A comprehensive multi-scale phenotyping framework for improving drought stress recognition, prediction and selection in WOSR

Winter oilseed rape suffered from severe crop losses in many regions of Germany and Europe in recent years as a result of repeated dry and hot spring and summer weather patterns. New, resilient cultivars are essential to overcome yield losses in the face of these challenges. However, targeted breeding to improve yield stability under drought stress is extremely difficult due to the low heritability of this trait complex and depends very much on high-quality recording of relevant target traits in large breeding populations. STRESS-STOPP pursues the hypothesis that an extremely accurate recording of previously ignored, drought-stress-associated physiological phenotypes, as well as related genotype-by-environment interactions in suitable breeding populations, using machine-learning methods for pattern recognition in performance-relevant phenotype, genotype and environmental data, can lead to increased selection accuracy for yield security of winter oilseed rape in drought-stress environments.
To address these limitations, STRESS-STOPP integrates high-resolution physiological phenotyping under controlled drought conditions using the DroughtSpotter XXL facility. In this system, oilseed rape plants are grown in large containers positioned on individual gravimetric scales, enabling automated weight measurements every five minutes. These data allow precise quantification of water use dynamics, transpiration patterns, and drought-induced physiological responses at the single-plant level. Complementary to this, all plants are repeatedly scanned using the PlantEye 3D multispectral scanner, generating dense point clouds from which vegetation indices, canopy structural parameters, and digital biomass traits can be extracted. The combination of gravimetric water-use data and 3D multispectral phenotypes provides a comprehensive physiological fingerprint of each genotype’s drought response.
A central goal is to link these high-resolution physiological traits with field performance. Therefore, phenotypic datasets from the DroughtSpotter XXL will be integrated with UAV-based remote sensing data collected in multi-environment field trials. By combining 3D PlantEye traits, temporal water-use profiles, and drone-derived vegetation indices, the project aims to establish a predictive framework that translates controlled-environment physiological phenotypes into field-level drought-stress prediction and yield performance. This integrative approach is expected to improve the reliability of genotype evaluation and enable the development of more drought-resilient winter oilseed rape cultivars.