Insights into biomass accumulation and challenges in grain yield prediction of elite breeding materials using UAV‐based vegetation indices in soft red winter wheat

Summary

This study evaluates the utility of unmanned aerial vehicle-derived vegetation indices for predicting grain yield in soft red winter wheat breeding programmes. Although cumulative vegetation indices reliably estimated biomass accumulation and improved predictive ability when used as secondary traits in genomic prediction models, their genetic correlation with grain yield was inconsistent across environments. The authors caution wheat breeders against treating vegetation indices as direct proxies for grain yield, owing to confounding by micro-environmental variation and environment-specific prediction variability.

UK applicability

The findings are potentially relevant to UK wheat breeding programmes, as soft red winter wheat is grown in the UK and high-throughput phenotyping is increasingly adopted by UK research institutes and breeding companies. However, the study's environmental specificity suggests that validation in UK-specific growing conditions would be necessary before routine adoption of these UAV-based approaches for yield prediction.

Key measures

UAV-based vegetation indices, above-ground biomass (ground truth validation on 22 breeding lines), grain yield, multivariate genomic prediction models, random regression with Legendre polynomials

Outcomes reported

The study evaluated 596 soft red winter wheat genotypes across six environments using UAV-based vegetation indices to assess biomass accumulation and predict grain yield. Cumulative vegetation indices reliably captured biomass, but showed low and inconsistent genetic correlation with grain yield across environments and growth stages.

Topic tags