Artificial individual encounter history file

[alyssam]

Hi,

As a newcomer to MARK, I have extensively searched the help forum and read the essential chapters (1-8, 15, 18, and A.2) of the manual, along with 2 of my colleagues. We have been unable to find instructions on how to generate artificial encounter histories for studies with animals that have no individually identifiable marks (Manual section: 18.2.1).

I have a mark-resight data set with all the parameters I need to generate population estimates using the mixed logit-normal mark-resight model, but I don't understand how to summarise the data into the artificial individual encounter histories.

Thanks in advance for any advice offered, it will be much appreciated.

Alyssa.

[bmcclintock]

Hi Alyssa,

Unfortunately, the instructions for creating artificial encounter histories in the absence of individually identifiable marks is limited to pages 18-5 and 18-6 of the The Book. The bottom line is that they can be generated any number of ways, as long as: 1) the sums of each column in the encounter history (when each individual entry, 1 or 0, is multiplied by the corresponding frequency) is equal to the total number of marked individual sightings for the respective sighting occasion; and 2) the sums of the frequencies with non-missing entries (i.e., not “....”) for each primary interval equals the known number of marked individuals in the population during the primary interval. For example, for 3 primary intervals each consisting of 4 sighting occasions:

11111111..... 5;
111011110111 3;
011011110110 1;
....11100110 1;

This would indicate that 5*1+3*1+1*0 = 8 marked individuals were sighted on the 1st sighting occasion of primary interval 1, and 3*1+1*1+1*1 = 5 marked individuals were sighted on the 3rd sighting occasion of primary interval 3. There were 5+3+1 = 9 marked individuals known to be in the population during primary interval 1, 5+3+1+1 = 10 marked individuals known to be in the population during primary interval 2, and 3+1+1 = 5 marked individuals known to be in the population during primary interval 3. Another way of presenting exactly the same data (makes no difference to MARK):

11111111..... 1;
11111111..... 1;
11111111..... 1;
11111111..... 1;
11111111..... 1;
111011110111 1;
111011110111 1;
111011110111 1;
011011110110 1;
....11100110 1;

or even

111111110111 1;
111011110110 1;
11111111..... 1;
....1111..... 1;
11111111..... 1;
111111110111 1;
111011110111 1;
11101111.... 1;
011011110110 1;
11111110.... 1;

It really doesn't matter as long as these 2 conditions are satisfied. It might be nice to put together a script to do this for us, but for the time being I don't know of one. Perhaps I'll get working on one if there's demand.

Cheers,
Brett

[alyssam]

Hi Brett,

I thought the sums of columns and frequencies was the clue, but couldn't quite get my head around the process. Your reply has made it clearer.

Thank-you very much for your response and a great chapter.

Cheers,
Alyssa.

[amandarg]

I was having the same problem as Alyssa and the explanation helped a lot so thank you. However, is there any chance someone has written a script yet? I notice it's been a year since this question was posted.

Thanks so much,
Amanda

[bmcclintock]

Hi Amanda,

To my knowledge, no such script has been written yet. Perhaps I'll get to it in the near future...

Cheers,
Brett

[amandarg]

Brett,
Thanks for the quick response,

Amanda