Table 2.
Advantages and disadvantages of common used high-throughput phenotyping techniques and their evaluated traits used in GWAS.
| Techniques | Traits used for GWAS | Advantages | Disadvantages |
|---|---|---|---|
| Visible light/RGB imaging | morphological traits (shape, color, size-related traits): (1) panicle traits: e.g. panicle length, rachis length, primary branch number; (2) leaf traits: e.g. green leaf area; (3) tassel traits: e.g. tassel weight, tassel length, spike length, branch number; (4) root traits: e.g. total root length, total surface area, convex hull area, adjust depth; (5) canopy traits: canopy coverage, biomass, radiation interception efficiency, radiation use efficiency; (6) cell traits: cell column number and width; (7) seed traits: e.g. germination rate at certain time, volume increase, mean germination time. (8) others: tiller number, projected shoot area, relative growth rate, transpiration rate, transpiration use efficiency, plant compactness, digital biomass, plant height, etc. |
low equipment expense, suitable for wide applications |
only allow appearance information acquisition; highly depend on image processing algorithms, |
| X-ray computed tomography | (1) tiller traits: tiller number, size, and shape related parameters, tiller angle; (2) tiller growth traits: absolute growth rate of total tiller area, relative growth rate of total tiller area, absolute growth rate of tiller number, relative growth rate of tiller number. (3) stem vascular bundles traits (micro-CT) |
sensitive | high expense |
| Visible and near-infrared spectroscopy | (1) spectral indices for crop and canopy: e.g. reflectance at specific wavelengths, normalized differential spectral index, differential spectral index, simple ratio index, the modified canopy adjusted ratio index 2, the MERIS terrestrial chlorophyll index, the normalized difference vegetation index, photochemical reflectance index, canopy spectral reflectance, the derived hyperspectral indices; (2) biochemical parameters: e.g. total phenol, proanthocyanidin, and 3-deoxyanthocyanidin concentrations in grain, glucosinolate content in seed. | biochemical component content can be estimated by modeling | point measurement (cannot be represent spatial information), background interference |
| Multispectral/hyperspectral imaging | (1) hyperspectral indices: e.g. total reflectance, average reflectance, the derived hyperspectral indices; (2) lodging traits: e.g. visual scores of lodging intensity, severity and lodging index. |
rich spatial and spectral information acquisition | complex data/image processing |
| Chlorophyll fluorescence | chlorophyll fluorescence induction parameters (e.g. Fv/Fm) | high sensitive to plant physiological changes | point measurement |
| Fluorescence imaging | biovolume estimations | high sensitivity | sensitive to interference, small field of view |
| Nuclear magnetic resonance | seed oil content | high resolution | high cost |