend of Figure 2. Together, these sections may be viewed as representing a depositional environmental transect from the slope to the centre of the platform. Based on stratigraphically continuous sampling of these successions, we have assembled detailed conodont data from all the platform localities, and δ13C values are now available from the Huangnitang, Xiaotan, Beigongli and Puxihe sections. We plan to present most of the conodont data in a separate paper and the focus of the present contribution is the δ13C chemostratigraphy.
Sketch-map of the Yangtze Platform and adjacent regions in China showing early Late Ordovician (Katian) facies belts and the locations of the study localities, which are marked by X. The Huangnitang section is situated 5 km SW of the Changshan County town; that of Xiaotan is 30 km N of the Hexian County town; that of Beigongli is 25 km E of the Jinxian County town; that of Puxihe is 29 km N of Yichang City; that of Baiguowan, Donggongsi is 10 km N of Zhunyi City.
One important result of these investigations is the discovery (Bergström et al. Reference Bergström, Xu, Young, Schmitz and Saltzman2007a) of a significant positive δ13C excursion that we identify as the GICE, based on detailed conodont biostratigraphy. The recognition of the GICE in China not only confirms that this is a global excursion, but it also has major implications for the interpretation of the previously uncertain chronostratigraphic relations between a classic interval in the Chinese Ordovician succession and the Baltoscandic and North American successions. The purpose of the present report is to describe the isotopic record of this δ13C excursion, as well as the basic conodont biostratigraphy, in some selected sections in southern China, and to assess their significance for detailed correlations based on biostratigraphy and chemostratigraphy between the Chinese successions and coeval strata in Baltoscandia and North America. Also, we discuss a possible occurrence of the GICE in the Tarim Basin in northwesternmost China.
2. Stratigraphic framework
To clarify the stratigraphic context of the δ13C excursion interval, we will first briefly review the lithostratigraphy and biostratigraphy of the upper Sandbian and lower Katian successions, which include the Pagoda Formation and associated strata on the Yangtze Platform and its adjacent regions. The general stratigraphy in the study area is illustrated in Figure 3. The stage terminology used in this paper follows the recently ratified global stage terminology (cf. Bergström et al. Reference Bergström, Finney, Xu, Goldman and Leslie2006, Reference Bergström, Chen, Gutttiérrez-Marco and Dronov2008).
Stratigraphic chart showing late Sandbian and early–middle Katian formations in the Yangtze Platform region and their correlation to the Swedish (Fjäcka) and Kentucky successions based on biostratigraphy and chemostratigraphy.
2.a. Lithostratigraphy
In the Yangtze Platform succession, most of the Sandbian Stage is represented by the Miaopo Formation, a black to dark grey graptolitic shale that is less than 5 m thick at most localities. In some sections, there are a small number of thin interbeds of argillaceous limestone, particularly in the upper part of the unit. The Miaopo Formation is restricted to three major areas on Yangtze Platform (Chen & Qui, Reference Chen and Qui1986; Zhan & Jin, Reference Zhan and Jin2007, fig. 36), which have been interpreted to represent deeper-water intracratonic basins on the otherwise rather flat carbonate platform. In other parts of the platform deposition of limestones prevailed. These carbonate rocks are now referred to as the Datianba Formation (Nanjing Institute of Geology & Palaeontology, 1974). The conodont fauna of the latter unit is virtually identical to that of the Miaopo Formation (Hao Shougang, unpub. M.S. thesis, Peking Univ., 1981; An, Reference An1987; Chen et al. Reference Chen, Rong, Wang, Wang, Zhang and Zhan1995; Zhang, Reference Zhang1998) and these units are considered correlatives. Although abundant and taxonomically diverse graptolites dominate the fauna of the Miaopo Formation, it also contains many brachiopods, trilobites, ostracodes and conodonts.
