Using Field-Based Aerial Imaging Platform to Improve Selection Efficiency in Breeding for Drought Tolerance in Maize.

Constraints in field phenotyping capability is currently a major limitation in conventional and molecular breeding. To overcome this limitation, the use of aerial phenotyping platforms is becoming an interesting option because they enable to simultaneously and cost-effectively collect data from large numbers of plots and upscale the measurements, from a single plot basis to dissecting an entire trial. In addition, dynamic traits can be monitored using time series. We present the use of an unmanned aerial platform equipped with sensor for multispectral imaging in spatial field variability assessment and phenotyping for drought stress tolerance in maize. Multispectral aerial images were used to characterize experimental fields for spatial soil variability and derive indices for crop performance under drought stress. Results showed that the aerial platform enables to effectively characterize spatial field variation that can be used to select appropriate experimental designs and improve repeatability. Selection index using the Normalized Difference Vegetation Index (NDVI) data derived from spectral imaging increased the predicted response of grain yield by 11 to 18% compared with direct selection for grain yield alone. This work suggests that the aerial sensing platform designed for phenotyping has the potential to effectively assist in maize breeding targeting stress environments where heritabilities for grain yield are often low (Bolafios and Edmeades, 1996).

Keywords: Maize, phenotyping platform, drought, remote sensing, selection efficiency