Defining Spectral Radiometric Indices for the High-Throughput Remote Sensing Selection of Wheat Germplasm with Higher Leaf Epicuticular Wax Content for Drought and Heat Tolerance Breeding.

Wheat (Triticumm aestivum L.) is a major source of calories in many developing countries, and represents the main source of food for 2.4 billion poor people in the world. However, wheat grain yield has been adversely affected by biotic and abiotic stress that decrease wheat metabolic and reproductive efficiency. An adaptive protection to these factors is through accumulation of leaf and glume epicuticular waxes, associated with tolerance to drought and heat stress by increasing water-use-efficiency (WUE), decreasing transpiration rate, and canopy temperatures by reflecting excess radiation. The primary objective of this project was to develop a spectral index for epicuticular wax phenotyping. This index will serve as a new tool to breed for drought and heat tolerance, making it feasible to rapidly and non-destructively analyze wax in conjunction with other physiological traits throughout development, during early generations, and at larger numbers of trial locations. A population of 266 landraces collected in Iran and Mexico, and products of intraspecific hybridization with wild relatives developed by CIMMYT was used to conduct this study. Phenotyping under heat and drought stress was conducted at the Norman E. Borlaug Experimental Station (CENEB) in Ciudad Obregon, Mexico in 2013 and 2014. Following published methodology, canopy reflectance with and without light source were recorded as close to solar noon as possible at 9-10 days after pollination (DAP). Ten wavelength signatures per single plot from canopy and leaf were registered using a FieldSpec 4 Hi-Res Spectroradiometer, 350 to 2400 nm. Quantitative analysis of wax was conducted by colorimetric assays to twelve punches (1 cm diameter) collected from flag leaves. The most relevant single wavelengths were extracted, a derivative analysis was conducted, and wavelength differences were determined by applying the RELIEF-F algorithm procedure. Analysis of wavelength differences and assessment of ratios were assessed through IDL programming.