Much of the promise for the sustainable intensification of agriculture needs to come from closing crop yield gaps across underperforming areas [1, 2]. Jägermeyr et al [3] assess the global potential of farm water management to close yield gaps, revealing that calorie production could be increased by 18%–60% on existing agricultural lands. Their analysis defines the biophysical realm of possibility for agricultural production related to water management—offering a new 'benchmark' to assess the potential for sustainable intensification. A growing body of research on the biophysical potential for sustainable intensification raises new questions about how such global supply-side interventions (including, but not limited to, farm water management) might interact with other dimensions of the food system. Here, we consider a series of potential interactions in the context of the findings from Jägermeyr et al for farm water management.

Water management is central to enhancing yields in many regions, with the aim being to get more 'crop per drop' [4, 5]. To this end, Jägermeyr et al [3] consider a mix of four soil-water water management interventions to minimize non-productive water consumption in agriculture (defined as 'integrated crop water management'): reducing soil evaporation for enhanced crop water productivity, capturing surface runoff for supplemental irrigation, increasing soil infiltration capacity for rainfed crops, and improving irrigation system efficiency in order to expand irrigated areas by using the saved water. Reducing soil evaporation has been widely discussed (see Falkenmark and Rockström [6, 7]), but assessment of its role in closing yield gaps globally is novel and informative. Nonetheless, reducing non-productive water consumption in irrigated systems seems to offer the greatest potential crop production increases by providing the water needed to expand irrigation [3]. This is logical given that irrigated yields typically exceed those of rainfed systems—for example, cereal crops in irrigated systems yield ~1.7× more globally than non-irrigated cereals [8].

Jägermeyr et al make important strides in pinpointing the potential to close yield gaps relative to the biological and hydrologic processes of agriculture, for which they outline myriad potential co-benefits, including increased capacity to buffer against the impacts of climate change. Such advancements in global agriculture research provide a solid foundation to consider how biophysical dimensions of sustainable intensification might link to broader social-environmental factors related to all stages of food production, distribution, and consumption. Using the findings of Jägermeyr et al [3] and their 'ambitious yet achievable' scenario of global crop water management interventions, we briefly outline four questions and avenues for future research related to the role of sustainable intensification targets in the global food system more generally (figure 1). These questions are intended to complement the findings of past research by identifying potential barriers to and facilitators of crop water management interventions at distinct scales.

How might ambitious farm water management interventions to achieve global sustainable intensification targets affect local smallholder farming systems?

Jägermeyr et al evaluate their global simulations using national or regional studies on water harvesting, mulching, and soil moisture conservation, but a closer look at local diversity in farming systems is a logical next step. Notably, their simulations involve large tracts of key crops (e.g., pulses, soybean, cotton, coffee, and sunflower) being converted to more efficient drip irrigation [9], which would carry large upfront capital costs. Simple drip irrigation and water harvesting methods may be accessible and greatly beneficial to smallholder farmers [10], but the coordination involved in the most ambitious scenario simulated by Jägermeyr et al may only be achievable with a transition away from small scale family farming toward larger, more commercialized agriculture in key areas where some of the greatest potential productivity gains exist. For example, national agricultural censuses commonly define smallholder and medium-sized farms as being <10>et al find among the largest potential to increase calorie production (e.g., Ethiopia, Pakistan, India, and China) [11]. Other soil water conservation practices (intercropping, mulching or no-tillage agriculture) require more modest investments and are likely more accessible to all farmers.

How might widespread integrated crop water management interventions compete with other uses of land and water at the river basin scale?

Widespread water harvesting for agriculture would likely require a large amount of land—as much as 1 hectare of ponds for every 10 hectares of farmland [12]—representing an important constraint when moving from biophysical models to application of water management interventions. Additionally, while capturing surface runoff for irrigation may be a good idea when viewed solely in terms of global crop production potentials, the appropriate environmental flow requirement thresholds could vary widely across basins (as shown by Gerten et al [13] using different methods to assess aquatic ecosystem needs). Maintenance of environmental flows could have further value downstream given the vital role of inland fisheries in regional food security, such as that of the Mekong River in Southeast Asia [14]. Moves to optimize water for crop production therefore may not serve all food security goals in the same way across all river basins globally.

Could global trade be a mechanism to help scale farm-level water management interventions?

Food trade is increasingly important to both food security and water resource management [15], with considerable populations fundamentally dependent on food imports due to local renewable water constraints [16, 17]. Global irrigation water consumption associated with agricultural exports, including animal and industrial products, is ~301 km3 yr–1 [18]—roughly one quarter of the irrigation water consumption (1268 km3 yr−1) simulated by Jägermeyr et al for their 12 crop functional types. Reducing non-productive water consumption for a few highly traded and water-intensive crops could dramatically reshape the global landscape of virtual water trade. More than half (55%) of virtual irrigation water consumption is linked to seed cotton alone [18]—a 'thirsty' crop in comparison to cereals and other food staples that, despite its economic value, has a negligible role in global food production. Additional research is needed to understand how trade patterns and total water consumption might change under widespread crop water management targets [19], particularly in sub-Saharan Africa, where yield gaps are pronounced and market access is a major barrier to sustainable intensification [20]. At the global scale, improved crop water management in export-producing countries could reinforce the water savings potential of international food trade resulting from relative gains in water-use efficiency among trade partners [21].

How does sustainable intensification compare to more demand-side interventions?

Food security is not guaranteed by increased food production alone—it encompasses multiple complex factors that determine accessibility to food at different scales, including purchasing power, socio-political context, and access to distribution channels [22]. Growing demand for more resource-intensive foods exerts increasing pressures on land and water [23], and contributes disproportionately to the 60%–100% projections of future crop calorie demand targeted by Jägermeyr et al [3]. Shifting diets away from animal protein in favour of local crop-derived proteins has the potential to reduce irrigation water consumption by as much as 14% [24] while substantially cutting supply chain food losses and waste could reduce it by 12% [25]. Diets, waste, and governance are therefore critical determinants of the extent to which crop production actually needs to increase in order to ensure global food security [1, 22].

A critical challenge for agriculture this century will be to do more with less. Global biophysical modelling studies, including that of Jägermeyr et al, offer vital benchmarks to help assess supply-side strategies toward a more sustainable and productive food system. A crucial next step to harness the immense potential of crop water management and other more supply-side strategies in achieving this goal is to assess location-specific pathways to sustainable intensification based on farming system characteristics [20]. Synthesis of local case studies could assess the appropriateness of specific farm management interventions and account for the cultural and economic diversity of smallholder farming systems [26]—possibly favouring less effective but more feasible soil water conservation techniques in some areas. Water funds could be a critical policy mechanism to help incentivize farm water management interventions within individual basins [27], and such markets could help to moderate the potential tradeoffs with downstream water users described above. The role of agricultural trade as both a facilitator and outcome of sustainable intensification targets also warrants further assessment. Since consumers in virtual water importing nations benefit from water resources in exporting countries, novel international water funds or payment for ecosystem services schemes could be a further policy consideration to integrate supply- and demand-side strategies in agricultural water management (figure 1). Improved crop water management will be essential to help address global food production and freshwater availability this century. Further untangling the diversity of local pathways needed to scale integrated crop water management is a key next step for global research.

For references, see Pathways to sustainable intensification through crop water management at environmentalresearchweb's sister journal ERL.