Strategies for selection and characterization of soybean lines with emphasis on rust tolerance

• Grant number: 17/24266-0
• Support Opportunities: Scholarships abroad - Research Internship - Doctorate
• Start date: March 01, 2018
• End date: August 31, 2018
• Field of knowledge: Agronomical Sciences - Agronomy
• Principal Investigator: Natal Antonio Vello
• Grantee: Elesandro Bornhofen
• Supervisor: Aaron Joel Lorenz
• Host Institution: Escola Superior de Agricultura Luiz de Queiroz (ESALQ). Universidade de São Paulo (USP). Piracicaba , SP, Brazil
• Institution abroad: University of Minnesota, St. Paul (U of M), United States
• Associated to the scholarship: 17/11235-0 - Strategies for selection and characterization of soybean lines with emphasis on rust tolerance, BP.DR

Abstract

The cost-efficiency for soybean production in Brazil has been challenged since the first occurrence of Asian Soybean Rust disease (ASR) in 2002 and billions of dollars are currently spent in its control. Hence, the aim of this research project is to select soybean elite lines tolerant to ASR from a recurrent selection program and understand genetic aspects behind the tolerance mechanism. The dataset is characterized by sixty-four F5 populations originated from an 8x8 partial diallel evaluated in two contrasting sites in 2015/16 crop season using a randomized complete block design. Phenotypic-based plant selection was performed to compose the 768 F5:6 inbred lines, which were tested in 2016/17 using augmented block design, with 12.5% of checks. In order to measure the level of tolerance of each genotype, we carried out each trial in duplicate, contrasting only for the ASR management (rust and rust-free conditions). Therefore, 3,286 experimental plots were measured for several traits during the two crop seasons. The statistical procedures to be implemented will be based on spatial analysis by modeling spatial trends using two-dimensional P-spline mixed model, fitting a linear mixed model to estimate variance components for multi-environment diallel and the use of multi-trait mixed model analysis combining correlated traits in order to increase the accuracy of predictions. In both crop seasons, the disease pressure was high and sufficient to differentiate the overall performance of adjacent trials and genotypes inside trials. Preliminary results show a difference of 18% in hundred seed weight and 36% in grain yield for F5 populations between treated and untreated trials, attesting the efficiency of natural infection. In relation to the lines, the differences were narrower proving the efficiency of selection. Also, high spatial variability was observed in the trials in augmented design. Our proposed topic addresses a significant problem and, therefore, will advance the state of knowledge in this field.