The Miaopo Formation is conformably overlain by the Pagoda Formation, which is 5–60 m thick and covers thousands of square kilometres on the Yangtze Platform. It has the character of a blanket-type deposit but there are some indications, also from our study, that this unit is somewhat diachronous across its vast distribution area (Chen et al. Reference Chen, Rong, Wang, Wang, Zhang and Zhan1995). The dominant lithology is a light-grey to purple, medium to thick-bedded, limestone, but locally the unit contains thin partings of green shale. A striking, and much discussed, lithological feature in many sections is the presence of abundant shrinkage cracks in the limestone that are filled by calcareous muds. Although fossils are not abundant at most sites, the Pagoda Formation has yielded a relatively diverse fauna of trilobites, brachiopods, and other shelly fossils, among which nautiloid cephalopods are particularly prominent. The depositional environment of this extremely widespread deposit has been the subject of vastly different interpretations, ranging from shallow subtidal to water depths of several hundred metres (Zhan & Jin, Reference Zhan and Jin2007). It appears that the prevailing current opinion is that it is a relatively deep water, rather than shallow water, deposit which is consistent with the distinctive deep/cold water character of its conodont fauna.
The Pagoda Formation is overlain, apparently conformably, by the Linhsiang Formation. This 1–25 m thick unit includes medium- to thick-bedded, nodular, impure limestones and thin mudstone beds. It contains a diverse shelly fauna and a conodont fauna that is very similar to that of the underlying Pagoda Formation. The Linhsiang Formation is overlain by the Wufeng Formation, a very widespread black graptolite shale that is at most localities about 5 m thick and ranges to the base of the Hirnantian Stage. It contains the taxonomically most diverse graptolite fauna known in the upper Katian anywhere in the world (Chen et al. Reference Chen, Melchin, Sheets, Mitchell and Fan2005).
2.b. Biostratigraphy
The Miaopo Formation contains a taxonomically diverse graptolite fauna of the globally distributed N. gracilis Zone (Chen et al. Reference Chen, Rong, Wang, Wang, Zhang and Zhan1995). The graptolite species association typical of the overlying C. bicornis Zone has been recorded only from the Shuanjiakou section in Hunan (Liu & Fu, Reference Liu and Fu1989; Wang, Chen & Erdtmann, Reference Wang, Chen, Erdtmann, Webby and Laurie1992) that represents the Chiangnan (Jiangnan) slope belt. The lower part of the Miaopo Formation has yielded diverse conodonts of the Pygodus anserinus Zone (Ni & Li, Reference Ni and Li1987; An, Reference An1987), which is in good biostratigraphic agreement with the graptolites. At several localities, calcareous, non-graptolitiferous, beds in the very uppermost part of the formation have yielded specimens of the conodont subzone index species Baltoniodus alobatus (in older Chinese literature referred to as Prioniodus lingulatus; see, e.g. An, Reference An1987). The B. alobatus Subzone, which has been the youngest subzone of the Amorphognathus tvaerensis Zone in the Atlantic conodont zone classification scheme (Bergström, Reference Bergström1971, Reference Bergström, Ebbestad, Wickström and Högström2007), elsewhere corresponds to the middle–upper part of the C. bicornis Zone in the standard Pacific graptoplite zone succession.
It should be noted that in Baltoscandia, the top of the range of B. alobatus Subzone is well below the top of the Am. tvaerensis Zone, and in recent years, this post-B. alobatus interval of the Am. tvaerensis Zone has in Estonia been recognized as the Amorphognathus ventilatus Zone (see, e.g. Männik, Reference Männik and Põldvere2003, Reference Männik, Hints and Ainsaar2004). At the present time, we do not know whether the B. alobatus Subzone is restricted to the uppermost Miaopo Formation or if it extends into the lowermost part of the overlying Pagoda Formation.
