Table 4. Results of traffic models analyzing habitat and road characteristics related to polecat road casualties using only data points associated with traffic counts (n = 150 presence, 1,058 pseudo-absence).

The conditional average model was created from the top subset (Δ AIC ≤ 2) of fourteen models. Regression coefficients (β) and 95% confidence intervals (±95% CI) are given as well as the significance of each variable where * P < 0.05, ** P < 0.01, and *** P < 0.001. The weight of variables included in conditional average model is also given, with the importance (number of candidate models that contain the variable) in parentheses.

	Best- approximating model		Full average model
Variable	β	95% CI	β	95% CI	Weight
Road density	4.58	1.70***	4.05	0.83***	1.00 (14)
Distance to rabbits	3.80	1.39***	3.86	1.39***	1.00 (14)
Traffic volume	2.77	1.99**	2.49	2.02*	1.00 (14)
Elevation	−2.42	1.13***	−2.41	1.13***	1.00 (14)
Patches	1.14	0.19***	1.13	0.19***	1.00 (14)
PC1	−1.12	0.32**	−1.13	0.32***	1.00 (14)
PC2	−0.20	0.26	−0.11	0.28	0.55 (7)
Waterway density	–	–	0.56	0.90	0.43 (6)
Moran’s I	–	–	1.12	4.11	0.38 (6)
PC3	–	–	−0.05	0.20	0.33 (6)

Notes.

^*P < 0.05.

^**P < 0.01.

^***P < 0.001.