3-D shape analysis of ELONGATE DENDRITIC: A tool to refine the botanical attribution of cereal phytoliths?

  • Hermans, R. M. (Speaker)
  • Mao Li (Contributor)
  • William H. Brightly (Contributor)
  • Timothy Gallaher (Contributor)
  • Caroline Strömberg (Contributor)
  • Nys, K. (Contributor)
  • Snoeck, C. (Contributor)

Activiteit: Talk or presentation at a conference

Description

The Poaceae (grass) family produces diagnostic phytoliths, including ELONGATE DENDRITIC and ELONGATE DENTATE/DENDRITIC (hereafter, both groups are referred to as ELONGATE DENDRITIC). These morphotypes are commonly formed in the inflorescences of grasses, especially the domesticated cereals. In archaeological samples, their presence may be attributed to cereals as well as to wild grasses because they very often share similar morphologies. Some scholars have sought to devise methods for improving the taxonomic resolution of the morphotype, but there is as of yet not a reliable method.
In this project, we focus on a selection of ELONGATE DENDRITIC within Triticeae and Aveneae from both cereals such as wheat, rye, oats, and barley, and wild grasses with ELONGATE DENDRITIC that overlap in morphology with those of the cereals. The aim is to investigate the discriminative power of ELONGATE DENDRITIC to distinguish cereals versus wild grasses, as well as the taxonomic level at which these cereals can be differentiated from one another.
Confocal images of ELONGATE DENDRITIC phytoliths were converted into digital 3-D phytolith surface models that can be quantified and analyzed. Because more traditional techniques such as distance-based and landmark-based morphometrics have limitations when it comes to analyzing complex phytolith shapes lacking clear landmarks, we instead use a persistent homology approach. This method is rooted in the field of algebraic topology and shows potential to quantify these complex phytolith shapes. This presentation aims to explore persistent homology as a new tool in phytolith research and to present preliminary results from applying it to discriminating ELONGATE DENDRITIC using a 3-D phytolith model dataset.
Periode5 sep 2023
Evenementstitel13th International Meeting on Phytolith Research
EvenementstypeConference
LocatieIsrael
Mate van erkenningInternational