problems with abundance estimates in closed population model

[amandarg]

I am using closed population models to estimate abundance over a 13 day trapping period. Most of my candidate models give me reasonable estimates of N, however, the model m(bh) gives me outragiously high estimates which I know are wrong. I am modeling the DM following White's 2008 paper describing how to do it. From reading others comments on the forum I know bh can have issues if there are not enough recaptures but I seem to have a decent amount. I am attaching my input file as maybe that will shed light on the problem to someone. The first day we opened traps we did not have any captures and i was wondering if that may be the problem.

Thanks,
Amanda

0111111111111 1 0 ;
0110100110101 0 1 ;
0100111111111 0 1 ;
0100001001000 1 0 ;
0011011111111 1 0 ;
0010101101111 1 0 ;
0010000100000 0 1 ;
0001111111111 1 1 ;
0001011111111 0 1 ;
0001000101000 1 0 ;
0001000001000 1 0 ;
0001000000001 1 0 ;
0001000000000 1 0 ;
0000111111110 0 1 ;
0000111101010 1 0 ;
0000111001000 1 0 ;
0000110111111 0 1 ;
0000100110100 1 0 ;
0000100011000 0 1 ;
0000010111110 1 0 ;
0000010111011 0 1 ;
0000010101000 2 1 ;
0000010001111 0 1 ;
0000010001011 0 1 ;
0000010001001 0 1 ;
0000010000011 0 1 ;
0000001100101 0 1 ;
0000001011101 1 0 ;
0000001011100 0 1 ;
0000001011001 0 1 ;
0000001010001 1 0 ;
0000001001001 1 0 ;
0000001000110 0 1 ;
0000001000000 1 0 ;
0000000111111 0 2 ;
0000000110110 0 2 ;
0000000110000 2 1 ;
0000000101100 1 0 ;
0000000101000 1 0 ;
0000000100101 0 1 ;
0000000100011 0 1 ;
0000000100000 1 0 ;
0000000011110 0 1 ;
0000000010101 0 1 ;
0000000010001 1 0 ;
0000000010000 0 3 ;
0000000001110 1 1 ;
0000000001100 0 1 ;
0000000001001 1 0 ;
0000000001000 1 1 ;
0000000000111 1 1 ;
0000000000110 0 3 ;
0000000000101 0 3 ;
0000000000100 1 2 ;
0000000000011 1 1 ;
0000000000010 0 2 ;
0000000000001 4 2 ;

[gwhite]

Yep, no captures on the first occasion of a removal model is a total kiss of death. Removal models require that the data show a downward trend in the number of captures. See page 108 of the Los Alamos Nerp Report for a formula from Seber and Whale (1970) that provides a failure criterion. No question that you data will fail!

Gary

[birdman]

Amanda,
What are the two columns following your encounter histories, a frequency and a covariate, or I presume, two frequencies indicating two groups? Either way, it doesn't look to me like you have a huge number of captures overall. Have a look at your capture probabilities for the Mbh model. If you have extremely low capture probabilities, you may not be able to accurately model N. Just a thought.
Cheers, S

[amandarg]

Hi, I am sorry it has taken me so long to respond, I am currently in the field. Gary, I am not conducting a removal study but a mark-recapture study. So, in a mark-recapture study would having no captures that first day still throw off all my analysis. Is there a way around this? S, the two columns after the capture history do represent two groups (males and females).

Amanda