Canopy temperature depression sampling to assess grain yield and genotypic differentiation in winter wheat

Canopy temperature depression (CTD = air temperature [Ta] - canopy temperature [Tc]) has been used to estimate crop yield and to rank genotypes for tolerance to heat and drought, but when to measure CTD for breeding selection has seldom been addressed. Our objectives were to evaluate the suitability of CTD for the Texas High Plains environment and to determine optimal measurement times in relation to growth stage, time of day, and weather. Three years of CTD and weather data were used to assess regression models of grain yield in three wheat (Triticum aestivum L.) lines. Under dryland agriculture, long-term mean CTD at noon and yield were correlated in 2000 and 2001. The relation of short-term CTD readings to grain yield was highly variable. Poor correlation was associated with days of low solar irradiance, high wind speed, and rain events. Genotype effects on CTD were detected for all hours of day and night. Genotype x hour interaction was insignificant at night, suggesting that nighttime measurements may provide more stable conditions for CTD comparison among genotypes. In general, tree regression assessed grain yield from short-term CTD measurements better than linear regression and suggested that the best times to measure CTD were 0900, 1300, and 1800 h. Tree regression models provided a heuristic interpretation of crop water status under different scenarios of soil water availability.