Geochemical fingerprinting of fossils with uncertain stratigraphic provenance: A case study from the Lower Jurassic Nishinakayama Formation (Yamaguchi, Japan)

Abstract

A fossil without provenance data is problematic because it cannot be placed into meaningful paleoecological and paleobiogeographic contexts. This problem is particularly acute when the fossil is suspected or known to have originated from a formation within which a major interval of biotic and/or paleoenvironmental change has been recognized that could change the paleobiological implications of the taxon in question. Two reptile fossil specimens (a testudinate and a crocodylomorph) preserved within ex situ mudstone cobbles of the Lower Jurassic (Pliensbachian-Toarcian) Nishinakayama Formation in Yamaguchi, Japan exemplify this problem. Both specimens are preserved alongside associated ammonoid fossils that biostratigraphically constrain them to the Toarcian interval of the Nishinakayama Formation, but it is unclear whether their stratigraphic provenance is below, within, or above the interval of the formation that preserves the chemostratigraphic markers of the Toarcian Ocean Anoxic Event (T-OAE, ∼182.5 ​Ma), which were first established at the nearby locality Sakuraguchidani. Herein, we used isotope ratio mass spectrometry and portable energy-dispersive X-ray fluorescence to investigate the geochemistry of sedimentary matrix sampled from each fossil and a new measured stratigraphic section of the Nishinakayama Formation close to where the fossil specimens were found. We interpret a ∼2‰ positive shift in δ¹³C_org at the base of the section as the recovery of the negative carbon isotope excursion associated with the T-OAE, providing additional evidence of the event from a new locality within the Nishinakayama Formation. Linear discriminant analysis (LDA) of the total geochemical dataset was then used to explore the multivariate separateness of binned intervals of the composite section and predict the provenance of each fossil. The results suggest with 93.33% confidence that both fossils were derived from strata above the T-OAE interval. This predictive method can be applied to any fossil collected ex situ with preserved rock matrix and for which the general provenance is known or suspected.