Proximal Sensing: Experiences from Arabidopsis and Brassica Rapa.

Efforts at relating genotypes to phenotypes are currently impeded by the inability to rapidly collect large amounts of plant trait data.  This talk describes experiences with two approaches.  The first is based on creating detailed three-dimensional models via stereophotogrammetry.  Individual Arabidopsis plants from a nested association mapping panel were multiply imaged under chamber and field conditions.  An automated pipeline constructs the models and extracts information on leaf morphology.  A second technology involves highly wavelength resolved spectral radiometry.  A set of Brassica rapa recombinant inbred lines were subjected to different levels of crowding in the field.  Single-plant reflectance spectrograms were obtained in the range of 350-2500 nm.  To begin the process of indentifying genetically controlled processes that might have distinct reflectance correlates, both individual waveband data and a suite of commonly used remote sensing indices were analysed for significant line, treatment, and interaction effects as well as QTL mapping.