model
{
	# j = 1: GLMM, no NA data
	# j = 2: GLMM, NA data
	# j = 3: CAR,  no NA data
	# j = 4: CAR,  NA data
	Tau.noninformative <- 1.0E-4
	S.max <- 1.0E+4
	for (i in 1:N.site) {
		for (j in 1:4) {
			Y[i, j] ~ dpois(mean [i, j])
			log(mean[i, j]) <- beta[j] + r[j, i]
		}
		for (j in 1:2) {
			r[j, i] ~ dnorm(0.0, tau[j])
		}
	}
	r[3, 1:N.site] ~ car.normal(Adj[], Weights[], Num[], tau[3])
	r[4, 1:N.site] ~ car.normal(Adj[], Weights[], Num[], tau[4])
	for (j in 1:4) {
		beta[j] ~ dnorm(0, Tau.noninformative)
		tau[j] <- 1 / (s[j] * s[j])
		s[j] ~ dunif(0, S.max)
	}
}