Unveiling the hidden: investigating the first animals of South America through phase-contrast SR-microCT and deep learning image processing

• Grant number: 20/11320-0
• Support Opportunities: Scholarships in Brazil - Post-Doctoral
• Start date: July 01, 2021
• End date: June 30, 2024
• Field of knowledge: Biological Sciences - Zoology - Paleozoology
• Principal Investigator: Miguel Angelo Stipp Basei
• Grantee: Bruno Becker Kerber
• Host Institution: Instituto de Geociências (IGC). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Associated research grant: 15/03737-0 - The Rio de La Plata Craton and the Western Gondwana Assembly, AP.TEM
• Associated scholarship(s): 22/06133-1 - X-ray diffraction tomography (XRD-CT) and pair distribution function tomography (PDF-CT): novel tools for paleontological investigations, BE.EP.PD

Abstract

The onset of animal life on Earth and its evolution is one of key biotic events that occurred at the end of the Neoproterozoic, mostly during the Ediacaran Period (635-539 Ma). During this time, a diverse and global biota of soft-bodied organisms of mostly unknown affinities developed in benthic ecosystems, setting set the stage for the major biotic transformations of the succeeding Cambrian radiation. Major evolutionary innovations occurred in the metazoan lineage during the Ediacaran, such as the origin of mobility, biomineralized skeletons and predatory activity. However, despite its importance, researchers still strive to understand the nature of the fossils suggested to mark this event. In its great part, the lack of a better comprehension is related to the taphonomical biases that affect this fossil record. Here, I propose to apply the zoom tomography, which will be available in the MOGNO beamline, to the high-resolution study of key Ediacaran fossils. Importantly, I will develop for the first time the use of deep learning (i.e. machine learning) for the automated segmentation of microtomographic data from fossils of the Ediacaran Period. This method will considerably increase the quantity of data able to be processed, consequently improving the quality and statistical significance of the final results. These neural networks also can be expanded and utilized to the study of fossils from other periods of the geological time scale. I will also investigate the fossils through other synchrotron-based techniques (Nano-XRF, XRD and XANES), identifying the possible biases present in these fossils and providing a clear picture of their original characteristics. Therefore, I selected fossils that hold the potential to answer major questions regarding the Ediacaran biota, such as the so-called 'Ediacaran-style' preservation, the first evidences of animal-substrate interaction, and the original morphology of the tubular fauna at the end of the Ediacaran. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

QUIROZ-VALLE, FRANCY ROXANA; LINO, LUCAS MARTINS; CITRONI, SERGIO BRANDOLISE; BASEI, MIGUEL ANGELO STIPP; HUECK, MATHIAS; VLACH, SILVIO ROBERTO FARIAS; BECKER-KERBER, BRUNO; PRADO, GUSTAVO MARCONDES EVANGELISTA MARTINS. Paleovolcanology, geochemistry, and zircon U-Pb-Hf isotopes of the volcano-sedimentary sequences from the Ediacaran Campo Alegre-Corupa Basin, southern Brazil: Linking volcanism and tectonics during Western Gondwana consolidation. GONDWANA RESEARCH, v. 115, p. 27-pg., 2022-12-20. (15/03737-0, 19/06838-2, 20/11320-0)