Huggins closed capture help!

[GNealeMT]

Dear MARKers:

We are analyzing fox mark recapture data via MARK. We have a 4 day closed capture session on 12 8x5 grids (40 traps each) with associated encounter histories for about 200 foxes. We have been running the Huggins closed capture model.

We are running into problems when we try to analyze foxes by groups or covariate. We would like to estimate abundance for each of the 4 habitat types we placed our 12 grids within. We get pretty decent estimates but the CIs are way off.

Our inpute file looks like this:

fox ID #; Encounter history; Group 1 (each record is unique individual so every row has a number 1); additional # for grouping variable (for habitats 1-4).

Example:

/*001 */ 0110 1 4;
/*002*/ 0001 1 3;
/*003*/ 1110 1 1;

etc.

We tested 4 models ( 1. p(t)c(t) not equal; 2. p ( c ) not equal; 3. p.c. equal; and 4. p.c(t) not equal).

Does it look like we have an input format error?

Thanks for any help you can offer!

[cooch]

Dear MARKers:

We are analyzing fox mark recapture data via MARK. We have a 4 day closed capture session on 12 8x5 grids (40 traps each) with associated encounter histories for about 200 foxes. We have been running the Huggins closed capture model.

We are running into problems when we try to analyze foxes by groups or covariate. We would like to estimate abundance for each of the 4 habitat types we placed our 12 grids within. We get pretty decent estimates but the CIs are way off.

Our inpute file looks like this:

fox ID #; Encounter history; Group 1 (each record is unique individual so every row has a number 1); additional # for grouping variable (for habitats 1-4).

Example:

/*001 */ 0110 1 4;
/*002*/ 0001 1 3;
/*003*/ 1110 1 1;

etc.

We tested 4 models ( 1. p(t)c(t) not equal; 2. p ( c ) not equal; 3. p.c. equal; and 4. p.c(t) not equal).

Does it look like we have an input format error?

Thanks for any help you can offer!

Your suspicion concerning the input file format seems correct - your use of a single column covariate to indicate group won't work as intended - instead, what you've got is (in effect) a regression on an ordinal number indicating group. There are two ways to handle group in simple ANOVA-like design (where the null is simple heterogeneity among groups).

Read Chapter 2 of the MARK book, and then have a look at section 12.6 in Chapter 12 (the individual covariates chapter). That should help.

[GNealeMT]

Thanks for the suggestion cooch. I looked at it awhile ago but think I forgot to thank you! g

[cooch]

Thanks for the suggestion cooch. I looked at it awhile ago but think I forgot to thank you! g

You're welcome. No thanks needed (although I do accept PayPal contributions