Bayesian Analysis of the Level of Resistance of Four Varieties of Maize to Downy Mildew Disease

Keywords: Credible interval, Gibbs sampler, Posterior distribution, Yield


This paper is based on Bayesian analysis of an experiment conducted by Kwara State Agricultural Development Project
to determine the level of resistance of four maize varieties to Downy mildew disease. The four varieties of maize namely
DMLSR-White, DMRESR-Yellow, 43DMR-SR-White and TZSR-White were planted in six different farms known to be
affected by the disease. Three variables were observed from each farm; the height of the plants, the weight of cobs, and
the weight of grains. The design corresponding to this experiment is a randomized complete block design, with the
varieties as the variable of interest while the farms serve as the blocks. Bayesian analysis was carried out on each of the
three variables, which involves estimating the values of parameters in the model from samples generated from the
posterior distribution of the parameters. Gibbs sampler was used to estimate the values of these parameters and
inference was made based on 95% equal tails credible interval. It was discovered that the four maize varieties are
equally resistant to the disease although the result revealed that the prevalent of the downy mildew disease is
significantly different among the farms.