The late Cretaceous and early Paleogene periods witnessed a series of hyperthermal events marked by transient warming and negative carbon isotope excursions (CIEs). Two prominent events, the Late Maastrichtian Warming Event (LMWE) and the Early Danian Dan-C2 event, straddle the Cretaceous-Paleogene (K-Pg) boundary. Both events are linked to significant climatic and biotic changes; the LMWE is potentially associated with the late Cretaceous mass extinction, while the Dan-C2 event influenced post-extinction biotic recovery. The Dan-C2 event's global significance and classification as a true hyperthermal remain debated due to inconsistencies in marine records. The primary drivers of these CIEs are proposed as Deccan volcanism and carbon pools regulated by orbital forcing. The timing of both events coincides with maxima in the 405-kyr long-eccentricity cycle, and Dan-C2 CIEs correlate with 100-kyr short-eccentricity maxima, indicating orbital forcing's role. The concurrent Deccan volcanism suggests significant volcanic carbon emissions, particularly during the LMWE; however, the causal link and the exact contributions of volcanism remain uncertain due to conflicting model results on CO2 emissions from Deccan volcanism. Most existing carbon isotope data stem from marine sediments, highlighting the need for terrestrial records to fully understand these events. This study utilizes high-resolution terrestrial δ¹³Ccarb data from the Nanxiong Basin to investigate the terrestrial records of LMWE and Dan-C2 events, assess the global significance of Dan-C2, and quantify the relative contributions of Deccan volcanism and orbital forcing to carbon cycle perturbations during both events.

Previous research has extensively documented the Dan-C2 event in various locations, including the Atlantic and Tethys Oceans, and Eurasia. However, marine records display inconsistencies: the negative δ¹³C excursions are mainly observed in planktonic foraminifera and bulk samples, while benthic foraminifera often lack these signals. Similarly, warming evidence from oxygen isotopes is confined to surface waters in parts of the North Atlantic, with limited evidence of bottom water warming. These inconsistencies challenge the event's global significance and hyperthermal classification. Studies have explored the roles of Deccan volcanism and orbital forcing in driving these hyperthermal events. The extremely negative CIE values for both LMWE and Dan-C2 events align with maxima of the 405-kyr long-eccentricity cycle, suggesting orbital forcing's contribution. The temporal overlap with Deccan volcanism implicates substantial volcanic carbon emissions, especially during the LMWE. However, the models simulating CO2 release from Deccan volcanism provide varying results, leading to uncertainty regarding the exact role of volcanism in these events. The lack of terrestrial carbon isotope records necessitates further investigation to fully understand these hyperthermal events and the interplay between volcanic activity and orbital forcing.

This study focused on the Nanxiong Basin in southeastern China, known for its continuous red fluvial-lacustrine clastics spanning the Upper Cretaceous to the lower Paleocene. The study area, the CGD-CGY section, contains abundant fossils and paleosol layers with pedogenic carbonate, indicating a hot and (semi)arid paleoclimate during the K-Pg boundary. A total of 274 fresh samples were collected from the upper Zhenshui Formation to the lower Xiahui part of the Shanghu Formation at approximately 1-meter intervals. The chronology of the section is based on a paleomagnetic framework combined with a chronological control point at the K-Pg boundary. The carbon isotopic composition of bulk carbonate was determined using a Thermo Fisher Isotope ratio mass spectrometer coupled with a GasBench II. The standard deviation of δ¹³C values was calculated from replicate analyses of an internal laboratory calcite standard and was better than 0.15‰. The measured δ¹³C values are reported relative to those of the Vienna Pee Dee Belemnite (V-PDB). To analyze the influence of orbital forcing on the carbon cycle, evolutionary power spectral analysis was performed on the δ¹³C records from both the Nanxiong Basin and ODP site 1262 using Acycle software. The δ¹³C records were interpolated and detrended using local regression smoothing (LOWESS). Fast Fourier Transform (FFT) was employed, utilizing 500-kyr and 150-kyr sliding windows for the complete and discrete time windows, respectively. The data was deposited and made available online.

