Mean haul-out duration in relation to predictor variables.

For continuous predictors (i.e., cloud cover) the effect is the change in haul-out duration for a 1 unit change in the predictor. Marginal means (i.e., SAS least-squares means; [41]) are the estimated mean duration of haul-out bout at each level of the categorical predictor variable, adjusted for other variables in the model.
1‘hr_start’ and ‘hr_end’ not used simultaneously.
2Results for binary vessel predictors (i.e., absence/presence); results for the binary vessel variables are easier to interpret, but show similar patterns to the continuous vessel predictor models.
Mean haul-out duration in relation to predictor variables.