Analysis approach - CJS / Multi-state / RD / secr?

questions concerning analysis/theory using program MARK

Analysis approach - CJS / Multi-state / RD / secr?

Postby SoConfused » Fri Mar 23, 2018 8:24 am

Hello,

I'm working on a dataset of fish mark-recap, with 5 years (and 2 seasons) of data, where a season (fall/spring) is the occasion (=10 sampling events). The sampling was done using gill-nets in a fully randomized design for each season, and each sampling took place over 2-3 weeks. The total sampling area is divided into 2 areas of interest. Fish move between those 2 areas freely (although they seem to like it better in one of those areas).

My main interests are overall survival and occasion-specific abundance within each of the two areas. It would be nice to estimate the probability of movement between the two areas. I'm ok with running different models to obtain the different outputs, but can't quite wrap my head around the options.

So far I've been ignoring within-occasion structure, and using CJS / POPAN on the combined dataset and the data separated into two subsets based on area, which isn't great, since I'm losing a lot of recap information. Data quirks: the within-occasion recapture rate is very low (<100 individuals out of ~10,000 tagged), while the CJS-estimated between-occasion recapture is around 0.05.

What approaches can I try to analyse the dataset as a single entity and still get what I'm after? It's my understanding that I can't get abundance estimates from multi-state models, and I think that the low within-occasion recapture would cause issues with robust design. I started looking into secr, but am not sure if it would provide the output I'm after or whether it can handle the non-repeat sampling locations (i.e., a net isn't reset at the same location after lifting).

Any advice would be appreciated.
SoConfused
 
Posts: 30
Joined: Wed Nov 05, 2014 8:25 am

Return to analysis help

Who is online

Users browsing this forum: Bing [Bot], Google [Bot] and 1 guest