Ecologists arguably have been remiss in not developing a formal underpinning for the epistemology of ecology, at least not until the 1980s. At that time, the rather forced imposition of deterministic or heavily constrained stochastic population and community models (see Roughgarden, 1979) drew fire, principally through the emergence of ideas of system “openness” (Wiens, 1984; Gaines and Roughgarden, 1985; Amarasekare, 2000; Hughes et al., 2000; Thrush et al., 2000), non-equilibria (DeAngelis and Waterhouse, 1987; Seastadt and Knapp, 1993) and, especially, “scale” (Wiens et al., 1987; Kotliar and Wiens, 1990; Holling, 1992; Levin, 1992, 2000; Pascual and Levin, 1999). Scales of measurement and observation have tremendous impact on the interpretation of what we think we know about systems and how they operate, which clearly has ramifications for most of the hotly contested areas in community ecology. One such dispute concerns the respective roles of “top-down” (large-scale patterns determine the possibilities for small-scale ones; Whittaker et al., 2001) and “bottom-up” (large-scales are emergent properties of small-scale processes; Wootton, 2001; Ludwig, this volume, Chapter 6) processes in pattern generation in ecological communities (Carpenter et al., 1985).
An increasing number of field studies (e.g., Bowers and Dooley, 1999; Orrock et al., 2000) and simulations (e.g., Bevers and Flather, 1999; Mac Nally, 2000b, 2001) conducted at multiple spatial scales show that outcomes depend upon how the study is constructed and conducted.