agridat (version 1.23)

george.wheat: Multi-location/year breeding trial in California

Description

Multi-location/year breeding trial in California

Usage

data("george.wheat")

Format

A data frame with 13996 observations on the following 5 variables.

gen

genotype number

year

year

loc

location

block

block

yield

yield per plot

Details

This is a nice example of data from a breeding trial, in which some check genotypes are kepts during the whole experiment, while other genotypes enter and leave the breeding program. The data is highly unbalanced with respect to genotypes-by-environments.

Results of late-stage small-trials of 211 genotypes of wheat in California, conducted at 9 locations during the years 2004-2018.

Each trial was an RCB with 4 blocks.

The authors used this data to look at GGE biplots across years and concluded that repeatable genotype-by-location patterns were weak, and therefore the California cereal production region is a large, unstable, mega-environment.

Data downloaded 2019-10-29 from Dryad, https://doi.org/10.5061/dryad.bf8rt6b. Data are public domain.

References

None

Examples

if (FALSE) {
  library(agridat)
  libs(lattice, reshape2)
  data(george.wheat)
  dat <- george.wheat
  
  dat$env <- paste0(dat$year, ".", dat$loc)
  # average reps, cast to matrix
  mat <- reshape2::acast(dat, gen ~ env, value.var="yield", fun=mean, na.rm=TRUE)
  lattice::levelplot(mat, aspect="m", 
                     main="george.wheat", xlab="genotype", ylab="environment",
                     scales=list(x=list(cex=.3,rot=90),y=list(cex=.5)))

}