TABLE 1.
Dependent variable | Parameter | Estimate | SE | p‐value | Relative importance |
---|---|---|---|---|---|
Body size | Intercept | 0.43 | 0.10 | <0.001 | — |
RA | −0.28 | 0.12 | 0.027 | 1.00 | |
PRED | −0.12 | 0.08 | 0.131 | 1.00 | |
DISP | −0.64 | 0.11 | <0.001 | 1.00 | |
RA × DISP | 0.43 | 0.16 | 0.007 | 1.00 | |
Movement morphology | Intercept | 0.21 | 0.10 | 0.029 | — |
RA | −0.08 | 0.08 | 0.321 | 1.00 | |
PRED | 0.18 | 0.13 | 0.167 | 1.00 | |
DISP | −0.37 | 0.12 | 0.002 | 1.00 | |
PRED × DISP | −0.27 | 0.15 | 0.084 | 0.61 | |
Basal activity level | Intercept | 0.24 | 0.09 | 0.008 | — |
RA | 0.08 | 0.10 | 0.429 | 1.00 | |
PRED | −0.07 | 0.09 | 0.408 | 1.00 | |
DISP | −0.45 | 0.10 | <0.001 | 1.00 | |
RA × DISP | 0.08 | 0.17 | 0.629 | 0.28 |
Notes: Parameter estimates, standard error and relative importance from the best model or from averaged models (when multiple models with ∆AIC < 2) of dispersal syndromes. Estimates for resource availability (RA), predation risk (PRED) and dispersal status (DISP, residents vs. dispersers) are given for standard resources, presence of predation risk and residents respectively. p‐values are estimated with z‐test for averaged best model (i.e. morphology and activity) and with t‐test for single best models (i.e. body size). The best models explained 25%, 23% and 25% of the marginal variance and 41%, 50% and 25% of the conditional variance, respectively, for body size, movement morphology and basal activity level (calculated on the best model or on the model with the same structure as averaged best models). Random intercepts are population identity within experimental block within species within taxon, population identity within treatments. See full model selection in Tables S2–S4.