The LFF hypothesis and the MRW-based variant both postulate a transition towards constrained area use with increased resource density. While LFF describes this constraint by relatively smaller “diffusion rate” from a transition towards BM, MRW describes this constraint by stronger site fidelity relative to a given sampling scale under infinite or constrained memory horizon. The memory horizon regards how far back in time the animal choose previous locations for return steps (infinite horizon means that all previous locations are considered). (A) Spatial fixes from MRW with very narrow memory horizon (LW-like, left display) and a correlated random walk emerging as BM by sampling fixes at tobs = 103 (right display). (B) Four scenarios under MRW and infinite memory horizon, where frequency of return steps is increasing by a factor of 10 per display from left towards right. The embedded red line in the lower part of each panel shows the spatial scale (for example, 1 km) for the respective arenas. For a given accumulated movement length (total path), BM covers a smaller area than a LW. Similarly, MRW with a small return interval on average covers a smaller area than under condition of a larger return interval.