Table 4. Resource hypothesis model formulation and coefficient estimates for cross sectional prevalence study design.

Model	Parameter	Coefficient estimate	Standard error	Z value	P value	Odds ratio
log [π÷(1−π)] = β1+β2CS+β3DS+β4 |ΔEVI|+β5HS+β6NWB+β7DW+β8DP+β9X+β10Y+r.eff.(location) Calibration: le Cessie-van Houwelingen goodness of fit test (Z = 0.2, P = 0.8) Validation: AUC 0.7 Pseudo r2 = 14%	(Intercept)	12.04	143.76	0.08	0.93	…
	Condition score (CS)	0.76	0.39	1.93	0.05	2.13
	Dist. to streams (DS)	−0.57	0.21	−2.73	0.01	0.57
	EVI decline (ΔEVI)	−0.38	0.15	−2.59	0.01	0.68 *
	Herd size (HS)	0.01	0.02	0.61	0.54	1.01
	No. water bodies (NWB)	0.06	0.05	1.26	0.21	1.06
	Wallaby herd density (DW)	0.38	0.38	1.00	0.32	1.46*
	Wild pig density (DP)	−0.84	0.40	−2.09	0.04	0.43 *
	X coordinate (X)	−0.34	0.90	−0.38	0.71	0.71
	Y coordinate (Y)	−1.60	2.29	−0.70	0.49	0.20

Random effects terms for herd, and fixed effect covariates for latitude and longitude were included to control clustering of data and spatial trends or autocorrelation.

^*These covariates were transformed (normalised z = (x−μ)÷σ) to yield more interpretable odds ratios.