Efficiency of augmented p-rep designs in multi-environmental trials

Area of Science:

• Agricultural Science
• Biometrical Genetics
• Plant Breeding

Background:

• Augmented designs with unreplicated entries are common in early generation testing for plant breeding due to seed limitations.
• Partially replicated (p-rep) designs, which replace check plots with replicated entries, offer an alternative for multi-environmental trials (METs).
• A gap exists in comparing the efficiency of augmented p-rep designs against traditional augmented and replicated designs within METs.

Purpose of the Study:

• To compare the efficiency of augmented p-rep designs with augmented and replicated designs in multi-environmental trials (METs).
• To evaluate the impact of different error models and entry effect assumptions (fixed vs. random) on design efficiency.
• To investigate the influence of correlated entry effects, relevant to genomic selection, on design performance.

Main Methods:

• Simulated genetic effects and allocated them to plot yields in triticale and maize uniformity trials using four designs: unreplicated, augmented, replicated, and augmented p-rep.
• Varied the number of environments while keeping the number of entries and total plots constant.
• Incorporated simulations with different error models (spatial vs. randomization-based) and entry effect assumptions (fixed vs. random), including correlated effects for triticale data.

Main Results:

• Unreplicated and augmented p-rep designs demonstrated an advantage in efficiency.
• A preference for using random entry effects was observed, particularly when effects were correlated, reflecting relationships among entries.
• Spatial error models offered minor advantages over purely randomization-based models.

Conclusions:

• Unreplicated designs are the most efficient for multi-environmental trials (METs) when seed is limited.
• Augmented p-rep designs are more efficient than augmented and replicated designs when replication per environment is required.
• The use of random entry effects, especially with correlated effects, is recommended for improved efficiency in plant breeding METs.