Royle Counts - inflated aic values

[Alan Dextrase]

Hello,

I am relatively new to Mark and have been using the Royle Count occupancy models and comparing outputs to those from Presence (Poisson only) and the Royle Nmixture nlm routine in R (Poisson and NB). My data have several sampling occasions where double digits of animals were counted (up to 76 at one site on one occasion). I find the aic value from Mark is about an order of magnitude higher than the outputs from Presence or nlm programs (Presence and nlm produce the same output for the Poisson model - NB is not currently in presence). I have tried simulated data with single digit counts and get comparable outputs from all three programs, but with simulated data that include double digit counts, the Mark aic values are again higher than the other programs. I couldn't find this issue described in the forum, so I am not sure what I am doing wrong. I have attached a sample of my data (3 sample occasions). Thanks for any help you can provide.

Al

000000 1;
000000 1;
000000 1;
000000 1;
020203 1;
070805 1;
291502 1;
010201 1;
000000 1;
000000 1;
010000 1;
010000 1;