www.phidot.org

Hi all,

I'm somewhat new to MARK and RMark and I have a question about the assumptions MARK makes in determining survival when ages are grouped.

I have two age classes of birds that are morphologically distinct (1yr, and 2+yrs). We regularly catch birds first when they are 2+year olds. I'm wondering if my analysis is assuming birds that were caught first as 2+ were 1 the previous year and using those birds to determine 1yr old survival, or if only birds caught first as 1yr old are used to determine 1yr old survival? I would prefer the latter as I am uncomfortable with the assumptions necessary for the former.

My RMark code is as follows:

Code: Select all

tsprocess <-process.data(datMark_F2,model="CJS",
                         begin.time=1990,
                         groups= ("age"),
                         initial.ages =c(1, 2))

ts.ddl <- make.design.data(tsprocess, parameters=list(Phi=list(age.bins=c(1, 1.5, 29)),
                                                     p=list(age.bins=c(1,1.5, 29))))
levels(ts.ddl$Phi$age)<- c("SY", "ASY")
levels(ts.ddl$p$age)<- c("SY", "ASY")

p.dot <- list(formula= ~1)
p.time <- list(formula= ~time)
Phi.dot <- list(formula=~1)
Phi.age <- list(formula=~age)

cml <- create.model.list("CJS")
female.results <- mark.wrapper(cml, data=tsprocess, ddl=ts.ddl, output=F, adjust=F)

Thank you very much for any and all assistance!

I'm not certain why you would think it is the former. With CJS models (presumably that is what you are using) the data are only model after the initial release (first 1 in capture history). Thus those that are age 1 at initial capture have to survive from age 1 to age 2 to be seen again. All those assigned to age 2 have a common survival with the binning you used and it is only for survival for age 2 and older.

It would probably be good for you to read the sections on CJS in the MARK book. Your question suggests you have either not read it or didn't understand the basic model structure for CJS models or known fate if that is the data you have. While RMark makes it easier to use MARK, you still need to understand the underlying models. RMark is simply creating the design matrix for you from formulas. You still need to understand what those formulas represent and how they are used in the model for the data. Your knowledge of both can be improved by reading the MARK book and examining the design matrix that RMark creates.

regards --jeff

Thanks! That's what I thought but I think I got all confused about the basics while talking to some people about my data. I will go re-read!

11arc4 wrote:Thanks! That's what I thought but I think I got all confused about the basics while talking to some people about my data. I will go re-read!

The key 'reading' would be chapters 4, and then the first half (or so) of chapter 7. When you get to a level of understanding to consider linear models and the like, then chapter 6, and last half of chapter 7.