Table 3.

GLM models – significant associations with prevalence or abundance of parasites in foxes

Parasite	Model Fit	Significant parameters	Value and 95% CIs	Parameter type	P value
E. multilocularis	Binomial GLM (logistic)	Age	0.82(0.65–0.99)	Odds ratio	0.043
E. multilocularis	Zero-inflated Negative binomial GLM	Age (zero-inflated)	1.28 (1.01–1.62)	Odds ratio	0.042
		k	0.26(0.19–0.37)
Dipylidium caninum	Negative binomial GLM	Age	0.57 (0.37–0.81)	Incidence rate Ratio	<0.0001
		Presence of E. multilocularis	3.81(1.35–10.16)	Incidence Rate Ratio	0.008
		Presence of T. leonina	0.13(0.02–0.80)	Incidence Rate Ratio	0.028
		k	0.13(0.08–0.17)

Incidence Rate Ratio. This is the parameter in a negative binomial generalised linear model. For a continuous variable such as age it is the proportional change in parasite count for each year. Thus, for each 1 year increase in age the abundance of D. caninum parasites decreases by 57%. For a binomial variable such as presence of E. multilocularis it is the ratio of D. caninum parasite counts between those foxes infected with E. multilocularis and those non-infected.