Abstract
<jats:title>Summary</jats:title> <jats:p>(1) In plant ecology, functional traits are widely used to predict plant performance in ecosystems. Little is known on molecular traits and how they integrate with the classical trait spectrum, in particular in non-vascular land plants. We conceptualize Essential Molecular Variables (EMVs) using complex-thallose liverworts in biocrusts as reference.</jats:p> <jats:p>(2) We combined morphometry and bioimaging with liquid chromatography high- resolution mass-spectrometry (UPLC/ESI-QTOF-MS) with data-dependent acquisition of tandem mass-spectra (DDA-MS). To identify EMVs, data mining was performed including regression trees, random forest and redundancy analysis combined with manual filtering and functional annotation using literature and chemical libraries.</jats:p> <jats:p>(3) We found the molecular trait spectrum to be highly structured and largely orthogonal to classical traits. We identified six major categories of EMVs: molecule structure, carbon metabolism, amino acids and peptides, shikimates and phenylpropanoids, terpenoids and alkaloids that were functionally related to biotic interactions and bioclimatic properties.</jats:p> <jats:p>(4) We find EMVs to be indicators that characterize plant performance in ecosystems. Molecular traits and EMVs represent the mechanistic link to form and function and extend the functional trait concept. They provide mechanistic insights into how global changes destabilize liverworts in biocrusts and represent a new tool set by informing functional ecology on molecular mechanisms.</jats:p>