The Pagoda and Linhsiang formations have not produced biostratigraphically diagnostic graptolites, but both of these units contain a moderately diverse conodont fauna (An, Reference An1987; Ni & Li, Reference Ni and Li1987), which seems to be quite similar at all study localities. The conodont species association represents the relatively deep-water Hamarodus–Dapsilodus–Scabbardella biofacies of Sweet & Bergström (Reference Sweet and Bergström1984). To illustrate the general composition of the conodont fauna of the Pagoda Formation, the conodont species ranges through the unit in the Puxihe section are shown in Figure 4. As far as we are aware, this is the first published diagram illustrating in some detail the conodont species succession through the Pagoda Formation.
Conodont species ranges in the Pagoda Formation in the Puxihe section. Note the occurrence of the biostratigraphically important species Amorphognathus aff. Am. ventilatus near the base of the Pagoda Formation.
A biostratigraphically significant, and geographically widespread, species that first appears within the Pagoda Formation, and is common in many samples, is Hamarodus europaeus. An (Reference An1987, p. 81) recognized a H. europaeus Zone in the Pagoda Formation, and other authors (see Chen et al. Reference Chen, Rong, Wang, Wang, Zhang and Zhan1995) have distinguished an Am. superbus–H. europaeus Zone, but it is outside the present account to discuss the merits of these rather loosely defined zone classifications. Nevertheless, it is appropriate to note that in Sweden, H. europaeus appears near the end of the GICE interval. This is slightly higher stratigraphically in relation to the GICE than is the case in our Chinese study sections, where this species first occurs near the middle of the GICE interval. However, the conodont fauna of most of the GICE interval in Sweden is sparse and includes mainly biostratigraphically non-diagnostic coniform taxa (Bergström, Reference Bergström, Ebbestad, Wickström and Högström2007), and it is quite possible that the currently known FAD of H. europaeus in Sweden is slightly younger stratigraphically than is the case elsewhere.
Specimens of the zone index Amorphognathus superbus have been reported from the upper Pagoda Formation at several localities (An, Reference An1987), but its precise range within the unit remains unknown. This is also the case with the reported occurrences of the morphologically characteristic and biostratigraphically useful species Amorphognathus complicatus. Specimens of Amorphognathus occur in some of our Pagoda samples but we have not yet recovered positively identifiable specimens of Am. superbus or Am. complicatus.
A sample from a level 2.5 m above the base of the Pagoda Formation in the Puxihe section has yielded several M elements of an Amorphognathus species that appears morphologically indistinguishable from specimens from the Mòjcza Limestone of Poland referred to as Am. aff. Am. ventilatus by Dzik (Reference Dzik1999, pl. 1, figs 6–10, 13–20). Based on direct comparison with Estonian specimens kindly supplied by Dr P. Männik, the Chinese specimens are also similar to the morphotype identified as Amorphognathus ventilatus by Männik (Reference Männik and Põldvere2003, Reference Männik, Hints and Ainsaar2004) and Männik & Viira (Reference Männik, Viira and Põldvere2005). Amorphognathus ventilatus was first described from significantly younger (late Katian; Am. ordovicicus Zone) strata (Ferretti & Barnes, Reference Ferretti and Barnes1997) than the early Katian units yielding the Polish and Estonian specimens. The species Am. ventilatus was based solely on M elements and the other elements of its apparatus have not been recognized in the type collection. According to Dzik (Reference Dzik1999, fig. 3), the Polish representatives of the dextral Pa elements have the extra postero-lateral process characteristic of Am. tvaerensis. This feature is unknown in stratigraphically younger representatives of the genus, such as Am. complicatus, Am. superbus and Am. ordovicicus, and it would be surprising if it reappeared in Am. ventilatus. In view of this, we believe that unquestioned reference of the early Katian specimens to Am. ventilatus may be premature. Pending a needed taxonomic revision of Amorphognathus, in the present paper we use the designation Am. aff. Am. ventilatus for the early Katian morphotype. Obviously, the taxonomic problems of this early Katian morphotype do not affect its stratigraphic utility.
