Table 2. Results from stepwise regression to quantify associations between virus occurrence probabilities and biotic covariates (BFF (Pteropus alecto, Black flying fox), GHFF (P. poliocephalus, Grey-headed flying fox) and LRFF (Little red flying fox, P. scapulatus)) and abiotic covariates (mean Normalized Difference Vegetation Index (NDVI) within 20 km over the preceding 3 months, average precipitation in the preceding month, and average water vapour pressure, and year). Virus abbreviations as per Table 1.

	TevPV	YepPV	GroPV	HeV	MenV	HerPV†	YBPV
Intercept	−7.38 ***	−13.25 ***	−21.23	−6.92 ***	−20.07	−57.7	−5.73 ***
Avg precipitation (1mo)	−2.72 ***	−4.55 **	–	−3.32 *	0.97 *	−19.78	−0.95
Vapour pressure	–	–	−0.6 *	–	−1.45 **	−3.22	−0.74 *
Avg NDVI (3mo)	–	−2.26 *	−1.47 *	–	–	–	–
BFF	0.94 ***	1.64	1.19 *	–	–	–	–
GHFF	−1.59 ***	−6.61 ***	−2.69 **	–	−1.05	−24.55	−1.22 **
LRFF	–	–	−115.85	–	−114.68	–	–
Year 2011	2.71 ***	–	1.58 *	2.35 **	–	–	0.9 *
Year 2012	–	–	–	–	–	11.78	–
Location VIC	–	–	−20.4	−18.85	−20.86	–	1.12

Table shows final terms after backward stepwise selection, along with their coefficients and significance (*p < 0.05; **p < 0.001; ***p < 0.0001). – represents that the term was not included in the final model. Note that coefficients are from a binary logistic regression and can be interpreted as effects on a virus’ log odds of detection. For year and location, 2010 and Queensland represent the reference category, respectively. Colour scale ranges from dark blue (strong positive effect size) to dark red (strong negative effect size). ^†Poor model fit.