Table 4.

Summary of model testing for differences in caterpillar numbers and weight across sites and day of the breeding season, after backward selection.

Response
Explanatory variable	Estimate	Std. Error	df	t value	P value
(a) Amount of caterpillars (log)
Intercept	0.31	0.17	53	1.84	0.071
Site	−0.60	0.07	132	−7.90	<0.001
Day	0.01	0.01	132	2.34	0.020
(b) Caterpillar weight (log)
Intercept	−0.02	0.01	130	−2.16	0.032
Site	0.02	0.01	130	1.74	0.084
Day	0.001	0.0002	130	3.54	<0.001
Site * Day	−0.001	0.0002	130	−2.76	0.007

Response variables were log-transformed to achieve normality of residuals. Initial explanatory variables were site, day and site*day interaction, and were dropped if not significant via backward selection (only significant variables are shown). Tree ID was always included as random factor.