The occurrence of this species in the Pagoda Formation is particularly important because it belongs to a plexus of rapidly evolving platform conodonts with great biostratigraphic utility. Both in Poland (Dzik, Reference Dzik1999) and Estonia (Fig. 5; see Männik Reference Männik and Põldvere2003, Reference Männik, Hints and Ainsaar2004; Männik & Viira, Reference Männik, Viira and Põldvere2005), Am. aff. Am. ventilatus has a very short stratigraphic range. Its range in the Estonian Mehikoorma (421) core is compared to that in the Puxihe section in Figure 5. Significantly, its first appearance datum (FAD) is just above the peak δ13C values of the GICE curve in both China and in Estonia. The presence of this conodont species in the δ13C excursion interval at Puxihe provides strong evidence that the Chinese δ13C excursion is indeed the GICE.
Stratigraphic chart illustrating the range of Amorphognathus aff. Am. ventilatus in relation to the GICE in the Puxihe section and the Mehikoorma (421) drill core, Estonia. The Estonian biostratigraphy and chemostratigraphy are based on Männik & Viira (Reference Männik, Viira and Põldvere2005) and Martma (Reference Martma and Põldvere2005), respectively. Note the close similarity in the first appearance datum (FAD) of this conodont species between the two sections. For explanation of lithological patterns in the stratigraphic columns, see Figure 4. The pattern in the 286.4–298.5 m interval in the Estonian drill-core denotes dolomitic marl with lenses and interbeds of carbonate.
Another conodont of potential stratigraphic significance is Protopanderodus insculptus, which appears in the upper Pagoda Formation at some localities and also is present in many samples from the overlying Linhsiang Formation. In North America, this species is not known below the upper part of the Amorphognathus superbus Zone, but it has recently been reported (although not figured) from what appears to be a much lower stratigraphic level in the well-known Upper Ordovician section along Mirny Creek in northeastern Siberia (Zhang & Barnes, Reference Zhang and Barnes2007). As a whole, the Pagoda–Linhsiang conodont fauna is so closely similar to those of the early–middle Katian Skagen, Moldå and Slandrom formations in Sweden (Bergström, Reference Bergström, Ebbestad, Wickström and Högström2007) that there is little doubt that they represent the same stratigraphic interval, which is consistent with the δ13C chemostratigraphy (cf. Fig. 3).
3. δ13C chemostratigraphy
3.a. Sample preparation and isotopic analyses
The lab preparation and isotopic analyses of the limestone samples followed the standard procedures described in Bergström, Saltzman & Schmitz (Reference Bergström, Saltzman and Schmitz2006) and Schmitz & Bergström (Reference Schmitz and Bergström2007).
3.b. Chemostratigraphy of the study sections
Figure 6 summarizes the chemostratigraphy of the four Chinese study sections as well as that of important sections in Sweden and Kentucky.
Correlation diagram based on δ13C chemostratigraphy and biostratigraphy showing the inferred relations between Chinese, Swedish and Kentucky formations. Note that the baseline and peak δ13C values are very similar in the illustrated sections. Also note that the GICE occupies a stratigraphically lower position in the Pagoda Formation in the central part of the Yangtze Platform (Puxihe section) than along the platform margin (Xiaotan and Beigongli sections). The Yanwashan Formation has frequently been correlated with the Pagoda Formation (Chen et al. Reference Chen, Rong, Wang, Wang, Zhang and Zhan1995), but the δ13C chemostratigraphy suggests that only its upper part is likely to be coeval with a part of the latter formation. The first appearance levels of key conodonts are shown by short horizontal lines in the Xiaotan and Puxihe sections. For explanation of lithological patterns in the stratigraphic columns, see Figure 4. The limestone pattern in the lower 4 m of the Kentucky column denotes the thick-bedded limestone of the Tyrone Formation.
