Scientists Develop New Method for Remote Sensing-Based Assessment of Cotton Verticillium Wilt

Scientists from the Aerospace Information Research Institute (AIR) of the Chinese Academy of Sciences (CAS), in collaboration with Shihezi University, have developed a new method for grading cotton verticillium wilt (VW). This method correlates with yield loss and is suitable for remote sensing monitoring to assess VW severity. Their research has been recently published in Agricultural and Forest Meteorology.

Accurately grading the severity of cotton VW is critical for managing outbreaks, monitoring disease spread, identifying resistant crop varieties, and estimating yield loss. However, China's current national grading standard (GB/T 22101.5—2009, known as GB), which categorizes VW severity into five levels based on the percentage of diseased leaves per cotton plant, faces major challenges in real-world fieldwork and remote sensing applications.

The GB standard, based on the proportion of symptomatic leaves, is decoupled from actual yield loss, making it difficult to evaluate the true impact on crop production. Moreover, manually counting all leaves is time-consuming and cumbersome. Additionally, the "bottom-up" progression of VW symptoms leads to a mismatch with the "top-down" observation perspective of remote sensing, further reducing monitoring accuracy.

To address these issues, the research team conducted a detailed analysis of the disease progression in cotton plants. They found that fruit branch leaf (FL) exhibited a stronger correlation with cotton yield than main stem leaf (ML). The disease status of MLs in the top 3 layers (L1-L3) and FLs in the top 5 layers (L1-L5) effectively indicated yield loss.

Building on these insights, the team developed the Eight-Position Grading (EPG) method. Based on the GB grading principle and the natural spiral arrangement of cotton branches (known as the "3/8" distribution), EPG focuses only on the most critical layers—L1–L3 for MLs and L1–L5 for FLs—thus streamlining disease assessment while maintaining high accuracy.

Compared with traditional field surveys, the EPG method simplifies data collection by assessing only a critical subset of leaves in the top canopy. This not only reduces manual labor but also enables large-scale, efficient monitoring. EPG also demonstrates a stronger correlation with yield loss, capturing yield reduction with a 12% gradient, whereas GB shows a weaker association with actual outcomes.

Furthermore, theoretical simulations and UAV-based experiments have verified that EPG is better suited for large-scale remote sensing applications. EPG achieved significantly higher estimation accuracy for VW severity (R² = 0.76) compared to GB.

By aligning disease severity grading with remote sensing capabilities, the EPG method provides a practical, scientific approach for disease assessment. It offers an efficient tool for monitoring cotton VW severity and enables more accurate estimation of yield loss across large agricultural areas.

The spatial morphology of cotton plant and the target leaf distribution involved in EPG: (a) cotton plant; (b)branch distribution; (c) theoretical canopy; (d) MLs in L1-L3 and the FLs in L1-L5 of a single cotton plant; (e) MLs in L1-L3 and the FLs in L1-L5 of the cotton plants that intersect with the canopy of (d). (Image by AIR)