Paper No. 16
Presentation Time: 9:00 AM-6:00 PM

THE TRIBRACHIATUS LINEAGE: CONDUCTING SHAPE ANALYSIS ON NON-BIREFRINGENT NANNOFOSSILS


BORD, David, Department of Earth and Planetary Sciences, Rutgers University, 610 Taylor Road, Piscataway, NJ 08854-8066 and AUBRY, Marie-Pierre, Department of Geological Sciences, Rutgers University, Wright Labs, 610 Taylor Rd, Piscataway, NJ 08854-8066, dbord@eps.rutgers.edu

A morphometric analysis was undertaken to characterize the Early Eocene Tribrachiatus lineage, one of the few unquestionable lineages among coccolithophores (Haptophyta). The salient features of the lineage unfold at North Atlantic DSDP site 550 over an 8.8 m interval (from the upper range of T. bramlettei through the range of T. contortus, up to the lower range of T. orthostylus). The evolutionary pattern observed is best described as punctuated gradualism. Tribrachiatus coccoliths are single hexa– or triradiate crystals. The stem species, T. bramlettei, is constructed of two stacked equilateral triplets, with evenly distributed arms. The similar T. contortus has unevenly distributed arms separated by alternate ~70º and ~40º angles. Tribrachiatus orthostylus consists of a single triplet. The T. lineage involves a rotation of the two triplets so as to produce a triradiated species.

This morphometric study aims at determining the tempo and mode of evolution in the lineage. Whereas the data collection of specimens digitally photographed at multiple focal planes is relatively easy the automated analysis is hampered by the nature of the analyzed material itself. Measurements of the main morphologic characters (e.g., total diameter, angle between arms) are made on digital composite micrographs using a combination of Zeiss software (AxioVision) and open-source imaging software (Imagej). The composites themselves are produced using Photoshop’s stacking function. Beyond this, the study cannot take advantage of automated software for shape analysis because the coccoliths analyzed here, being non-birefringent and poorly delineated from adjacent particles in brightfield, make binary segmentation nearly impossible, which prevents fluid automated shape analysis. We have thus conducted the binary segmentation by hand in Photoshop and analyzed them with software from the Stony Brook morphometric website (e.g., tps software by James Rohlf). Although an extremely tedious procedure this allows digitizing landmarks and creating outlines for each specimen of Tribrachiatus. Whereas preparation techniques may be improved for analysis in brightfield, programs maybe developed for solving the difficulties faced here. A similar problem is encountered by SYRACO (Système de Reconnaissance Automatique de Coccolithes).