Confronting spatial capture–recapture models with realistic animal movement simulations