The Huangnitang section represents a deeper-water slope environment. In most of the Yenwashan Formation, the δ13C baseline values are between +0.5‰ and +1‰, but in the uppermost 2–3 m of the formation, the δ13C values increase to near +2‰. We interpret this excursion, which is of admittedly rather small magnitude, as the GICE. At the present time, there is no reliable biostratigraphic control of the upper part of the Yenwashan Fomation, but conodonts of the Pygodus anserinus Zone and the B. alobatus Subzone of the Am. tvaerensis Zone have been recorded from the lower part of the formation (An, Reference An1987). This suggests that this interval is partially equivalent to the Miaopo Formation. It should also be noted that very recently, the Nemagraptus gracilis graptolite fauna was recorded from the top part of the Hulo Formation just below the base of the Yenwashan Formation at Huangnitang (Chen et al. Reference Chen, Zhang, Bergström and Xu2006b). The available biostratigraphic data are consistent with the idea that the identified GICE interval in the upper part of the Yenwashan Formation is coeval with the GICE interval in the Pagoda Formation on the Yangtze Platform. The shape of the δ13C curve suggests that only the lower part of the excursion is present in the Yenwashan Formation and, provided there is no stratigraphic gap at the formational boundary, we expect that the upper part of the GICE is represented in the lower part of the overlying Huangnekang Formation, the δ13C chemostratigraphy of which has not yet been investigated.
The Xiaotan (Hexian) and Beigongli (Jinxian) sections in Anhui represent sites near the eastern platform margin, the latter site being the one closest to the Chiangnan (Jiangnan) slope belt. At the former site there was in the past a complete exposure of the approximately 14 m thick Pagoda Formation and an exposed thickness of about 12 m of the underlying Datianba Formation. Unfortunately, recent farming activities have resulted in much of the formerly excellent outcrop now being covered. At Xiaotan, the δ13C baseline δ13C values are ~+0.5‰ in the upper Datianba Formation and shift toward more positive values through the lower Pagoda Formation, with δ13C values reaching a maximum of ~+2.3‰ in the middle Pagoda Formation near the level of the first appearance of H. europaeus. We interpret this excursion as the GICE. The δ13C values return to baseline values in the upper part of the Pagoda Formation. The thickness of the excursion interval is about 10 m.
Several metres of the lower Pagoda Formation are covered in the Beigongli section, but the δ13C data from the current exposure shows that an excursion identified as the GICE is present in the same interval of the Pagoda Formation, and reaches the same peak values, as at Xiaotan. Interestingly, the Beigongli GICE curve appears to have an incipient two-peak shape that approaches its appearance in many GICE localities in North America and northern Europe (cf. the curves from the Fjäcka and Kentucky successions shown in Fig. 6). The incompletely exposed excursion interval at Beigongli has a thickness of about 5 m.
The entire Pagoda Formation, as well as immediately underlying and overlying strata, are excellently exposed at the Puxihe Quarry N of Yichang in Hubei (Fig. 7). As in the other study sections, the δ13C baseline values are ~+1‰ and the GICE excursion peak values a little more than +2‰. However, as shown in Figure 6, the interval of GICE is present in the lower, rather than the upper, part of the Pagoda Formation. This is consistent with the appearance of Hamarodus europaeus and Amorphognathus aff. Am. ventilatus at 1.5 and 2.5 m, respectively, above the base of the formation. This suggests that the base of the Pagoda Formation is somewhat diachronous from the platform margin to the central part of the platform. The significance of the presence of the biostratigraphically important species A. aff. Am. ventilatus just above the peak of the excursion has been discussed in Section 2.b.
The exposure of the entire Pagoda Formation at the Puxihe Quarry near Puxihe River about 29 km N of Yichang, Hubei. Note the black shale of the Miaopo Formation at the bottom of the quarry and the nodular, thin-bedded limestone of the Linhsiang Formation at the top of the quarry wall. The Pagoda Formation is about 11.5 m thick at this locality.
