GrowScreen-Rhizo 3 - automated large-scale high throughput greenhouse phenotyping of plant root and shoot development

Roots play a pivotal role for plant performance, but they are difficult to access, which hampers quantitative measurements. Repeated imaging of rhizotrons, flat growth containers with a transparent side, has proven suitable to assess dynamics of root traits in indoor experiments. However, measuring hundreds of soil-grown plants with high temporal resolution remains a laborious challenge. We introduce a novel whole-plant phenotyping platform with a capacity of almost 900 rhizotrons, which we named GrowScreen-Rhizo 3. This platform was designed to image shoots and roots of individual plants simultaneously and derive digital proxy traits for biomass and growth. In addition, built-in weighing and watering stations deliver water use data for each rhizotron. To achieve the desired throughput (image all 896 plants once a day) a high degree of automatization and standardization was required. We realized a modular plant-to-sensor solution, using a fleet of automated guided vehicles (AGVs) to transport large rhizotrons (80 × 40 × 5 cm) to four measurement chambers for daily imaging, weighing, and watering. Simultaneous imaging of the root system with a high-resolution camera (116 μm per px) and the shoot from six different viewing angles allows to monitor plant growth with high spatial and temporal accuracy. First, we verified that moving plants to the measurement chambers did not significantly affect above- or belowground plant growth. Next, we measured phenotypic variation in root and shoot traits of 24 barley genotypes, parents of a nested association mapping population. Our analysis revealed that heritability of root traits such as root system depth and seminal root length was moderate to high (r² = 0.52 and r² = 0.93, respectively), enabling further assessment of increasing numbers of recombinant genotypes. The results demonstrate the suitability of GrowScreen-Rhizo 3 to phenotype a range of plant species characterized by various growth habits, including crop, niche, and wild plant species. We conclude that GrowScreen-Rhizo 3 will contribute significantly to the development of phenotyping pipelines for the identification of candidate genotypes with improved resource use efficiency and to pre-breeding processes of climate-resilient crops.