Found dataset and survival of Arctic grayling

[DPeterson]

I’m trying to provide some analysis suggestions on a “found” dataset for Arctic grayling, a freshwater salmonid. The dataset contains 15+ years mark-recapture information along an 80+ km river segment – thousands of fish marked and a few hundred recaptures (at most). The data were not collected under an operational experimental design, i.e., effort and spatial extent of sampling varied across years, sites haphazardly selected, etc. Only small reaches/segments of the river are sampled in any given year, and there was never any population closure at any site.

The dataset may have its problems, but it appears remotely possible to test hypotheses about (apparent) survival rates in relation to various abiotic and biotic factors over time.

Model: Of the models in MARK, the CJS appears to be only one that would be nominally appropriate. Am I missing any other potentially appropriate models?

Data preparation: The existing dataset may require extensive filtering or stratification which may substantially reduce the number of recaptures. Any suggestions for a general strategy to make these data usable? Is there a threshold (# recaps?) below which the attempt should be dropped?

Thanks,
Doug