Objectives
- Evaluate elite breeding lines, candidate lines, and check cultivars under shorter season environments for adaptation traits.
- Evaluate elite breeding lines, candidate lines, and check cultivars in multi-location tests for stability of yield and maturity.
Project Description
We have adapted the linear mixed model described in previous reports to include a hierarchical nested structure that considers the relatedness of blocks in the RCBD design from a shared location. We have updated the mixed model to analyze multiple years of data and have adapted the model to assess the variance of individual location-year combinations (i.e. ‘environments’). Using this modified linear mixed model, we are developing a user-friendly pipeline in the R environment, using ASReml-R, and have used it to analyze the 2018 and 2019 Linseed Coop and SVPG datasets. The pipeline is being developed mainly to model the complex relationship between genotypic yield performances across varying environments, but a less complex model can also be used for traits with less variation, such as determinate habit and stem dry down, and in cases where only a small amount of data are available.
We determined that genotype, environment (the unique combinations of locations and years), and the genotype × environment interaction significantly affect the response of five important agronomic traits (yield, height, maturity, determinate habit, and stem dry down). We identified entries with yield ranks that vary between zones as well as entries with stable high yield ranks across zones. Importantly, we identified entries with superior performance in Zone 3, which includes the shorter growing season locations. We assessed the heritability (a measure of the accuracy of phenotypic selection for a crop characteristic within a population) for yield, maturity, height, and northern adaptation traits, with determinate habit being the least (41.9%) and height being the most (81.1%) heritable trait; yield had a heritability of 64.5%. We tested the correlation between all traits using the combined data from all zones and found several highly significant associations. We looked at the correlation between traits within zones and found them to be zone-specific in several instances. Taken together, the results of this project will help us to identify and characterize candidate lines with superior performance in the northern grain belt of the Canadian prairies.
Grower Benefits
- We developed a statistical pipeline that uses a linear mixed model to improve estimation of cultivar agronomic trait performance.
- We have used the linear mixed model to analyze multi-environment data from the 2018 and 2019 Linseed Coop and SVPG trial and have assessed the performance of yield, height, maturity, determinate habit, and stem dry down.
- Using the mixed model, we have identified cultivars with superior yield performance in the northern growing zone compared to the southern zone.
- We are continuing to develop and expand our statistical pipeline to address a variety of statistical analyses, and to improve usability so that the mixed model can be easily implemented by users with diverse statistical backgrounds.
