Ground and Aerial Remote Sensing for Peanut Drought Tolerance Phenotyping.

Phenotyping for drought tolerance has always been a challenge for physiologists and breeders in the unpredictable environment of the mid-Atlantic. Ground remote sensing has been available for decades but the long time required to screen hundreds of lines under constant weather change rendered phenotyping inefficient in this environment. Unmanned aerial systems (UAS) require much less time to screen large breeding populations and, thus, could be more effective phenotyping choices. Twenty three peanut genotypes, planted in replicated plots at the Tidewater Agricultural Research and Extension Center in Suffolk, VA, were half irrigated and half maintained dry from the beginning pegging (July 19) through full seed (Aug 29) growth stages. Plots maintained dry were covered with rain exclusion shelters to prevent rainfall. During this time, ground measurements were collected on canopy temperature depression (CTD), Normalized Difference Vegetation Index (NDVI), and RGB color space indices. Plant wilting was visually assessed. After shelters removal, canopy temperature (CT), NDVI, and RGB indices were measured with sensors available on a Falcon 8 UAS platform from 10 m above the plots. At maturity, pod yield, Sound Mature Kernel (SMK) content, and crop value were measured and compared with ground and aerial vegetation indices, and the wilting scores. Wilting was significantly correlated with all agronomic characteristics with coefficients of correlation ranging from -0.62 to -0.77; and it was well estimated by all remotely sensed vegetation indices, i.e., coefficients of correlation up to 0.91 depending on the date and time of the day. CTD was negatively associated with the agronomic characteristics starting two weeks after soil drying , i.e., hot canopies produced less yield, SMK, and value. Before that, positive relationships were observed, i.e., cooler canopies produced less yield, SMK and value. This suggests that water saving physiological traits may be responsible for peanut drought tolerance in the mid-Atlantic region. Greener canopies, evaluated either by NDVI and RGB indices, were higher yielding than less green plots. Agronomic characteristics of optimum irrigated and water stressed plots were equally well estimated with both, ground and aerial sensing technologies.