Protocol that includes live encounters and dead recoveries

[Fish_Boy]

Greetings! I am planning a sampling protocol for a capture-recapture study. There are two parts to the study

1) a capture-mark-release phase that is repeated daily for five days.
2) a dead recovery phase that is done after each day of capture-mark-release.

Between each capture-mark-release phase there is a manipulation followed by a dead recovery of marked and unmarked individuals. This will be repeated five times during the season. The questions that need answers include:

1) What is the population size?
2) How many die as a result of manipulation?
3) What proportion of the population is affected by manipulation?

[Fish_Boy]

Has anyone tried an approach like this before? I was thinking with five primary occassions each with five secondary occasions a Robsut Design would work for the live recapture periods. But how would the dead recoveries portion work? It seems to me that there are two probelms each with its own solution. It would be ideal if there some way to get an abundance estimate for both the live recaptures and the dead recoveries. I am stepping outside of my comfort zone by trying to include an abundance based on the dead recoveries. ANY advice would be greatly appreciated.

At this particular location there is plenty (1000's) of animals to tag so I do not think that sample size will be an issue.