Abstract:
The data from a uniformity triai on banana was analysed for finding
the optimum plot size for conducting field experiments. The results showed
that single plant plots were most efficient. The co-efficient of variation of yield
decreased steadily with increasing plot size. The empirical law suggested by
Fairfield Smith gave a satisfactory fit to the data. The number of replications
required for providing estimates with 5% standard error decreased with increased
plot size, but the total experimental material required was minimum when single
plant plots were used. However, as banana plants are liable for disease incidence,
three plant plots are suggested for experiments.
