Table 1.

Selection of models fitted to investigate variation in brucellosis seroprevalence in the Alpine ibex Capra ibex population from the Bargy massif (northern French Alps). Only the models with AIC_c weight ≥ 0.05 are provided here; for the full list of candidate models, see Supplementary Information 4. In model acronyms, “ + ” corresponds to additive effects and “×” to the interaction between the 2 factors. k is the number of parameters, LL is the maximum log-likelihood, ΔAICc is the difference in the Akaike information criterion between the model with the lowest AIC_c and the other models, and AIC_c weight is Akaike weight. “Age” and “sex” are ibex age (quadratic term) and sex, respectively. “UnitsMFall” are sociospatial population units (see Fig. 1 and Supplementary information 1). “Periods” opposed data collected before and after the slaughtering operations that occurred during autumn 2013 and early spring 2014.

Model	k	LL	AICc	ΔAICc	AICc weight
unitsMFall + sex x age	10	−129.53	279.99	0.00	0.27
sex + age + unitsMFall	8	−131.84	280.28	0.29	0.23
age + unitsMFall	7	−133.17	280.82	0.83	0.18
unitsMFall + period + sex x age	11	−129.17	281.47	1.47	0.13
unitsMFall + age + period	8	−132.66	281.94	1.94	0.10