Table 3:

Summary of the relative importance of each explanatory variable in each set of candidate generalized additive mixed models and the area under the receiver operating characteristic curve of each full model, for each response modelled

Variable	Probability walking	Probability resting	Probability drinking	Probability wetting	Probability feeding	Probability shade use	Duration shade use
Black globe	1.0	1.0	0.67	1.0	1.0	1.0	1.0
Time of day	0.99	0.44	1.0	1.0	0.92	0.35	0.15
Season	0.31	0.33	0.38	0.98	0.97	0.48	0.30
Family group	0.77	0.47	n/a	0.49	n/a	0.54	0.94
Age class	n/a	0.48	0.48	0.48	0.47	0.48	0.33
State	n/a	n/a	n/a	n/a	n/a	n/a	1.0
ROC	0.62	0.79	0.65	0.76	064	0.74	0.53*

The relative importance of each explanatory variable was assessed and ranked by summing the Akaike weights $(w_i)$ across all candidate models in which the particular variable appeared. Between 0 and 1, the larger the value the more important the variable is relative to other variables within the set of candidate models (Burnham and Anderson, 2002). Variables significant in at least one of the selected best or alternative best models are shown in bold. Variables not included in the analysis for a particular behaviour are illustrated as n/a. ROC, receiver operating characteristic. *Adjusted R² value, not ROC value.