The δ¹³Ccarb values from the Nanxiong Basin show a consistent decrease from 66.4 to 66.2 Ma, reaching a minimum at approximately 66.2 Ma, with a negative CIE of ~1.5‰, corresponding to the LMWE. After the K-Pg boundary, δ¹³Ccarb decreases sharply, exhibiting several negative excursions, including double CIEs at ~65.8 and ~65.7 Ma representing the Dan-C2 event (first CIE ~3‰, larger than the second CIE ~2‰), and another excursion at ~65.3 Ma corresponding to the Lower C29n event. Comparison with δ¹³Cbulk values from ODP site 1262 shows muted CIEs during the LMWE in both terrestrial and marine records. The LMWE's duration (200-300 kyr) is significantly longer than the Dan-C2 CIEs (~100 kyr). However, the CIE magnitudes are greater in the Nanxiong Basin (1.5-3‰) than in ODP Site 1262 (0.5-1‰). Evolutionary power spectral analysis revealed significant 405-kyr long eccentricity cycles in both terrestrial and marine δ¹³C records from 66.6 to 65 Ma. The dramatic negative δ¹³C excursion after the K-Pg boundary obscured the 100-kyr short eccentricity cycle in both records within a 500-kyr sliding window. Analysis of discrete time windows without dramatic excursions showed significant 100-kyr cycles above and below the K-Pg boundary in the Nanxiong Basin, while these cycles were less consistent at ODP site 1262, even disappearing during the LMWE. The terrestrial record of the Dan-C2 event, combined with inconsistencies in marine records (limited negative δ¹³C excursions mainly in bulk samples and planktonic foraminifera, restricted distribution to the Atlantic and Tethys Oceans, and lack of bottom-water warming), raises questions about its global significance as a hyperthermal event. Comparison with the PETM suggests that carbonate dissolution and sedimentation rate influence CIE magnitudes, explaining the discrepancies between marine and terrestrial records. Post K-Pg boundary, spatial heterogeneity in primary/export productivity could contribute to the inconsistent marine δ¹³C records for the Dan-C2 event due to the mass extinction and ecosystem changes. Mercury (Hg) concentrations in the Nanxiong Basin are anomalous from 66.4 to 65.6 Ma, attributed to Deccan volcanism. Both LMWE and Dan-C2 events overlap with central Deccan volcanism and 405-kyr long eccentricity cycles, but differ in CIE magnitudes, warming patterns, and short-eccentricity cycle significance. Model simulations suggest that greater CO2 release occurred before the K-Pg boundary, supporting the hypothesis that Deccan volcanism played a larger role in the LMWE than in the Dan-C2 event.

The findings suggest that Deccan volcanism and orbital forcing played distinct roles in the carbon cycles of LMWE and Dan-C2 events. The muted and prolonged δ¹³C excursion during the LMWE, coupled with the absence of short-eccentricity cycles in the marine record, indicates that Deccan volcanism, potentially through passive degassing of early Deccan magmas, triggered the LMWE, amplified the climate's sensitivity to orbital forcing, and disturbed the carbon cycle. In contrast, the Dan-C2 event appears primarily driven by orbital forcing, with reduced volcanic influence. The inconsistencies in marine records of the Dan-C2 event are likely linked to drastic post-extinction ecosystem changes and spatial heterogeneity in ocean productivity, but further research is needed to fully elucidate the underlying mechanisms. The difference in the CIE magnitudes, warming patterns (surface vs. deep-sea warming), and the presence or absence of short eccentricity cycles between the two events support the conclusion that the dominant mechanisms differed, with Deccan volcanism playing a more significant role in the LMWE and orbital forcing dominating the Dan-C2 event.

This study presents a terrestrial δ¹³C record of the LMWE and Dan-C2 events from low-latitude East Asia. The terrestrial and marine records provide insights into the global distribution of these events, showing inconsistencies for the Dan-C2 event likely related to mass extinction-induced ecosystem changes and ocean heterogeneity. The analysis suggests that Deccan volcanism and orbital forcing played different roles in these events, with volcanism primarily influencing the LMWE and orbital forcing dominating the Dan-C2 event. Further research is needed to understand the complexities of the spatial heterogeneity in ocean productivity and its impact on carbon cycling after mass extinction events.

While this study provides valuable terrestrial data, limitations exist. The interpretation of the data relies on the accuracy of the chronological framework and the assumption that diagenetic effects on the δ¹³Ccarb data are minimal. Further studies with more detailed records are needed to improve our understanding of ocean heterogeneity and its role in shaping carbon cycle dynamics following the mass extinction.