Geobiology Seminar


Thursday, December 6, 2018, 11:00am


Haller Hall (Geological Museum 102)

Jessica Creveling
Assistant Professor, CEOAS
Oregon State University

A Library of Early Cambrian Chemostratigraphic Correlations from a Reproducible Algorithm

d13Ccarb chemostratigraphic correlation facilitates the interpolation of age models developed at radiometrically calibrated, carbonate-bearing stratigraphic sections to others lacking temporal constraints. For Neoproterozoic––Paleozoic applications, this practice commonly assigns ages to fossil first appearances, thereby constraining the tempo of early animal evolution. The robustness of the interpolated chronologies and the resulting evolutionary insight relies, in part, on the identification of a single, unique alignment solution between a time-calibrated and time-uncertain section, yet visual-based correlation methods cannot evaluate this premise. Here, I review the dynamic programming (‘slotting’) method for obtaining optimal alignments (in a least squares sense) between two d13Ccarb records subject to penalties on the insertion of hiatuses (that stretch and squeeze the time-axes) and the similarity of the duration of the two records. I employ dynamic programming to assess alignments between a radiometrically calibrated Cambrian Series 1–2 d13Ccarb reference curve (Maloof et al., 2010) and intrabasinal sections that lithostratigraphy predicts overlaps a subset of this reference curve, but for which auxiliary time-stratigraphic data is absent (Maloof et al., 2005). I present a library of d13Ccarb alignments that both support the statistical significance of published visual alignments between the d13Ccarb records (relative to the null hypothesis of no true correlation) and also yield additional, significant correlation solutions that accord with lithostratigraphic constraints. I repeat this exercise to report the uniqueness of interbasinal d13Ccarb correlations between Morocco, Mongolia, and Siberia (Kouchinsky et al., 2007; Smith et al., 2016). I conclude that each valid d13Ccarb alignment between a reference and time-uncertain section predicts a distinct age for a fossil first appearance at the latter; for this Cambrian Series 1–2 case study, this range of ages can exceed ~10 Myr. I advocate that the age uncertainty in a fossil datum arising solely from multiple valid chemostratigraphic alignments should propagate into error estimates for inferred tempos of animal evolution.