Genotype × trait biplot and canonical correlations for spectral and agronomic traits in corn

Abstract

The current challenge of corn (Zea mays L.) crops is to reach high yield to supply the world's demand for food, especially under different fertilization regimes and using the latest technology. We hypothesized that wavelengths and vegetation indices have a linear relationship with agronomic variables in corn. Our objectives were to verify the formation of super-traits and to study the association between wavelengths and vegetation indices with agronomic traits in corn cultivated under high and low topdressing N. The experiments were carried out in two crop seasons using a randomized block design with three replicates in a factorial scheme (11 genotypes × 2 contrasting levels of N: low, 60 kg ha–1 and high, 180 kg ha–1). At full bloom, the spectral variables green (550 nm), red (660 nm), red-edge (735 nm), and near-infrared (790 nm) and the vegetation indices normalized difference vegetation index (NDVI), normalized difference red-edge index (NDRE), green normalized difference vegetation index (GNDVI), and soil-adjusted vegetation index (SAVI) were evaluated. Agronomic traits evaluated were leaf N, plant height, first pod height, stem diameter, cob length, rows per cob, grains per row, and grain yield. Results showed that SAVI, GNDVI, NDVI, and NDRE and red-edge and red were associated with agronomic traits in corn. The association between the agronomic traits evaluated here can be used to estimate the leaf N content and corn yield.