Table 3.
Cox-regression for prey species (influence of covariates on survival). Regression analysis 1 includes all specimens, analysis 2 was calculated without three Gobiodon histrio specimens with unusually high condition factors (Cf). A negative value in B and a value below 1 in exp(B) indicates that the covariate reduces the likelihood of being eaten
| Term | B | d.f. | Significance | exp(B) |
|---|---|---|---|---|
| Analysis 1 | ||||
| Cf of prey | 1.40 | 2 | 0.007 | 4.058 |
| Gobiodon histrio | –3.59 | 1 | 0.001 | 0.028 |
| Gobiodon sp.3 | –2.46 | 1 | <0.001 | 0.085 |
| Analysis 2 | ||||
| Gobiodon histrio | –2.39 | 1 | 0.002 | 0.092 |
| Gobiodon sp.3 | –1.63 | 1 | 0.003 | 0.197 |