4 Patterns and Impacts of Movements at Different Scales in Small Mammals JAMES E. DIFFENDORFER, MICHAEL S. GAINES, AND ROBERT D. HOLT Introduction Understanding how patterns of movement are affected by the spatial structure of an environment is a key question for landscape ecologists (Wiens 1995). All organisms exist in spatially heterogeneous environments and movement through these mosaics clearly has impacts on individual fitness (Fretwell and Lucas 1970, Morris 1992), population demography (Pulliam 1988, Hanski and Gilpin 1991, Pulliam and Danielson 1991) and community structure (Connell 1961, Danielson 1991, 1992, Holt 1993, McLaughlin and Roughgarden 1993, Abramsky et al. 1994, Brown 1996). Ostfeld et al. (1996) consider heterogeneity to be a unifying hypothetical foundation in ecology and conservation. Despite recognizing heterogeneity, we do not fully understand how move- ments interact with spatial mosaics to influence population-level processes (Wiens 1995). Theoretical studies have shown that movements over hetero- geneous landscapes, even as simple as two distinct habitat types, can gener- ate a multitude of dispersal patterns and population dynamics (Holt 1985, Pulliam 1988, Pulliam and Danielson 1991, McPeek and Holt 1992, Dias 1996). Movement is a common denominator in the understanding of how spatial heterogeneity influences ecological processes. Thus, landscape ecolo- gists need to focus on how spatial heterogeneity influences movement pat- terns and how movements, in turn, influence spatial patterns of abundance across space. A number of authors have delineated conceptual frameworks for studying the interaction between spatial heterogeneity and movement pat- terns (Senft et al. 1987, Gautestad and Mysterud 1993, Ims 1995, Morris 1995), with all emphasizing that movements are influenced by factors at different spatial scales. We should expect that as scales change, different processes cause movements and impact population demography in a variety of ways. In fragmented or patchy landscapes, movement may reflect the size and spacing of habitat patches.. We suggest studying movement at three spatial 63