Richness Occupancy Estimation

[campbelb]

Dear forum members:

I've seen a few discussions regarding richness estimation from occupancy models but I still have some difficulty interpreting Psi in the context of a species x site design matrix with each row corresponding to a unique sitexspecies record.

My first question; perhaps the most basic. To estimate occupancy among a range of sites encounter histories where no species where recorded would be included if estimates for each species was required. This becomes a rather cumbersome exercise when the species pool is large (in my case, 224 total species). What I am really interested in is site specific estimates of occupancy for particular bird guilds where effort (number of surveys) varies per site as well as measured landscape covariates. How should I proceed to construct the proper design matrix for this level of analysis?

As an example of what I'm trying do, I have two individual covariates for occupany (PF-forest cover, ECON - edge contrast) , one individual covariate for detection probability (effort) and one group covariate I want to model (Guild). However, I'm uncertain about how to incorporate site into this design.

<-process.data(cabruca.in, model="Occupancy",groups=c("GUILD"))


p.dot<- list(formula=~1)
p.effort<- list(formula=~effort)

Psi.dot<- list(formula=~1)
Psi.PF500<- list(formula=~PF500)
Psi.ECON<- list(formula=~ECON)
Psi.PF500.ECON<- list(formula=~PF500+ECON)
Psi.PF500.ECONINT<- list(formula=~PF500*ECON)

Any recommendations would be appreciated.

Brian Campbell