Table 2.
GLM models illustrating how other hypotheses about killer whale reproduction were evaluated.
| Comparison | Error | Model | n | nA | BIC | Pr [M|Y] |
| Repro. lifespan | Gaussian | Null | 21 | 751.2 | 0.824 | |
| Repro. lifespan | Gaussian | mom's survival status | 21 | 11 | 754.3 | 0.176 |
| IBI | Poisson | Null | 62 | 107.1 | 0.645 | |
| IBI | Poisson | mom's survival status | 62 | 28 | 108.4 | 0.345 |
| juvenile φ(1–5) | Binomial | Null | 890 | 419.4 | 0.963 | |
| juvenile φ(1–5) | Binomial | grandma's survival status | 890 | 674 | 426.6 | 0.027 |
| juvenile φ(1–5) | Binomial | grandma's survival status + maternal age | 890 | 432.9 | 0.001 |
This table includes comparisons for reproductive lifespan, females' interbirth intervals, and survival of grandoffspring (φ). In each case, the null model is the simplest possible (grandmothers have no effect). The sample size for each comparison (n) represents the number of whales, except in the case of juvenile survival, where it represents whale-years (an individual may be recorded for several years). The subset of sample size with grandmothers alive is also included (nA).