Multi-Sensor and Multi-temporal High-Throughput Phenotyping for Monitoring and Early Detection of Water-Limiting Stress in Soybean

Sarah E. Jones,Timilehin Ayanlade,Benjamin Fallen,Talukder Z. Jubery,Arti Singh,Baskar Ganapathysubramanian,Soumik Sarkar,Asheesh K. Singh

2024-02-29

Abstract:Soybean production is susceptible to biotic and abiotic stresses, exacerbated by extreme weather events. Water limiting stress, i.e. drought, emerges as a significant risk for soybean production, underscoring the need for advancements in stress monitoring for crop breeding and production. This project combines multi-modal information to identify the most effective and efficient automated methods to investigate drought response. We investigated a set of diverse soybean accessions using multiple sensors in a time series high-throughput phenotyping manner to: (1) develop a pipeline for rapid classification of soybean drought stress symptoms, and (2) investigate methods for early detection of drought stress. We utilized high-throughput time-series phenotyping using UAVs and sensors in conjunction with machine learning (ML) analytics, which offered a swift and efficient means of phenotyping. The red-edge and green bands were most effective to classify canopy wilting stress. The Red-Edge Chlorophyll Vegetation Index (RECI) successfully differentiated susceptible and tolerant soybean accessions prior to visual symptom development. We report pre-visual detection of soybean wilting using a combination of different vegetation indices. These results can contribute to early stress detection methodologies and rapid classification of drought responses in screening nurseries for breeding and production applications.

Machine Learning,Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The paper primarily focuses on the stress monitoring and early detection methods for soybeans under drought conditions. Specifically, the research aims to address the following key issues:

1. **Developing methods for rapid classification of soybean drought stress symptoms**: The study combines multi-sensor information (such as various sensors mounted on drones) and machine learning algorithms to achieve rapid classification of soybean drought stress symptoms. The goal is to establish an efficient and automated process that can quickly identify the performance of soybean plants under drought conditions.

2. **Exploring early detection methods for soybean drought stress**: In addition to classifying already apparent stress symptoms, the research also focuses on finding methods that can detect soybean drought stress before visual symptoms appear. This means that researchers hope to predict or identify which plants may face water limitation issues before they show obvious stress signs.

To achieve the above goals, the research team employed a series of technical means, including using drones to collect high-throughput phenotypic data, analyzing these data with machine learning methods, and attempting to extract features from data at multiple time points to achieve early warning functions. The study found that the red-edge band and green band are most effective for classifying canopy wilting stress, and the Red Edge Chlorophyll Vegetation Index (RECI) can distinguish between sensitive and drought-tolerant soybean varieties before visual symptoms appear. Additionally, the research demonstrated how combining different vegetation indices can improve classification accuracy and explored how to achieve early detection of stress at different growth stages. These results help improve drought stress screening methods in breeding and agricultural production.