Research Scientist Agriculture and Agri-Food Canada
Disclosure: Disclosure information not submitted.
In field plot experiments, replication provides an estimate of error in ANOVA and is used to separate differences among treatments. The effectiveness of a replicated design is limited by field homogeneity to a square or rectangle of about 50 varieties replicated three or four times. Phenomic selection can screen hundreds, or even thousands, of experimental varieties in one experiment. The daily duration of phenomic measurements increases with replication, resulting in temporal differences in canopy traits. The Modified Augmented Design (MAD), developed by Lin and Poushinsky (1983), uses main plot checks arranged in a grid to measure field heterogeneity; and sub-plot checks to estimate efficiency of the data correction method. Each test variety is repeated once reducing the total number of plots. An RCBD trial and a MAD trial were compared for phenomic traits and yield. Twenty-one varieties were planted in an RCBD with 4 replications (84 plots) and in a MAD design with 1 main check and 2 sub-checks (40 plots). The PlotCam was used to capture RGB images on a weekly basis to calculate vegetative cover (VC). After analysis and data adjustment, yield and seed weight from the MAD design correlated with the RCBD, while the VC between the two designs were only similar during senescence. Further work is needed to validate the use of the MAD design in phenomic trials.