Spatial-dependent single-season model--- why is p = 1.00?

[lpetracca]

Hi everyone,

I'm currently analyzing data from a lion survey in West Africa. We have completed 15-km transects in 98 sampling units of 225 km2, recording presence/absence of lion spoor every 500 m.

After running a null model, we find that detection probability (p) is estimated at 1.00. This estimate does not change, even when we change our segment length. Surely we do not have perfect detection probability.

I should mention that we do not have an equal number of segments in each grid, as we superimposed the grid following the survey--- this is obviously a flaw in study design. We therefore have some cells with 2 segments and others with 50-70 segments. Could this be the culprit for such a strange result?

Thanks in advance for any help. Still trying to learn the intricacies of these models.