Table 2.

AIC model selection table for movement and body temperature models after a summer capture event

Model specification	dAIC	dLogLik	df	weight
Movement
s(DsC, by = capture category)	0.0	124.5	74.8	0.95
s(DsC, by = surgery category)	5.9	125.1	78.3	0.05
s(DoY)	72.0	87.2	73.4	0.00
Null model	239.9	0.0	70.2	0.00
Body temperature
s(TsA_hours, by = capture category)	0.0	1839.5	46.0	1.00
s(DoY)	537.6	1566.7	42.0	0.00
Null model	3638.7	0	25.9	0.00

All movement models additionally included an interaction term of body size and sex as well as a random intercept for the individual bear at the capture event and therefore only varied in the terms specified in the table below. All body temperature models additionally included an interaction term of body size and sex, a smoother term of hour of the day and a random intercept for the individual bear at the capture event. The prefix ‘s’ indicates that the term was included as a smoother to account for non-linearity. dAIC = delta AIC, logLik = log-likelihood, df = degrees of freedom, weight = AIC weight. DsC = Days since capture, capture category = 2 levels (Winter/Summer capture vs Control group), DoY = day of the year, TsA_hours = Time since Antidot (h), surgery category = 3 levels (Only muscle biopsy vs Muscle biopsy + abdominal surgery vs Control group).