Capture Mark Recaptures with mark

[Isaël]

I am a trainee in M2 Conservation and the Restoration of the Biodiversity ( CRB) in the nature reserve of the Island of Platière ( 38, France ). I'm sorry for my bad English.

The purpose of my training course is to see if the fragmentation of the landscape plays in the fragility of Rhopalocériques populations (diurnal butterflys) noticed with the previous data of monitoring(at least to contribute to it this year).

I shall thus like estimating the capacity of movement of a species (Lysandra bellargus) and to obtain demographic data as emigration and immigration rate, mortality during the dispersal...
The objectives are rather simple but I shall like making something solid, and it is for that reason that I seek you.

I am thus going to make of the Capture Mark Recaptures on 10 for 15 lawns (according to necessary time) during two periods of flight ( 2 generations) of this species. The interesting lawns are the yellow polygons in the big red polygon on the map as attachment.

Lawns except the red polygon may less be connected but if the sampling allowed it, it would be interesting to me to visit them also.

I have to begin the CMR very soon (when the weather report will allow it) but I shall to be certain to make a correct sampling to be able to use an good model analyse.

I think of using the Mark software to analyse the data. After a lot of bibliography I saw that there is a wide range of useful models.

Robust Design Model seemed to me very good but the secondary sessions require to pass at least 3 times in the same day on every site. I wonder if such a pressure of capture
shall not impact on a butterfly so small by perturbing seriously its dispersal behavior so biasing all the study. I would prefer if possible to visit only once a day in every lawn and to visit more lawns.

Set apart the rates of survival and the probability of recapture, the model CJS ( Cormack-Jolly-Sebber) does not seem to be able to give information as immigration and emigration or the mortality in the course of dispersal.

Not being a specialist, I remain rather indecisive in front of the panel of possibilities and a scientific opinion of experienced person shall be very precious so that I can fix a sampling method adapted to a good model of analysis.

Very cordially, I thank you in advance

[sbonner]

Isael,

This sounds like a perfect application of a multi-state model to me. You have 10 to 15 different states (i.e. locations) and can use the multi-state model to estimate capture probabilities and apparent survival probabilities. Details of the multi-state model can be found in a lot of books, e.g. Williams, B. K., Nichols, J. D., & Conroy, M. J. (2002). Analysis and management of animal populations., and an Chapter 8 of the GIM (http://www.phidot.org/software/mark/docs/book/pdf/chap8.pdf).

One possible complication is that you won't be able to estimate actual survival probabilities with any CJS based model unless all individuals stay within the study locations. If individuals can move outside of the study areas -- which seems likely -- then you will only be able to estimate the apparent survival probability. That is, the probability that an individual remains alive and within the study area from one occasion to the next.

The robust design would allow you to estimate the probability that individuals emigrate outside of the studies areas, but this would require capturing individuals multiple times in short periods and I can see why you are afraid that this might affect behaviour.

Some questions to think about at this stage are how many individuals you think you can mark and how high the probabilities of recapture might be. Using MARK you can analyse simulated data to see if the type of data you expect will produce adequate estimates of the quantities you are estimated in. It will also give you practice so that you'll be ready to roll when the real data arrives!

Anybody else have other suggestions?

[Isaël]

I thank you very much for your informations.
If I want to test any models simulations with mark, I have to built a .txt to simulate real data?
Is anybody know a good model to obtain demographic data (emigration, immigration, population size...) with a sampling phase less intensive than RDM?
CJS is not very interessant for me because apparent survival and recapture probability have no real use for my study.

[sbonner]

Isael,

You don't need to input any data to run simulations. From the initial window select File > Simulations > Set Up Simulations. Then double-click anywhere in the window that opens. This will bring you to new window in which you can specify the basics of your model, but you won't have to specify a data file. Choose the Multi-state recovery model and specify the number of occasions and states for your model.

This will bring up a new window in which you can specify different 'true' models for generating data and further models for fitting the data. You input the demographics from using the beta values. Unfortunately, I can't help with selecting these parameters. You will have to choose values that you think are representative of your study. I suggest trying a range to see how the estimates change.

More information on running simulations is included in Appendix A of the GIM (http://www.phidot.org/software/mark/doc ... /app_1.pdf).

Cheers,

Simon

[Isaël]

Thank you again,

I'm using mark in this time with my data . I use the Multistates model.
In my inp, I've 10 states (10 Capture-release sites) and 2 groups (males and females). Here is a part of my .txt.

/*ID historique M F */
/*001*/ AB0000000000 0 1 ;
/*002*/ A00000000000 1 0 ;
/*003*/ AJ0JA0000000 1 0 ;
/*004*/ A00000000000 1 0 ;
/*005*/ A00000A00000 1 0 ;
/*006*/ A0A0A0A0A000 1 0 ;
/*007*/ AA0A0AA0B000 1 0 ;

I have read a big part of the markbook but I've just 2 questions that I am not sure:

-In the matrice index chart, the number of parameters is very big (220), it is logic (Phi site 1 to site 2, 1 to 3...10 to 1...) but do you think if I can use my data like this ?
-Is it possible with this sort of data and Multisatates model to obtain another estimations that recapture probability and apparent survival? (for example size of population, movements...)

I thank you in advance,

Isaël