Spectral-Based Classification of Genetically Differentiated Groups in Spring Wheat Grown under Contrasting...

Area of Science:

• Genetics and Breeding
• Plant Physiology
• Agricultural Science

Background:

• Global food security necessitates advancements in cereal crop genetics and breeding.
• Understanding wheat genetic structure is crucial for effective germplasm management and crop improvement.
• Previous studies assessed germplasm sustainability, but spectral-based genetic classification remains underexplored.

Purpose of the Study:

• To develop and evaluate a spectral-based classification approach for assigning wheat cultivars to genetically distinct subpopulations.
• To assess the accuracy of machine learning models in predicting genetic structure using foliar spectral data under varying water regimes.
• To determine the stability and reliability of spectral-based classification across different environmental conditions.

Main Methods:

• A panel of 316 spring bread wheat cultivars was analyzed.
• Foliar spectral data and genetic information were collected from cultivars grown in rainfed and fully irrigated environments.
• Machine learning models, including Convolutional Neural Network (CNN), multilayer perceptron, and partial least squares discriminant analysis, were trained and compared.

Main Results:

• Convolutional Neural Network (CNN) demonstrated superior accuracy (92-93%) in classifying wheat cultivars into subpopulations across both water regimes.
• Spectral differences between genetically differentiated groups were less pronounced under rainfed conditions, impacting clustering accuracy.
• CNN showed stable prediction performance irrespective of water availability, unlike other models.

Conclusions:

• Foliar spectral variation is a reliable indicator for inferring cultivar belonging to genetically differentiated groups.
• Spectral-based classification offers a stable and accurate method for crop genetic resources management, adaptable to diverse environments.
• This approach holds significant promise for enhancing the efficiency of wheat breeding programs and germplasm utilization.