Free Access
Table II
Comparison of candidate multi-environment data models for statistical analyses, for all analysed traits (survival subjected to angular transformation). In these variance-component models, site is a fixed effect. The predictive quality of the models was evaluated by means of their AIC (Akaike’s Information criterion) in the smaller-is-better form.
| AIC | ||||||
|---|---|---|---|---|---|---|
|
|
||||||
| No. of | Total | Stem | Survival | |||
| Model | Description | covariance | height | DBH | form | |
| parameters | ||||||
| Uniform across-site population × site | ||||||
| 1 | variance, site × replicate variance and | 3 | 17823.4 | 28037.9 | 1579.1 | –4527.2 |
| residual variance. | ||||||
| The same model as in 1, replacing site × | ||||||
| 2 | replicate by the nested effect site × | 3 | 16532.1 | 27175.8 | 1569.0 | –5737.9 |
| iBlock within replicate. | ||||||
| Heterogeneous across-site population | ||||||
| 3 | variances (5), iBlock within replicate | 15 | 16424.7 | 27042.5 | 1544.2 | –6269.0 |
| variances (5) and residual variances (5). | ||||||