robust design applied to PIT tag data

[zajanc]

I am a fisheries biologist currently working on applying a robust design to our PIT-tag data for bull trout in the McKenzie basin in Oregon. Our study area includes a small reservoir and 3 tributary arms. We have numerous trapping occasions within the reservoir, however relatively few recaptures (<30). Do you think this type of analysis (based on trapping only) would be worthwhile given the small number of recaps?

And also, do you think it would be reasonable or desirable to include PIT antenna data (from all 3 arms and 1 reservoir station) in the same analysis, perhaps in some multi-strata model?