Site covariates in a 2-species model

[Orien Richmond]

I'm running the 2-species model with a single covariate--patch area (Z-transformed). When I run the "occupancy is independent" model with area as a covariate the output looks OK. When I run the default model (with the extra parameter for phi), I get convergence problems; I also noticed that parameter estimates for phi and area are listed as detection parameters, not occupancy parameters in the model output. When I run the default model with the alternative parameterization (psiBa parameterization) there don't seem to be problems with convergence.

Are the convergence problems likely due to the fact that phi is restricted in the range of values it can possibly take due to limits in degree of overlap possible between 2 species?

What is the alternative parameterization (psiBa) set up to estimate? Is it treating the species interaction factor as an odds ratio of occupancy for the two species, the reparameterization described in MacKenzie et al.'s occupancy book (p. 236)?

The parameter names differ between the design matrix and the model output file (both untransformed estimates or beta's, and transformed estimates).

Are the following parameter names equivalent?
pA' = p*A = rA
pB' = p*B = rB
pAB = p*AB = lam

Thank you!