Mixed spatial models for data analysis of yield on large grapevine selection field trials.

Area of Science:

• Agricultural Science
• Biometrics
• Genetics

Background:

• Field trials are essential for selecting superior genotypes in crops like grapevines.
• Spatial variation in soil fertility and environmental factors can introduce spatial correlation, masking true genotypic differences.
• Classical statistical models may not adequately address spatial complexities in large field trials.

Purpose of the Study:

• To introduce and evaluate mixed spatial models for improving statistical data analysis in grapevine selection trials.
• To compare the efficiency of mixed spatial models against classical randomized complete block models.
• To assess the impact of spatial modeling on selection decisions and genetic gain prediction accuracy.

Main Methods:

• Utilized mixed spatial models to analyze yield data from grapevine field trials.
• Compared the performance of mixed spatial models with a classical randomized complete block model.
• Data comprised yield from three large experimental populations of Arinto, Aragonez (Tempranillo), and Viosinho grapevine clones.

Main Results:

• Mixed spatial models demonstrated a significantly better fit to the yield data compared to the classical approach.
• The application of mixed spatial models positively impacted selection decisions.
• Improved accuracy in predicting genetic gain was observed when using spatial mixed models.

Conclusions:

• Mixed spatial models offer a more precise statistical approach for analyzing grapevine field trial data.
• These models effectively mitigate the effects of spatial correlation, leading to more reliable identification of superior genotypes.
• Implementing mixed spatial models enhances the accuracy of genetic gain predictions, crucial for effective breeding programs.