LB 39-330 - Different tools for different trades: Contrasts in diversity and phylogenetic conservatism of secondary metabolites in fruit and leaves of Psychotria (Rubiaceae)

Animal-mediated seed dispersal represents a distinctive adaptive landscape for plant secondary metabolites, encompassing a variable combination of antagonistic and mutualistic interactions. Fleshy fruit tissue must remain attractive or at least palatable to seed dispersers, while at the same time defending against non-dispersing frugivores and seed predators. Fruit and seeds involved in such interactions can contain a diverse array of secondary metabolites, but distinguishing the compounds associated with fruit-centric interactions remains a challenge. Focusing on the species-rich pantropical genus Psychotria, we use a metabolomics approach to characterize the chemical diversity, tissue specificity, and phylogenetic relationships of fruit secondary metabolites in contrast to leaf secondary metabolites. We analyze leaf, pericarp, and seed tissue of 21 Psychotria species, using untargeted UPLC-MS2 metabolomic profiling and structural similarity-based molecular networking to quantify chemical diversity and classify compounds for phylogenetic comparisons. Molecular networking was conducted using the GNPS Feature-Based Molecular Networking workflow, with compound classification added using SIRIUS fragmentation tree computation and isotope pattern analyses. Chemical similarity was compared with phylogenetic similarity through hierarchical clustering, phylogenetic signal (Blomberg’s K) tests, and Mantel tests for each tissue type. Across Psychotria species, fruit-specific compound richness exceeded that of leaf tissue by 2.8-fold, though species-level compound richness of fruit vs. leaves was highly variable. Phylogenetic conservatism varied by compound class and by tissue type, with the most chemically diverse compound class, alkaloids, exhibiting K < 1 in leaves and pericarp and K >1 in seeds. Interspecific chemical similarity of leaves and seeds was correlated with phylogeny (Mantel tests: P< 0.05), while pericarp was uncorrelated with phylogeny. Thus, our study provides evidence for evolutionary divergence in fruit- and leaf metabolomes, while demonstrating the efficacy of metabolomics for such purposes.