The unprecedented impacts of humans on the Earth (e.g. Rockström et al., 2009; Sachs et al., 2010; Foley, 2011) have prompted a debate on the future of land use and, with agriculture as the major anthropogenic use, on new models of agricultural production. In addition to concerns about the eects of current agricultural land use, the debate is fuelled by the predicted growth of the human population, at least until 2050, and the associated need to nourish an extra 3 billion people in the face of a limited unused area of agriculturally exploitable land (e.g. Godfray et al., 2010). Leadley et al. (2014) predict rapid shifts in the state of ecosystems at regional scales within the next decades that it will be dicult to reverse. Assuming business-as-usual socioeconomic development pathways, these shifts will drive coupled human-environment systems to highly degraded states in terms of biodiversity, ecosystem services, and human well-being. Steen et al. (2015) describe how planetary boundaries that delimit a ‘safe space’ for global societal development have been exceeded in the domains of climate change and land-system change and are reaching well into high-risk zones for biosphere integrity (operationalized as genetic diversity) and biogeochemical ows of P and N. For various indicators, it is not the absolute global levels that are a cause for concern, but the speed at which current levels are being reached, which is unprecedented in the Anthropocene (Kidwell, 2015).