Automatic morphology phenotyping of tetra- and hexaploid wheat spike using computer vision methods

Area of Science:

• Plant Science
• Genetics
• Bioinformatics

Background:

• Intraspecific classification of cultivated plants is crucial for biodiversity conservation and understanding plant origins.
• Modern wheat species (diploid, tetraploid, hexaploid) evolved from wild ancestors through genome doubling.
• Ploidy level identification is a key taxonomic step for wheat species.

Purpose of the Study:

• To investigate morphological spike characteristics of hexaploid and tetraploid wheat species using high-performance phenotyping.
• To identify quantifiable traits for differentiating wheat ploidy levels and species.
• To lay the groundwork for automated plant classification systems.

Main Methods:

• Phenotyping of 17 wheat species (595 plants, 3348 images) focusing on spike morphology.
• Quantitative analysis of nine spike traits (shape, size, awns area) using the WERecognizer program.
• Cluster analysis of spike traits to compare tetraploid and hexaploid species variability.

Main Results:

• Hexaploid wheat species exhibited greater trait variability compared to tetraploid species.
• Cluster analysis revealed two main groups based on spike morphology, largely separating hexaploids from tetraploids.
• Specific exceptions were noted, with some tetraploids clustering with hexaploids and vice versa.

Conclusions:

• Morphological spike characteristics, analyzed via computer imaging, provide discernible differences between hexaploid and tetraploid wheat species.
• These findings support the development of automated methods for wheat classification based on ploidy level and species.
• High-performance phenotyping offers a robust approach for taxonomic studies in cultivated plants.