|
Calf survival
|
Juvenile survival
|
---|
Model | BIC | Pr [M|Y] | Pr [M|Y] v. null | BIC | Pr [M|Y] | Pr [M|Y] v. null |
null | 177.7 | 0.29 | - | 419.4 | 0.14 | - |
age | - | - | - | 416.9 | 0.51 | 0.78 |
age + age2 | - | - | - | 418.1 | 0.28 | 0.66 |
time | 183.3 | 0.02 | 0.06 | 426.2 | 0.01 | 0.03 |
moms > 35 (F) | 176.5 | 0.51 | 0.64 | 424.2 | 0.01 | 0.09 |
birth order | 180.9 | 0.06 | 0.17 | 423.7 | 0.02 | 0.11 |
first born (F) | 181.9 | 0.03 | 0.11 | 422.1 | 0.04 | 0.21 |
momAge | 180.5 | 0.07 | 0.19 | - | - | - |
grandmaAlive | 182.6 | 0.02 | 0.08 | - | - | - |
- Models of calf survival (from age 0 to 1) were applied to 329 whales to evaluate maternal effects (binomial response) and for juvenile survival (ages 0 to 5) for 1411 whale-years (337 animals). In the latter comparison, juvenile survival was held constant (models that allowed survival to vary as a function of juvenile age were also considered, but the results did not change). All ages in the table refer to the age of the mother; birth order was treated as a continuous variable. Model selection tools like BIC are preferable over traditional significance tests, because they can be used to estimate model probabilities, Pr [M|Y]. For each model, we also calculated the probability for model i relative to the null model (constant survival, intercept only), Pr [M
i
|Y]/(Pr [M
null
|Y] + Pr [M
i
|Y]).