Role of covariates in determining detection probability of leopard sign (Pt) on 2‐km‐long replicates of west Chure, based on covariates for probability of occurrence of leopard from the global model, $\widehat{\mathrm{\Psi }}$ (Global) = IO+R+N+PD+PS+WB+L+T

Model	AIC	ΔAIC	w	Model Likelihood	K
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(IO+N+L) θ 0 pi (·)	614.8	0	0.5317	1	31
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(IO+N+L+R) θ 0 pi (·)	616.68	1.88	0.2077	0.3906	32
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(IO+N+L+Samp_Eff) θ 0 pi (·)	616.79	1.99	0.1966	0.3697	32
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(IO+N) θ 0 pi (·)	621.02	6.22	0.0237	0.0446	30
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(IO+N+R) θ 0 pi (·)	622.55	7.75	0.011	0.0208	31
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(IO+R) θ 0 pi (·)	622.77	7.97	0.0099	0.0186	30
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(IO+N+Samp_Eff) θ 0 pi (·)	623.02	8.22	0.0087	0.0164	31
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(IO+L) θ 0 pi (·)	623.68	8.88	0.0063	0.0118	30
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(IO) θ 0 pi (·)	625.59	10.79	0.0024	0.0045	29
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(IO+Samp_Eff) θ 0 pi (·)	625.94	11.14	0.002	0.0038	30
$\widehat{\mathrm{\Psi }}$(Global) θ 0(·) θ 1(·) p(R) θ 0 pi (·)	662.42	47.62	0	0	29
$\widehat{\mathrm{\Psi }}$(Global), θ 0(·) θ 1(·),p(L) θ 0 pi (·)	668.29	53.49	0	0	29
$\widehat{\mathrm{\Psi }}$(Global), θ 0(·) θ 1(·),p() θ 0 pi (·)	670.11	55.31	0	0	28
$\widehat{\mathrm{\Psi }}$(Global), θ 0(·) θ 1(·),p(N) θ 0 pi (·)	671.97	57.17	0	0	29
$\widehat{\mathrm{\Psi }}$(Global), θ 0(·) θ 1(·),p(Samp_EFF) θ 0 pi (·)	672.08	57.28	0	0	29
$\widehat{\mathrm{\Psi }}$, θ 0(·) θ 1(·),p() θ 0 pi (·)	676.43	61.63	0	0	20

$\widehat{\mathrm{\Psi }}$: model‐averaged leopard occupancy; p = replicate‐level detectability; AIC = Akaike's information criterion, ΔAIC = difference in AIC value between the top model and the focal model; w = AIC weight; Model likelihood is −2 logarithm of the likelihood function evaluated at maximum; θ ₀ = Pr (leopard presence in a replicate/grid occupied and which was absent in the previous replicate) and “θ ₁” = Pr (leopard presence in a replicate/grid occupied and was present in the previous replicate); k = number of model parameters; Covariates: IO: management regime (grids inside and outside of the protected areas); R = terrain ruggedness averaged across each grid; N = nondifferent vegetative index averaged across each grid; PD = averaged human population density in each grid; PS = prey species (rhesus, barking deer, chital); WB = wild boar; L = livestock presence; Samp_Eff = sampling effort; T = tiger; + = covariates modeled additively; (·) = parameters are held constant. β‐coefficient estimates for IO, N, L = −4.97 (SE 1.34), −1.46(SE 0.47), 1.327 (SE 0.49), respectively.