. 2018 Apr 19;8:6285. doi: 10.1038/s41598-018-24627-1

Table 2.

Ecological factors explaining variation in cuckoo parasitism among passerine species in UK and Germany.

Count model coefficients	UK			Germany
Count model coefficients	Estimate	SE	p	Estimate	SE	p
Intercept	−20.030	4.742	0.000	−27.758	6.054	0.000
log (nest.depth + 1)	−0.330	0.885	0.709	−0.995	1.707	0.560
Crevice vs. Cavity	4.707	0.730	0.000	3.750	1.147	0.001
Dome vs. Cavity	5.405	1.140	0.000	5.033	1.228	0.000
Open vs. Cavity	4.505	0.777	0.000	3.766	1.070	0.000
log(Population size + 1)	0.283	0.067	0.000	0.524	0.090	0.000
log(Nest height + 1)	−0.041	0.131	0.756	0.257	0.168	0.125
Grassland vs. Forest	1.105	0.575	0.055	1.770	0.868	0.041
Rocky areas vs. Forest	−1.228	1.272	0.334	−1.764	1.172	0.132
Shrubland vs. Forest	1.596	0.508	0.002	1.933	0.816	0.018
Wetlands vs. Forest	3.074	0.752	0.000	3.177	1.163	0.006
Plant material vs. Invertebrates	−1.194	0.624	0.056	−4.326	0.907	0.000
Breeding overlap	−0.008	0.022	0.721	0.023	0.025	0.345
log(Female body size)	10.625	2.487	0.000	13.683	3.207	0.000
log(Female body size²)	−1.544	0.367	0.000	−2.146	0.465	0.000
Log(theta)	−0.291	0.224	0.194	−0.971	0.283	0.001
Zero hurdle model coefficients
Intercept	−5.696	14.438	0.693	−20.092	7.837	0.010
log(nest.depth + 1)	−1.939	2.110	0.358	1.144	1.095	0.296
Crevice vs. Cavity	4.252	2.353	0.071	3.650	1.610	0.023
Dome vs. Cavity	0.069	2.703	0.980	3.972	1.870	0.034
Open vs. Cavity	5.489	3.157	0.082	4.148	1.562	0.008
log(Population size + 1)	0.895	0.385	0.020	0.543	0.143	0.000
log(Nest height + 1)	−1.359	1.313	0.301	0.027	0.209	0.896
Grassland vs. Forest	9.059	5.601	0.106	3.307	2.933	0.260
Rocky areas vs. Forest	−0.144	2.264	0.949	−1.540	1.450	0.288
Shrubland vs. Forest	12.633	6.433	0.050	0.391	0.870	0.653
Wetlands vs. Forest	−4.319	2.932	0.141	−0.199	1.316	0.880
Plant material vs. Invertebrates	16.245	4479.8	0.997	−2.481	1.146	0.030
Breeding overlap	0.078	0.063	0.213	−0.029	0.029	0.319
log(Female body size)	3.798	5.730	0.508	7.897	3.925	0.044
log(Female body size²)	−0.919	0.853	0.281	−1.249	0.573	0.029

The estimates presented are from hurdle regression models with two components: a truncated negative binomial component for the positive counts (Count model coefficients) and a binomial component for the zero vs positive counts (Zero hurdle model coefficients). Estimates in bold have p-values below 0.05 and are considered significant. See Materials and Methods for details regarding each factor.