Movement patterns of a keystone waterbird species are highly predictable from landscape configuration

Table 3 Statistics of linear mixed-effects models to test the effects of landscape metrics on the maximum flight distances between roost and foraging sites and home range sizes of mallards in Western Europe

	Dependent variable (log-transformed)	Fixed effect	Intercept (±SE)	Estimate	SE	p-value	Marg. R²	Cond. R²
A	Maximum flight distance	Water surface area	8.41 (0.33)	−0.00118	0.00043	<0.001	0.29	0.47
A	Maximum flight distance	Total shore length		−0.00521	0.00077
B	Maximum flight distance	Water surface area	7.77 (0.52)	−0.00113	0.00085	0.344	0. 11	0.71
		Number of water bodies	7.69 (0.43)	−0.00104	0.00046	0.074	0.33	0.76
		Total shore length	7.73 (0.23)	−0.00436	0.00069	0.001	0.41	0.58
	Home range size	Water surface area	5.73 (0.51)	0.00097	0.00116	0.329	0.03	0.41
		Number of water bodies	6.82 (0.39)	−0.00145	0.00049	0.009	0.29	0.45
		Total shore length	7.55 (0.20)	−0.00989	0.00074	<0.001	0.54	0.54

Test results are shown for the top model for maximum flight distances, containing both water surface area and total shore length (A), and for univariate models for maximum flight distance and home range size (B). P-values were computed using likelihood ratio tests between the null models and the models containing the metric of interest as only independent variable. Marginal and conditional R² values were calculated following [52]

ISSN: 2051-3933