3.c. Regional chemostratigraphic correlation
Based primarily on δ13C chemostratigraphy, the suggested stratigraphic relations between the Chinese study sections, the Fjäcka section in Sweden and the composite Kentucky section in the USA are shown in Figure 6. As shown in this figure, the correlation of the δ13C curves between these geographically widely separated regions appears straightforward and consistent with the conodont biostratigraphy. It is appropriate to note that because of pronounced faunal differences, it is currently not possible to correlate directly between China and Kentucky by means of conodont biostratigraphy, but such a correlation is possible between Baltoscandia and Kentucky. Interestingly, the base of the Pagoda Formation on the Yangtze Platform appears to be broadly at the same stratigraphic level as the base of the Skagen Formation in Sweden and the base of the Lexington Limestone in Kentucky. At least in the Puxihe section, the Miaopo/Pagoda Formation contact, although lithologically distinct, appears gradational, as is the case with the Dalby/Skagen formational contact in Sweden. However, the Tyrone/Lexington Formation contact in Kentucky is marked by a prominent unconformity that represents the M4/M5 sequence boundary of Holland & Patzkowsky (Reference Holland and Patzkowsky1996). Although there is little doubt that the studied Chinese and Baltoscandian depositional environments below and into the GICE interval represent a shallowing episode, those in Kentucky and several other North American regions (Oklahoma, Upper Mississippi Valley, New York State–Ontario; cf. Bergström et al. in press; Barta et al. Reference Barta, Bergström, Saltzman and Schmitz2007) seem to be the opposite and reflect an early phase of a transgression. Hence, the relations between the GICE and eustasy are not straightforward and require further study.
4. Possible occurrence of the GICE in the Tarim Basin succession
As noted above, the present study is the first to positively identify the GICE in China. However, there is a published δ13C record that suggests that this δ13C excursion may be present in another part of China. This record can be found in an internationally widely overlooked pioneer study of carbon, strontium and oxygen isotopes in marine carbonates from the important Ordovician succession of the Tarim Basin in Xinjiang, northwesternmost China, by Jiang et al. (Reference Jiang, Zhu, Chen, Zhang and Qiao2001). In this investigation they presented δ13C data from 86 samples collected from four drill-cores in the Tazhong region in the Tarim Basin. Unfortunately, their report includes very few biostratigraphic data that can be used for assessing the precise age of their samples. A few conodont species suggest a Tremadocian–Floian (Early Ordovician) age for their stratigraphically oldest samples, most of which show baseline δ13C values of <+1‰. However, many samples from a stratigraphically younger interval in the Tazhong 12 and Tazhong 29 drill-cores show δ13C excursion values of ~+2‰ (Fig. 8), which are comparable with those of the GICE in the Pagoda Formation recorded above. The few conodont and coral species listed from this interval by Jiang et al. (Reference Jiang, Zhu, Chen, Zhang and Qiao2001) are not useful for precise biostratigraphic dating, but Zhao, Zhang & Xiao (Reference Zhao, Zhang and Xiao2000) reported Plectodina n. sp. B, Pl. n. sp. C, Panderodus gracilis and Taoqupognathus tumidus from the suggested excursion interval in the Tazhong 12 drill-core. Slightly higher stratigraphically, these conodont species are followed by Belodina confluens. The evolutionary stage of the two Plectodina species is comparable to that of Pl. aculeata and related Plectodina taxa that first appear in the North American Midcontinent succession in strata of late Sandbian age just below the GICE interval (Bergström & Sweet, Reference Bergström and Sweet1966). Significantly, in the latter region, B. confluens first occurs just above the GICE interval (Young, Saltzman & Bergström, Reference Young, Saltzman and Bergström2005). The morphologically distinctive and stratigraphically short-ranging species Taoqupognathus tumidus was first described from the Dicranograptus kirki Zone (Ea3) in southeastern Australia (Trotter & Webby, Reference Trotter and Webby1994). The precise international correlation of this graptolite zone is somewhat