Phylogenetic Inference from Atomized 3D Morphometric Data: A Case Study using Kangaroos

Area of Science:

• Evolutionary Biology
• Morphometrics
• Phylogenetics

Background:

• Traditional phylogenetic reconstruction from morphological data is prone to investigator bias and analytical challenges.
• Geometric morphometrics offers objective tools, but direct landmark analysis has limitations for phylogenetic inference due to correlated variation and saltational evolution.
• Existing methods often struggle with non-independence of landmarks, multimodal variation, and missing data.

Purpose of the Study:

• To develop and apply an objective method for discretizing morphometric data for phylogenetic analysis.
• To address limitations of traditional morphological data in phylogenetics, including investigator bias and violated model assumptions.
• To improve the integration of phenotypic and genomic data for more accurate phylogenetic inference.

Main Methods:

• Developed an "atomized" character approach to discretize 3D surface scan data of Macropodiformes mandibles and postcranial elements.
• Partitioned landmarks into co-varying functional units and applied novel discretization techniques to continuous shape variation.
• Constructed discrete morphological character matrices using clustering algorithms, minimum evolution criteria, and tree sampling procedures.

Main Results:

• Phylogenetic analyses using the novel matrices generally recovered genera and deep-level macropodiform clades.
• Intergeneric relationships within the rapidly diversifying macropodine subfamily were not accurately reconstructed, mirroring issues with continuous data, traditional characters, and DNA data.
• The atomized characters showed promise for improving objectivity, accuracy, and clocklikeness in morphological phylogenetics.

Conclusions:

• The developed atomized character approach offers a promising pathway for more objective and accurate morphological phylogenetics.
• This method can accommodate correlated homoplasy and improve the estimation of morphological evolution rates.
• The approach facilitates better integration of phenotypic and genomic data for robust phylogenetic inference.