Community ecology is experiencing a resurgence, driven in part by its central importance in addressing critical applied problems, ranging from the control of pest and invasive species, to the wise harvest of natural resources, to projecting the impact of global climate change. A fundamental tenet of community ecology is that species do not exist in isolation: they are directly, and more importantly, indirectly interconnected with myriad other species. The essential ‘glue’ that holds communities together and that makes them more than the haphazard sum of individual species is the nexus of indirect interactions among three or more species that emerges from direct interactions such as predation, competition and mutualism between pairs of species. The recognition of indirect effects has triggered a rapid growth of empirical and theoretical research that aims to predict community-level dynamics under different contexts.

Indirect effects occur when the impacts of one species on another are influenced by one or more intermediate species. Indirect effects are diverse, but can be classified broadly as either (1) density-mediated or (2) trait-mediated. Density-mediated indirect effects (DMIEs) result from numerical responses of species to each other. For instance, a fox may kill rabbits, reducing rabbit population size, and so relaxing herbivory upon herbaceous plants. Hence, the fox’s indirect effect on plants is mediated by density changes in rabbits; this is known as a trophic cascade. DMIEs, such as depicted by this trophic cascade, and other mechanisms such as apparent competition between prey mediated by the numerical response of a shared predator, have been well studied and have contributed to our understanding of community organization and ecosystem functioning in both terrestrial and aquatic systems (Holt and Lawton 1994; Polis et al. 2000; Terborgh and Estes 2010). However, the fox may not only kill rabbits, it may alter their behaviour and other traits. For instance, rabbits may hide more in the presence of foxes and so have less opportunity to feed on herbs. This indirect effect of the fox upon the plants could be strong, even though rabbit abundance remains high in the presence of foxes. In this case, it is a change in a rabbit’s trait – altered behaviour to avoid predation – that determines the outcome of the predator–prey direct interaction and the nature of the indirect effect of foxes on plants.