MSCRD Behavioral Response between Primary Sessions

[brp]

We have been trapping foxes for the past 4 years. We are using a Multi-strata Closed Robust Design with 4 primary sessions (years: 2007, 2008, 2009, 2010) and 4 secondary occasions (days) within each year. Age (pup, non-pup) define the states. Apparent survival rate is the main interest, but we would also like to assess the behavioral effects of trapping.

There are 32 capture probabilities (4 days in each of 4 years for 2 ages). There are 24 recapture probabilities (3 days in each of 4 years for 2 ages). It appears that capture and recapture probabilities are defined within a primary session (year). These can be used to assess behavioral effects of trapping within a primary session (year).

But is there a way to estimate capture and recapture rates between primary sessions (in hopes to assess behavioral effects of trapping between primary sessions (years)? Or do we assume, since there is a time lapse of a year between primary sessions, that the behavior of a fox in later years will not be affected by whether or not s/he has been captured in previous years (no long-term memory)?

Thank you,
Becky

[cooch]

We have been trapping foxes for the past 4 years. We are using a Multi-strata Closed Robust Design with 4 primary sessions (years: 2007, 2008, 2009, 2010) and 4 secondary occasions (days) within each year. Age (pup, non-pup) define the states.

So, a pup in year (i) becomes and adult (non-pup) in year (i+1), conditional on surviving?

[brp]

Yes. I have fixed the parameters as follows: Psi Pup to Non-pup = 1 and Psi Non-pup to Pup = 0.

[jlaake]

So why are you using MSCRD? Why not standard robust design? MS models are intended for cases where the transitions between states are not deterministic like an age model.

--jeff

[jlaake]

Sorry I meant to answer your question. You could model a behavioral effect across primary sessions by using an individual covariate that was 0/1 if not caught/caught in previous session or you could use the proportion or number of times caught in the previous session. Whatever you think may be releavant. That would be equivalent to a trap dependence effect which is modelled with the lagged value of the capture history. --jeff

[brp]

I want to calculate apparent survival rates for pups and non-pups separately.

Kendall recommended the following (see post "Age in RD models"):

Normally age is dealt with through manipulating the diagonals of the PIMS. However, that does not work when dealing with p's in the RD, because there is a separate PIM of p's for each primary period. Therefore, if you believe age affects p, I would use the multi-strata(state) closed robust design module, and treat age as a stratum. In this case state transitions (pup -> juvenile; juvenile -> adult) are deterministic, and some are not permitted (adult -> juvenile; adult -> pup; juvenile -> pup), so there are no extra parameters to estimate (just fix transition probabilities appropriately). I'm assuming individuals only remain juveniles for one year, but if it is longer that could be accounted for.

[Bill Kendall]

BRP,

You are correct that my previous posting could be pertinent here, but at least thinking about it quickly, I believe it would be more limiting than Jeff Laake's suggestion. With his approach, you can hypothesize that this trap effect (e.g., trap response persists into the next season) applies each primary period the individual is captured. I believe the approach I suggested (creating trapping frequency states) would only be applicable to following the first capture of an individual.
Bill

[brp]

So why are you using MSCRD? Why not standard robust design? MS models are intended for cases where the transitions between states are not deterministic like an age model.

--jeff

The following is my response to jlaake's question above "why are you using MSCRD?" Kendall's recommendation (below) has worked well for me thus far, but now the question about trap dependence between primary sessions has come up. I will take jlaake's advice on using a covariate.

My response:
I want to calculate apparent survival rates for pups and non-pups separately.

Kendall recommended the following (see post "Age in RD models"):

Normally age is dealt with through manipulating the diagonals of the PIMS. However, that does not work when dealing with p's in the RD, because there is a separate PIM of p's for each primary period. Therefore, if you believe age affects p, I would use the multi-strata(state) closed robust design module, and treat age as a stratum. In this case state transitions (pup -> juvenile; juvenile -> adult) are deterministic, and some are not permitted (adult -> juvenile; adult -> pup; juvenile -> pup), so there are no extra parameters to estimate (just fix transition probabilities appropriately). I'm assuming individuals only remain juveniles for one year, but if it is longer that could be accounted for.