problematic, although Williams (Reference Williams1982) considered it to correspond to the upper Dicranograptus clingani Zone and lower Pleurograptus linearis Zone in northern Europe. It should also be noted that in another well, the Tazhong 29 well, Zhao, Zhang & Xiao (Reference Zhao, Zhang and Xiao2000) recorded Dicranograptus cf. clingani resicis from the excursion interval. Assuming that this graptolite is correctly identified, it is indicative of the Dicranograptus clingani Zone (Williams & Bruton, Reference Williams and Bruton1983), a graptolite zone that contains most of the GICE in Baltoscandia (Bergström et al. Reference Bergström, Huff, Saltzman, Kolata and Leslie2004, fig. 4). Hence, these limited biostratigraphic data suggest that the Tazhong δ13C excursion is comparable in age to that of the GICE in North America and northern Europe. However, the biostratigraphically incomplete data at hand do not exclude the possibility that this excursion rather represents the slightly younger Kope δ13C excursion of Bergström et al. (Reference Bergström, Young, Schmitz and Saltzman2007b), even if we believe that the evidence now available favours the identification of the excursion as the GICE. Clearly, more detailed chemostratigraphic and biostratigraphic studies are needed in this remote part of China to confirm fully the presence of this intriguing isotopic perturbation in the carbon isotope cycle.
δ13C curve through the thick Ordovician succession of the Tazong 12 well in the Tarim Basin based on sample data from Jiang et al. (Reference Jiang, Zhu, Chen, Zhang and Qiao2001). To illustrate the excursion in a smoother curve, three-point average δ13C values were used to construct the isotope curve. The GICE is interpreted to be in the remarkably thick interval between approximately 4700 m and 5100 m. For explanation of the lithological pattern in the stratigraphic column, see Figure 4.
5. Conclusions
The principal results of the present study may be summarized as follows:
(1) For the first time, the early Katian δ13C excursion known as the GICE is documented from eastern Asia, where it is recorded from several sections of the Pagoda and Yanwashan formations on the Yangtze Platform and the Chiangnan (Jiangnan) slope belt in southeastern China.
(2) In terms of conodont biostratigraphy, this excursion occurs in the same stratigraphic interval as in Baltoscandia. This interval can be correlated with the global early Katian Stage, as well as with the early part of the North American Chatfieldian Stage.
(3) The discovery of the GICE in the Chinese Ordovician succession confirms that the GICE was a global perturbation of the carbon cycle, and in this respect it is comparable with the latest Ordovician Hirnantian excursion (HICE).
(4) The Pagoda Formation is one of the most distinctive, well-known and geographically widespread formations in the Ordovician of China, but its precise age and international correlation have remained uncertain. The new evidence from δ13C chemostratigraphy and conodont biostratigraphy constrains the age of most of the formation, although the age of its topmost part still remains uncertain. There are also strong indications that the Pagoda Formation is somewhat diachronous across the Yangtze Platform, but further studies are needed to assess this interpretation fully.
(5) Somewhat inconclusive data suggest that the GICE may be recognized also in the subsurface of the Tazhong area in the Tarim Basin in northwesternmost China.
(6) The present study illustrates the usefulness of δ13C chemostratigraphy for precise correlations over ultra-long distances. It confirms that in the Upper Ordovician, not only the HICE but also the GICE is a unique global correlation tool that can be successfully employed for clarification of stratigraphic relationships that are difficult, if not impossible, to work out using ordinary biostratigraphic methods.
Acknowledgements
Support to the senior author, Chen Xu and Rong Jia-yu from the Chinese Academy of Sciences (Project KZCX2-YW 122) and the National Nature Foundation (Project 40532009) made this study possible and is gratefully acknowledged. We thank Dr P. Männil for providing Estonian specimens of Am. aff. Am. ventilatus and Mr Chen Qing and Mr Brad Cramer for drawing the text-figures. We are also indebted to the reviewers, Drs D. Kaljo and L. Ainsaar, for valuable manuscript comments.
