Timing is critical: consequences of asynchronous migration for the performance and destination of a long-distance migrant

Table 3 Influence of timing of migration on white stork migratory performance

Response	Leg	Predictor	Estimate (SE)	t/z value	p value	R² marginal	R² conditional
Beeline distance	1	Intercept	449.4 (88.0)	5.11	–	0.01	0.80
		Date	− 0.32 (0.39)	− 0.82	0.412
		Age (juv.)	− 6.59 (18.5)	− 0.36	0.722
	2	Intercept	499.3 (230.5)	2.17	−	0.04	0.59
		Date	0.27 (0.99)	0.27	0.787
		Age (juv.)	− 55.92 (41.07)	− 1.36	0.173
	3	Intercept	2470.4 (328.2)	7.53	–	0.10	0.23
		Date	− 3.35 (1.37)	− 2.45	0.014*
		Age (juv.)	47.00 (51.84)	0.91	0.365
Route straightness	1	Intercept	0.91 (0.12)	7.89	< 0.001***	–	–
		Date	− 0.06 (0.05)	− 1.10	0.270
		Age (juv.)	− 0.39 (0.13)	− 3.06	0.002**
	2	Intercept	1.79 (0.25)	7.24	< 0.001****	–	–
		Date	0.18 (0.10)	1.77	0.077
		Age (juv.)	− 0.40 (0.26)	− 1.58	0.115
	3	Intercept	1.03 (0.10)	9.90	< 0.001***	−	–
		Date	0.15 (0.05)	2.70	0.007**
		Age (juv.)	− 0.25 (0.12)	− 2.13	0.034*
Flight ODBA	1	Intercept	0.202 (0.018)	11.32	−	0.04	0.59
		Date	0.007 (0.003)	2.08	0.038*
		Age (juv.)	0.014 (0.008)	1.62	0.104
	2	Intercept	0.200 (0.019)	10.30	–	0.01	0.81
		Date	0.003 (0.004)	0.96	0.340
		Age (juv.)	− 0.005 (0.010)	− 0.53	0.594
	3	Intercept	0.150 (0.011)	13.39	–	0.04	0.89
		Date	0.005 (0.002)	2.23	0.026*
		Age (juv.)	0.002 (0.006)	0.44	0.661

Results of the LMMs (and GLMM, for route straightness), testing the influence of timing of white stork migration and age on migration beeline distance, route straightness and flight ODBA, using bird ID and year as random factors, and on the flight ODBA model, also adding logger model as a random factor. For the route straightness and flight ODBA models, the variable “Date” has been scaled by subtracting the mean date and dividing by the standard deviation. Pseudo-r squared values are not available for GLMMs with beta distributions

ISSN: 2051-3933