Table 2.

a Nestling diet (binomial); and b adult starling provisioning rates (zero-inflated Poisson GLMMs with log-link function) in relation to day status (high human presence, HHP; or low human presence, LHP), time block (morning, early afternoon or late afternoon) and age of nestlings (weeks one, two or three)

(a) Response: nestling diet (proportion anthropogenic food provisioned) (n = 82 nest watches from 15 nests)
		AICc	ΔAICc	ώi
Top model set
Day status + Time block + Age		269.3	0.00	0.940

	Estimate	SE	2.5%	97.5%
Effect size of explanatory terms
Intercept	− 0.42	0.53	− 1.46	0.62
Day statusϮ ‘LHP’	− 1.32	0.35	− 2.00	− 0.64
Time block† ‘late afternoon’	0.72	0.33	0.06	1.37
Time block† ‘morning’	− 0.93	0.32	− 1.56	− 0.29
Ageǂ ‘week two’	0.74	0.38	− 0.02	1.49
Ageǂ ‘week three’	− 0.29	0.47	− 1.21	0.63

(b) Response: provisioning rate (n = 85 nest watches from 15 nests)
		AICc	ΔAICc	ώi
Top model set
Day status + Time block + Age		469.5	0.00	0.516
Day status + Time block		470.8	1.34	0.264

	Estimate	Adj. SE	2.5%	97.5%
Effect size of explanatory terms
Intercept	1.71	0.17	1.38	2.04
Day statusϮ ‘LHP’	0.04	0.13	− 0.21	0.29
Time block† ‘late afternoon’	− 0.38	0.14	− 0.66	− 0.10
Time block† ‘morning’	− 0.06	0.12	− 0.29	0.17
Ageǂ ‘week two’	0.30	0.14	0.03	0.57
Ageǂ ‘week three’	0.08	0.16	− 0.22	0.39

AICc, ΔAICc and model weights (ωi) are presented for all models within Δ2AICc for each analysis. Estimates, standard errors and 95% confidence limits presented are for the top model in the case where only one model within Δ2AICc was returned, or are model-averaged coefficients in the case of more than one competing model within Δ2AICc (in which case adjusted SEs are presented). Estimates are not back-transformed. Factors highlighted in bold have confidence intervals which do not contain zero

Ϯ Day status: ‘HHP’, † time block: ‘early afternoon’, and ǂ age: ‘week one’ were used as reference categories