Table 3.
HV ~ OW category | Intercept | Presence of eddy | P intercept /P variable | ∆AIC Null |
M1 – Outward phase | 3.57 ± 0.20 | −0.14 ± 0.40 | <0.0001/NS | +1.87 |
M2 – Central phase | 1.62 ± 0.11 | 0.12 ± 0.16 | <0.0001/NS | +3.30 |
M3 – Inward phase | 3.43 ± 0.25 | −0.635 ± 0.29 | <0.0001/< 0.05 | - 2.14 |
HV ~ FSLE (>0.1) | Intercept | FSLE (>0.1) | P intercept /P variable | ∆AIC Null |
M4 – Whole trip | 3.60 ± 0.51 | −7.29 ± 2.78 | <0.001/< 0.05 | - 6.40 |
TV ~ Current speed | Intercept | Current speed | P intercept /P variable | ∆AIC Null |
M5 – Whole trip | 3.51 ± 0.19 | −1.51 ± 0.24 | <0.001/<0.001 | −27.38 |
Linear mixed models were independently built with individual bird included as a random effect (n = 6) for each explanatory variable. Response variables are heading velocity (HV) and travel velocity (TV). The Okubo-Weiss parameter is a binary factor coding for the occurrence of eddies (0: absence, 1: presence). Current speed and FSLE are continuous variables. Only FSLE values >0.1d−1 were selected to test for the influence of filaments on HV when penguins were within a filament. Significant coefficients (mean ± se) are in bold. P p.value, NS non-significant. ∆AIC Null shows the AIC deviation from AIC of the null model