Link of laminated crusts associated with pisolites to sea-level change: A case study from the Ladinian (Middle Triassic) Longtou Formation, Guizhou, China

The nature (depositional or diagenetic) of laminated crusts associated with pisolites is still in debate, which hampers the interpretation of sea-level change on carbonate platforms. The Ladinian Yangtze platform margin, formed by the Longtou Formation, contains abundant laminated crusts associated with pisolites, providing an opportunity to decode their formation in response to sea-level variations. There, individual laminae are formed of isopachous radiaxial fibrous calcite (RFC) crystals. The RFC displays strong undulose extinction and orange-red or dull red cathodoluminescence. The pisolites (vadoids) coexisting with the laminated crusts are characterized with inverse graded bedding, stalactitic cements, and red luminescence or non-luminescence. The laminated crusts and associated pisolites have delta O-18 values ranging from -10.56 parts per thousand to -3.83 parts per thousand and from -7.12 parts per thousand to -633 parts per thousand, respectively, which are more negative than that of the Ladinian normal seawater. They also share similar rare earth element and yttrium (REE + Y) distribution patterns, with enriched heavy rare earth element (HREE), enriched Y element, true negative Ce anomalies, and Y/Ho ratios between 34 and 51. Detailed petro-graphic and geochemical analyses indicate that the laminated crusts and associated vadoids formed in caves in response to alternating meteoric water and seawater, with the fluctuating water table resulting in their mixed deposition. The preservation of laminated crusts and vadoids in the Longtou Formation implies that they formed during a period of overall sea-level transgression, which was punctuated by sea-level regression and allowed formation of cave fillings. Our results suggest that laminated crusts present in ancient carbonate successions may indicate the positions of subaerial exposure related to sea-level regressions. (C) 2019 Elsevier B.V. All rights reserved.