Table 1.
Model comparison results for the effects of sex, age class (A) and year (B) on Florida panther survival. Year was included as a random effect using the top-ranked sex and age model as a base. For each model, we present the number of parameters, the difference in Akaike’s Information Criterion (ΔAIC) and the Akaike weight (wi)
| Model | Parameters | ΔAIC | wi |
|---|---|---|---|
| A. Sex and age class models | |||
| Sex × Age1a + Olderb | 4 | 0 | 0·658 |
| Sex × Age2c | 5 | 1·98 | 0·245 |
| Sex × Age1 | 3 | 4·69 | 0·063 |
| Sex | 1 | 5·95 | 0·034 |
| Constantd | 0 | 15·25 | 0·000 |
| B. Additive effect of year (random) | |||
| Sex × Age1 + Older | 4 | 0·00 | 0·522 |
| Sex × Age1 + Older + rand(Year)e | 5 | 0·18 | 0·478 |
Age1 divides panthers into subadults (1–2·5 and 1–3·5 years for females and males, respectively) and adults (≥2·5 and ≥3·5 years for females and males, respectively).
Older refers to older adults (≥10 years); the first model therefore has the same older adult effect for both sexes, whereas the second model allows for different older adult effects between sexes.
Age2 divides the panthers into subadults (same as Age1), prime adults (2·5–10 and 3·5–10 years for females andmales, respectively) and older adults (≥10 years).
No predictor variables.
rand(Year) refers to a random effect of year as a categorical variable, with 1981–1986 grouped together (temporal variance).