Basic openCR questions about primary and secondary sessions

[Sarah Lydia]

Hi,

I am hoping to get some advice as I am just starting a mark recapture analysis, using openCR to estimate density. I have read lots of the vignettes and guides but before I get too deep into the models I'd like some basic help with how to structure the data. 

I have data from 15 years of trapping a species at a single site. The site was divided into three sections and each year the sections were trapped 4 times (e.g. all sections trapped once in Jan, once in April, July and October; the exact dates changed each year but stuck to the seasons). The areas are not distinct in terms of home ranges - the species' home ranges are likely to overlap between different sections.

I've organised the data so the years are the primary sessions with each trapping event being a secondary session (15 primary sessions (years), 12 secondary sessions per year). However, the population isn't closed between the trapping events, so should the trapping events (12 a year) be the primary sessions instead, each with one secondary session? That would be 180 primary (and secondary) sessions. Or alternatively, should the months be the primary sessions (4 a year) each with 3 secondary sessions (representing the three areas of the site trapped in sequence in that month) - but I can't assume closure between the secondary sessions in this case.

Also, because the three areas of the site were trapped in sequence,  the number/position of traps changed with each trapping session (i.e. we started with the traps in section 1, then moved to section 2, then 3, each month). Would that mean I'd need a different detector file/mask for each primary session, or can you specify which traps were used with a 'usage' argument?

I am interested in looking at the change in density over time but most interested in the most recent density estimate. Am I overcomplicating it? Should I just look at the most recent years of data?

Thanks so much,

Sarah

[Murray Efford]

I would have thought each month was a primary session (allowing turnover between primary sessions, presumably modelled with a 'season' covariate). It seems you have no opportunity for within-month recaptures, except maybe between sections. Did you get within-month movement between sections? If not then you could save some trouble by treating sampling within each month as synchronous across sections. Otherwise there will be three sequential secondary sessions in each month, and you will need to identify the active section on each with the usage attribute. In any case, the same mask can be used for all (it doesn't hurt if it's larger than needed for a particular section).

It seems the data don't lend themselves to closed population analysis, so I won't push that. I suggest starting with just annual samples (e.g., Jan in each year), assuming you have a long-lived species.

Murray