Help with a bird migration study

[brid0030]

I am new to MARK, but somewhat familiar with the capture/recapture paradigm. I recently did a field project with the intent of using MARK to analyze the data, and I was hoping to get some feedback about what particular analysis would be best.

The experiment was pretty straightforward. First, I captured about 85 Dark-eyed Juncos on their wintering grounds in Oklahoma. slightly more than half of these birds were color banded and released. Twenty individuals were color banded, held in cages with all the food they could eat, and then released. Another twenty birds were banded and held in cages for a week on a food restricted diet before release. My question is whether these feeding treatments affected spring departure.

When the birds were released, I started doing resighting surveys nearly every day (usually at least two surveys per day). Resighting rates were pretty good. Most birds were seen nearly every day or every other day until they were (presumably) gone. I only recorded seeing an individual once during each survey (I don't know if the same bird was seen once or a dozen times in any one survey). The birds were individually banded, so I have individual resighting histories for every one, and during capture we collected some potentially informative covariate data (size, body condition, sex, age, etc).

Most analyses using MARK seem to be concerned about population size. For my analysis, I don't care about that. I want to know if "apparent survival" (note I'm regarding migration departure as "death"). differs among the treatment groups and how it might relate to sex, age and body condition as well. So what I need is an analysis that tests for differences among the treatment groups (high-food, low-food, and control) while considering various other factors and covariates. It would also be nice to be able to take survey effort into account as the amount of survey time varied from day to day. Also, weather affected the surveys, so maybe the total number of resights would be a good effort index.

I've been reading through the relevant parts of the Gentle Introduction, and I think I have some idea of where to start. But I would like to see if the member of this forum have any advice or cautions first.

Thanks.

[cooch]

I am new to MARK, but somewhat familiar with the capture/recapture paradigm.

<snip>

Most analyses using MARK seem to be concerned about population size.

Not sure where you get this idea from - in fact, the vast majority of MARk users focus on estimation of vital rates (say, survival, recruitment), and not generally on abundance itself. As a simple indicator, notice that the vast majority of the online 'tutorial material' for learning MARK (including 'the book') has relatively little on abundance estimation.

For my analysis, I don't care about that. I want to know if "apparent survival" (note I'm regarding migration departure as "death"). differs among the treatment groups and how it might relate to sex, age and body condition as well. So what I need is an analysis that tests for differences among the treatment groups (high-food, low-food, and control) while considering various other factors and covariates. It would also be nice to be able to take survey effort into account as the amount of survey time varied from day to day. Also, weather affected the surveys, so maybe the total number of resights would be a good effort index.

I've been reading through the relevant parts of the Gentle Introduction, and I think I have some idea of where to start. But I would like to see if the member of this forum have any advice or cautions first.

Thanks.

Regardless of where you end up, read/study Chapters 1 -> 7. Then, you're ready to tackle whatever analysis you're interested in. Once you've worked through that material, I suspect you'll be able to answer your question yourself.

[brid0030]

OK. Maybe I'll get some help if I am more specific.

Here is a typical encounter history:

0011110110111011000000000000000

The bird was banded sometime before the first zero (we did the banding before the daily resighting surveys), and was seen on the 3rd, 4th, 5th, 6th, 8th, etc... surveys. It was last seen on the 16th survey and presumably migrated shortly thereafter. So, I need a model that will count those first two zeros as "bird present but not seen" when calculating p (resighting probability). I think the original Jolly-Seber model will do that, but I need to somehow turn off the births (net new entrants) parameter. I know for certain that there are no births. I just want apparent survival. Is it possible to constrain B (births) to zero? Or maybe the POPAN formulation of JS. I see in the gentle intro (section 13.3.2 about POPAN): "How are death-only models (e.g., all Bi known to be zero) or birth-only models (all φ=1) or closed models be obtained by constraining the likelihood function." So there is the question, but where is the answer.

[brid0030]

Just found the way to fix parameters in the "run model" window. Way too easy.