Mark resight_Poisson log normal model?

[Lyndon]

I've been presented with estimates, data and a draft manuscript. The person who did the analysis has moved on and is not available to answer my questions. I'm trying to replicate the analysis so I can rewrite some parts of the manuscript.
The sampling protocol seems to be that a boat with a photographer on each side made two passes through a large group of dolphins. The photographs were allocated to two classes, those that identified distinctively marked individuals (DMI photographs) and those that depicted non-distinctively marked individuals (NDMI photographs). Matching of DMI photographs to individuals determined the number of times each was 'captured' in DMI photographs.
The objective is to fit a model to these data to estimate the size of the group. The model employed was described as a Poisson log normal mark resight model.
My short questions are 1) is this appropriate and 2) which of the mark resight alternatives in Mark should be used to estimate it.
I have other questions that I'm seeking comment on.
Normally the number of marked animals in a population is known prior to the first resighting survey. That is not the case here - an animal is marked when a photo of it is judged to depict a distinctively marked individual. After the photo assessment and matching of photos to individuals are done, we know the number of marked indiviuals in the population even if we don't know how many distinctively marked (markable) individuals there are.
The two passes through the group seem to have served to permit resights of the same individuals but are not separate surveys. There's only one group in the data.
79 individuals were resighted a total of 93 times with between one and five sightings (including the initial marking) per individual.
The data file has Unmarked Seen = 136 (I assume this is the number of NDMI photographs), Marked Unidentified = 0, and Known Marks = 0.
I welcome your comments.
Thank you,
Lyndon.