Global urbanization and food production are in direct competition for land. This paper carries out a critical review of how displacing crop production from urban and peri-urban land to other areas – because of issues related to soil quality – will demand a substantially larger proportion of the Earth’s terrestrial land surface than the surface area lost to urban encroachment. Such relationships may trigger further distancing effects and unfair social-ecological teleconnections. It risks also setting in motion amplifying effects within the Earth System. In combination, such multiple stressors set the scene for food riots in cities of the Global South. Our review identifies viable leverage points on which to act in order to navigate urban expansion away from fertile croplands. We first elaborate on the political complexities in declaring urban and peri-urban lands with fertile soils as one global commons. We find that the combination of an advisory global policy aligned with regional policies enabling robust common properties rights for bottom-up actors and movements in urban and peri-urban agriculture (UPA) as multi-level leverage places to intervene. To substantiate the ability of aligning global advisory policy with regional planning, we review both past and contemporary examples where empowering local social-ecological UPA practices and circular economies have had a stimulating effect on urban resilience and helped preserve, restore, and maintain urban lands with healthy soils.

The term “soil health” has captured the interest of government, and land managers, whilst the academic community has struggled to rationalise its use and wider benefit. It has proved a powerful tool in conveying best practice to a lay audience. However, the widespread adoption of the “metaphor” has resulted in calls for tools that facilitate the measurement of soil health, preferably quantitatively, and often as a single figure, for ease of use/communication and cost of monitoring. The insurmountable problem is that soil health is neither a readily quantifiable nor measurable object. Only organisms can have ‘health’, which manifests as characteristics of a living system—true of complex systems exhibiting “emergent” properties such as resilience in the face of perturbation. We pose the key question: is soil really a system capable of exhibiting “health”, or any other property emerging from a complex, connected, self‐regulating system? We argue that if you cannot detect emergent properties, you are: (i) looking at the wrong dynamic parameter; (ii) not considering the entire system; or (iii) not evaluating a system at all. We suggest that our focus should instead be on the relationships between components, complexity, and function. Using this as a basis for a new framework will allow us to assemble and align disparate threads of soil science into a cogent and coherent “new theory of soil health”, which is an essential and practical step forward for the sustainable management of global soil resources, across all land uses. Highlights: The term “soil health” is widely used, but understood and used in different ways. Health is a characteristic of an identifiable entity – usually an organism or discrete system. To identify and measure soil health we need to discover its system properties and their dependence on complexity and connectivity. The scale at which system properties emerge is currently unclear. We suggest a broad programme of research to identify dynamic emergent properties, especially resilience, in order to determine soil health.

The importance of soils to society has gained increasing recognition over the past decade, with the potential to contribute to most of the United Nations’ Sustainable Development Goals (SDGs). With unprecedented and growing demands for food, water and energy, there is an urgent need for a global effort to address the challenges of climate change and land degradation, whilst protecting soil as a natural resource. In this paper, we identify the contribution of soil science over the past decade to addressing gaps in our knowledge regarding major environmental challenges: climate change, food security, water security, urban development, and ecosystem functioning and biodiversity. Continuing to address knowledge gaps in soil science is essential for the achievement of the SDGs. However, with limited time and budget, it is also pertinent to identify effective methods of working that ensure the research carried out leads to real‐world impact. Here, we suggest three strategies for the next decade of soil science, comprising a greater implementation of research into policy, interdisciplinary partnerships to evaluate function trade‐offs and synergies between soils and other environmental domains, and integrating monitoring and modelling methods to ensure soil‐based policies can withstand the uncertainties of the future. Highlights: We highlight the contributions of soil science to five major environmental challenges since 2010. Researchers have contributed to recommendation reports, but work is rarely translated into policy. Interdisciplinary work should assess trade‐offs and synergies between soils and other domains. Integrating monitoring and modelling is key for robust and sustainable soils‐based policymaking.

Arable soils are critical resources that support multiple ecosystem services. They are frequently threatened, however, by accelerated erosion. Subsequently, policy to ensure their long-term security is an urgent societal priority. Although their long-term security relies upon a balance between the rates of soil loss and formation, there have been few investigations of the formation rates of soils supporting arable agriculture. This paper addresses this knowledge gap by presenting the first isotopically constrained soil formation rates for an arable (Nottinghamshire, UK) and coniferous woodland hillslope (Shropshire, UK). Rates ranged from 0.026 to 0.096 mm yr⁻¹ across the two sites. These rates fall within the range of previously published rates for soils in temperate climates and on sandstone lithologies but significantly differed from those measured in the only other UK-based study. We suggest this is due to the parent material at our sites being more susceptible to weathering. Furthermore, soil formation rates were found to be greatest for aeolian-derived sandstone when compared with fluvially derived lithology raising questions about the extent to which the petrographic composition of the parent material governs rates of soil formation. On the hillslope currently supporting arable agriculture, we utilized cosmogenically derived rates of soil formation and erosion in a first-order lifespan model and found, in a worst-case scenario, that the backslope A horizon could be eroded in 138 years with bedrock exposure occurring in 212 years under the current management regime. These findings represent the first quantitative estimate of cultivated soil lifespans in the UK.