Hi back mate.
I am not sure of what is your issue with RMark. It should be great if you could provide a RePrex (see the link below). The error tht you have can be due to so many things, not necessarily to fitting the re-sighting probability.
But, You should make a vector with the covariate value, no need for any naming long it, just a vector (numeric, s is number of pictures for each secondary occasion). But you cannot use it on primary occasions. There is not such a p (recapture/reshigting probability) there, just survival an emigration/immigration. The p translates to the secondary occasions and is there where you have to decide if use it somehow globally for each primary occasion (and then expanded to all secondary occasions inside, as in my previous example) or let it change for each secondary occasion (that seems sensible as you are taking pictures)..
But anyway, as you link your p to number of pictures, are they of cetacean schools? (dolphins, pilot whales etc). That have an issue of lack of independence on reshigting probabilities, and depending on the final model to be fitted or the actual data, you may have to move to HMM (in R, TMB or Rcpp, hard coding, unfortunately, or outside R, E-Surge, with an user manual quite un-helpful).
Please check the link below
Cheers
Fer