Table 4.

Model averaging results from binomial generalized linear mixed models of heartworm infection risk in coyotes (n = 146)

Model	Predictor	Estimate	SE	z	Pr(> |z|)	Mean OR	95% CI
Resident and transient coyotes (n = 146)	(Intercept)	−0.59	0.45	1.33	0.18	0.55	(0.23–1.33)
	Age class (subadult)	−0.75	0.46	1.64	0.1	0.47	(0.19–1.16)
	Age class (pup)	−2.37	1.02	2.33	0.02	0.09	(0.01–0.69)
	Prop. low developed in home range	−0.22	0.27	0.82	0.41	0.8	(0.48–1.36)
	Prop. medium developed in home range	−0.63	0.33	1.9	0.06	0.53	(0.28–1.02)
	Prop. mosquito habitat in home range	0.33	0.23	1.46	0.15	1.4	(0.89–2.19)
	Year	−1.74	1.53	1.14	0.26	0.18	(0.01–3.54)
	Year (quadratic)	0.75	0.65	1.15	0.25	2.12	(0.59–7.64)
	Age class (subadult) * proportion medium developed	1.48	0.48	3.06	0.002	4.39	(1.7–11.3)
	Age class (pup) * proportion medium developed	−1.26	1.33	0.95	0.34	0.28	(0.02–3.84)
Resident coyotes only (n = 107)	(Intercept)	−0.49	0.3	1.62	0.1	0.61	(0.35–1.08)
	Age class (subadult)	−0.75	0.5	1.5	0.13	0.48	(0.18–1.27)
	Age class (pup)	−3.49	2.66	1.31	0.19	0.02	(0.00–4.48)
	Proportion medium developed in home range	−0.56	0.34	1.64	0.1	0.57	(0.29–1.12)
	Proportion mosquito habitat in home range	0.28	0.25	1.14	0.26	1.32	(0.82–2.14)
	Age class (subadult) * proportion medium developed	1.16	0.56	2.07	0.04	3.19	(1.06–9.53)
	Age class (pup) * proportion medium developed	−3.36	3.83	0.88	0.38	0.03	(0.00–62.73)

Predictors were obtained from the top-ranking models (ΔAICc < 2; Additional file 1: Table S4). Coyote home ranges were estimated by calculating and plotting 95% minimum convex polygons (MCPs). Results using 95% adaptive local convex hulls (a-LoCoH) are summarized in Additional file 1: Table S4 and S5)

SE is the standard error, Pr(> |z|) the P-value associated with the z statistic, and mean OR the mean odds ratio