Within ecology animal movement is often studied in disjunct sub-disciplines narrowing on specific problems such as experiments on homing behavior or theoretical considerations of optimal search strategies. We outline a unifying framework integrating individual-level mechanistic movement behaviors and population-level spatial distributions in relation to underlying spatio-temporal dynamics of resources. We distinguish among (1) non-oriented movements based on diffusion and kinesis in response to proximate sensory stimuli, (2) oriented movements relying upon perceptual cues, and (3) memory mechanisms that assume prior knowledge about target locations. The relative importance of these mechanisms should depend not only on life-history traits but also on resource dynamics, which ultimately shape the population-level patterns. For example, well-dispersed resources should facilitate sedentary ranges, whereas resources that are predictably distributed in space but vary seasonally should cause migratory patterns.  A third pattern is 'nomadism', which should emerge when resource distributions are unpredictable. Drawing upon recent advances in relocation data from animal tracking and analyses techniques for animal movements we suggest a combined program to build stronger links between underlying movement mechanisms and broad-scale patterns of population distributions.