Table 3.

Linear mixed effects model for the timing of breeding (lay date) for successful nests only in relation to (a) the urban gradient (1–78% urban cover); and (b) nest type (nest boxes versus other), nearest neighbour distance (log transformed) and weather conditions during the egg-laying period (spring). Estimated lay period of 52 and 24 peregrine territories, respectively, over 26 years (1989–2014) on the Cape Peninsula.

(a) all nests n = 316 lay date	estimate	SE	t-value	χ2	df	P-value
Intercept	0.62	0.80	0.78	0.60	1	0.439
Urban gradient	-0.36	0.08	-4.45	19.84	1	<0.001
Nearest neighbour distance	-0.13	0.23	-0.59	0.35	1	0.555
Pc1 lay period (cold/wet–warm/dry)	-0.13	0.04	-3.32	11.00	1	0.001
Pc2 lay period (warm/wet–cold/dry)	0.62	0.80	0.78	6.24	1	0.013
(b) urban nests n = 159 lay date	estimate	SE	t-value	χ2	df	P-value
Intercept	-2.00	1.27	-1.58	2.48	1	0.115
Urban gradient	0.26	0.14	1.84	3.38	1	0.066
Nest type (nest box) †	-0.36	0.22	-1.63	2.64	1	0.104
Nearest neighbour distance	0.64	0.35	1.81	3.29	1	0.070
Pc1 lay period (cold/wet–warm/dry)	-0.14	0.06	-2.48	6.15	1	0.013
Pc2 lay period (warm/wet–cold/dry)	-0.32	0.09	-3.61	13.00	1	<0.001
Urban gradient x nest type †	-0.40	0.22	-1.78	3.16	1	0.075
Pc2 lay period x nest type †	0.28	0.09	3.08	9.46	1	0.002

† Nest type ‘other’ were used as a reference category. Pc1: negative values: cold and wet weather conditions; positive values: warm and dry weather conditions; Pc2: negative values: warm and wet weather conditions; positive values: cold and dry weather conditions. See full matrix of factor loadings of PCA in supplementary material S3.