Sarcoptic mange in the Scandinavian wolf Canis lupus population

BMC Veterinary Research

Table 3 Model selection for predictors of seroprevalence of sarcoptic mange in adult wolves captured in Scandinavia

Model	Parameters	K	AIC_c	ΔAIC_c	ω_i	r_s(SE)
Combined	Lat + Sex * Pack Size	5	78.16	0	0.38	0.86(0.02)
+ Density	Lat + Sex * Pack Size + Territory Density	6	79.00	0.84	0.24	0.86(0.03)
+ Body Condition	Lat + Sex * Pack Size + Body Condition	6	79.04	0.88	0.24	0.59(0.03)
– Pack Size	Lat + Sex	3	80.57	2.41	0.12	0.62(0.03)
Submodel intrinsic	Sex * Pack Size	4	83.36	5.20	0.01	-
Submodel density	Lat + Pack Size	3	84.83	6.67	0.01	-
Density	Lat + Long + Pack Size + Territory Density	4	88.47	10.31	0	-
Intrinsic	Sex * Pack Size + Body Condition + Age + Repro	6	90.41	12.25	0	-
Null model	1	1	92.00	13.84	0	-

Top models are validated by k-fold cross validation (r_s). Lat / Long: Latitude and longitude of the territory centre point, Repro: Reproducing or non-reproducing pair, Pack Size: Number of wolves within the territory, Territory Density: Mean Euclidian distance to the next three territory centre points, Body Condition: Individual residual distance to the linear regression line of log body weight and log body length. Presented are the two main models (Intrinsic, Density), the top models for each variable group, the model combining the top submodels, the combined model with variables ranking within the cut off in the variable group model selection and the null model

ISSN: 1746-6148