High-throughput chlorophyll fluorescence image-based phenotyping for water deficit stress tolerance in wheat

As global populations increase and economies expand, the demand for freshwater is surging, exacerbated by the effects of climate change and shifting lifestyles. It is resulting in widespread water stress and straining food production systems, a challenge anticipated to intensify in the coming decades. One potential solution to mitigate the impact of water scarcity, particularly in water-deficient regions, is the cultivation of water deficit stress-tolerant crop varieties. This study explores the simultaneous assessment of photosynthetic machinery and plant growth responses using chlorophyll fluorescence (ChlF) image based high-throughput phenotyping (HTP) for water deficit stress tolerance on 184 RILs in a controlled environment phenotyping facility. Under stress, recombinant inbred lines (RILs) displayed a diminished variable to maximum fluorescence ratio (Fv/Fm) compared to the control. However, stress-tolerant lines maintained higher Fv/Fm ratio and projected Fv/Fm area, mitigating water stress-induced yield losses. Machine learning using K-Nearest Neighbor, Support Vector Classifier and Random Forest, classified wheat RILs intro stress tolerance classes using sensor derived parameters with high accuracy of 0.56, 0.58 and 0.60 respectively. This study demonstrates the full potential of ChlF image-based phenotyping for enhanced throughput, identifying stress tolerance RILs as well as sensor derived traits, and novel indices. It emphasizes the importance of utilizing innovative data analytics techniques like PCA, clustering and machine learning to alleviate the data analysis bottleneck of HTP, for accelerating the pace of crop improvement for stress tolerance and sustainable food production.