Table 2. Regression coefficients, adjusted standard error, values of t or z (for full averaged model) and p values from linear regression models between wintering greylag geese (log-transformed) as dependent variable and year, surface of crops and January temperatures as predictors.

Only best sub-models are represented. When there is more than one sub-model with ΔAIC < 2 (see S2 Table) full model averaged coefficients are shown (see methods). Sample sizes vary greatly because of missing data for predictor variables, especially for land use.

		B	Adjusted SE	t/z	p
Sweden	Intercept	–598.48	37.85	15.81 1	<0.001
n = 24 years	Year	0.30	0.02	16.14 1	<0.001
	Temperature	0.18	0.05	3.35 1	<0.001
	Sugar beet	-0.01	0.02	0.64 1	0.525
Denmark	Intercept	–374.62	56.26	–6.66 2	<0.001
n = 19 years	Year	0.19	0.03	6.63 2	<0.001
	Winter cereals	0.01	0.00	4.93 2	<0.001
	Temperature	0.39	0.06	6.21 2	<0.001
Germany	Intercept	–282.40	71.75	3.94 1	<0.001
n = 16 years	Year	0.15	0.04	4.04 1	<0.001
	Temperature	0.11	0.04	3.10 1	<0.01
	Oilseed rape	-0.00	0.00	0.93 1	0.354
Netherlands	Intercept	–235.70	11.01	21.31 1	<0.001
n = 30 years	Year	0.12	0.01	22.64 1	<0.001
	Temperature	-0.01	0.01	0.49 1	0.625
	Winter cereals	-0.00	0.00	0.57 1	0.569
	Potaoes	0.00	0.00	0.37 1	0.711
Belgium	Intercept	–528.65	88.25	–5.99 2	<0.001
n = 12 years	Year	0.27	0.04	6.09 2	<0.001
France	Intercept	–340.60	58.02	5.87 1	<0.001
n = 23 years	Year	0.18	0.03	5.98 1	<0.001
	Winter cereals	-0.00	0.00	0.77 1	0.439
Spain	Intercept	–56.30	10.93	5.14 1	<0.001
n = 28 years	Year	0.03	0.00	6.18 1	<0.001
	Temperature	–0.09	0.03	3.47 1	<0.001
	Rice	-0.00	0.00	0.71 1	0.478

¹z value;

² t value