Table 1.
Seaside Sparrow THg concentration data of the ten best-performing general linear mixed models (with site and individual identity as random effects)
| Model | df |
log likelihood |
AICc | ΔAICc | wi |
|---|---|---|---|---|---|
| Year + Oiling | 7 | 107.83 | –201.0 | 0.00 | 0.588 |
| Year + Oiling + [PAHs] | 8 | 107.97 | –199.1 | 1.92 | 0.225 |
|
Year + Oiling + Year × Oiling |
9 | 108.44 | –197.8 | 3.46 | 0.119 |
|
Year + Oiling + Year × Oiling + [PAHs] |
10 | 108.73 | –196.1 | 4.87 | 0.052 |
| Year | 5 | 101.44 | –192.5 | 8.47 | 0.009 |
| Year + [PAHs] | 6 | 101.44 | –190.4 | 10.61 | 0.003 |
| Oiling | 6 | 101.44 | –190.4 | 10.61 | 0.003 |
| Oiling + [PAHs] | 7 | 101.88 | –189.1 | 11.89 | 0.002 |
| Intercept only | 4 | 94.83 | –181.4 | 19.57 | 0.000 |
| [PAHs] | 5 | 95.00 | –179.6 | 21.34 | 0.000 |
The models test various hypotheses about factors and covariates explaining inter-individual variation in THg
AICc are Akaike’s Information Criterion scores adjusted for small sample sizes (Burnham and Anderson 2002. ΔAICc is the difference between the AICc score of a given model and the lowest AICc of all models (i.e., the model including the effect of oiling history and year, in this case). wi is the Akaike’s weight of each model. The best models are in bold. [PAHs] stands for sediment concentrations of polycyclic aromatic hydrocarbons