The effect of spatial and temporal heterogeneity on the design and analysis of empirical studies of scale-dependent systems.

Processes interacting across scales of space and time influence emergent patterns in ecological systems, but to obtain strong inference and empirical generalities, ecologists need to balance reality with the practicalities of design and analyses. This article discusses heterogeneity, scaling, and design analysis problems and offers potential solutions to improve empirically based research. In particular, we recommend bridging the dichotomy between correlative and manipulative studies by nesting manipulative studies within a correlative framework. We suggest that building on variation, by designing studies to detect variability, rather than fighting it often leads to an increase in generality. We also emphasize the importance of natural history information for determining likely scales of spatial and temporal heterogeneity and the probable occurrence of feedback loops, indirect effects, and interacting processes. Finally, we integrate these concepts and suggest planned iterations between multiscale studies to build up natural history information and test the strength of relationships across space and time. This offers a way forward in terms of heuristically developing models and determining ecological generalities.