multi-state w/ early (nest) and late season (juvi) surveys

[Sunday]

I’m looking to run a single season multi-state model with a bit of a twist. I have survey data for loons that include early season surveys for nesting loons (replicated twice for most sites) and late season surveys for juveniles (replicated twice for most sites). My question is, can I use a regular multi-state model and constrain the detection probability of the first surveys to be zero for juveniles and fix the detection probability for the second surveys to be zero for nests? Or should I use the more general multi-season multi-state parameterization somehow as Jim Hines recommended in this previous post:

viewtopic.php?f=11&t=2411&p=7745&hilit=multi+state#p7745

If it’s the multi-season, multi-state parameterization, I’d very much appreciate some guidance on how to apply it in this situation because I'm struggling to figure it out.

FYI...I'm most interested in the three states: unoccupied, nesting, juvenile. However, I was also considering including non-breeding adults to make 4 states total if it doesn't over-complicate things and if I have the data to run it.

Many thanks!
Trevor