Table 2. Justification of variables incorporated into candidate models for estimating daily nest survival rate of five forest passerines on Navarino Island, Chile.

Model	Variable	Predictions	Rationale
Null	Intercept only	Nest survival is random	Assumes daily survival rate (DSR) is constant.
Temporal effects	Day of year	Negatively associated with DSR	Late nesters will have lower nest survival because of the overlap with increased depredation pressure in the forest interior (i.e., N. vison), due to their breeding dynamics.
	Nest age (linear vs quadratic effects) and nest stage	Negatively associated with DSR	Nest age and stage influence adult behavior around the nest (increased nest visitation for food provisioning), and increased noise and odor from nestlings. These cues could be detected by predators.
Habitat effects	Concealment	Positively associated with DSR	Under the ‘total-foliage’ hypothesis, more nest concealment not only protects the nest and its content from predators, but also the adults entering and leaving it.
	Nest height off the ground (linear vs quadratic effects)	Positively associated with DSR	Under the ‘predator proximity’ hypothesis, nests closer to the ground will be more susceptible to ground predators
	Ground predator index	Negatively associated with DSR	Under the ‘predator proximity’ hypothesis, nests with higher index score will be more susceptible to predation.
	Canopy cover, canopy height, understory cover and understory height	Variables associated with nest-site selection will have equivalent effect on DSR.	Rationale of these variables’ effect on DSR is equivalent to that described in nest-site selection (Table 1).