Introduction
The ways in which productivity, stability, population interactions, and community structure are regulated in ecosystems have been a central focus of ecology for over a century. At large spatial scales, major insights into these dynamics have been principally derived from analyses of changes induced from hunting, harvesting, and agricultural practices – so-called “natural experiments.” In terrestrial ecosystems estimates of the fraction of land transformed or degraded by human activity fall within the range of 39 to 75% (Vitousek et al., 1997; Ellis et al., 2010). Equally profound is the reality that up to 75% of the global oceans and in particular the continental shelf, transitional slope water areas, and reef habitats have been strongly impacted by human activity (Halpern et al., 2008).
One of the most widely studied human impacts has been the over-exploitation of large-bodied species. Berger et al. (2001) estimated that the spatial distribution of large mammalian carnivores that once played a dominant role in terrestrial ecosystems has declined by 95–99%. In the global oceans large predatory fish biomass may be as low as 10% of pre-industrial levels (Myers and Worm, 2003). These changes have created a vertical compaction and blunting of the trophic pyramid (Duffy, 2003; Chapter 14, this volume). On a global scale, these losses are attributable to a positive association between body size and sensitivity to population declines experienced by larger species which exhibit a greater susceptibility to decline or collapse as a consequence of their lower population densities, greater times to maturity, lower clutch sizes, and larger home ranges (Schipper et al., 2008). This reduction in the abundance of apex predators has led to abnormally high densities of their former prey in a wide range of ecosystems, which has, in turn, resulted in sometimes catastrophic changes in the ecosystems occupied. This has led some to conclude that large-bodied species are essential to the maintenance of ecosystem structure and stability (Hildrew et al., 2007; Estes et al., 2011).