The analysis of variance for grain yield for a rainfed lowland rice variety trial conducted at Champassak, Laos in the wet season of 2004 is presented below:
|
Source of variation |
Df |
Mean square |
Expected mean square |
|
Replicates |
3 |
688426 |
|
|
Genotypes |
21 |
759729 |
rσ2G + σ2e |
|
Residual |
66 |
159692 |
σ2e |
1. Estimate σ2G and σ2e
2.Calculate the predicted SEM, SED, LSD.05 , and H for trials with 1-8 reps. The appropriate t-value is 1.997.
|
Number of replicates |
SEM |
SED |
LSD.05 |
Reduction in LSD due to adding last replication |
H |
|
1 |
|
|
|
|
|
|
2 |
|
|
|
|
|
|
3 |
|
|
|
|
|
|
4 |
|
|
|
|
|
|
5 |
|
|
|
|
|
|
6 |
|
|
|
|
|
3. If two 3-replicate variety trials like the one described above were conducted adjacent to each other in the same field and in the same season, what would be the expected correlation in variety means between the 2 trials?
4. What are the 2 possible ways to increase the precision and repeatability of a variety trial?