Table 4.
Selection of models according to the Akaike information criterion (AIC) for investigating the effect of seed provenance and site of transplant on seedling and sapling adaptive traits assessed in reciprocal transplants with Euterpe edulis in different forest types of the Brazilian Atlantic Forest
| Forest types | Adaptive trait | ΔAIC | ||||
|---|---|---|---|---|---|---|
| Seed provenance | Site of transplant | Both | With interactiona | Seed mass inclusionb | ||
| Semideciduous Forest versus Rainforest | Seedling emergence | 100.53 | 28.26 | 19.30 | 0.00 | |
| Seedling mortality | 267.34 | 5.39 | 1.19 | 0.00 | ||
| Seedling density | 33.88 | 39.37 | 35.77 | 0.00 | ||
| Seedling aboveground dry mass | 28.15 | 38.13 | 3.81 | 0.00 | Yes | |
| 49.20 | 53.21 | 24.91 | 22.31 | No | ||
| Sapling survival | 0.00 | 1.00 | 1.91 | 0.21 | ||
| Sapling root–shoot dry mass ratio | 5.13 | 8.34 | 0.00 | 1.78 | Yes | |
| 5.79 | 8.45 | 0.45 | 2.18 | No | ||
| Sapling total dry mass | 7.71 | 0.00 | 1.59 | 3.57 | Yes | |
| 11.57 | 3.54 | 5.21 | 7.20 | No | ||
| Rainforest versus Restinga Forest | Seedling emergence | 0.00 | 3.17 | 1.27 | 3.06 | |
| Seedling mortality | 12.97 | 0.00 | 1.42 | 2.36 | ||
| Seedling density | 13.70 | 6.99 | 0.00 | 1.56 | ||
| Seedling aboveground dry mass | 79.24 | 85.34 | 0.00 | 1.67 | Yes | |
| 103.35 | 105.58 | 25.66 | 27.27 | No | ||
| Sapling survival | 0.00 | 1.00 | 1.91 | 0.21 | ||
| Sapling root–shoot dry mass ratio | 9.44 | 2.47 | 1.68 | 3.54 | Yes | |
| 8.02 | 0.87 | 0.00 | 1.84 | No | ||
| Sapling total dry mass | 3.79 | 11.54 | 1.38 | 1.28 | Yes | |
| 2.20 | 9.93 | 0.00 | 0.06 | No | ||
Bold letters indicate models with ΔAIC < 2, or when more than one model has ΔAIC < 2, bold letters indicate the model with the lowest ΔAIC, indicating the most parsimonious model. aInteraction: considers that the effect of one predictor on the response variable is different at different levels of the other predictor, that is, the influence of seed provenance is different if seeds are in a site or another. bSeed mass inclusion: inclusion (yes) or not (no) of seed mass in the model depending on the significant effect of this covariable in the response variable (adaptive trait).