Chemostratigraphic correlations across the first major trilobite extinction and faunal turnovers between Laurentia and South China

During Cambrian Stage 4 (~514 Ma) the oceans were widely populated with endemic trilobites and three major faunas can be distinguished: olenellids, redlichiids, and paradoxidids. The lower–middle Cambrian boundary in Laurentia was based on the first major trilobite extinction event that is known as the Olenellid Biomere boundary. However, international correlation across this boundary (the Cambrian Series 2–Series 3 boundary) has been a challenge since the formal proposal of a four-series subdivision of the Cambrian System in 2005. Recently, the base of the international Cambrian Series 3 and of Stage 5 has been named as the base of the Miaolingian Series and Wuliuan Stage. This study provides detailed chemostratigraphy coupled with biostratigraphy and sequence stratigraphy across this critical boundary interval based on eight sections in North America and South China. Our results show robust isotopic evidence associated with major faunal turnovers across the Cambrian Series 2–Series 3 boundary in both Laurentia and South China. While the olenellid extinction event in Laurentia and the gradual extinction of redlichiids in South China are linked by an abrupt negative carbonate carbon excursion, the first appearance datum of Oryctocephalus indicus is currently the best horizon to achieve correlation between the two regions.


Analytical methods for carbon and oxygen isotopes Carbonate carbon and oxygen isotope analysis
All samples for isotope analyses were cleaned and weathered crusts and calcite veins were avoided. Samples analyzed at the University of California, Davis (UCD) were micromilled and were roasted at 375°C under vacuum for 30 minutes to remove organics and subsequently reacted in 105% phosphoric acid at 90°C in either a common acid bath on a GVI Optima Stable Isotope Ration Mass Spectrometer or a Gilson Multicarbon Autosampler system interfaced with an Elementary Isoprime Mass Spectrometer. CO2 gas was analyzed in dual inlet mode and the resulting values were corrected using an internal standard. Samples obtained at the University of Münster, Germany were pulverized using a tungsten carbide ring and puck mill. Material was analyzed for δ 13 Ccarb and δ 18 Ocarb using a Gasbench II connected via a ConFlow-III Interface to a Finnigan MAT DeltaPlusXL. Samples analyzed at the Las Vegas Isotope Science Laboratory of the University of Nevada, Las Vegas (UNLV) were microdrilled and reacted with orthophosphoric acid for 10 minutes at 70°C in a Kiel-IV Device automatically connected to a Finnigan DeltaPlus dual-inlet mass spectrometer. Precision monitored by NBS-19 calcite and an internal standard (UNLV, UCD & Münster) was < 0.1‰ for both δ 13 Ccarb and δ 18 Ocarb with the exception of δ 18 Ocarb (≤ 0.2‰) at the University of Münster lab. All delta values are reported relative to the VPDB standard.
Additional samples from the Wuliu-Zengjiayan section and the Miaobanpo section were collected using both a Dremel drill and/or a dental drill from clean, cut surface of each sample. Approximately 200 μg of sample powders were weighted out for carbonate carbon (δ 13 Ccarb) and oxygen (δ 18 Ocarb) analyses.
Isotopes were analyzed with a Finnigan DeltaPlusXP system at the Earth Systems Center for Stable Isotopic Studies of Yale University (ESCSIS) and a Finnigan MAT252 at the Nanjing Institute of Geology and Palaeontology (NIGP). Reproducibility monitored by NBS-19 is better than 0.1‰ for both δ 13 Ccarb and δ 18 Ocarb. Samples from the Miaobanpo section were also analyzed for organic carbon (kerogen) isotopes.
Approximately 30 mg powders of each sample were collected and sent to a commercial lab for analyses.

Organic carbon isotope analysis
Carbon isotope analyses of organic matter (δ 13 Corg) occluded within the bulk rock (shales to shaley limestones) were carried out at two facilities at the University of California, Davis. Prior to powdering the bulk rock samples by tungsten ball mill, the billets were trimmed to remove any observable weathering surfaces or fracture/secondary void filling calcites. In order to assess whether contamination from modern or extra-formational (i.e. hydrocarbons) organic matter has affected the measured δ 13 Corg values, 40 shale samples were separated into two aliquots. One aliquot was treated solely with trace-element-grade hydrochloric acid (HCl), whereas the second aliquot was pretreated with dimethylchloride (DMC) following decarbonation using hydrochloric acid. Pretreating sedimentary rock samples characterized by low level TOC (total organic carbon) contents has been shown to be effective to remove labile, post-depositional organic carbon from kerogen. The majority of δ 13 Corg values (72%) exhibit minimal difference (of  1‰ around zero) between samples extracted using HCl and samples prewashed with DMC prior to extraction with HCl indicating minimal effect of the DMC treatment. The difference between the two methods ( 13 Corg) can be expressed in the following equation (1): The remaining samples (28%) of the  13 Corg values range between 0.07 and 4.84‰ but do not show a consistent trend toward negative or positive  13 Corg values. Based on this experiment, subsequent samples were not pretreated with DMC and solely decarbonated using HCl. Sample Powders (~100 g) were repeatedly reacted with 10% trace-element-grade HCl until no reaction was observed (typically repeat the same procedure 2 to 4 times). Insoluble residues remaining after decarbonation were washed with deionized water (3 to 6 times) until solution pH was neutral, filtered, dried, and wrapped in tin capsules and analyzed interlaboratory variability. Long-term reproducibility for δ 13 Corg is ≤ 0.3‰ based on repeated analysis of multiple internal standards and replicates.

Chemostratigraphy of four Cambrian sections in Nevada, USA
In the Groom Range section ( Supplementary Fig. S3 Fig. S1D).

Restudy of Oak Springs Summit section
Of particular note is the section for Oak Springs Summit presented by Faggetter et al. 1  However, given the lack of δ 13 Ccarb data in this portion of the section, such differences may be due to the limited δ 13 Ccarb data.

Absolute age estimate
According to Zhao et al. 3 , the 206 Pb/ 238 U age estimate of 509.1±0.62 Ma for the base of Miaolingian Series and Wuliuan Stage is based on the refined dating of ash beds from the Comley Sandstone of United Kingdom (Ref. 4 ). This date is older than the one we used for Laurentia (Ref. 5 ) (Fig. 2) based on zircons recovered from Cambrian sections in the Grand Canyon. As a result, the maximum date for the extinction of olenellids and the traditional "lower−middle" Cambrian boundary of Laurentia is younger than 508 Ma.
How much younger cannot be accurately ascertained. The next radiometric date above the 508 Ma are approximately 503 Ma from the Drumian Stage in Germany (Ref. 6 ), Tasmania (Ref. 7 ), and the United Kingdom (Ref. 4 ).