High-throughput phenotyping for soybean plant maturity date and wheat plant height using unmanned aerial system

• Ano de defesa: 2020
• Autor(a) principal: Volpato, Leonardo
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Viçosa
• Palavras-chave em Português: Melhoramento genético; Glycine max; Triticum sp; Drone; Imagens fotográficas; Sensoriamento remoto; Melhoramento Vegetal
• Link de acesso: https://locus.ufv.br//handle/123456789/28715

Improvements in agronomical practices and plant breeding are paramount responses to the present and future challenges imposed by biotic stress and abiotic food production factors. On what concerns breeding currently, constraints in-field phenotyping capability limit our ability to dissect quantitative traits' genetics, especially those related to yield and environmental stress tolerance. Advances in phenotyping technology are critical to ensure that crops' genetic improvement meets future global demands for food and fuel. Progress in sensors, aeronautics, and high-performance computing combine with robust statistical analysis is paving this way. Throughout history, improving traits of interest depends on the ability to quantify phenotypes across genotypes replicated over multiple environments. Therefore, potentially valuable traits may have been neglected due to costly phenotyping and technological limitations. Field-based high throughput phenotyping (HTP) platforms will combine non-invasive remote sensing methods and automated environmental data collection. The current work aims to handle phenotypic data establishing UAS-based HTP platforms to improve data quality and provide a feasibility pipeline implementation applied to plant breeding programs. More specifically, chapter one deals with spring wheat (Triticum sp.), while chapter two, with the methodologies used in soybean breeding programs [Glycine max (L.) Merrill]. Plant height from two spring wheat breeding cycles was assessed using plot-level information from aerial imaging-derived 3D crop surface models at different phenological stages. From five contrasting environments, soybean maturity date with 53 experimental trials was collected by time-course images using RGB cameras equipped with low-cost drones. These studies showed that the phenotypic data obtained from a time-series imagery estimation via the UAS-based HTP platform could improve the collected data field's efficiency and quality and scalable to tens-of-thousands of plots into a modern plant breeding program. Keywords: Drone imagery. RGB camera. Data collect. Crop surface model. Plant maturity. Glycine max. Triticum sp.