Table 2. Model selection using AICc.
Results from model testing of squamate biogeography via BioGeoBEARS, with the bestsupported model (DEC+ j) in bold. The addition of the ‘j ’ parameter did not absorb a disproportionate amount of the dispersal estimated in the other models, a caution expressed by some previous authors (Ree & Smith, 2008; see Matzke, 2021).
| Model | lnL | k | d | e | j | AICc | Wt |
|---|---|---|---|---|---|---|---|
| DEC | −7,319 | 2 | 0.0016 | 1.00E−12 | 0 | 14,642 | 9.00E−10 |
| DEC+J | −7,297 | 3 | 0.0015 | 1.00E−12 | 0.0008 | 14,601 | 1 |
| DIVALIKE | −7,885 | 2 | 0.0019 | 1.00E−12 | 0 | 15,774 | 1.40E−255 |
| DIVALIKE+J | −7,819 | 3 | 0.0017 | 1.00E−12 | 0.0019 | 15,645 | 1.90E−227 |
| BAYAREALIKE | −12,365 | 2 | 0.01 | 0.01 | 0 | 24,735 | 0 |
| BAYAREALIKE+J | −12,278 | 3 | 0.01 | 0.01 | 0.001 | 24,561 | 0 |