LB 26-222 - Functional phytochemical diversity increases throughout the growing season in a rare swamp tree species: Demonstrating the accessibility of modern untargeted metabolomics

Phytochemicals mediate many plant-organismal interactions, such as pollination and herbivory. Identifying and understanding changes in the concentration of specific phytochemicals can provide ecologists a clearer picture of the drivers of many of these interactions. However, analysis of most of these compounds is time-consuming and challenging, and requires analytical and technical expertise. It can take weeks to correctly identify a few dozen compounds, especially if the chemical ecologist is unfamiliar with the system. In the past decade, open-source and free software applications for analyzing large datasets of chemical information have advanced greatly in their availability and ease of use.Using a combination of these programs and methods, we constructed a pipeline for untargeted metabolomic analysis of phytochemical diversity throughout the growing season in a common garden of second-year Populus heterophylla trees. Chemical features were identified and aligned from LC-MS chromatograms using MZmine3 and associated spectra and quantification tables were submitted to GNPS for ion-identity-molecular-networking (IIMN) and compound library screening. Additionally, SIRIUS, a machine-learning software, was used to predict the structures of over 3500 compounds. GNPS-IIMN and SIRIUS results were combined and the R package chemodiv was used to generate a phytochemical dissimilarity matrix and assess phytochemical diversity in old and young leaves throughout the growing season.We found that Hill’s functional phytochemical diversity increased as the season progressed, suggesting P. heterophylla invests in a greater diversity of compounds, possibly to mediate an increased number of interactions later in the growing season. This increase in phytochemical diversity was explored using CANOPUS, a chemical identification algorithm that predicts biosynthetic pathways and compound classes based on structure in a systematic manner analogous to organismal taxonomy. For example, while phenolic glycosides (including anti-herbivory salicinoids) and quinic acid derivatives, two groups known to be abundant in Populus, decreased in August following a July peak, other organooxygen compounds, such as disaccharides, cyclohexanols, and glucuronosides accumulated from June to August. Each identified compound or class can be interpreted in a similar manner, utilizing library matches when available in addition to CANOPUS assignments to predict or assign metabolic or ecologic function.These results add to a growing body of research demonstrating that untargeted plant metabolomics has become more accessible. These methods provide chemical specificity unavailable with assays such as total phenolic content and we encourage ecologists to explore the use of such pipelines in their work to tease apart specific drivers of ecological interactions.