www.phidot.org

Hi,
I fit some models via the conditional likelihood to data collected on linear and curvilinear trap arrays. In many cases, CVs around estimates of esa (a-hat) and density reported by "derived" are high (>0.40), but CVn is reasonable (<0.25).

I assume that the high CVs around the esa are partially attributable to small samples of recaptures at different traps, and possibly also variable movement rates among individuals. However, I was wondering if the curvilinear shape of the trap array (very high edge:area ratio) might also contribute to the high CVesa.

Thanks,

Eric

I would guess it's not the shape per se, but more the limited number of recaptures per animal. If you want to pin down the effect of shape on precision of esa you could try a few simulations with conditional likelihood linear vs rectangular trap arrays at the same spacing. (sim.capthist with e.g. a linear 1 x 20 array and a 4 x 5 array, followed in each case by secr.fit)
Murray