Table 10. Estimated prediction accuracy (correlation between predicted and adjusted grain yield, averaged over 50 TRN-TST partitions, cycle 2011-2012 (W2), CV2.
Models/Environmentsa | Correlation (SE) | Change%b | Numberc | ||
---|---|---|---|---|---|
Single environment | |||||
0iFN | 0.473(0.052) | − | − | ||
5iBH | 0.414(0.063) | − | − | ||
5iBN | 0.510(0.052) | − | − | ||
5iFN | 0.448(0.054) | − | − | ||
Pairs of environments | Interaction Model | Across-Env Model | |||
0iFN | Cor = 0.34 | 0.512(0.042) | 0.436(0.052) | 8.3%; 17.5% | 50; 47 |
5iBH | 0.451(0.056) | 0.386(0.054) | 8.9%; 16.8% | 50; 48 | |
0iFN | Cor=-0.05 | 0.467(0.043) | 0.271(0.060) | −1.2%; 72.2% | 13; 50 |
5iBN | 0.502(0.048) | 0.279(0.056) | −1.6%; 80.0% | 6; 50 | |
0iFN | Cor = 0.31 | 0.520(0.043) | 0.406(0.056) | 9.8%; 27.9% | 50; 49 |
5iFN | 0.502(0.050) | 0.412(0.060) | 12.0%; 21.9% | 50; 49 | |
5iBH | Cor = 0.33 | 0.477(0.053) | 0.426(0.061) | 15.2%; 11.8% | 48; 47 |
5iBN | 0.546(0.047) | 0.483(0.051) | 7.2%; 13.2% | 47; 49 | |
5iBH | Cor = 0.41 | 0.500(0.049) | 0.465(0.050) | 20.9%; 7.7% | 50; 45 |
5iFN | 0.520(0.051) | 0.490(0.055) | 16.0%; 6.1% | 50; 40 | |
5iBN | Cor = 0.41 | 0.558(0.050) | 0.541(0.052) | 9.6%; 3.2% | 49; 43 |
5iFN | 0.501(0.052) | 0.490(0.055) | 11.8%; 2.2% | 50; 37 | |
All environments | |||||
0iFN | 0.513(0.044) | 0.298(0.051) | 8.5%; 72.6% | 45; 50 | |
5iBH | 0.536(0.050) | 0.481(0.047) | 29.6%; 11.5% | 50; 49 | |
5iBN | 0.531(0.044) | 0.401(0.044) | 4.1%; 32.4% | 37; 50 | |
5iFN | 0.561(0.046) | 0.523(0.049) | 25.3%; 7.4% | 50; 43 |
TRN-TST, training-testing.
Environments are described by a sequence of codes: 0i, 2i, and 5i denote the number of irrigation cycles; B/F denotes whether the planting system was “bed” (B) or “flat” (F); N/H denotes whether planting date was normal (N) or late (H, simulating heat); Z indicates no tillage.
Change in prediction accuracy of the M×E model relative to the prediction accuracy of the single-environment (before semicolon) and relative to the prediction accuracy of the across-environment model (after semicolon).
Number of partitions (of 50) for which the M×E model gave greater accuracy than the single-environment (before semicolon) and the across-environment model (after semicolon).