Dominick Noeker
Research Mentor(s): Theodore Matel
Mentor Department: Earth and Environmental Science
Authors: Dominick Noeker, Theodore Mattel
Session: Session 6 (3:00pm – 3:50pm)
Presentation Type: Poster 72
Abstract
The Lauraceae family encompasses between 2500 – 3500 species, is distributed mostly in tropical regions with a few species occurring in temperate regions, and forms a major fraction of woody biomass and species diversity in the Neotropics. Neotropical Lauraceae are notoriously difficult to identify to species but exhibit recognizable features at the family level such as leaves with a ranalean odor and anthers with 2-4 flap-like valves. Previous systematic studies and dichotomous keys have mainly utilized floral characteristics as a means to identify lower level taxa. Nonetheless, botanists and paleobotanists conducting forest inventories or studying species diversity frequently collect foliage that contains sufficient variation to differentiate species even without knowledge of higher taxonomic affinities. The aim of this study is to systematically describe the variation in leaf morphology among a set of Neotropical Lauraceae species in order to (1) quantify variation in architectural traits among species with known taxonomy, and (2) test for combinations of characteristics that diagnose higher taxa. Here, we describe the variation in leaf architecture among 90 species of the family Lauraceae co-occurring in a 50-hectare mapped forest plot in Colombian Amazonia and evaluate this variation in respect to their recognized species boundaries using 28 vein characteristics outlined in the Manual of Leaf Architecture. We show that the clearest features of a Lauraceae vein network include a pinnate primary vein framework with ascending secondary veins and percurrent tertiary veins, and this suite of characters occurs in all species examined except for some members of the genus Caryodaphnopsis which produce three basal veins. Most species cannot be diagnosed from a change in a single characteristic, rather a combination of multiple. For the most part, Lauraceae leaf venation displays a looping behavior in veins of all but those of the highest degree. Going forward, the data generated for this study will be used as training data for a convolutional neural network to score leaves for the character states listed in the Manual of Leaf Architecture and applied to better understand the present and historical composition of Colombian rainforests.
