Table 2.
The predictive abilities for the three selection criterions; CV2, CV1 and CV0 analyzed to predict grain yields under stress (S) conditions using the calibration sets viz., stress dataset alone and Combined [All (stress and non-stress)] (CSN) datasets using seven models.
| Mixed models | Cross-validation scenarios | ||||||
|---|---|---|---|---|---|---|---|
| CV2 | CV1 | CV0 | |||||
| Calibration sets | CSN | S | CSN | S | CSN | S | NS → S |
| M1: E + A | 0.242 | 0.254 | 0.170 | 0.230 | 0.185 | 0.073 | 0.171 |
| M2: E + A + AE | 0.381 | 0.355 | 0.340 | 0.349 | 0.165 | 0.047 | 0.141 |
| M3: E + GCA | 0.254 | 0.281 | 0.197 | 0.256 | 0.157 | 0.060 | 0.140 |
| M4: E + GCA + SCA | 0.274 | 0.328 | 0.209 | 0.306 | 0.162 | 0.125 | 0.138 |
| M5:E + GCA + SCA + GCA × E + SCA × E | 0.374 | 0.364 | 0.345 | 0.359 | 0.146 | 0.108 | 0.122 |
| M6: E + GCA + SCA + GCA × E + A × E | 0.371 | 0.364 | 0.339 | 0.355 | 0.149 | 0.114 | 0.118 |
| M7: E + GCA + A + GCA × E + A × E | 0.388 | 0.362 | 0.344 | 0.355 | 0.176 | 0.073 | 0.149 |
Also, the extreme right column of the table contains the predicted grain yield under stress conditions using the non-stress (NS) dataset in CV0 scenario. The highlighted values represent models harboring highest predictive abilities in each of the cases.