Modelling spatial trends in sorghum breeding field trials using a two-dimensional P-spline mixed model

Area of Science:

• Agricultural Science
• Biometry
• Genetics

Background:

• Spatial variation in field trials significantly impacts genotype evaluation.
• Current spatial analysis methods involve complex, multi-step modeling processes.
• Efficient genotype selection requires accurate adjustment for field trends.

Purpose of the Study:

• To introduce and evaluate a novel, user-friendly spatial analysis method called SpATS.
• To compare the performance of SpATS against traditional spatial models in sorghum breeding trials.
• To demonstrate SpATS's ability to account for continuous field variation in a single modeling step.

Main Methods:

• Application of a novel spatial method using two-dimensional P-splines with anisotropic smoothing within a mixed model framework (SpATS).
• Analysis of large, partially replicated sorghum breeding trials.
• Comparison of SpATS model with standard spatial models employing autoregressive correlation of residuals.

Main Results:

• SpATS performed comparably to elaborate, trial-specific spatial models.
• The method achieved equivalent improvements in precision and genotypic value predictions.
• SpATS simultaneously modeled spatial trends and genetic effects in a single, flexible model.

Conclusions:

• SpATS is an efficient and easy-to-use alternative for routine plant breeding trial analyses.
• The method simplifies model selection and reduces parameter identification issues.
• SpATS provides a flexible approach to adequately adjust for field trends, enhancing genotype evaluation.