Why ecologists describe all food webs as open systems with regard to energy flow

In recent years, food web ecologists have recognized that food webs are open systems – that are influence by processes in adjacent systems – and spatially heterogeneous (Polis et al. 1996). This influence of adjacent systems can be bottom-up, due to allochthonous inputs of resources (Polis & Strong 1996, Huxel & McCann 1998, Mulder & De Zwart 2003), or top-down due to the regular or irregular presence of top predators (e.g., Post et al. 2000, Scheu 2001). However, without a clear understanding of the size of a system and a definition of its boundaries it is not possible to judge if flows are internal or driven by adjacent systems. Similarly, the importance of allochthony is only assessable when the balance of inputs and outputs are known relative to the scale and throughputs within the system itself. At the largest scale of the food web – the home range of a predator such as wolf, lion, shark or eagle of roughly 50 km2 to 300 km2 –the balance of inputs and outputs caused by wind and movement of water may be small compared to the total trophic flows within the home range of the large predator (Cousins 1990). Acknowledging these issues of space, Polis et al (1996) argued that progress toward the next phase of food web studies would require addressing spatial and temporal processes. Here, we present a conceptual framework with some nuclei about the role of space in food web ecology. Although we primarily address spatial aspects, this framework is linked to a more general concept of spatio-temporal scales of ecological research.