Advanced search
Start date
Betweenand


High-throughput phenotyping during soybean plant development and seed production under water deficit

Full text
Author(s):
Carlos Henrique Queiroz Rego
Total Authors: 1
Document type: Doctoral Thesis
Press: Piracicaba.
Institution: Universidade de São Paulo (USP). Escola Superior de Agricultura Luiz de Queiroz (ESALA/BC)
Defense date:
Examining board members:
Silvio Moure Cicero; Denise Cunha Fernandes dos Santos Dias; Jose de Barros França Neto
Advisor: Silvio Moure Cicero
Abstract

The aim of this study was to observe the effect of water stress on plant development, production components and seed physiological quality, as well as to verify the effectiveness of using hyperspectral images in conjunction with chemometric techniques to assess phenotypic dissimilarity between soybean cultivars subjected to water deficit. To this end, two cultivars were used, which differ in their responses to water stress: BRASMAX COMPACTA IPRO (susceptible) and DM 66I68 IPRO (tolerant). The trial was carried out in pots and grown in a greenhouse, where the following water restriction conditions were applied: control (65 % of the substrate\'s water retention capacity - CRAS); moderate deficit (50 % of CRAS) in V2; severe deficit (30 % of CRAS) in V2; moderate deficit in R6; severe deficit in stage R6; moderate deficit in V2 and R6; severe deficit in V2 and R6. In the first experiment, the soybean plants from the different water restriction conditions were sampled at different phenological stages and the following analyses were carried out: dry matter mass; soluble sugar content; protein content; lipid content; evaluation of production components; analysis of water content and seed physiological quality. In the second experiment, plants (leaves and seeds) collected at different stages of development were used to acquire hyperspectral images using the Sisuchema sensor. The images were then subjected to exploratory data analysis and used in chemometric models, including Support Vector Machines and K-farthest neighbor. In summary, the results showed that water stress during the vegetative phase negatively affects the growth and development of the plants of the susceptible cultivar, as well as the yield. In contrast, the tolerant cultivar maintained its yield despite the adverse conditions. However, when water deficit occurred in both the vegetative and reproductive stages, plant development, production components and seed quality were affected, regardless of the susceptibility of the cultivars. In addition, seed quality was more impaired when water deficit occurred in the reproductive stage, also influencing the chemical composition of the seeds. On the other hand, hyperspectral images proved promising for plant phenotyping; the prediction models based on leaf data showed excellent classification results, allowing phenotypic dissimilarity to be assessed using hyperspectral sensors, while the models drawn up using data from seeds were not consistent in classifying cultivars. (AU)

FAPESP's process: 21/01266-0 - High performance phenotyping during plant development and soybean seed production under water deficiency
Grantee:Carlos Henrique Queiroz Rego
Support Opportunities: Scholarships in Brazil - Doctorate