Deciphering the genetic basis of yield components in wheat by integrating hyperspectral-based phenomes.

Genome-wide association studies (GWAS) have advanced crop genetics by the detection of loci controlling complex traits; however, their power is often constrained by the quality and the throughput of phenotypic data. In this study, we integrated hyperspectral and genomics data to investigate the genetic architecture of spectral signatures associated with yield components in wheat. A diverse panel of 341 soft wheat lines was evaluated over three years, and hyperspectral data were collected using a UAV-mounted sensor. Among 273 spectral bands, those most strongly correlated with grain yield (GY), thousand-grain weight (TGW), and grains per unit area (GN) were selected. Principal component analysis was used for dimensionality reduction, and the first principal component (PC1), here defined as the hyperspectral phenome, accounted for 78.9%-97.1% of overall variance. The GWAS using both manual phenotypes and hyperspectral phenomes identified 31 significant marker-trait associations (MTAs), including several pleiotropic loci shared across traits and data types. A notable SNP on chromosome 1A, associated with all three hyperspectral phenomes, was located within a gene specifying a chlorophyll a-b binding protein, a key component of photosynthesis and stress response. Additional MTAs were linked to genes involved in cytochrome P450 metabolism and LRR proteins, highlighting their roles in yield and environmental response. Overall, this study shows that hyperspectral imaging serves as a valuable, high-throughput secondary correlated trait for uncovering novel loci and dissecting the genetic basis of complex yield traits in wheat.