A guide to plant morphometrics using Gaussian Mixture Models

Abstract

Plant morphology is crucial in defining and circumscribing the plant diversity around us. Statistically speaking, the study of morphology is done using morphometry, that in the context of plant systematics is used to verify hypotheses of morphological independence between taxa. Nevertheless, methods currently used to analyse morphological data do not match with the conceptual model behind species circumscription on morphological grounds. Here we 1) provide a step-by-step guide to perform linear morphometric analyses in the context of plant systematics and 2) we develop a new conceptual, statistical, and probabilistic framework for analyzing morphometric data using Gaussian Mixture Models (GMMs) in plant taxonomy to compare alternative taxonomic hypotheses.