Abstract

Vegetative indices (VIs) collected from an unoccupied aerial vehicle (UAV) equipped with a multi-spectral camera can be used to study growth and development of alfalfa throughout each growth cycle. Random regression models are well suited to fit longitudinal phenotypes such as VIs collected over time to estimate growth curves using covariance functions. Using these functions genetic variation in growth through time can be estimated and the relationships between VIs and end-use traits, like forage yield and quality, can be assessed. The main objectives of this project are (1) to incorporate aerial high-throughput phenotyping to predict performance and genetic merit of the breeding materials, (2) to fit longitudinal random regression models to estimate genotype-specific growth curves and estimate the heritability of key growth parameters. Univariate and multivariate models were used to estimate heritability of image features for alfalfa trial of Helfer, 2020 and 2021. The heritability of different image features in alfalfa ranged from 0-0.78. The preliminary results showed the strongest correlation for Green NDVI and biomass yield (0.4053, 0.7875, and 0.6779), followed by Red edge NDVI and biomass yield (0.417, 0.7898, and 0.6417) for the first, second and third cuttings respectively of the experimental trial located at Helfer, Ithaca for 2020, while the genetic correlations for 2021 were strongest for Red edge NDVI and biomass yield (0.76, 0.74, and 0.66) followed by Green NDVI and biomass yield (0.75, 0.76 and 0.60) for the first, second and third cuttings. The potential of random regression models was investigated using Legendre polynomial functions. Random regression model converged for most of the time points and showed potential for modeling genetic parameters associated with growth and development.