A mathematical landmark-based method for measuring worn molars in hominoid systematics

Area of Science:

• Paleontology
• Primatology
• Anthropology

Background:

• Worn teeth in hominoid fossils lack clear anatomical landmarks, hindering systematic studies.
• Discarding worn teeth reduces sample sizes and skews data for fossil and extant hominoid research.
• This issue is critical for low-sample groups, imbalanced datasets, and key fossil specimens.

Purpose of the Study:

• To propose a novel methodology for measuring molar size and shape using mathematically-derived landmarks.
• To enable the inclusion of worn teeth in hominoid systematic studies, irrespective of wear.
• To improve accuracy and sample size in the analysis of hominoid evolution.

Main Methods:

• Utilized 110 lower second molar specimens from five extant hominoid species (chimpanzees, bonobos, gorillas, and humans).
• Applied a methodology based predominantly on mathematically-derived landmarks for molar measurement.
• Performed principal components analysis (PCA) and discriminant function analysis (DFA) for classification.

Main Results:

• Achieved good species separation in morphospace using PCA.
• Attained high classification accuracy: 96.4% at species level and 88.2% at subspecies level via DFA.
• Demonstrated comparable or superior accuracy to traditional anatomical landmark methods, even on cross-validation.

Conclusions:

• Mathematical landmarking is a rapid and uncomplicated method for analyzing worn molar teeth.
• This approach effectively maximizes sample sizes and balances datasets in hominoid systematic studies.
• Supports the broader inclusion of worn dental specimens in evolutionary and systematic research.