analysing special data from censuses

[garciap]

Dear people in the forum,
I’m trying to analize in MARK a difficult kind of data to get a robust estimation of population density. This data is about counts of otters during vigils in freshwaters. The basically method is to distribute people along the stretch to be sampled, and the people stay quiet during 1 to 2 hours recording otter’s observations. Nevertheless, otter lack any natural identifiable mark so we decide the identity of an animal observed by different people on the basis of the direction, size, speed and sex of the animals observed (as done in other canrivores lacking natural marks).
Thus I couldn’t do Jolly-Seber models, but I think there are some ways to analyse the data, as I’ve seen that there are works in which mark-recapture models are used for multiple observers of birds lacking natural marks (i.e. Allredge et al. The Auk, uses program CAPTURE to estimate the probability of detection of birds by two or multiple observers in point counts).
I think that these kinds of analyses could be performed with my data, maybe considering each person a different session? Using the closed capture designs obviously, and perhaps estimating how many people has seen the same animal one could get a closed captures analysis.
What do you think? I’m very interested in estimating capture probability under our design, at least as a way to evaluate the reliability of the method. Any help, please?
Many thanks for any comment in advance,

Pablo