Single release cohort

[feliona]

Dear MARK forum,
I’d like to rise a question that had been asked some years ago (viewtopic.php?f=1&t=87&p=146&hilit=batch#p146) but had remained unanswered. I’d like to estimate apparent survival and recapture probability in a study with individually tagged fish (PIT-tags) in a small, shallow stream. The fish were all marked the same day in autumn and followed throughout the winter by remote detection using mobile tracking equipment (26 weekly trackings). How to deal with a single release cohort in MARK? On page 4 in chapter 4, this type of a mark-recapture-scheme is shortly introduced (as batch-marking), but I cannot find further inputs. Searching for “batch-mark” is a little misleading as it often refers to group-markings (i.e. no individual marks).
Many thanks for your help!
christine

[cooch]

Dear MARK forum,
I’d like to rise a question that had been asked some years ago (viewtopic.php?f=1&t=87&p=146&hilit=batch#p146) but had remained unanswered. I’d like to estimate apparent survival and recapture probability in a study with individually tagged fish (PIT-tags) in a small, shallow stream. The fish were all marked the same day in autumn and followed throughout the winter by remote detection using mobile tracking equipment (26 weekly trackings). How to deal with a single release cohort in MARK? On page 4 in chapter 4, this type of a mark-recapture-scheme is shortly introduced (as batch-marking), but I cannot find further inputs. Searching for “batch-mark” is a little misleading as it often refers to group-markings (i.e. no individual marks).
Many thanks for your help!
christine

There is nothing explicitly in the book about batch marking, since, under the default PIM structure (which assumes differences over time, but not among release cohorts), there is nothing 'special' you need to do with the PIMs. The assumption (in the book) was that if you thoroughly understood how MARK handles parameter indexing, then you'd have realized this, without explicit example of same.


Take for example, the following example .inp file, showing a single release cohort on occasion 1, 6 occasions.

Code: Select all

100000 249;
110000 89;
110111 2;
110001 4;
100010 6;
101010 8;
100100 11;
101000 24;
111100 9;
101100 12;
100111 4;
111000 23;
101110 4;
100101 3;
100011 3;
110010 8;
111111 3;
110100 13;
110110 6;
101111 3;
100001 2;
110011 3;
100110 3;
101011 1;
111011 1;
111110 3;
110101 2;
111010 1;


The PIM structure you need to analyze these data is the same as you would use if in fact there were individuals releases on each occasion -- under the assumption that apparent survival is a function of time, and not release group. So, the PIM would be the standard default time-dependent PIM for 6 occasions:

Code: Select all

1 2 3 4 5
  2 3 4 5
    3 4 5
      4 5
        5


Note that with a single release cohort, you can't partition time and 'time since marking' effects, since time and TSM would be co-linear for a single release cohort. All of this is discussed in chapter 7.