Unraveling plant phenotype to genotype associations with daily hyperspectral traits in Populus trichocarpa .

Abstract

ABSTRACT Hyperspectral remote sensing is a powerful, high-throughput phenotyping tool that quantifies physiologically and structurally relevant wavelengths across diverse genotypes and over varying temporal scales. In this study, we combined tower-based continuous hyperspectral sensing with genome-wide association studies to analyze 1,423 wavebands (400-900 nm) and derivative vegetation indices across 505 genotypes and the genetic architecture of hyperspectral phenotypes over time in Populus trichocarpa Torr. & Gray grown under field conditions. Wavelengths related to chlorophyll and carotenoid absorption spectra exhibited the strongest genetic variation resulting in 98 significant SNP associations. Notably, we found substantial overlap in genetic association between the blue and red spectral regions, indicative of carotenoids and chlorophyll, respectively, and identified more than 10 candidate genes associated with chloroplast function, underpinning photosynthetic activity. Furthermore, fluctuations in associations for vegetative indices, such as the chlorophyll:carotenoid index (CCI), across the growing season reveal a temporally dynamic genetic architecture of physiological traits associated with fall senescence of this temperate tree species. Finally, we also observed correlations (⍴=0.3, p 0.5, p<1x10 -16 ), reinforcing the value of hyperspectral measurements for predicting traits linked to tree productivity. These findings highlight the potential of high-throughput, rapid, hyperspectral genome wide association studies GWAS to uncover physiologically meaningful genetic variation and offer promising insights for future acceleration for plant breeding.

Plant phenotyping relevance

タワー型連続ハイパースペクトルセンシングを用いて多数の遺伝子型の生理・構造形質を時系列で取得し、表現型解析とGWASに substantively 適用しているため、フェノタイピング手法が中心的である。

Code and data availability

The paper's hyperspectral phenotype dataset (tower-based hyperspectral traits for 505 Populus trichocarpa genotypes) is explicitly stated to be publicly available through the Oak Ridge National Laboratory LabKey data portal with DOI 10.25983/CBI/3012775. This is a paper-specific, public, actionable phenotype dataset. A