Quantifying the Diversity of Agronomic Traits of Cicer Milkvetch (Astragalus cicer L.) Germplasm Collections Using UAV-Based Imagery

Cicer milkvetch (Astragalus cicer L.) is a non-bloat perennial forage legume well suited for stockpile grazing because of its persistence, high late season quality and yield in the moist areas of the Canadian Prairies. Remote sensing via unmanned aerial vehicles (UAV) has the potential to accurately phenotype forage germplasm with less manual labour. The objectives of this study were: 1) Evaluate the diversity and relationship among 27 populations of cicer milkvetch using agro-morphological traits, 2) Differentiate a sub-set of cicer milkvetch populations from June to mid-October using UAV-based multispectral imaging, and 3) Identify UAV-based traits with significant correlations to the forage DMY and growth of cicer milkvetch. Near Clavet Saskatchewan, a completely randomized nursery of 27 cicer milkvetch populations was characterized nine times from June to mid-October in 2020 and 2021. Measurements included the agro-morphological traits maximum stem length, leaf number per stem, stem density, rhizome spread and plant area scores, and the UAV-based traits NDVI green area and NDVI canopy volume. The first forage harvest occurred in late June and the stockpile harvest in mid-October. Excluding leaf number per stem (first harvest) and plant area (stockpile harvest), significant differences (p<0.05) were identified between populations at both harvest times for the agro-morphological traits. Based on the agro-morphological traits at both harvest times, the first three principal components described 89% of the variation in the data. NDVI green area and NDVI canopy volume were able to differentiate (p<0.05) high and low vigour populations across the growing season. Among the agro-morphological traits, maximum stem length had the highest correlation with forage DMY at the first (r 0.74) and stockpile (0.83) harvests. NDVI green area had the highest correlation with forage DMY among the UAV-based traits at the first harvest (r 0.91) and the correlation with stockpiled forage DMY was improved when NDVI green area was recorded in mid-September (r 0.92). Across the growing season, NDVI green area accurately modelled changes in maximum stem length (r2 0.42-0.60). With a more user-friendly image analysis process, UAV measurements could be an accurate tool for the precision phenotyping of forage germplasm.

Name thesis:degree_name

Master of Science (M.Sc.)

Level thesis:degree_level

Masters

Discipline thesis:degree_discipline

Plant Sciences

Grantor

University of Saskatchewan

Year dc:date.issued

2023

Author dc:creator

• MacTaggart, David Robert

Advisor dc:contributor.advisor

• Biligetu, Bill

Committee members dc:contributor.committeemember

• Sharbel, Timothy F
• Beattie, Aaron D
• Lardner, Herbert (Bart) A
• Shirtliffe, Steve J
• Asselin, Sean R

• Genetic diversity, forage breeding, phenotyping, gene bank, multi-spectral imaging, extended grazing, drone, biomass, NDVI

Language dc:language.iso

en

Handle dc:identifier.uri

https://hdl.handle.net/10388/14703

OAI identifier oai:identifier

oai:harvest.usask.ca:10388/14703