Evaluation of yield and drought using active and passive spectral sensing systems at the reproductive stage in wheat

Active and passive sensors are available for ground-based, high-throughput phenotyping in the field. However, these sensor systems have seldom been compared with respect to their determination of plant water status and water use efficiency related parameters under drought conditions. In this study, five passive and active reflectance sensors, including a hyperspectral passive sensor, an active flash sensor (AFS), the Crop Circle, and the GreenSeeker, were evaluated to assess drought-related destructive and non-destructive morphophysiological parameters (ground cover, relative leaf water content, leaf temperature, and carbon isotope discrimination of leaves and grain) and grain yield of twenty wheat (Triticum aestivum L.) cultivars. Measurements were conducted in a 2-year study, including a drought stress and a control environment under field conditions. A comparison of the active sensors at the heading, anthesis and grainfilling stages indicated that the Crop Circle provided the most significant and robust relationships with drought-related parameters (relative leaf water content and leaf and grain carbon isotope discrimination). In comparison with the passive sensor, the five water and normalized water indices (WI and NWI-1 to 4), which are only provided by the passive sensor, showed the strongest relationships with the drought stressrelated parameters (r = -0.49 to -0.86) and grain yield (r = -0.88) at anthesis. This paper indicates that precision phenotyping allows the integration of water indices in breeding programs to rapidly and cost-effectively identify drought-tolerant genotypes. This is supported by the fact that grain yield and the water indices showed the same heritability under drought conditions.