A NEW HIGH-RESOLUTION NEOGENE CHRONOSTRATIGRAPHIC MODEL FOR THE SOUTHERN OCEAN: APPLICATION OF THE CONSTRAINED OPTIMIZATION METHOD TO DIATOM BIOSTRATIGRAPHY

Developing a coherent Neogene history of the Antarctic ice sheet requires a high-resolution biostratigraphic age model, applicable to the variety of existing onshore and offshore paleoenvironmental records. Constrained Optimization (CONOP), a computer-assisted, multidimensional version of graphic correlation, allows incorporation of comprehensive biostratigraphic, magnetostratigraphic, and tephrostratigraphic datasets, and eliminates much of the subjectivity inherent in traditional zonal biostratigraphic methods. Furthermore, CONOP provides quantitative results, including detailed analyses of all adjustments made to observed sequences, and confidence intervals on all age assignments. This study employs thorough testing of CONOP model settings and input data to establish sensitivity of the solution to these factors, cull unreliable data and produce a robust, precise and reproducible age model.

The new quantitative biostratigraphic model incorporates records of 78 diatom taxa, 29 paleomagnetic reversals, and two radiometrically dated ash layers from 27 Plio-Pleistocene sections in the Southern Ocean and around the Antarctic continental margin. A CONOP analysis of this dataset yields precise ages for biostratigraphic events as old as 8 Ma, bracketing event ages to within an average of 0.2 Ma. The model provides up to an order of magnitude greater temporal resolution than traditional zonal biostratigraphic models, identifies many new potentially useful data, and calls into question the reliability and isochroneity of some commonly utilized zonal marker events. The majority of age assignments for current zonal events are within the range of published age estimates; others differ from previous estimates by 2 m.y. Last appearance data for zonal taxa are generally assigned younger ages by the new model, indicating that traditional biostratigraphic methods identify average taxon ranges rather than total taxon ranges. Results of this phase of study will be invaluable for dating new core material in the future, and the planned incorporation of additional sections, fossil groups, and other proxy records will further improve the resolution of the model.